PDA

View Full Version : Mathhammer/Theoryhammer



ColonalKlink
01-03-2010, 05:18
I was on the Thread for the venomthrope, and i expressed my opnion that i dislike math/theoryhammer, and how neither factor into how i play my games, i do not look at an enemy unit and go " i have blah blah blah chance of hitting that, and blah blah chance to woulnd etc.".

so i was wondering what others views on these were?

go for it ladies and gents :P

WinglessVT2
01-03-2010, 05:22
They both factor, but not as much as some people pretend.
The entire point of them is that if you roll a large number of dice a large number of times, you'll see patterns appear, which you can use to base your plans around, as well as gauge 'average' impact of your units - but this doesn't mean that the first 10 times you roll four six-sided, they'll always come up as two 4+, and two below.

Juggalo
01-03-2010, 05:25
So you look at that Land Raider, and go "I don't care what the odds are of my boltguns killing it, I'm going to shoot away! DAKKA DAKKA DAKKA! Damn, failed to kill it. Stupid dice, luck must not favor me today."?

Whatever.

SandQueen
01-03-2010, 05:27
Im somewhat bad at math. I operate on the priciple that "If I dump 10 death spinners into that for three turns and I only need 2s to wound then it should be dead"

Not really mathhammer so much as Inevitablehammer

vladsimpaler
01-03-2010, 05:27
-The catbarf symbol spotlight is turned on-

carldooley
01-03-2010, 05:31
Im somewhat bad at math. I operate on the priciple that "If I dump 10 death spinners into that for three turns and I only need 2s to wound then it should be dead"

Not really mathhammer so much as Inevitablehammer

except that death spinners don't have an AP value, which means that you are going to be more killy versus kroot in the open than against a SM command Squad with an Apothecary.

velkore1134
01-03-2010, 05:36
If I have a weapon that can cause an effect on something, I shoot, I don't go "hmm 33.3, repeating of course, chance of survival. Oh **** Leeroy just ran in" etc.

But I won't send my gaunts on a Leman Russ

azimaith
01-03-2010, 05:39
Everyone uses "mathhammer" if we take mathhamer as a synonym for basic statistical analysis in 40k. Whether we consciously do it or not is irrelevant. You would be unable to function at all without the ability to evaluate basic risk vs reward.

The moment I ask someone "Why are you shooting those four lascannon devastators at my gaunts rather than at my Hive Tyrant?" And he answers seriously:
"There's no difference between me shooting my lasguns vs my lascannons at the hive tyrant I can see!" That will be the first person to be addled enough to truly not use "mathhammer."

Without the concept of probability you wouldn't be able to tell the difference between a space marine or a guardsmen. They'd be identical as far as you were concerned.

WinglessVT2
01-03-2010, 05:41
Mathhammer and theoryhammer aren't substitutes for common sense.
No one who's qualified to even play the game is going to shoot their basic guns at a land raider.

MasterDecoy
01-03-2010, 05:45
While im fairly decent at math, I try not too look into it too hard at any one time.

Seriously you dont need to be a genious to know that you fire the high strength stuff at tough things and the mass fire at more nemerous things.

As long as there is a statistical chance I can actually get the result I want, Ill take those odds every time. Boltguns at that walker, yup no problems, rending genestealers vs a dreadie, bring it on. lasguns vs a wraithlord, your an idiot.


Also this:

Mathhammer and theoryhammer aren't substitutes for common sense.
No one who's qualified to even play the game is going to shoot their basic guns at a land raider.

carldooley
01-03-2010, 05:45
everyone uses mathhammer, even if it is only used at an instinctual level. Am I going to move in this direction because I can escape this unit that I am extremely unlikely to kill or extremely likely to wipe out. TH\SS terminators assaulting a unit of pariahs is STUPID, as they will likely strike first and kill the terminators before the terminators kill the pariahs. the previous example is why we use vindicators - no need to close and can kill the entire squad in one shot.

but wait, why use a vindicator when you can use a whirlwind? less points for greater range AND the ability to fire indirectly. Why is the Vindicator better? mathhammer.

Loki73
01-03-2010, 05:55
So you look at that Land Raider, and go "I don't care what the odds are of my boltguns killing it, I'm going to shoot away! DAKKA DAKKA DAKKA! Damn, failed to kill it. Stupid dice, luck must not favor me today."?

Whatever.

Lame response as that falls under common sense really. I think the OP intention is regarding the amount of x will kill x statistics and percentages and statistical analysis regarding dice rolling.

It gets old and is lame. The thing I hate is when people say X unit does this better than X unit. Take what you want and find a use for it. I only use average die rolls for things during battle thats it. Even then its rough and im right 90% of the time.

Occulto
01-03-2010, 05:56
Everyone uses math-hammer. I just find those who base every single decision solely around math-hammer to be incredibly dull.

I also find it laughable when someone says that one option is crap because a different weapon has a 2% better chance of success.

Iracundus
01-03-2010, 05:59
Lame response as that falls under common sense really. I think the OP intention is regarding the amount of x will kill x statistics and percentages and statistical analysis regarding dice rolling.


All "common sense" that people keep mentioning in this thread really is is unconscious rough and error prone statistical analysis. Just because people are unaware they're analyzing the probabilities under the guise and name of "common sense" doesn't mean they are not doing so.

SandQueen
01-03-2010, 06:10
except that death spinners don't have an AP value, which means that you are going to be more killy versus kroot in the open than against a SM command Squad with an Apothecary.

If you force any unit to make more saves than it has models over and over and over then its going to start hurting.

Most of the guns in my army dont have really good AP values but I have no issue against Marines. 3 Guided Warwalkers, each with dual scatterlasers can erase the better part of a tactical squad each and every turn just by raw weight of firepower. I dont need AP3 and AP2 if I get more than twenty twin-linked shots that need 2s to wound. Sooner or later you'll roll bad and things will start dieing. And if you dont then that's what the other three warwalkers are for.

Hence 'inevitablehammer'

ColonalKlink
01-03-2010, 06:20
i think most of oyu have a point, it may be unconcious(sp?) but i dont base my tactics or playing style around it, and saying "would you shoot a land raider with a boltgun" is tripe thats not odds, thats simple maths even i can do, str4 cant hurt av 14, im talking 20 boltguns shot at marine unit A 20% hit blah blah sort of thing then baseing my force around the stuff that the odds say are the best, just boring.

i know this works for some just not me :)

SandQueen
01-03-2010, 06:25
i think most of oyu have a point, it may be unconcious(sp?) but i dont base my tactics or playing style around it, and saying "would you shoot a land raider with a boltgun" is tripe thats not odds, thats simple maths even i can do, str4 cant hurt av 14, im talking 20 boltguns shot at marine unit A 20% hit blah blah sort of thing then baseing my force around the stuff that the odds say are the best, just boring.

i know this works for some just not me :)

alot of mathhammer, in my experience, goes hand in hand with rules lawyering and generally being a ******. If you're crunching numbers insteead of enjoying the 41st millennium then you're doing it wrong

WinglessVT2
01-03-2010, 06:30
I don't need to do it consciously.
It's second nature, and requires no effort - like knowing the profiles of every single thing in the game.

ColonalKlink
01-03-2010, 06:34
ah rule lawering... now thats a can of worms, kinda like the bs that is RAW...

Juggalo
01-03-2010, 06:35
Lame response as that falls under common sense really. I think the OP intention is regarding the amount of x will kill x statistics and percentages and statistical analysis regarding dice rolling.
Oh really, and why is that common sense? Because you recognize the statistical probability of succeeding in that action is 0. Let's not play with semantics.


ah rule lawering... now thats a can of worms, kinda like the bs that is RAW...
Now people that know the rules are rules lawyers? So you don't like to consider the odds of a particular action succeeding, and you don't like to know the rules of the models you're playing with. You're the poster child for terrible 40K players. I bet you cry cheese whenever you lose, too.

Fobster
01-03-2010, 06:51
If you don't use math and probability your a bad player.

ColonalKlink
01-03-2010, 07:19
Now people that know the rules are rules lawyers? So you don't like to consider the odds of a particular action succeeding, and you don't like to know the rules of the models you're playing with. You're the poster child for terrible 40K players. I bet you cry cheese whenever you lose, too.

Eh? you lost me there, im talking rules lawering in general, not math hammer, as i have already said "if it works for some people fine" just not for me.. and i didnt call anyone a rules lawer. as to not considering the odds of something suceeding, i said i dont build my game around the odds before hand (i.e choosing a unit just because it is statiscally(sp?) better)
and none of this has anything to do with not knowing the rules for the game or units... i mean where the hell did you get that from??? as a poster child for terrible 40k players.. nope i won sportsmanship awards 2 tourny's in a row, precisly because i dont care and am there to have a good time not compete.

just to clarify lol, rules lawering= Mawloc cant DS into enemy units, monolith cant shoot the turn it DS's Doom can kill things in transports (these are examples not my opinion)
Mathhammer= stats, odds, probability, effectivness (i think this is what it means anyway, reading the posts so far?)
Rules= Fleet of foot, rending, 24" single shot while standing still with rapid fire weapons.

WinglessVT2
01-03-2010, 07:42
But the mawloc can deep strike precisely wherever it wants to - even on top of units.
If a normal unit ends up on top of units, it has to roll for mishap, but if the mawloc does this, it triggers a special ability.

The monolith can't shoot when it appears, because the rules forbid this.

carldooley
01-03-2010, 07:54
I suppose that you could try for one of those cinematic moments - a tactical marine climbs on top of an enemy Land Raider, passes his strength test, opens one of the top hatches and drops a frag grenade into the crew compartment. . .

but this is a game, and if you want to try that at home that is fine - just don't expect anyone to allow it in an actual game.

enigma-96
01-03-2010, 08:01
Wow some of the mathammer supporters oddly enough seem incapable of making a logical deduction of the OP's original intent, ironic ain't it ;), yes the OP kind of worded things badly but it is VERY obvious that he isn't asking about everyday, subconscious mathhammer but about the conscious, thought out mathhammer, I mean really if anyone brings up the idea that we all use mathhammer, albeit subconsciously, you aren't necessarily wrong but that isn't answering the OPs question so what's the point of constantly bringing it up?

Likewise the way I see it mathhammer doesn't equal logic, because that is what subconscious 'mathhammer' would be, because then people would just call it logic, but no they call it mathhammer because it has a characteristic that makes it distinct from logic, in this case it only applies to warhammer (Obviously) and is on a conscious level where you are aware it is being performed.

Anyways, my view of it is that a general ability to 'range find' a percentage, I.E. to look at a situation and get the general gist of your chances in a matter of seconds is a good skill too have and is very helpful for when you have to decide between two courses of action. On the other hand their are people who mathhammer the crap out of way too many, read: all, situations which really slows the game down and kind of makes them out to be a douche. Now if they can do the math quick in their head go for it congratz on having a quick brain :p but I see too many people take a minute or more per situation, just to do the math not to do anything else, and it REALLY slows a game down to the point that it becomes very frustrating.

So no I don't particularly like mathhammer users unless they don't mind a few errors or they are really quick at it, so like 10% of the total ;). Though in list building they can use it all they want it is their time after all.

ColonalKlink
01-03-2010, 08:14
thank you enigma, reading my OP you are right.. bad wording didn't help.

As to wingless, the monolith can shoot its gauss blasters at any unit within 12" as per its rules, the big book forbids it shooting is it moves at cobat speed, which DS counts as, but its entry says it can.. ad infinatum...

now the mawloc is just something i saw being argued on warseer i do not remember the exact details, but the rules lawers were arguing that it could not burst up into a unit.

same with the doom, rules lawers think it can affect units in transports.

like i said these 3 are just examples of the rules lawering i have seen, NOT my opinion, and frankly the latter 2 are crazy.. in my opinion

Ironmonger
01-03-2010, 08:14
Really? One of these threads didn't turn up in a search? We were just talking about this, like... last week...

Anyways, 'mathhammer' is brilliant on paper, pants when that 2+ you needed turns up a 1. As for the 'mathhammer' partisans, I only get annoyed when they whine about why they 'shoulda' won but didn't 'for some reason...'

Seriously... am I caught in a Lost-esque time-warp, or were we just talking about quantum mechanics?!:wtf: I keep waiting for Jeremy Davies and Jeff Goldblum to burst through my door...:shifty:

EDIT: @SandQueen: Inevitablehammer=BRILLIANT! You can read about it right next to 'Commonsensehammer,' 'DUR!hammer,' and 'Letsworkonyouhavingfuninsteadoftryingtodefendyour mathshammer.'

Jagd44
01-03-2010, 08:15
I have been playing 40K for almost 2 years now. I think that most people, to a certain degree, use some applications of Mathhammer and/or Theoryhammer. An Imperial Guard player, like myself, generally uses lascannons against heavily armoured vehicles and monstrous creatures and lasguns against infantry. However, one could use lascannons against infantry and lasguns against monstrous creatures depending on the situation.

It never ceases to amaze me when I hear the more competitive players at my gaming store complaining that some shooting attack, close combat, cover saves, etc. did not comply with what should have happened statistically. They will mumble that it was lower than average, higher than average or that should not have happened as bad as it did. Causing six wounds to a Terminator Squad does not necessarily mean a dead Terminator, although a 67% chance that it will is still in your favor.

The problem is when certain players become adamant about certain units over others. Even more so when they start to question your choices for your army (The Basilisk vs the Medusa vs the Colossus is one of those arguments that I just cannot seem to get away from). They treat the results of Mathhammer and/or Theoryhammer as the be-all, end-all answers for their decisions and you have no idea what you are talking about. However, probability is more reliable than being able to roll dice well.

SandQueen
01-03-2010, 09:13
EDIT: @SandQueen: Inevitablehammer=BRILLIANT! You can read about it right next to 'Commonsensehammer,' 'DUR!hammer,' and 'Letsworkonyouhavingfuninsteadoftryingtodefendyour mathshammer.'

I do have fun with it, the shocking thing to me is that no one at my LGS seems to know how it works. To me its common sense, you lay down more shots then your opponent has models then something is going to die sooner or later. My usual army has the warwalkers, some dire avengers, warp spiders, guardians...all things that can lay down waves of firepower and then run away. Almost every other Edlar army at my LGS is Wraithguard/Wraithlord heavy or a circus of aspect warriors.

WinglessVT2
01-03-2010, 09:19
A blunt approach is often the most elegant and functional, both in execution and reliability.

You can count on people countering your tricks with tricks of their own, but there are no tricks available that counter superior volumes of firepower, or mêlée choppyness.

yabbadabba
01-03-2010, 09:30
I suppose that you could try for one of those cinematic moments - a tactical marine climbs on top of an enemy Land Raider, passes his strength test, opens one of the top hatches and drops a frag grenade into the crew compartment. . .
but this is a game, and if you want to try that at home that is fine - just don't expect anyone to allow it in an actual game. What you have described is an actual game, just with added house rules.

I think Math-hammer is distinct from just playing 40K in that people design armies based on maximising the chance of a thing in the gaming doing an action. The difference for me, is in taking 1 lascanon as an anti-tank threat, to taking an all missile launcher army, to taking a mech/melta army. In this you see the rise of one-trick-pony armies and extreme-theme armies. I can use Math-hammer but I prefer just taking an army I like the look of.

Sgt John Keel
01-03-2010, 09:53
Likewise the way I see it mathhammer doesn't equal logic, because that is what subconscious 'mathhammer' would be, because then people would just call it logic, but no they call it mathhammer because it has a characteristic that makes it distinct from logic, in this case it only applies to warhammer (Obviously) and is on a conscious level where you are aware it is being performed.

As far as I know, math-hammer is a derogatory term for statistical calculations as applied to the Warhammer rulesets. It is a sub-set of mathematical logic.

They way I see it, if you're breaking out a calculator during the game or stalling it while doing calculations in your head, you're doing it wrong. On the other hand, using it when discussing on Warseer or while making army list is an excellent application. That your gut feeling should be more accurate than correctly done statistical analysis is unlikely (pun indended), and if it is, good for you.

It is also imperative to understand that the application of 'math-hammer' needs to defer to the tactical situation, in the sense that most statistical analysis performed here is very limited in scope.

Obviously, as evidenced in anecdotes, people who are saying 'it shouldn't have gone that way' haven't really understood anything. Feel free to lecture them.

That said, I'm certainly no advocate for letting just 'math-hammer' dictate choices, whether it's in the army list stage or when choosing what to charge. Æsthetics is in many situations at least as important, and I think that needs to be recognised*. Taking a unit or an option because you think it's cool is just as valid a reason as taking it because it performs well.

*Yes, taking a Punisher because 'firing a lot of shots is awesome' is an æsthetic decision.

Sekhmet
01-03-2010, 10:28
No one who's qualified to even play the game is going to shoot their basic guns at a land raider.
Necron players do.


A blunt approach is often the most elegant and functional, both in execution and reliability.

You can count on people countering your tricks with tricks of their own, but there are no tricks available that counter superior volumes of firepower, or mêlée choppyness.
T8 is a very good trick to counter superior S4 and lower volumes of firepower, as is Av14 vs S7 and below. That being said, I almost always utilize volumes of firepower because I have no low AP weapons to speak of.

The way I see it, "mathhammer" is the usage of statistics to dictate your every move in 40k, from army selection to unit selection to what you shoot at in each phase. The usage of statistical information is simply "statistics." Mathhammer is what let us know that 4th ed Assault Cannons are better than Lascannons, assuming both are in range, against Av14 (besides the Monolith) and infantry, and thus a better buy in every way, which led to the nerfing of rending. Sure, people could have figured it out after years of play time, but all it took was 2 minutes of calculations.

Statistics plays a part in everything you do in life. Playing 40k using statistical information is the best way to play the game, but you cannot rely on it. If your Marine has a 50% chance to kill something, you don't simply throw two Marines at the target. If your living god has a 80% chance to kill that enemy HQ model in one round of combat, place a good deal of faith in it happening, but have a backup plan just in-case that 20% occurs.

How I personally play 40k is by a basic and quick statistical analysis (what unit should I use against what other unit to maximize my risk:reward ratio each turn) which includes a pretty good idea of what each unit in my opponent's codex does, followed by two to three layers of backup plans and goals set in motion. Unlike many people on Warseer, I don't play 40k with a pint in-hand for a good laugh with my mates. I paint when I feel like it, but I don't let people pressure me to doing it. I partake in the fluff by reading the books and occasionally playing Dark Heresy. I play 40k for the tactical challenge. Unless the game is already decided and the last turn or two is a chore of mopping up, I get extremely annoyed if someone knowingly makes a tactical blunder just "for fun," regardless of who's winning.

neko
01-03-2010, 10:32
As Sgt John Keel says, math-hammer is a useful tool. Most of the time I see math-hammer taking place is when people are analysing units, and if you ask me that's exactly the right time to be making use of statistical analysis. If anything, the real problem is that people try to use math-hammer without understanding how statistics actually work.

Quick'n'dirty math-hammer also has a place during the game, but really only if you're any good at working these things out on the fly.
An example would be firing something like a Punisher cannon at a squad of Teminators. It's tempting to think "It's a Heavy 20 weapon! It'll tear through them just with weight of fire!!", but it only takes me a few seconds to come up with:
20 shots > 10 hits > ~6-7 wounds > ~1 kill

It's worth noting that this uses what are called expectancies, and is just used to give a basic idea of what I can expect to average. If someone is doing this whilst thinking "I will hit with 10 shots, I will wounds with 6-7 shots", they're doing it wrong.

Archangel_Ruined
01-03-2010, 10:33
Everyone uses probability, all the time. It's how we're programmed, if you don't think you're good enough at maths then great news, you're brighter than you thought. You might not think in neat little 16.66% steps but you do know what's more likely. People who overanalyse these things usually aren't that good at the game anyway (shock horror, prepare for incoming abuse...), if you focus in on each action to that degree then your chances of keeping an eye on everything else rapidly decrease. That's the beauty of chess, no probability as such to worry about, any piece can take any other so you're just left to wonder where to put the damn things, that's what makes it the par excellence of tactical gaming.

Corrode
01-03-2010, 11:02
A lot of people in this thread seem to be under the impression that 'mathhammer' means 'sitting and working out the odds of every action you are about to make during a game'. That's not only silly, it should also be totally unnecessary, because anyone who uses mathhammer should already have done the maths before the game ever starts. It's a useful technique at the list-building stage, when choosing whether to take one choice or another - if they have similar roles, then compare them when performing that role and see which is more likely to succeed. This isn't the only thing one should consider when list-building, of course, but it's one thing.

Coming back to what I said about having done the maths already, what I mean by that is that it shouldn't be necessary for me to work out the odds of a melta vs. a lascannon in-game. I should already know that a multi-melta in 12" is more likely to pop a Land Raider than a lascannon is; I don't need to know the exact statistical probability at that point because it's irrelevant.

A lot of people also seem to have a real beef with mathammer whilst only having seen it misapplied. As has been said above, probability doesn't work like 'if I roll 20 shots at BS3 that means I'll hit 10 times', and people saying that are misguided. Equally, it's no good saying 'omg my melta will kill a Rhino 90% of the time, but this autocannon will only do it 50% of the time, the melta is better!' if the melta platform is a walking Tactical Marine and the autocannon platform is a Dreadnought. Mathhammer usually offers either best-performance comparisons or equivalent-situation comparisons (e.g. we assume the meltas are in short range, which lets us see them at their most useful, or we assume that both the melta and lascannon are at equal range, which lets us see them in equivalent positions), tempered with a nod towards the things you can't model like 'is this weapon more likely to get into range because it's mounted on a Land Speeder rather than this one that's on a walking Tactical Marine?'

tl;dr theoryhammer is a good way of working out expected performances and for matching weapons to targets. The results are an average and therefore demonstrate likelihoods, not certainties. It should be tempered with real-world considerations and is no substitute for actually knowing how to play the game. People who cry because their six meltas failed to even hit a Land Raider and 'at least 4 of them should have!' don't understand probability and have failed to use theoryhammer correctly.

ColonalKlink
01-03-2010, 11:04
-------------->ABUSE<-------------------

didnt want you disappointed angel

that actually made sense corrode, anyway, my beef was that when i expressed my disinterest, i had some people giving me abuse and generally being douches, even after i stated "its ok for some just not for me", these people give the whole hobby a bad name not just mathhammer

Epicenter
01-03-2010, 11:18
There is nothing inherently wrong with Mathammering or Theoryhammering stuff; especially for discussions on boards when you're making estimates for strategy when not actually playing the game (I think most of us posting here are not playing 40k and posting at the same time). It provides a common platform of "hard" statistics when comparing different units instead of the use of anecdotal evidence which frequently involves the citing of outliers and similar extreme cases, or on occasion outright lies. So for me, it's the only way to discuss the effectiveness of units; but it has limits that have to be kept in mind.

Are there people who take it extremes that I find ridiculous? Sure.

When people vehemently argue that something is "utter garbage" when the second place choice is 0.21% better, people need to step back and take a breath. People who think that "statistically unlikely" means "never" or "statistically likely" means "always" are asking for disappointment and frustration as well. I find that during any 40k game, at least one thing that is "statistically unlikely" will happen or "statistically likely" will not happen.

Gaargod
01-03-2010, 11:33
. People who overanalyse these things usually aren't that good at the game anyway (shock horror, prepare for incoming abuse...),



You're damm straight!


It takes mere seconds to work out basic probability (i.e. if i charge my unit, what will it do, on average). Everyone who has any idea of tactics does that. Even if they don't actually work out the probability, they still have a rough idea of what to expect.

However, if you really want to go overboard you start working out the probability of clumps of results. In other words, i know a BS4 meltagun within 6" has a (2/3 x 21/36 x 1/2) + (2/3 x 5/36 x 1/6) = 42/216 + 10/648 = 17/81 = 21.0% chance of destroying outright a land raider (although that's more complex maths than i usually would do in game!). Therefore i know relying on one to do it is a bad idea. 2 have odds of 37.6% of killing it, etc etc. However i could also work out the probability curve, so to speak - if i fire 2 meltas at it, most of the time i'm going to get a good result.


Perfect example from a while ago, in a tournament (although in fantasy). I remember it because probability backfired, quite badly :D

My opponent had 3 undivided beastmen chariots middle, from the old book obviously. I put my oldblood on carnosaur lord (lizardmen) out in front to tempt them to charge me. Considering i cause terro against their LD7, i reckoned at least one would fail terror (and run off the board). Further considering the high resilience of my lord, he and his big dino should have happily survived the impact, and proceeded to munch through the remaining chariots without issues (trust me on that one, its a very killy unit versus chariots!). I worked out the average to some accuracy, and the odds of it going badly wrong. It was somewhere under 10%, with a high reward (although the risk was also fairly high, if it went wrong)

It went wrong. Upgrade that, it went REALLY wrong.

Not only did all the chariots charge, he proceeded to roll very well for impact hits and attacks, ganking my dinosaur and putting a wound on my lord. Who then failed to kill any of them with Blade of Realities. Managed to be saved only by the fact my lord passed the resultant leadership test.



Now unfortunately, it did not go to plan. I actually had a backup, thank god, but the point is that the vast majority of the time it would have been awesome and more or less handed me the game on turn 2 (the chariots were his entire centre, and he had basically nothing left that could deal with my lord after that, who would have spent the rest of the game terrorizing/obliterating his units. Was looking forward to it :( )

catbarf
01-03-2010, 11:42
but i dont base my tactics or playing style around it

Well, you really can't base a playing style around it.

Math's just a tool. You can't base your battle plan around probability. But you can use it to make decisions- 'Which am I likely to do more damage to, those Tacticals in the open or Scouts in cover?'. Or, before you reach the table, 'Which is more likely to penetrate AV10, a Heavy Bolter or a Lascannon?'. You can use probability to make tactical decisions in-game, or figure out how to build your army. It's the latter application that people tend to cite on these boards.

Someone insisting that an option is crap because another is 2% better in a single situation is being a bit silly. And, in-game, someone who's shocked at their battle plan failing utterly in apparent defiance of the law of averages is also being a bit silly. Both are examples of misapplication of what amounts to a very simple tool, not a magic game-winning device. But a little bit of statistical analysis will give you a much more concrete view of the battlefield than wishy-washy personal (and fallible) experience and anecdotes of others.

Lord Asgul
01-03-2010, 11:49
I have never played to win because what is the point of winning if everyo who is involved in not having fun? My army is built around the way it looks and feels. And what parts come in the boxed sets haha. I have never thought before a game that I should this, this and that to counter them, I pick up what I want to use. I've seen things that still make me laugh, like a grot somehow wounding a khornate daemon prince, a force commander taking out a defiler in CC...in one turn, zombies munching on harlequins and a Archon. What is the point of playing tactically when you probably won't get memorable moments like that?

Bunnahabhain
01-03-2010, 11:52
Math-hammer is a good way of evaluating outcomes without using unreliable memory, and one off anecdotes.

For me, it is only simple arithmetic,( science and engineering at university level does rather require the ability to handle rather more maths than the basic stats here, and I've not felt the need to break out Differential equations for wargaming yet...) so for most things it takes a few seconds to do in my head.


Of course, as 95%+ of the math-hammmer here goes for expected values, rather than average ones- the distinction will exist for most rolls - it is a bit rough and ready, so not suited to looking at the last few percent diffrence between things, but it is good enough for what we want.

You'll also notice if you look carefully that some people round here ( usually including myself, unless I've got carried away in one of those threads with an argumentative type...) generally those for whom the maths involved is trivial, will give clear answers upon request for math-hammering of something, and then have a sentence of disclaimers saying what assumptions and approximations we've used, ie for transport busting, we've only counted immobilised or better as useful, and ignored staked weapon destroyed.

catbarf
01-03-2010, 11:55
and then have a sentence of disclaimers saying what assumptions and approximations we've used, ie for transport busting, we've only counted immobilised or better as useful, and ignored staked weapon destroyed.

Absolutely true. If you're calculating odds in a specific situation, they're only applicable in that situation, and it would be a mistake to assume they hold true elsewhere.


I have never played to win because what is the point of winning if everyo who is involved in not having fun? My army is built around the way it looks and feels. And what parts come in the boxed sets haha. I have never thought before a game that I should this, this and that to counter them, I pick up what I want to use. I've seen things that still make me laugh, like a grot somehow wounding a khornate daemon prince, a force commander taking out a defiler in CC...in one turn, zombies munching on harlequins and a Archon. What is the point of playing tactically when you probably won't get memorable moments like that?

Personally I use a few crappy units just because I like the looks and enjoy painting them. I use math as a tool to help me make decisions, not as a be-all and end-all of list construction and gameplay.

Corrode
01-03-2010, 11:59
When people vehemently argue that something is "utter garbage" when the second place choice is 0.21% better, people need to step back and take a breath. People who think that "statistically unlikely" means "never" or "statistically likely" means "always" are asking for disappointment and frustration as well. I find that during any 40k game, at least one thing that is "statistically unlikely" will happen or "statistically likely" will not happen.

To demonstrate this, I had a plasma gun fail to fire twice last night, and then kill himself on the second, whilst his plasma-cannon toting buddy failed a magnificent 4 shots; the only time he fired, he rolled double sixes for scatter. On average a plasma gun at long range should fail once per game, as should a plasma cannon, and managing to fail so spectacularly for so long and then roll double sixes at the one time they wouldn't be useful to me was such a vanishingly small sequence that it should never have happened.

Unfortunately, probability has no regard for the word 'should' and treats it like a prison bitch, so my plasma gunners managed to foul themselves up spectacularly regardless of the likelihood of my rolling sequence up there.

Sir_Turalyon
01-03-2010, 12:08
Statistical analysis neither replaces nor is replaced with common sense and game experience; it just gives more data for common sense and makes it easirt to relate to less typical or complex situations.

Example: Ezakiel's Mind Worm power - when succesfuly cast, target must take Ld test or is slain outright.

Bare common sense: Mind worm doesn't look powerful.

Bare statistical analysis: Chances of succesfuly cast mind worm killing target is, depending on target leadership:
Ld10 - 1/12
Ld9 - 1/6
Ld8- 5/18
Ld 7 - 5/12.

Common sense applied to statistical analysis: If psychic test is succesful and not nullfied, we can expect mind worm to be as effective (have the same chance of killing), based on target's Ld, as:
Ld 10: same as bolter hit against a terminator
Ld9: same as bolter hit against a regular marine
Ld8: same as a demolisher hit against a captain with 3++ stormshield or an autocannon hit against a marine
Ld7: same as a demolisher hit against a chaplain or a str4 chainsword hit against an Ork boy

Egaeus
01-03-2010, 12:19
To demonstrate this, I had a plasma gun fail to fire twice last night, and then kill himself on the second, whilst his plasma-cannon toting buddy failed a magnificent 4 shots; the only time he fired, he rolled double sixes for scatter. On average a plasma gun at long range should fail once per game, as should a plasma cannon, and managing to fail so spectacularly for so long and then roll double sixes at the one time they wouldn't be useful to me was such a vanishingly small sequence that it should never have happened.

Unfortunately, probability has no regard for the word 'should' and treats it like a prison bitch, so my plasma gunners managed to foul themselves up spectacularly regardless of the likelihood of my rolling sequence up there.

I think part of the issue here is that people tend to look at their rolling as specfic sub sets of all the rolls they make during a game. That is, I suspect that most people have a pretty "average" set of rolls for a game, it's just that certain rolls are given more weight than others.

I am guessing those plasma shots were just some of the rolls you made during the game...I am assuming there were lots of shots you took where you didn't roll ones. Plus rolls to wound, rolls to save and all the other rolls that make up a game. But simply due to the fact that a greater importance was placed on those particular rolls it gives the appearance of a statistical anomaly.

Lord Asgul
01-03-2010, 12:40
Absolutely true. If you're calculating odds in a specific situation, they're only applicable in that situation, and it would be a mistake to assume they hold true elsewhere.



Personally I use a few crappy units just because I like the looks and enjoy painting them. I use math as a tool to help me make decisions, not as a be-all and end-all of list construction and gameplay.

I still use my possessed :) in the last game I played they ate a squad of grey knight termies :) but were killed by 2 incinerators burnings :( I just like the models though :)

Archangel_Ruined
01-03-2010, 12:43
I agree with the above, nobody really notices when their lasgun toting models shoot out of their skins as they're using crappy guns they expect to fail (gross simplification, I know). Over several games things do even out, that's what statistics is about, trends.

Lord Asgul
01-03-2010, 12:50
I agree with the above, nobody really notices when their lasgun toting models shoot out of their skins as they're using crappy guns they expect to fail (gross simplification, I know). Over several games things do even out, that's what statistics is about, trends.

Hell even my traitor guard did good, a missile launcher destroyed a grey knight dreadnought :) it was his first kill in 2 years, I was proud :D

Archangel_Ruined
01-03-2010, 12:52
Give that man a shiny new skull.

Kurisu313
01-03-2010, 13:22
I really hate the basic false dichotomy threads like this generate.

Either you're an evil power-gaming waac math-hammerer, or you're a fluff gamer who cares nothing about winning and looks after homeless puppies in your spare time.

Math-hammer is a nonsense word if you ask me. 40k is a game based around statistical probabilities. It revolves around that most humble of random number generators - the die. Of course it follows basic statistics, there is no need to give it a new name.

We all know that a die will generate a number from one to six, and assuming that it's fair, should get every number about one sixth of the time. We can all extrapolate from that that a single attack needing fours to hit will succeed about 50% of the time.

So why is it that when we start talking in larger numbers of dice, all of it becomes useless junk? It's not an evil conspiracy theory, we know that ten dice rolling fours should generate five successes on average.

That's it! That's all! Statistics does not say that 10 dice cannot come up with 10 sixes or 10 ones. Indeed it says that it should happen, but vanishingly rarely. Lictors can destroy landraiders in one shot. A single gretchin can beat a unit of assault terminators in combat and run them off the board.

When I design my armies, I design them around what I usually want to take that week. Yesterday I built a really poor army that used lots of models that I rarely use, just for a chance to take them. This thursday I'm gonna use scout bikes and terminators because they are the things on my painting table.

But when I'm in a game, I cannot help but perform basic risk analysis. I know that my S7 weapons can't hurt that raider. I know that my lascannon is better off firing at a tank rather than a guardsmen. Why do I know that? Because of basic probability that we all know.

When you have a choice between shooting/assaulting one of two units, which do you choose? You base it on a combination of factors - where the units are, chance of being in range, closeness to an objective, etc, and the chance of succeeding in your goal. Do you need to wipe the unit or lock it for a turn? Probability is the ONLY tool to help you decide that.

It's not a substitute for being a good general, it one of a good general's tools, next to experience, which is used for other things. If I experience a lictor destroy a raider, I still know it's not likely to happen again and should have a look for a better target, or that that gretchin will probably be squishified by those termies.

I apologize for the rant, but I hope of made a valid, clear point (though I doubt it)

Lord Asgul
01-03-2010, 13:25
Give that man a shiny new skull.

He shall gain access to a shiny gold missile launcher :D

VonManstein
01-03-2010, 15:54
Mathhammer can be a shortcut to be being an very good player in a short time.

For some people who dont intentionally use mathhammer, they learn whats an effective decision to make and what not mainly through experience.
Others (including myself) dont need the experience, mathhammer can make up for that for a large part.

Yes, I mathhammer a lot, also during battles.
Example: Wraithlord has 2 flamers and 1 scatter laser and 1 EML. Which weapons do I use on that 8 man marine squad? I matthammer it during a battle (I can do it to 1 decimal precise pretty easily if its not too complicated) and make the most efficient decision because of that.

Other example: I have 2 plasmaguns and 1 Lasscanon. I need to inflict 1 wound on that MC and want to damage a rhino too. Which shoots at what? Again, I matthammer it and make the most efficient decision on that outcome.

It doesnt matter if the real result is close to the 'average' result or not, mathhammer only shows what is efficient and what not. It becomes also more interesting when you take average deviations and things like that in consideration too.
Example: whats the chance I dont hit at all with my MC's charging that tank which moves 14''.

Its very nice for list building too: You can check if you think you have enough Long range AT fire for the first rounds to reliably damage 2 rhinos or not. Again, the actual outcome may be different, but again that irrelevant. Its about making the most efficient decisions on average.

It never sets you back, thats what some people dont realize. Mathhammer has no drawbacks, it solely tells you what each thing is capable off in theory. It doesnt prevent you from making decions which have low chances of succes.

Netfreakk
01-03-2010, 15:59
Even those that don't do the actual math is still doing the mathhammer from experience. Why do you shoot meltas more readily at landraiders rather than lascannons? It's because you're chances of destroying it is higher. Why shoot bolters at other infantry with lower Toughness? You have better chances. You might not do the math right there while playing, but you're still employing the mathhammer that you've gained through experience. And guess what, it was the averages from all your experience that gave you the impressions of what's good against what, which falls in-line with the math hammer.

Sekhmet
01-03-2010, 16:08
I have never played to win because what is the point of winning if everyo who is involved in not having fun?

...

What is the point of playing tactically when you probably won't get memorable moments like that?
Fallacy: Playing to win and having fun are mutually exclusive.
Examples: the Olympics, professional sports, Magic: the Gathering tournaments, World Series of Poker, etc.

That's all I have to say on this topic, besides that I always play to win, and I have memorable moments in every game.

incarna
01-03-2010, 16:10
Mathhammer is a valuable guide.

Simple situation, this past Friday I was playing my mechanized Eldar against a Tyranid army at 1850. Turn 3 a Mawloc dropped into play. Since I kept everything in its transport knowing full well that that Mawloc was going to be trouble when it came into play I had some choices to make with the simple goal – I HAVE to deal with this Mawloc because it’s camping my objective.

Here’s the situation – two squads of 6 outflanking genestealers, 1 Mawloc, and 1 squad of 2 Lictors not in cover. How do I allocate the forces I have available; 2 fire dragon squads of 6 and 1 Dire Avenger squads of 10 all in Wave Serpents with twin linked shuriken cannons and twin linked shuriken catapults as well as 1 squad of 8 Dire Avengers who had their transport sliced and diced by the Lictors on the previous turn?

The answer may seem simple but a little mathhammer can guide the decisions I make – The Fire Dragons clearly have to target the Mawloc. The MC has 6 wounds so one squad of 6 Fire Dragons will probably miss twice and has about a 50/50 chance of doing 3 or 4 wounds. For this reason I know that I need to send BOTH squads of Fire Dragons to deal with the Mawloc instead of zipping one of the squads across the battlefield to deal with a squad of Warriors camping an objective.

The Lictors are a tad more resilient than they were in a previous incarnation so I need to allocate the appropriate force to deal with them. Should I use my squad of 10 or 8 Avengers? Well, 10 will bladestorm 32 shots, hit with ~22, and wound with ~11. If they go to ground they’ll have a 4+ armor save which means about 5 or 6 wounds will get through – so, I need to use the 10-large squad and play cleanup with a Wave Serpent if I don’t inflict enough wounds to kill both Lictors.

The Genestealers are the issue – I can vaporize ONE squad with the 8-large squad of Dire Avengers and the other squad with the two remaining Wave Serpents – but it’s cutting it a little close. Ah well – gotta do what I gotta do and hope I don’t tank a roll.

The decisions I made were guided by mathhammer – and things didn’t go exactly as planned, but pretty close. The 8-large Dire Avenger squad failed to wipe the genestealers but that was my fault – I elected not to bladestorm in hopes I would be able to shoot at some Warriors that were chewing through my sear counsel in close combat next turn. 18 shots, ~12 hits, ~6 wounds SHOULD do it – but the to-wound roll tanked and two remained alive. I could have cleaned up the remaining 2 with my 2 wave serpents but elected not to because the other squad of genestealers was in a position to eat my Fire Dragons who would be MUCH more valuable against the warriors when they finally broke loose. The two wave serpents killed all but 1 genestealer from the other squad, the fire dragons killed the mawloc, and the other squad of dire avengers killed the lictors.

It is this type of mathematical decision making that drives my competitive play. Of COURSE there’s a margin of error, but that’s something you have to live with.

Freakiq
01-03-2010, 16:14
Mathhammer is great for planning moves in advance.

Don't take it too seriously though, I've grown tired of hearing players whine when they don't roll averages.

sliganian
01-03-2010, 16:16
I'm reposting this from the Venomthrope thread because it seems relevant to this thread (perhaps even moreso)...

********
On Probablilities and Warhammer and the Venomthrope May not Suck:

This is somewhat of a similar post to one I did a while ago (it was for 4th edition I think), so if it bores you don't shout 'repeat!', just get on with life.

Before Turn 1 begins, actually before a die is rolled, there are some things that have already happened:

- Does the terrain layout help you, hinder you, or is your army neutral?
- Is the opponent's list generally easy for you, hard for you or neutral?
- Is the scenario rolled up easy for you, hard for you or neutral?
- Is the Deployment style in your favour, against you, or are you neutral?
- The Go First roll: was the result good for you, bad for you, neutral?

The upshot of the above exercise is that there are numerous conditions in play before the game even gets rolling. You may have Favourable Terrain, Opponent, Mission, Deployment and Go First roll -- You Rule!

Or, you may have Unfavourable Terrain, Opponent, Mission, Deployment and Go First roll -- the Game hates you with the passion of 1000 burning suns.

Mathhammer is wonderful for telling you -- in a vacuum -- whether to assault those Terminators with Fire Warriors is a Bad Idea or a Good idea.

Where is fails is accounting for situational weighting: e.g. Perhaps if you DON'T take the chance on assault those Terminators they will claim an Objective. Therefore, sucess or no sucess, momentum and metagame dictates that you SHOULD Assault with the Fire Warriors rather than Rapid Firing -- as it will pull the FW into 'contest' range of the objective. Oh yeah, what Turn is it? Who has what left on the table? Where is it on the table? Will it get to the fight to make a difference?

So what does this have to do with Venomthropes? Well, re-read the pre-Game condition situational analysis. It may be that depending on Mission, Terrain, Opponent, Deployment and Go First that having a Venomthrope would be a nifty and useful thing to have.
Or, it could be that you could be deep in stupid-land for taking one over something else. And all of this is happen before Turn 1 even starts.

Where Mathhammer fails is not at a tactical level but a strategic one -- namely in accounting for player skill and other battlefield variables that have occured outside of the "unit X assualts unit Y" stuff.

If a player knows EXACTLY what a Venomthrope is good at and applies it correctly, it will likely be a sound choice for the list. If a player is indecisive or picks the wrong targets for it, he will fail. X vs. Y calculations cannot help with this.

EDIT: By the way, from the previous thread, it was generally agreed that the strongest armies overall are the ones that are 'neutral' to the pre-game conditions.

VonManstein
01-03-2010, 16:20
Don't take it too seriously though, I've grown tired of hearing players whine when they don't roll averages.
Tell them the chances are higher you dont roll average in almost every situation. The chances of getting 12 hits with 18 BS4 shots are less than not rolling them for example.

marv335
01-03-2010, 16:48
I use "mathhammer" although I dislike the term.
I can calculate the probabilities of a unit doing damage to another, and my battlefield decisions are influenced by that, however it isn't the only factor in the process.
In my view, to win at 40k, you need to know what a unit can and cannot do, and how likely it is to do it.
Also you need the same info on your opponents forces.
How else can you judge where the greatest threat lies? How else can you choose how to deal with it?

It's just another tool to be used by the successful general.

Nezalhualixtlan
01-03-2010, 18:58
I use mathhammer extensively, but in terms of actually playing it's definitely an art of blending understanding probability and statistics with understanding the larger game strategy, and opposing player psychology than pure mathhammer.

It's extremely useful to know: If I attack unit W with X then I have Y chance of getting Z wounds on them, and the reverse for what they can do back. If you have that basic math hammered out, and know Y & Z vs the inverse they are likely to do back, vs the Y & Z of other units vs. unit W, then you can start to weave together a better overall chance of success in terms of how you efficiently allocate your points against your opponent.

But this alone will not win you games. If you don't do this against an opponent who does, they will have a massive advantage over you. But while it's necessary to play well, it is not sufficient.

You also have to realize that on occasion, you are better served by doing things that don't have the best overall probability of success because given the board conditions, a certain objective is far more important *right now* to base later contingency plans which have an even greater chance of success on. That may be counterintuitive at first, but it's kind of like in chess where you might sacrifice a pawn to take out a better piece of your opponents the following turn. The overall strategy has to be that you look for the best overall success for the duration of the game, not simply on a turn to turn basis. It does you little good to have great success in a small action this turn if it sets your opponent up for a huge success two turns later particularly at the start of the game. As the game progresses current success takes on an ever greater appeal and urgency. Early on though you may sacrifice for the greater overall strategy. If you fail to do so, and your opponent is, they may be setting things in motion to their advantage in this way before you even realize you are in for a beating. The best chess players see the potential many moves ahead, this game is little different.

This brings me to understanding opponent player psychology...

You can sometimes draw you opponents into doing things they shouldn't. And I don't mean by talking to them and subtlely trying to influence their decisions, and absolutely not by lying to them or misrepresenting what you have on the table or in a particular unit. You can sometimes do this by placing a lure. Be honest about what is there in your unit and what your opponent is facing, but sometimes if you have a great plan for what comes up in 2 turns, but it revolves around your opponent taking a particular action next turn, you might want to make a tactical feint with a unit that maybe looks like you made a mistake and left yourself out and exposed so that your opponent really *wants* to set themselves in a position that they don't realize will put you in even greater position to exploit their weakness.

But this is not chess, there are no guarantees in terms of set up. Everything just about has a chance of failure, and so it brings us back around to point 1: Mathhammer is important. It's important to know what your chances of success at a particular action so that you can have alternative contingency plans in place should it fail. But you also aren't likely to be doing rigorous statistical calculations in the middle of a game, particularly with so many pieces to move and so much time involved on just getting things where they need to be and rolling dice. Decisions based on this need to be made relatively fast. So it's important to do a little homework first and start learning the general chance of certain things happening with what you have in your army. It's good to know the chance of success for a melta vs AV14, and the relative level of success for other things and other levels of armor. It's why when I have a free chance I bang out the mathhammer on things like my new Tervigons breaking their gaunt spawning ability, and how many gaunts I should expect on average throughout a game when making a list. Or things like psychic failure chance, and how that gets modified by Shadow in the Warp, or other mitigating factors. I memorize a few key numbers, keep a general sense of others while having a fairly decent understanding of general probability and statistics so that I can make quick gut decisions based on sound mathhammer. It might not always be me making decisions on the exact numbers, but given a general knowledge of chance for success/failure it's a massive help.

So you are left with trying to chart your best overall path of contingency for success on an ever changing board condition in which certain contingency paths close off while others open. Mathhammer isn't everything. But probability and statistics are extremely useful to the seasoned general in terms of helping to chart a successful course.

catbarf
01-03-2010, 20:55
You also have to realize that on occasion, you are better served by doing things that don't have the best overall probability of success because given the board conditions, a certain objective is far more important *right now* to base later contingency plans which have an even greater chance of success on.

Strictly speaking, that can be factored into the equation. Your determiner of success doesn't need to be 'points value of stuff killed'. But otherwise I agree entirely.


Where Mathhammer fails is not at a tactical level but a strategic one -- namely in accounting for player skill and other battlefield variables that have occured outside of the "unit X assualts unit Y" stuff.


Just like how where chemistry fails is in knitting, where astronomy fails is in biology, and where psychology fails is in cooking. If you try to use math to conduct strategy, i.e. play the game for you, you're doing it wrong. It's very accurate in specific scenarios, but how you get there is entirely up to you.

Murdoch
01-03-2010, 21:50
why use a vindicator when you can use a whirlwind? less points for greater range AND the ability to fire indirectly. Why is the Vindicator better? mathhammer.

Because it is dead killy and shooty.... :D

Mathammer is fine if you are rolling lots and lots of dice... Guard, guants etc as a large number of dice will tend to favour a standard outcome...

Fewer dice = greater chance of um chance

catbarf
01-03-2010, 21:58
Mathammer is fine if you are rolling lots and lots of dice... Guard, guants etc as a large number of dice will tend to favour a standard outcome...

You wouldn't say that you have a good chance of winning the lottery and discount all probability simply because you're only buying one ticket, now would you?

Murdoch
01-03-2010, 23:19
You wouldn't say that you have a good chance of winning the lottery and discount all probability simply because you're only buying one ticket, now would you?

No and, for fear of seeming foolish, is that not what I am saying :confused:

What I am saying is you can't mathammer reliably on a small number of dice as the chance of the incorrect result is higher than rolling lots of dice...

Long time since I read probability and I have imbibed much OH since then... Quite prepared to say I am wrong but I think your lottery analogy and mine are the same... One ticket = one dice... chance of you rolling the right number reasonably low, several thousand tickets = lots of dice, chance of rolling the right numbers substantially improved. I know that is real laymen stuff but I am to old and addled to go through the calcs to work out all of the variables in drawing the correct 6 numbers from a bucket of 48 as opposed to rolling the probable outcome on 40 dice :D

Archangel_Ruined
01-03-2010, 23:20
Actually... The biggest criticism coming out here is how probability is no excuse for strategy and that there are several things that happen before and during a game that are out of your control. These are true for the most part but I think you'll find the best players have used probability, rules, skullduggery and general gittishness to get around these. It's the fundamental, basic tennet of the alphastrike model of play. If you're deep striking everything in, or have scout on the majority of units then you get a massive say over these 'variables'. It's also why the dual lash/oblit list is (was) pure filth, it takes variables away by allowing you to remove the tactical positioning of your opponents units and then open fire with the most effective weapon from a pretty useful list. It's all about removing as much luck as possible, which is surely the point of all probability, you're doing whatever you can to stack the odds in your favour, thus minimising the risk of failure. Just a point about how advanced bastardry in the field of maths can ruin your run of the mill players day.

ColonalKlink
01-03-2010, 23:40
another good point angel, but some of what you described as "stacking the odds in your favour" seems an excuse to abuse the rules in favour of the abuser, not cheating as such, but close...

also i was talking to a mate of mine about this thread and he said "just remember, opinions are like ********s, everyone has one, and they all stink" lol

Corrode
01-03-2010, 23:45
I think part of the issue here is that people tend to look at their rolling as specfic sub sets of all the rolls they make during a game. That is, I suspect that most people have a pretty "average" set of rolls for a game, it's just that certain rolls are given more weight than others.

I am guessing those plasma shots were just some of the rolls you made during the game...I am assuming there were lots of shots you took where you didn't roll ones. Plus rolls to wound, rolls to save and all the other rolls that make up a game. But simply due to the fact that a greater importance was placed on those particular rolls it gives the appearance of a statistical anomaly.

Yes and no. When talking specifically about the plasma cannon, the chance of the particular roll for that weapon being a 1 four times in a row is vanishingly small. Basic statistical knowledge lets me know that that sequence is unlikely to occur again (though it's no less likely than any other sequence, even if it's less likely than all the combined sequences where no 1s are rolled, or where there's only a single 1 somewhere, etc.), which means I can say 'my plasma cannon was unlucky' and not 'my plasma cannon sucks!'.

As you said though, it can often lead to people whining about their 'bad luck' when they're just focusing on things which deviated from the norm - a friend of mine had a terrible habit of throwing units into positions which were win big or lose big, failing a critical roll, and then whining about his 'luck' when he'd had pretty good rolls all game. Over the course of that game my luck was pretty average, but the plasma cannon stuck with me as an example for this thread - it 'shouldn't' do that, but it did anyway, and it may well do so again from now until the end of time (or until the model hits the wall, whichever comes first!). That doesn't make plasma cannons bad, it just means that my luck was poor as regards that particular sequence of rolls (or good, if you value getting unlikely sequences more than you do winning the game!).

It's always been interesting that a set of lucky rolls on the part of a favoured character or battle tank is always down to how 'good' they are, and a set of bad rolls is down to how 'unlucky' the player was. Very silly business.

e: Klink, you're not really understanding the point that's being made. The rules aren't being 'abused' in any way when one reduces the variables. It's a dice based game and ultimately every plan can fail; a lot of the tactics in the game are about making sure your probability of succeeding at a given action is as close to 1 as possible. This is why we drive meltas into short range rather than plugging away from 24", and why we shoot Predators with lascannons rather than autocannons. Every time you say 'hey, lascannons are better than autocannons for killing AV13' (I cannot remember if this is true or not), you're reducing variables in your favour. Some lists take this to an extreme, but that's the sign of a well-built list - if it can deal with everything with a high probability of success, it's a good list. Unless you literally play the game shooting your guns at the closest target or selecting by divining the entrails of a goat or something, you're 'reducing the variables' in one way or another.

Archangel_Ruined
01-03-2010, 23:55
Sign of a well built list or a sign of advanced bastardry? If you're at a tournament then it's the former, in most other situations it's definitely the latter. Messing with the odds to that extent does detract from your opps game, I know I've been on both ends, feeling that you've won or lost before a dice is thrown isn't great.

Murdoch
02-03-2010, 00:00
Actually... The biggest criticism coming out here is how probability is no excuse for strategy and that there are several things that happen before and during a game that are out of your control. These are true for the most part but I think you'll find the best players have used probability, rules, skullduggery and general gittishness to get around these. It's the fundamental, basic tennet of the alphastrike model of play. If you're deep striking everything in, or have scout on the majority of units then you get a massive say over these 'variables'. It's also why the dual lash/oblit list is (was) pure filth, it takes variables away by allowing you to remove the tactical positioning of your opponents units and then open fire with the most effective weapon from a pretty useful list. It's all about removing as much luck as possible, which is surely the point of all probability, you're doing whatever you can to stack the odds in your favour, thus minimising the risk of failure. Just a point about how advanced bastardry in the field of maths can ruin your run of the mill players day.

Isn't the art of 'controlling' variables at the very essence of strategy?

from laser guided to blanket bombing you are controlling the chance of the least favourable outcome from occurring.

I have no issues with people trying to model the outcome of a series of moves prior to performing them as they are thinking about what they are doing... Which in turn encourages me to think about what I am doing... In the same way I never had issues with running across the occasional min maxed list in 4ed and I don't have a problem with the any lash list i have come across... They don't make for fun games of high fantasy and heroics, they do, however make for a cerebral game of cat and mouse.

I like both types so I am happy either way :D

catbarf
02-03-2010, 00:03
Long time since I read probability and I have imbibed much OH since then... Quite prepared to say I am wrong but I think your lottery analogy and mine are the same... One ticket = one dice... chance of you rolling the right number reasonably low, several thousand tickets = lots of dice, chance of rolling the right numbers substantially improved. I know that is real laymen stuff but I am to old and addled to go through the calcs to work out all of the variables in drawing the correct 6 numbers from a bucket of 48 as opposed to rolling the probable outcome on 40 dice :D

It's not just about the likelihood of the rolls conforming to the mean outcome, it's about the individual rolls. Mathhammer works just as well for one die as it does for 40. More dice means that the results are more likely to conform to the standard deviation, but you can still ascertain odds with just a few dice with a great deal of precision. You don't need to play Guard to benefit.

In other words, if I roll one die, I can say with certainty that the chance of rolling a 4+ is 50%. If I have one Guardsmen fire one shot at a Space Marine, the chance of scoring a kill is 5.5%. The more dice you roll, the more likely it is that exactly 5.5% of the shots will result in kills. That's standard deviation. But you can still use that chance (5.5%) to make tactical decisions- which unit to shoot at, for example- whether you're firing one lasgun or one hundred.

Murdoch
02-03-2010, 00:25
It's not just about the likelihood of the rolls conforming to the mean outcome, it's about the individual rolls. Mathhammer works just as well for one die as it does for 40. More dice means that the results are more likely to conform to the standard deviation, but you can still ascertain odds with just a few dice with a great deal of precision. You don't need to play Guard to benefit.

In other words, if I roll one die, I can say with certainty that the chance of rolling a 4+ is 50%. If I have one Guardsmen fire one shot at a Space Marine, the chance of scoring a kill is 5.5%. The more dice you roll, the more likely it is that exactly 5.5% of the shots will result in kills. That's standard deviation. But you can still use that chance (5.5%) to make tactical decisions- which unit to shoot at, for example- whether you're firing one lasgun or one hundred.

I totally agree with all of that... :D

What I was trying to get across was that you are more likely to achieve your goal with a weight of numbers... Increased population tends to produce a flatter result with a more evenly spread SD... Law of large numbers and all that jazz ;)

You can use your example above to determine where to shoot the guardsmen but you are going to have to be lucky to achieve your goal... However if your goal is to kill one space marine then shoot 12 guardsmen at it and you are likely to have a dead marine...

Unless of course I am being really thick and missing something... Which is a major possibility :angel:

Corrode
02-03-2010, 00:30
Sign of a well built list or a sign of advanced bastardry? If you're at a tournament then it's the former, in most other situations it's definitely the latter. Messing with the odds to that extent does detract from your opps game, I know I've been on both ends, feeling that you've won or lost before a dice is thrown isn't great.

I don't agree with the assertion that to play a friendly game one should simply crap a list onto some paper. Is dual lash appropriate for playing a game against a first-timer who barely knows the rules? Probably not; a first-time player is unlikely to even come close to having the knowledge, ability, and diversity of collection to be able to counter such a strong list. Is it appropriate for a pick-up game? I'd say yes. I'd always prefer to take a strong list than a weak list against an unknown opponent; taking a strong list says 'I thought this would be a challenging game and thus I have brought a good list, which will enable me to perform well.'

The problem always lies in 'playing for fun' - some people don't find it fun to sit and theoryhammer something out, or even do basic analysis. Some like to play with cool looking things, and some make esoteric choices based almost entirely on their own idiosyncratic interpretation of a unit's value - these are the guys who run Flash Gitz because they're super-awesome, or Lictors because they like the fluff, or whatever. This is fine (though it's not how I play the game - I make choices like 'Land Speeders or Attack Bikes?' based on which I prefer visually, but I'd probably include one of the two instead of spending points on Vanguard, ever). Very often, though, these kinds of players view themselves as being morally superior - their refusal to try and make a good list means that they are somehow better than the people who do. It's the exact same way in WoW, with the exact same stupid personal digs and jibes about 'taking it too seriously' or 'not playing for fun'. Raiders who spend an hour with a spreadsheet organising their gear so that they can clear as much content as quickly as possible are 'sad', but people who raid 3 nights a week and clear two bosses because they're busy jumping around like they're high on sugar are 'playing for fun' even if they expend more time and effort than the raider. Both groups have the same goals, but the one which does it by careful thought and analysis is labelled as 'too serious' whilst the ones who seem to desire to complete the task by flinging crap until it sticks view themselves as being not only 'more fun' but superior people. It's a dichotomy that lies within people's inherent attitudes to what constitutes 'fun' - personally I enjoy games even if I spend most of it barely cracking a smile, but other people don't seem to enjoy them at all unless they get the chance to be WILD and CRAZY and it's hard to reconcile the two.

Personally I've always thought it to be part of the general anti-intellectual streak in most of the West, but that's just a passing thought and holds no real water.

neko
02-03-2010, 01:13
I sometimes love some of the surprises math-hammer springs on me.
One unit I'm throwing together is a unit of veteran guardsmen with 3 meltaguns. I plan to mount this unit in a Valkyrie, so I should be able to jump it out within 6" of a target and be confident of at least 1 dead tank right?

The expectancy doesn't look too bad vs AV14, although not as remarkable as I was expecting from 3 BS4 meltaguns: The above ambush would give 0.63 dead Land Raiders.

More important though is the probability that the ambush will kill AV14: The answer comes out as 51%, so similar to just rolling a d6 and hoping for a 4+.

Well, AV14 is meant to be the ultimate nightmare to beat, so lets try AV13: 58% doesn't look much better.

AV12? 64%, so still less chance than rolling a 3+ on a d6.

Before I started this, I would have thought that something like AV11 would have near certainly fallen over: 67% chance of being destroyed, so about the same as rolling a 3+.

AV10, being attacked by 3 BS4 meltaguns at close range? I'd have written that off as dead before, but it's only a 70% chance. Barely better than rolling a 3+ on a d6, and no where near as likely as rolling a 2+.


So, math-hammer has taught me that the planned unit would be nowhere near as reliable as I was expecting it to be, even if I was to pull off an attack against AV10. This means that I can now be more realistic with my battle plans, and won't spend my first dozen games thinking that I must be really unlucky and possibly not even then noticing that I'm expecting too much of them.

Disclaimer: I enjoy running the maths on various things, if only to give my mind some exercise. I wouldn't expect player to go to the above extent if they didn't know how to or didn't enjoy it.

catbarf
02-03-2010, 01:50
What I was trying to get across was that you are more likely to achieve your goal with a weight of numbers... Increased population tends to produce a flatter result with a more evenly spread SD... Law of large numbers and all that jazz ;)

Ah, I see, but still it's a bit yes and no. Large numbers of dice aren't more reliable- you're more likely to get the mean result, but the outliers are more extreme. Sure, if you fire ten shots instead of just two the chance of getting about half hits goes up, and the chance of all missing goes down. But when they do all miss, it's a much more pronounced effect. So while you should always meet threats with enough force to ensure that they're not threats anymore, there's no reason to be wary of single die rolls. After all, statistically speaking, there's no difference between 10 shots at BS4 and 20 shots at BS2.

Morgrad
02-03-2010, 03:16
Playing to win vs. playing for fun is a ridiculous false dichotomy (aren't they all ridiculous?) that I reject completely. I *always* play to win - even if I didn't build my force to win. I also *always* play to have fun.

Applying game theory to a game, and probability to dice, is not somehow making an ****** out of oneself. It's called playing well.

I feel cheated if my opponent doesn't play well.

That said, one of the BEST uses for "mathhammer" is making ballsy, "stupid" decisions where the risk v. reward skew is so huge that failure isn't that bad, but success is insane. If you have a .000000000001% chance of pulling it off (a grot charging 10 TH/SS termies), then it's pretty pointless. If you have a 10% chance, then it's very often worth the attempt.

You cannot make intelligent risk v. reward decisions without knowing the risk. Probability tells you that risk. Generals who make smart risk v. reward decisions win close games far more often than those who do not.

Of course, none of this takes into account tactical flexibility, positioning, setup, use of cover, use of rules, and the like. Those who account for these things will stomp on those who don't.

Znail
02-03-2010, 04:06
While you can manage to play the game without that much Mathhammer/Theoryhammer so is quite difficult to create good army lists without it. The alternative to using math is anekdotal statistics from your games, ie unit x did well in my last game so lets keep using that etc. The same is true for using wich units for what in the game. You may end up with mistaking Grots as the uber assault unit as they killed a Bloodthirster in one game. Ah, the another option here is to only copy army lists from the net. I much prefeer if more people use their own brains some.

Its worth remembering that while random deviation means that everage results wont happen everytime or even most often so is the reverse even more true, just because something happened in a game so does that not mean it was the average result. This is why its a bit silly to dissrespect math and claim that actual game experience is somehow better for judging how good or bad a unit is.

Ofcourse, a combination is best as math cant be used to predict actual battlefield condition and situations, so that is best to use experience for. While experience is poor at judging actual events like how good a unit performs a specific task under given conditions.

Dead Man Walking
02-03-2010, 04:21
"If you can get a quarterback who can throw a football down the field to an open man who can run really fast for a touchdown then thats a good thing." - Madden, the king of Theoryhammer.

"Go ahead and charge my Defiler with your two genestealers, I'd love the over-run and then move and fleet into combat with the back of your army since you only have a 6% chance to- Damn It I can't believe you blew up my Defiler!" - One of my mathhammer opponents who does the mathhammer everytime I grab up a set of dice.

"Sure its overcosted and ineffective but if used corectly the T-fex can win games for you!" - A member of Warseer who I don't reply to anymore.

When it comes right down to it I look at what the mathammer guys have to offer but I don't put too much stock in it unless its a noticable difference (like the difference between 1.83 marine kills and 1.65 marine kills) and I don't play theoryhammer. I play Vet-hammer, I try something out over and over again until I decide if it works and then I tell people about it.

Dead Man Walking
02-03-2010, 04:30
"If you can get a quarterback who can throw a football down the field to an open man who can run really fast for a touchdown then thats a good thing." - Madden, the king of Theoryhammer.



BTW, the problem with theory hammer is that your forgetting that your opponent has a brain and a plan too. Imagine the shock on Madden's face when his opponent says, "Okay I drive my Chimera 6 inches, hop out all my guard veterans and shoot 4 plasma guns and double tap my lazrifles at your open man with the ball."

Sygerrik
02-03-2010, 04:36
I played with someone lately that opined he would enjoy a game where no dice at all were rolled and players simply referred to statistical charts to determine outcomes.
I didn't argue the point because I was a bit floored by the insanity of that statement, not to mention the logistical problems it entails (if you have a BS3 lascannon, just one, how do you determine when it hits? does it hit once every two turns, or one half of one time per turn? argh.)

Mathhammer is a good way of determining matchups. As someone whose work involves a great deal of statistical modeling and inference, I find the idea that mathhammer can tell you anything about how to play is laughable. One thing mathhammer can do well is suggest the probable outcome of a given matchup. You can't use it say "squad X is superior to squad Y" or "gun X is superior to gun Y." You can use to say "Squad X with loadout x, shooting at Squad Y with loadout y, ceteris paribus, will likely accomplish impact Z."

People spend too much time trying to do the first thing.

AngryAngel
02-03-2010, 04:53
Everyone uses math-hammer. I just find those who base every single decision solely around math-hammer to be incredibly dull.

I also find it laughable when someone says that one option is crap because a different weapon has a 2% better chance of success.

I agree with this response 100 percent. While everyone use some small bits of mathhammer. It's when its taken to its extreme its gone too far. When someone is actually calling someone stupid for unit choices with but a small diffrence in percentage of success.

Everyone uses mathhammer alot or a little. Though all the mathhammer in the world won't help sometimes.

As for Theoryhammer. Thats just pointless, get out there and use/face the units to see what they will or won't do. No better way to judge something then by experience.

massey
02-03-2010, 05:25
BTW, the problem with theory hammer is that your forgetting that your opponent has a brain and a plan too. Imagine the shock on Madden's face when his opponent says, "Okay I drive my Chimera 6 inches, hop out all my guard veterans and shoot 4 plasma guns and double tap my lazrifles at your open man with the ball."

That is Pass Interference. 15 yard penalty, first down. :)

Mathhammer is a very useful tool. I will give an example from a game I played Saturday. I was playing against a 16 year old kid. I knew he had a fairly limited arrangement of models. He's started like 3 or 4 different armies, switches every couple of months, and never builds up a really good selection of miniatures before moving on to a new cool toy. He was bugging me for a game, and usually I just tell people "nah, I don't wanna play today", because I'm incredibly lazy. I don't like to dig out my army, build a list, play a few hours, and then put everything back up. I'm likely to get bored in the middle of the game and wander off to see what other people are doing.

Plus, if a cute girl walks into the store, I'm going to dive behind the cash register and hide under some boxes of old Spelljammer cards. Actually playing a game might slow my reaction time.

Anyway, on this day I decided to go ahead and just play the kid. He only wanted to play a 1000 point game, so I figured it would go pretty fast. Plus, we were going to use the Battle Missions rules, and I wanted to try out their new scenarios. But I was irritated that he kept pestering me, so I decided I'd build a nasty list and whoop his ass pretty hard. I can be a dick sometimes. And by sometimes, I mean, a lot. But I don't make secrets about it, and everybody knows that I play to win. If you don't want to bring your A-game, don't pester me into playing.

So at some point in the game, he moves his 5 man combat squad up to one of my Razorbacks. He shoots with a plasma gun, gets a penetrate, and rolls a 5. Destroyed. My 5 man squad inside (4 guys with bolters and a sergeant with bolt pistol and chainsword) hop out. On my turn, I have a choice to make. Do I shoot at them with my bolters, or shoot bolt pistols and charge?

Now mathhammer comes into play. He's got a plasma gun. I don't. I decide that I can negate the benefit of his squad upgrade by shooting pistols and charging. He's got more of a chance to kill me at range than I do to kill him. In assault we are equal, except I'll get the charging bonus. So that's what I do. I roll my attacks, crap out on the rolls, and don't kill anyone (which is okay, I understand there is a chance that this will happen). Then he says "I have 4 attacks with my regular guys and 2 with my power fist." Uh huh wha? Power fist? Where? "Oh, this guy with a chainsword actually has a power fist." Thanks a lot, jerk, for not telling me. So I ask "how many of your guys have power fists?" He points to every sergeant on the table. Gee, that changes a whole lot of my strategies.

How does mathhammer come into play? Well, I was making decisions based on the apparent fact that we were roughly equal in close combat. With my new knowledge that his guys had power fists, that changed the percentages by quite a bit. Knowing this, I would not have charged into combat. Nor would I have stayed and shot it out with a plasma gun facing me. Instead, my marines would have turned and moved behind cover, knowing they were facing a superior enemy.

That is how you use mathhammer. Which action gives me the greatest chance of success? Which gives me the least? Knowing this allows you to make wise choices. Those choices may not always work out (even before I found out he had a power fist, I crapped out on the rolls), but at least you can make them with eyes wide open.

I still ended up beating him to a pulp, though.

Pink Horror
02-03-2010, 06:02
You can't use it say "squad X is superior to squad Y" or "gun X is superior to gun Y."

What if gun X is superior to gun Y against any potential target in any potential situation? Wouldn't it then be superior? And what if the only difference between squads X and Y is that each uses the respectively named gun?

Archangel_Ruined
02-03-2010, 09:46
I think there's a difference between taking a good list and taking a top tier tournament list. I play to win, pure and simple, but I'm not a bad loser and I don't feel the need to wrangle every tiny advantage in friendly gaming. Running the numbers on a unit is a basic tool for army selection, and even if you don't actually calculate everything you'll still have a good idea of what's better per point. I just think taking it too far when you're not in the rarified air of a GT is plain unpleasant, at its logical conclusion probability determining strategy isn't fun against an unprepared player. Yes, I agree that tactics does mean minimising risk, controlling variables, but this is a game based on chance. If you do as much as possible to remove that chance you've altered not only your game but your opponents. This is not cool in pick up games. I should note that this isn't a limitation to 'mathhammer' as a tool or discipline, just where players should leave it be in most cases.

Also, looking at the above post I fail to see why I can't say a pulse rifle is better than a lasgun. It is, in every way. Because this can be safely relied upon then you can decide how best to cost such a weapon, it's a cornerstone of the whole pts/effect system of 40k. Perhaps GW would be better off looking a little closer at the numbers but at least they've got an idea.

willow560
02-03-2010, 11:23
Mathammer is important for one main reason. Humans are not good at estimating probabilities well. I've heard studies about weathermen predicting precipitation. They use a calibrated approach, and are actually pretty close to their predictions. That means, their model adapts to new information to fit the probability curve. That means, when they say there is a 30% chance of rain, there really is a 30% chance of rain. However, how many times do you get angry at the weatherman and think, "Why did it rain today, the weatherman sucks."

This is because a human brain does not process this information correctly. In fact we overestimate or underestimate by a lot. When they said 30% chance of rain, your brain said "No Rain" and when they said 80%, your brain said "Bring an Umbrella it is Certain." Human evolution has programmed us this way. If you are running from a noise because you thought it was a lion, it doesn't matter if you're wrong 95% of the time and it was not something dangerous. It doesn't pay off to make the probability estimate right. You brain short circuits and says, "Run Stupid" and it actually is the best choice.

Another good example is those things at the Roulette table that tell you what the past winning numbers were. Everyone here initially thinks "It can't possibly be Red again, Black is the safe bet" when there is a whole lot of Red. Or maybe you think the opposite and think "It can't fail to hit Red again, I know Red is the safe bet." They are equally probable all the time. Your brain is what is not thinking correctly.

However, as Humans we can also think. Math will help you break your intuitions faulty conclusions. It can help you figure out that shooting 2 meltaguns, despite what you initially thought, actually are not that likely to blow up that Land Raider. Mathammer is to deliberate calculated actions that we should be using on the tabletop to "Experienced based" calculations are to running from a lion because you heard something weird. Evolve. You're not being chased by lions anymore, use the math, it actually works. Unless of course you actually are in danger, then that crappy probability estimator in your brain is pretty darn good at telling you to get the hell out of that situation.

catbarf
02-03-2010, 12:00
You can't use it say "squad X is superior to squad Y" or "gun X is superior to gun Y." You can use to say "Squad X with loadout x, shooting at Squad Y with loadout y, ceteris paribus, will likely accomplish impact Z."

So then, how can you say that one gun or squad is better than another? If you've played many games and come to the conclusion, aren't you just anecdotally citing many specific situations in which the squad or gun in question was effective?

While the numbers you will get are strictly the results of a very specific scenario, you can go beyond the hard numbers and make a subjective evaluation. It's no less valid than an evaluation based on gameplay experience.

massey
02-03-2010, 15:16
I think there's a difference between taking a good list and taking a top tier tournament list. I play to win, pure and simple, but I'm not a bad loser and I don't feel the need to wrangle every tiny advantage in friendly gaming. Running the numbers on a unit is a basic tool for army selection, and even if you don't actually calculate everything you'll still have a good idea of what's better per point. I just think taking it too far when you're not in the rarified air of a GT is plain unpleasant, at its logical conclusion probability determining strategy isn't fun against an unprepared player. Yes, I agree that tactics does mean minimising risk, controlling variables, but this is a game based on chance. If you do as much as possible to remove that chance you've altered not only your game but your opponents. This is not cool in pick up games. I should note that this isn't a limitation to 'mathhammer' as a tool or discipline, just where players should leave it be in most cases.


I don't think you can actually do this. You can't remove every element of chance from the game. You just can't do it.

We're getting into some high level abstraction here. When you talk about "probability determining strategy" or "controlling variables", what exactly are you talking about in a game? How do you do something like this is a 2000 point game of Space Marines versus Eldar?

You can't. It's all abstraction.

Strategy is about understanding risk versus reward. You want to know the probability of something happening. If I have a Landspeeder Typhoon, and I'm behind cover, do I pop out of cover and shoot at that Land Raider across the board? I could blow it up. Or I can run the numbers, and see that my chance of destroying it with that 2 shot missile launcher is actually pretty slim. So instead I play the odds and stay out of sight, looking for a different target. Now, I should know this from the start of the game. So I'm going to plan the moves of my Landspeeder so that I can avoid things I can't kill. But which path I take, when I choose to turboboost from one bit of cover to the next, mathhammer doesn't help with that at all. It is merely a tool to assist you.

Sygerrik
02-03-2010, 15:17
What if gun X is superior to gun Y against any potential target in any potential situation? Wouldn't it then be superior? And what if the only difference between squads X and Y is that each uses the respectively named gun?

Theoretically possible, but in trade economics we cleave to the principle of "comparative advantage." That also applies in Warhammer. It runs as follows:
If Squad A does everything better than Squad B, that does not mean Squad B is worthless. Say Squad A is superior to every other squad (or just one specific other squad, Squad B) in every phase, but is limited in how many you choose (by force org or points value). This limitation is quite often the case. Also assume Squad A is much better at one thing (say, tank-busting) but only a bit better in another thing (say, board control). This is also often the case due to the specialization of units. Comparative advantage says you can maximize your army's utility by taking a few Squad A to break tanks and a few Squad B for board control-- thus you are at the point of maximum efficiency in output.

The only time mathhammer/theoryhammer can instruct you as to which squad is "better" is if it is, as you say, better at everything, in every situation. Even then, mathhammer doesn't tell you not to take the inferior squad, it merely tells you what to use each squad for.

And catbarf, I'd argue that you cannot say one gun is better than another except in the specific circumstances I outlined above (which, to be fair, do come up), mostly because guns do not exist in a vacuum. All mathhammer modeling has to take place in the framework of a real-game situation. Despite being "clearly superior" you cannot state that the Pulse Rifle is superior to the Bolter, because you will never see a Bolter on a Fire Warrior or a Pulse Rifle on a Space Marine. You can crunch the numbers and say "A squad of ten Fire Warriors, armed with Pulse Rifles, is X% more likely than a squad of ten Space Marines, armed with bolters, to wipe out a squad of Ork Boyz when rapid firing."

People who use statistics to declare a choice "superior" or "inferior" do not bother me, but it does bother me when they pretend "superior" means "a better choice" and not "more likely to achieve my desired goal in a specific set of circumstances." You're right, it's certainly valid to mathhammer these matchups, and if your matchups represent scenarios that you face frequently on the table then your mathhammer can help you make decisions, but it is fallacious to go from that to claiming that mathhammer somehow proved that one squad is the "best" choice.

Sekhmet
02-03-2010, 18:29
BTW, the problem with theory hammer is that your forgetting that your opponent has a brain and a plan too. Imagine the shock on Madden's face when his opponent says, "Okay I drive my Chimera 6 inches, hop out all my guard veterans and shoot 4 plasma guns and double tap my lazrifles at your open man with the ball."
My friend Napoleon would like to have a word with you and your assertions.

Dead Man Walking
02-03-2010, 18:34
My friend Napoleon would like to have a word with you and your assertions.

You mean the Napoleon that said, "If I drive 700,000 troops into the heart of Russia I will roll all of the russian army up and be the emperor of europe." Napoleon?

Unfortunately he found out that the Russians had brains and tactics too, they call it 'Scorched Earth' and to this day makes a land invasion of russia impossible.

Just more reason why theory hammer doesn't work.

Corrode
02-03-2010, 18:43
You mean the Napoleon that said, "If I drive 700,000 troops into the heart of Russia I will roll all of the russian army up and be the emperor of europe." Napoleon?

Unfortunately he found out that the Russians had brains and tactics too, they call it 'Scorched Earth' and to this day makes a land invasion of russia impossible.

Just more reason why theory hammer doesn't work.

Napoleon failed to take into account significant factors beyond 'my army is the best in the world', like the fact you have to actually reach the enemy to hit them. He used a tried and tested strategy (forced march and forage) which had worked well all over Europe; unfortunately for him, the population density in Russia (even in European Russia) is/was much lower than Western and Central Europe and this method failed him.

You seem to have some opposition to 'thinking about what you're doing' and I'm not sure where it stems from.

Lord Malorne
02-03-2010, 18:54
OP: Any good player uses mathhamer for combats, attacks an dif anything can kill anything, heck you even use it before you even put models on the table, in the 'making an army list phase'.

Don't be silly.

SPYDER68
02-03-2010, 18:57
You mean the Napoleon that said, "If I drive 700,000 troops into the heart of Russia I will roll all of the russian army up and be the emperor of europe." Napoleon?

Unfortunately he found out that the Russians had brains and tactics too, they call it 'Scorched Earth' and to this day makes a land invasion of russia impossible.

Just more reason why theory hammer doesn't work.

Thats his fault for doing his mathhammer wrong and not taking into effect he was charging into difficult terrain without frag grenades.

WinglessVT2
02-03-2010, 19:13
Dangerous and difficult terrain.

Badger[Fr]
02-03-2010, 20:15
Mathhammer is a reliable tool, but had been given a bad name by people who misused and failed to understand it.

First and foremost, most players have at best a tenuous grasp of Statistics and tend to rely far too much on average values, with little regards to actual probabilites and standard deviation. As a consequence, their calculations make little sense and are of dubious use during an actual game. Worse, these very same players do not understand that Mathhammer cannot predict the outcome of a single game, and only describes vague trends.

Sekhmet
02-03-2010, 22:07
You mean the Napoleon that said, "If I drive 700,000 troops into the heart of Russia I will roll all of the russian army up and be the emperor of europe." Napoleon?

Unfortunately he found out that the Russians had brains and tactics too, they call it 'Scorched Earth' and to this day makes a land invasion of russia impossible.

Just more reason why theory hammer doesn't work.
Are you seriously making that argument, that because someone can make a mistake or two that the ENTIRETY of strategy is useless?

And it was a little over half the number you listed, Napoleon never had an army even close to 700,000.

catbarf
02-03-2010, 23:05
Despite being "clearly superior" you cannot state that the Pulse Rifle is superior to the Bolter, because you will never see a Bolter on a Fire Warrior or a Pulse Rifle on a Space Marine.

That's ridiculous. You're telling me that I can't say that the engine on a Ferrari is more powerful than the engine on a Model T, simply because you could never see a Model T with a Ferrari's engine and vice versa. A Pulse Rifle is better than a Bolter, it's got higher S and range. This is obvious and the two can be directly compared.

Here's an example of a statement I feel perfectly justified in making: Devourers, on Tyranid Monstrous Creatures, are better than Deathspitters. Period. Bottom line. Run the numbers yourself, there is no scenario in which the Deathspitters are better, and they're the same cost. How's that invalid?

And even if there were such a scenario, if it happened to be obscure or unlikely, I would still be justified in saying that Devourers are better. 'Better' and 'worse' are subjective concepts. The math only serves as examples that inform such a judgment. It is just as valid to judge a unit or weapon based on how it ranks up mathematically as drawing from gameplay experience. Either way is a subjective decision informed by gathered data on performance in specific circumstances.

Archangel_Ruined
02-03-2010, 23:16
There are builds that seriously cut down on variables, or that stack the odds so far in your favour that it's crooked at anything other than top-tier play. A Vulkan marine army that deepstrikes everything takes a lot away from your opponent, you can't be shot at if your opp gets first turn, half of your infantry will turn up on turn one, you can't deepstrike into trouble with drop pods, you can field multiple units for each role and your primary tank killers are twin linked so missing is unlikely (1/9 unlikely, that's very low by GW standards). It's nasty and it controls a lot of variables right off the bat. That's just one example of the alphastrike lists out there, they exist purely to land a crippling blow so early in the game a player can't recover from it. If that isn't controlling the variables, and hence the game I'm not sure what is. At the top levels there are counters, like presenting so many targets in a guard army they can't be dealt with in one massive strike. However, in pick up games against your average army, a fully optimised list will just dominate to the point of boredom.

massey
02-03-2010, 23:39
There are builds that seriously cut down on variables, or that stack the odds so far in your favour that it's crooked at anything other than top-tier play. A Vulkan marine army that deepstrikes everything takes a lot away from your opponent, you can't be shot at if your opp gets first turn, half of your infantry will turn up on turn one, you can't deepstrike into trouble with drop pods, you can field multiple units for each role and your primary tank killers are twin linked so missing is unlikely (1/9 unlikely, that's very low by GW standards). It's nasty and it controls a lot of variables right off the bat. That's just one example of the alphastrike lists out there, they exist purely to land a crippling blow so early in the game a player can't recover from it. If that isn't controlling the variables, and hence the game I'm not sure what is. At the top levels there are counters, like presenting so many targets in a guard army they can't be dealt with in one massive strike. However, in pick up games against your average army, a fully optimised list will just dominate to the point of boredom.

That's a bunch of crap. You go ahead and drop in your auto-win list. Your opponent doesn't have to deploy anything. So you can land wherever you want, and watch as he drives in his vehicles somewhere else. You're deluding yourself if you think you are controlling variables.

Pete_x
02-03-2010, 23:44
Mathhammer works, it does exactly what it is supposed to do. Math, when applied correctly, are truth, period. The problem is, and i'll quote myself : "when applied correctly". One of the first theorem of probability is that if the probability of an event is 50% it will happen 50% of the time if you have x number of tries where x approches infinity. So if you see someone who moans that they didn't get their expected results, ask him if they threw an infinite number of dices.

I agree with whoever said that probability where fine but mathhammer got a bad name from people who missuses it. Bottomline is mathhammer is a tool and if you don't want to use it fine, but please either you accept it works or you try and understand it before giving it a bad rep.

Oh and btw 10 marines will kill as many orks as 10 firewarrior. Just so you guys know ;)

Worsle
02-03-2010, 23:58
That's ridiculous. You're telling me that I can't say that the engine on a Ferrari is more powerful than the engine on a Model T, simply because you could never see a Model T with a Ferrari's engine and vice versa. A Pulse Rifle is better than a Bolter, it's got higher S and range. This is obvious and the two can be directly compared.

Here's an example of a statement I feel perfectly justified in making: Devourers, on Tyranid Monstrous Creatures, are better than Deathspitters. Period. Bottom line. Run the numbers yourself, there is no scenario in which the Deathspitters are better, and they're the same cost. How's that invalid?

And even if there were such a scenario, if it happened to be obscure or unlikely, I would still be justified in saying that Devourers are better. 'Better' and 'worse' are subjective concepts. The math only serves as examples that inform such a judgment. It is just as valid to judge a unit or weapon based on how it ranks up mathematically as drawing from gameplay experience. Either way is a subjective decision informed by gathered data on performance in specific circumstances.

Problem is people don't always use the maths properly. The devourers vs deathspitters is a good use of maths for decision making and tells you something you can use. While knowing the difference between pulse rifles and bolters is valid maths and can be good to know can lead to wonky decisions.

The maths can only take you so far and some people can't follow the maths or be bothered to that is the real problem. Knowing the difference between the bolter and the pulse rifle does not help you know if firewarriors are good in a Tau army. Knowing the possible damage your rapid firing tau could do does not tell you how good the firewarriors are. Comparing a firewarrior directly to a marine taking into account all their basic equipment is still not giving you information you should be building an army with. This does not make any of the maths behind it bad or even the information its self useless but it can lead to people making decisions based on false premise. So yes you can compare the engines between to cars and it is not wrong to do so but that alone does not give you the full picture of the cars or their performance.

While the problem does not lie in the maths it self but peoples application of it that would be the downfall of theory hammer. Can be useful but it tends to give you a lot of information that does not work in real situations. Not saying mathhammer here either because it is really just maths and maths is never a bad thing in its self.

Oh and as a side note, firewarriors are bad but you have to take a squad anyway :(.

Edit, just realised I said fish of fury you know what I know why tau are so disliked, it is all these silly little names. I mean really the number of them just for the suits is obscene. I will change it to some thing that annoys me less.

Commissar von Toussaint
03-03-2010, 00:13
I see a lot of people beating up on mathhammer, but it seems that some of this may be envy.

Let's face it: some people build killer armies and seem to have all the bases covered.

I'm a dedicated mathhammer-er. From the very first release of the "new" 40k in October '98, I dove in with mathematical analysis and compared points vs killing power, probable outcomes and thereby determined what equipment my army would use.

At that time, heavy weapons in devastator squads cost more than those in tactical squads. Same killing power, much higher price. Solution: Use lots of tactical squads.

As other posters have pointed out, probability isn't a guarantee. The one game I lost with my marines demonstrated this in spades: I had Imperial guardsmen walk across open ground without loss in the teeth of heavy bolter fire and eradicate my army. My dreadnought arrived on the field and probably blew up its own assault cannon. All of my plasma weapons self-immolated in the first two turns. Horrific.

But that's the game. We roll dice, it happens. A few years ago I was playing a 2nd ed. game and was blasting away at a carnifex that made 13 consecutive 4+ saves. :eek:

What can you do? Make him keep rolling. He failed the 14th. :skull:

Real mathhammer (that is to say, using proper statistical analysis) requires you to know the probabilities of certain actions, gauge the benefits and risks of the possible outcomes, and then make a choice.

If you are a good player, you will be doing this constantly. Each move of the enemy causes you to recalculate so that when it finally is your turn, you waste no time in thought because you already know what you are going to do.

When you play against someone like this (and I have) it can be very frustrating, especially if you see yourself as a more "fluff-based" player, or if you have an emotional connection with your army or units.

At that point, a more emotional player might get upset. It's bad enough to lose, but to do it in such an unsentimental way really sucks. And it doesn't help when your opponent, in a gesture of condolence, proceeds to spell out exactly why the odds were always in his favor and the battle pretty much had to go that way.

It is certainly annoying when the odds defy you, though I figure that luck is a sort of transferable thing: there's a finite amount of it out there and if my games suck up all the bad parts, that means that I'll be spared nasty stuff in other parts of my life.

To put it another way, it's better to lose a game of 40k you should have won easily than have your car break down in traffic or find out you have a brain tumor.

One thing that makes mathhammer so frustrating is that GW's games are so dependent on luck and randomness. We all know that there are big problems with balance and it doesn't take a lot of effort to exploit them. And if you put even a little bit of thought into it, you can REALLY exploit them.

Which is honestly the larger problem.

Sygerrik
03-03-2010, 00:51
That's ridiculous. You're telling me that I can't say that the engine on a Ferrari is more powerful than the engine on a Model T, simply because you could never see a Model T with a Ferrari's engine and vice versa. A Pulse Rifle is better than a Bolter, it's got higher S and range. This is obvious and the two can be directly compared.

Here's an example of a statement I feel perfectly justified in making: Devourers, on Tyranid Monstrous Creatures, are better than Deathspitters. Period. Bottom line. Run the numbers yourself, there is no scenario in which the Deathspitters are better, and they're the same cost. How's that invalid?

And even if there were such a scenario, if it happened to be obscure or unlikely, I would still be justified in saying that Devourers are better. 'Better' and 'worse' are subjective concepts. The math only serves as examples that inform such a judgment. It is just as valid to judge a unit or weapon based on how it ranks up mathematically as drawing from gameplay experience. Either way is a subjective decision informed by gathered data on performance in specific circumstances.

Your first paragraph, while technically true, is utterly meaningless. It's like the old joke-- a man is blown off course in his hot air balloon, and passing a stranger on the ground, calls out "Where am I?" The man on the ground replies "You're in a hot air balloon!"
"Knowing" that Pulse Rifles are superior to Bolters (which, in a pure vacuum, they are) tells you absolutely nothing. It's like knowing that AV 14 is higher than AV 13. It tells you absolutely nothing about how to equip your tac marines. It tells you absolutely nothing about the likely outcome of a firefight between a tac squad and a FW squad. It tells you nothing about anything except the vaguely reassuring sensation that you have remembered lessons about which numbers are larger than other numbers.

I urge you to read my posts carefully. I've stated what Mathhammer is good for, and I've even agreed that there are some circumstances in the game where one option is better in all reasonable circumstances than another. These situations are the exception rather than the rule, however, and my most vehement argument has been against people who seem to think that mathhammer can teach you how to build an "optimal" list. It can certainly help you build the optimal list to defeat hordes of Termagaunts, or the optimal list to kill Chimera spam, but there's no such thing as the "optimal list" full stop.

Statistics is one of the most complex and misunderstood fields in mathematics. I direct you to Zed Shaw's article, found here. (http://www.zedshaw.com/essays/programmer_stats.html) I have taken extremely advanced courses in statistics and I apply it on a daily basis, and while my perspective is colored by the field in which I work, I feel extremely confident in making judgments about what statistics can and cannot do.
If you are telling me that, (as an example) taking into account the prevalence of vehicles in the metagame, you understand the value of anti-tank and are able to crunch numbers to find the squad loadout that, per point, gives you the highest % chance of destroying or incapacitating vehicles of any given AV, I will tell you that you are applying excellent mathhammer, and you are likely to be making well-informed judgments about how to commit your forces.

If you tell me that you are able to use pure statistics to determine which choices to take without seeing the table or the opponent's army, and more importantly, without an idea of what each choice will specifically be doing, I will gently remind you that what you're doing may be good generalship, but it sure as hell ain't statistics.

catbarf
03-03-2010, 01:33
If you are telling me that, (as an example) taking into account the prevalence of vehicles in the metagame, you understand the value of anti-tank and are able to crunch numbers to find the squad loadout that, per point, gives you the highest % chance of destroying or incapacitating vehicles of any given AV, I will tell you that you are applying excellent mathhammer, and you are likely to be making well-informed judgments about how to commit your forces.

This is exactly what I advocate. What I disagree with is the notion that one cannot use probability to judge which of two options is the better. It's not going to be true in all situations, and there are usually assumptions made in the calculations, but in the end it's a subjective decision informed by probability and I see no reason why statistics cannot be applied in deciding which of two options is, in general, as a whole, superior.

I can work out the effectiveness of meltaguns and plasma guns for vehicle killing. There are going to be situations in which each is better. But I would feel justified in saying that meltaguns are better on the basis of calculations done, not because they reflect the game as a whole, but because in my opinion the scenarios I've used for comparison best reflect reality.

Sekhmet
03-03-2010, 01:54
This is exactly what I advocate. What I disagree with is the notion that one cannot use probability to judge which of two options is the better. It's not going to be true in all situations, and there are usually assumptions made in the calculations, but in the end it's a subjective decision informed by probability and I see no reason why statistics cannot be applied in deciding which of two options is, in general, as a whole, superior.

I can work out the effectiveness of meltaguns and plasma guns for vehicle killing. There are going to be situations in which each is better. But I would feel justified in saying that meltaguns are better on the basis of calculations done, not because they reflect the game as a whole, but because in my opinion the scenarios I've used for comparison best reflect reality.
You're disagreeing with a notion that no one in this thread has put forth.

He wrote:

You can't use it say "squad X is superior to squad Y" or "gun X is superior to gun Y." You can use to say "Squad X with loadout x, shooting at Squad Y with loadout y, ceteris paribus, will likely accomplish impact Z."

People spend too much time trying to do the first thing.

His wording implies he meant "strictly superior in every way," as evidenced by his later posts:


If Squad A does everything better than Squad B, that does not mean Squad B is worthless. Say Squad A is superior to every other squad (or just one specific other squad, Squad B) in every phase, but is limited in how many you choose (by force org or points value). This limitation is quite often the case. Also assume Squad A is much better at one thing (say, tank-busting) but only a bit better in another thing (say, board control). This is also often the case due to the specialization of units. Comparative advantage says you can maximize your army's utility by taking a few Squad A to break tanks and a few Squad B for board control-- thus you are at the point of maximum efficiency in output.

The only time mathhammer/theoryhammer can instruct you as to which squad is "better" is if it is, as you say, better at everything, in every situation. Even then, mathhammer doesn't tell you not to take the inferior squad, it merely tells you what to use each squad for.

And catbarf, I'd argue that you cannot say one gun is better than another except in the specific circumstances I outlined above (which, to be fair, do come up), mostly because guns do not exist in a vacuum. All mathhammer modeling has to take place in the framework of a real-game situation.


The only situation I've ever seen a strictly superior option in 40k that made any sense in a comparison was the old Kroot Merc list, in which you could take either an ork slugga or a bolt pistol for the same cost (i think) and on the same models, but the bolt pistol had ap5 while the slugga did not. Or maybe it was an ork slugga vs shuriken pistol.

In every other case, while overall in your current metagame, option 1 might be better, there will be places and times in which option 2 is superior, thus option 1 is not strictly better. If that situation ever came up, it would render option 2 completely and utterly useless to the point that no one in their right mind should take it, and it shouldn't have even been included in the game.

Pink Horror
03-03-2010, 07:29
Edit, just realised I said fish of fury you know what I know why tau are so disliked, it is all these silly little names. I mean really the number of them just for the suits is obscene. I will change it to some thing that annoys me less.

It's one of my biggest reasons for disliking them. I can talk about army lists with anyone except Tau players.

Archangel_Ruined
03-03-2010, 10:27
Sorry, getting back to controlling variables, things like the alphastrike list aren't so bad on the tournament scene precisely because your opponent wont deploy their forces to get dropped on turn one, in pick up games I've never seen anyone decide not to start on the board (deepstriking aside, obviously). This is why I think fully optimised armies just aren't cricket in casual play. And by the by, 10 marines will kill far fewer orks than the fire warriors at 25".

EldarWolf
03-03-2010, 12:14
alot of mathhammer, in my experience, goes hand in hand with rules lawyering and generally being a ******. If you're crunching numbers insteead of enjoying the 41st millennium then you're doing it wrong

This! Far too many people on here spend time doing just this!

The Highlander
03-03-2010, 12:42
Mathhammer is a tool. It can be a valuable tool at times, but it is still just a tool and not a replacement for judgment and insight built up from experience.

I normally use the same core to each of my armies for a long time. By doing this I am able to build up a detailed understanding of what they can and can’t do, which I can then use to decide what units to take and how to use them.
I also know the importance of flexibility, and having redundancy in my units to take advantage of any unexpected lucky (or unlucky) breaks, and of always having a backup plan.

Pete_x
03-03-2010, 13:18
Mathhammer is a tool. It can be a valuable tool at times, but it is still just a tool and not a replacement for judgment and insight built up from experience.

If, by insight build up from experience, you mean stuff like how my opponent will react then yes you are exactly right


I normally use the same core to each of my armies for a long time. By doing this I am able to build up a detailed understanding of what they can and can’t do, which I can then use to decide what units to take and how to use them.
I also know the importance of flexibility, and having redundancy in my units to take advantage of any unexpected lucky (or unlucky) breaks, and of always having a backup plan.

however this leads me to think you really mean that mathhammer will not tell you how a unit perform in a given situation. And that is where you are wrong. For exemple you know about redundancy because you failed an easy roll once and it screw you up. I know about redundancy because I can do the math and see that my fire dragons squad has about 8% chance of not destroying that that Land Raider. While small 8% of failure is no guarantee of success either. Add to that the fact it's mostly a suicide unit, etc. you get the point. It served no purpose that they failed to destroy my friends hammerhead last time I played him because i already knew it could happen even if that was a first. I didn't learn anything then because Math already told me it could happen.

Sekhmet
03-03-2010, 13:56
alot of mathhammer, in my experience, goes hand in hand with rules lawyering and generally being a ******. If you're crunching numbers insteead of enjoying the 41st millennium then you're doing it wrong
If you think there is only one valid way of enjoying a game, you're doing it wrong.

A lot of fluff players, in my experience, go hand in hand with generally being a ****** as well. Fluff players are often guilty of telling their opponent what they should and should not take in their own army. They feel perfectly justified in saying that you should not take two minimum veteran squads and load up on tanks and valks because it's not in the fluff, even though it was their opponent who spent the money to buy the minis and the time to paint them.

Live and let live, let people enjoy 40k how they want to enjoy it. Some people only collect minis, some people only paint, some people play with the intention of creating epic battles in the fluff, and some play with a competitive mindset. Any one of those people can be *******, and any one of them, assuming they play the game, can cheat, be rules lawyery, or post on an internet forum and argue about random things that don't matter in real life.

And just because your only experience with statistics is with "*******" doesn't mean that all, or even many, people who use statistics are. Just because in my experience, people who play entirely for the fluff make the game horrifically boring and unfun doesn't mean all fluff players are like that either.

tl;dr - don't tell me what to do and stop making assumptions.

Tactical Retreat!
03-03-2010, 14:13
Keep your government hands of my Medicare!

Sygerrik
03-03-2010, 15:21
This is exactly what I advocate. What I disagree with is the notion that one cannot use probability to judge which of two options is the better. It's not going to be true in all situations, and there are usually assumptions made in the calculations, but in the end it's a subjective decision informed by probability and I see no reason why statistics cannot be applied in deciding which of two options is, in general, as a whole, superior.

I can work out the effectiveness of meltaguns and plasma guns for vehicle killing. There are going to be situations in which each is better. But I would feel justified in saying that meltaguns are better on the basis of calculations done, not because they reflect the game as a whole, but because in my opinion the scenarios I've used for comparison best reflect reality.

What you are describing is not statistics. It is good generalship, and it is the kind of tactical thinking that wins tournaments. But it is not statistics. Whether it is "mathhammer" (a term that has yet to appear in the lexicon of higher math, as much as we may lobby for it) or not is debatable, given that there is no formal definition of "mathhammer." Probability is a very specific field that involves calculating the likelihood of given outcomes given one, fixed set of parameters. When are are calculating that meltaguns are superior to plasma guns when held by BS 4 Marines, at the single task of destroying a Chimera from 12" away, given that you have moved, and are facing the front armor... that is statistics. When you are saying that this makes meltaguns "better" than plasma guns, or a superior choice when you are facing a mechanized list, you are no longer doing pure statistical analysis but have instead verged into the field of generalship. You are making a subjective choice based on a numerical analysis, which is what you've been saying, but you cannot claim that the math is on your side. The math is on nobody's side. The math is impartial.

The Highlander
03-03-2010, 17:15
If, by insight build up from experience, you mean stuff like how my opponent will react then yes you are exactly right

Not quite, I meant things like what damage my units are likely to be able to do, and how much they will be able to absorb. For example, I know that if I charge my truck boyz into a single monstrous creature then there’s a very good chance I’ll win. I don’t need to sit down and do the maths to work out the odds, I can way up their chances based on past experience.


however this leads me to think you really mean that mathhammer will not tell you how a unit perform in a given situation. And that is where you are wrong. For exemple you know about redundancy because you failed an easy roll once and it screw you up. I know about redundancy because I can do the math and see that my fire dragons squad has about 8% chance of not destroying that that Land Raider. While small 8% of failure is no guarantee of success either. Add to that the fact it's mostly a suicide unit, etc. you get the point. It served no purpose that they failed to destroy my friends hammerhead last time I played him because i already knew it could happen even if that was a first. I didn't learn anything then because Math already told me it could happen.

Once again, that’s not what I was meaning by redundancy. What I was talking about was having extra units/wargear in my army to cover unforeseen circumstances. For example, all my shoota boy units have 2 rocket launchers and a Nob with a power claw to enable them to target tanks and hard targets. They’re not my first choice for dealing with these sort of things, but because I’ve given them some extra gear then can still attempt to deal with them if my preferred units can’t (because they are not in position, are dead or have failed). There are far too many variable in warhammer for mathhammer to be anything more than a tool.

Pete_x
03-03-2010, 19:23
@ the highlander

That is also what I meant as redundancy. I was in a hurry and didn't make it clear enough sorry. However you might be able to remember that your truck boyz will do good against a monstreous creature but your systems is ultimately flawed. At some point you might find yourself in a situation you have no precedence on, or worst one where your only reference are erronous as they standed out of the norm. Mathhamer has no such situation, only situation too complicated to calculate. That is where I'll concede your experience is a huge asset.

And yes I've been saying from the start mathhammer is a tool

catbarf
03-03-2010, 20:50
What you are describing is not statistics. It is good generalship, and it is the kind of tactical thinking that wins tournaments. But it is not statistics. Whether it is "mathhammer" (a term that has yet to appear in the lexicon of higher math, as much as we may lobby for it) or not is debatable, given that there is no formal definition of "mathhammer." Probability is a very specific field that involves calculating the likelihood of given outcomes given one, fixed set of parameters. When are are calculating that meltaguns are superior to plasma guns when held by BS 4 Marines, at the single task of destroying a Chimera from 12" away, given that you have moved, and are facing the front armor... that is statistics. When you are saying that this makes meltaguns "better" than plasma guns, or a superior choice when you are facing a mechanized list, you are no longer doing pure statistical analysis but have instead verged into the field of generalship. You are making a subjective choice based on a numerical analysis, which is what you've been saying, but you cannot claim that the math is on your side. The math is on nobody's side. The math is impartial.

I agree entirely, it's a bit of miscommunication. I interpreted a different meaning from your initial post-
I find the idea that mathhammer can tell you anything about how to play is laughable.
But all the same, it's a bit trivial. Of course there's a subjective component to any evaluation. You never see people simply cite statistics in isolation for no reason, it's always to provide evidence that supports their conclusions. I suppose there is a distinction between the act of applying calculations and the calculations themselves, but we just end up using a catch-all term... Mathhammer ;)

Sekhmet
03-03-2010, 21:03
Actually I do think statistics and "mathhammer" (how I do hate that word) can be used to create an effective tournament list and is not limited to providing specific unit v unit performance.

You can gather data across many high-level tournaments, mostly what lists were taken and how well those lists did. You'll have to weigh the value of some tournaments more than others, which is probably a subjective process, but not entirely inaccurate, based on how high the level of competition is, how close it is (geographically and skill-level) to the tournament/place you're playing in, etc.

Then you'll have some data on the amount of, say, low armor (av10, 11), medium armor (12) and high armor (13, 14) transports and the effectiveness of the units they're transporting, how much 2+, 3+ and 4+ armor units you'll see on average, etc etc. Then if you have to build a single list, a "take all comers" list for a tournament metagame, you'll have the statistical data necessary to add the perfect amount of anti-tank (melta guns, lascannons), anti-transport (autocannons/missile launchers/plasma), anti-light infantry (heavy bolters), anti-heavy infantry (plasma, missiles, lascannons, power weapons), anti-monstrous creature (plasma, power fists), anti-psyker (hoods via psykers, grey knights, or inquisitors), etc.

Most people don't use raw statistical data, they use anecdotal evidence on what are the most common units on the table (MEQ, GEQ and light transports) and build a list around killing those units, with an eye to dealing with ridiculous threats via melta and power fists.

But you could use statistical data to make a very scientific list. The formulas would be exceedingly complex and require a lot of thought into how to categorize your data, but it could be done.

Sygerrik
03-03-2010, 22:16
Actually I do think statistics and "mathhammer" (how I do hate that word) can be used to create an effective tournament list and is not limited to providing specific unit v unit performance.

You can gather data across many high-level tournaments, mostly what lists were taken and how well those lists did. You'll have to weigh the value of some tournaments more than others, which is probably a subjective process, but not entirely inaccurate, based on how high the level of competition is, how close it is (geographically and skill-level) to the tournament/place you're playing in, etc.

Then you'll have some data on the amount of, say, low armor (av10, 11), medium armor (12) and high armor (13, 14) transports and the effectiveness of the units they're transporting, how much 2+, 3+ and 4+ armor units you'll see on average, etc etc. Then if you have to build a single list, a "take all comers" list for a tournament metagame, you'll have the statistical data necessary to add the perfect amount of anti-tank (melta guns, lascannons), anti-transport (autocannons/missile launchers/plasma), anti-light infantry (heavy bolters), anti-heavy infantry (plasma, missiles, lascannons, power weapons), anti-monstrous creature (plasma, power fists), anti-psyker (hoods via psykers, grey knights, or inquisitors), etc.

Most people don't use raw statistical data, they use anecdotal evidence on what are the most common units on the table (MEQ, GEQ and light transports) and build a list around killing those units, with an eye to dealing with ridiculous threats via melta and power fists.

But you could use statistical data to make a very scientific list. The formulas would be exceedingly complex and require a lot of thought into how to categorize your data, but it could be done.

Ultimately the inescapable conclusion is that any consideration of how to build your list, while informed by statistics, cannot be done by pure statistics-- because certain unknowns come into play, and you have to simply put your best guess, based on past experience and anecdotal evidence, into these unknowns.

Catbarf, if I sounded a bit harsh, I do apologize; I think I misunderstood your position on the whole mathhammer thing. It bothers me to see statistics abused and the most flagrant and prolific abusers are wargamers. You do seem to really understand the power and limitations of statistical analysis.

Statistics cannot substitute for skill as a general. A good understanding of probability can help a good general make good decisions, but a lack of understanding of the tactical principles (force allocation and the like) underpinning the game means that regardless of how mathematically perfect a list is, it will not perform, since that player is unable to set up the situations where his statistically optimized troops are able to perform to capacity.

While it is possible to survey tournaments, separate unit types into groups, determine the most likely circumstances of a matchup against those groups, mathematically determine the optimum performer in your Codex for those matchups, then build a list by weighting the likelihood of encountering those matchups and including the precise fraction of counter units... by the time you're finished, you've made so many ungrounded or partially grounded assumptions that what you're doing is no longer pure statistics.

Pete_x
03-03-2010, 22:25
While it is possible to survey tournaments, separate unit types into groups, determine the most likely circumstances of a matchup against those groups, mathematically determine the optimum performer in your Codex for those matchups, then build a list by weighting the likelihood of encountering those matchups and including the precise fraction of counter units... by the time you're finished, you've made so many ungrounded or partially grounded assumptions that what you're doing is no longer pure statistics.

Or a new codex came out, the meta changed and all your hard work is obsolete before you even had a chance to put it too good use :D

catbarf
04-03-2010, 03:39
Apologies all round, Sygerrik. As you touched on, there's a reason wargames use dice- randomness. It takes generalship to deal with the fact that the dice will not always follow the averages, and the best mathematician in the world can't win if he can't deal with the fact that no plan survives hostile contact.

I've seen the dice-less wargame done before, but it is a chess-like affair, and not my cup of tea. There's just something about it that seems too direct- like arm wrestling versus an actual sport. They might both be testing the same abilities (physical strength and endurance), but the game is part of the challenge too.

Hive Fleet Bahamut
04-03-2010, 07:45
I don't believe in Mathhammer. :evilgrin: Why might you ask? Well statistically improbable as it may be (I've taken a 2nd yr university stats class so I know about statistical probability) I am a crappy roller. I've been playing all kinds of games that involve dice for about 20yrs. I'm still friends with some of the people I've been playing games with since back in the day and they know it too. I can't roll dice to save my life. I have terrible luck and so since stats don't believe in me I don't believe in stats.

:shifty: Some of this is a lie, some isn't.