PDA

View Full Version : Probable n00b question

PaulmanMN
02-11-2009, 15:09
Got a question for all you old beardy players..

Being relatively new and all (about a year) to the hobby, I have already built up a couple of 40K armies to about 5000 points each, SM's and 'Crons.

What I would like to know is, (and I think I already know the answer), is there any "number crunching" software out there for you army to see how it would fare against X army?

I have most all of the spreadsheets and things for building up the actual force for game plat, but what I am looking for, is a piece of software, where I can punch in the number of points I am using, the untis themselves, and then give it X amount of points for a battle against X army, and let it run all the random numbers to determine my odds versus any particular force?

Clear as mud isnt it... thats the best way I have to explain it..

Thanks dudes..

PaulmanMN

IronNerd
02-11-2009, 15:13

Slightly longer answer: The game would cease to be interesting in any way if this was possible. There are SO many variables above and beyond the army lists that make up the game (like, oh, let's say... Player Skill). If the game was built where we could make armies and determine odds, we could just build armies and roll a D100... good game. Yes, there are armies that will fare better against a particular enemy, but it is IMPOSSIBLE to determine some sort of odds.

Hypaspist
02-11-2009, 15:16
Not that I am aware of, and the inherent problem you have here, is that mathhammer is all well and good, but how do you define the tactics and deployment in any given game, into a piece of software.

To put it more succinctly you could certainly mathammer 6 turns of Squad A shooting at Squad B and the eventual results. However there is no game I have ever played where two armies sit opposite each other, within each others optimal firing ranges and just blast away. (with no other interactions)

Factor in the complexity of Army builds (ie Ork Codex as a case in point, just *what* is a 2000 point Ork army..... so many variables.
In fact even if you have nailed down 2000 points of Orks, how can you guarantee what units will be brought to bear against each other in any given game.

If that helps?

Whitehorn
02-11-2009, 15:18
Roll a d6.

PaulmanMN
02-11-2009, 15:18
I figured as much, and I think my player skill isnt hideous, but... I really struggle against some stuff, and other forces i can just OWN, on a regular basis. I guess maybe I am just looking for a random "odds" generator.

ie:

10 Sm's rolling 20 d6's against toughness 5, what are my odds of rolling X.

*sigh*

Maybe I should write it myself, but I am far too lazy.

mightymconeshot
02-11-2009, 15:37
it is called mathhammer for that

Necro Angelo
02-11-2009, 15:41
...I recall some guy who was a computer expert-kind-of-guy who made a program to find the best ranked single combat models in 40k; it just confirmed what we already knew.

Karhedron
02-11-2009, 15:42
10 Sm's rolling 20 d6's against toughness 5, what are my odds of rolling X.

Well for shooting it is a bit of simple maths

20 x (2/3) x (1/3) = 4.44 wounds (before armour saves)

That is the average (mean) number of wounds you will cause with 10 Marines in close range firing at T5 targets.

The problem quickly becomes that you might have a flamer in there. How many wounds does the flamer cause? It depends on how bunched up the enemy squad is might be nothing to worry about or you may hose the entire unit.

It is these sort of uncertainties which cannot be represented in a program.

If you find you are having consistent problems against certain armies then take a wander over to the Tactics section of the forum as there is always plenty of good advice there for beating army X with army Y. If you find some of your opponents are loosing interest because you always beat them, you could even be sporting and suggest they do the same. ;)

CraftworldsRus
02-11-2009, 17:33
I figured as much, and I think my player skill isnt hideous, but... I really struggle against some stuff, and other forces i can just OWN, on a regular basis. I guess maybe I am just looking for a random "odds" generator.

ie:

10 Sm's rolling 20 d6's against toughness 5, what are my odds of rolling X.

*sigh*

Maybe I should write it myself, but I am far too lazy.

Someone did, I just can't find it. It was on my old laptop, but that things hard disk failed. It ran in the command prompt.

PaulmanMN
02-11-2009, 17:35
Well, I can regularily *POUND* on CSM's with SM's (He doesnt run a double lash/obliterator list), hold my own against Tau and IG. But if I play Necrons, they phase out every time against IG, Eldar, Tau and CSM's. I was just looking for a mathmatical "edge", if you know what I mean.

Sojourner
02-11-2009, 17:38
'Solving' the game of chess would take longer than the present age of the universe given the most powerful single computer ever built. 40k is many, many orders of magnitude more complex than chess, so any computational solution of the sort that you're implying would be inadequate to the point of being meaningless. Devising the appropriate approximations would be many PhDs worth of work.

So...no.

PaulmanMN
02-11-2009, 17:43
I wasnt really worried about the whole tactics thing, I am good with that, I just suck at math.

I just was looking for something that I could plug numbers into and get random results, and keep running it and looking for a trend. Always of course disregarding things like possible cover saves etc etc. Just purely something to run odds. Thanks dudes..

MrBims
02-11-2009, 18:56
I just was looking for something that I could plug numbers into and get random results, and keep running it and looking for a trend. Always of course disregarding things like possible cover saves etc etc. Just purely something to run odds. Thanks dudes..

Roll a d6. Repeat until you have as much data as you need to come to a conclusion. Then roll some more.

CrownAxe
02-11-2009, 19:28
I just was looking for something that I could plug numbers into and get random results, and keep running it and looking for a trend. Always of course disregarding things like possible cover saves etc etc. Just purely something to run odds. Thanks dudes..
If that's what you think you need you must suck at math

If you want the probably of an event, just assume infinite samples and multiply everything by the known probability (which is what you should do)

If you just want a finite sample, roll dice

bigcheese76
03-11-2009, 17:41
I dont actually know of any software that will do this at present, but it could be fun to try writing some of your own. Id write a program like that if i didnt spend all my spare time playing actual games.

sigur
03-11-2009, 17:51
I dont actually know of any software that will do this at present, but it could be fun to try writing some of your own. Id write a program like that if i didnt spend all my spare time playing actual games.

.....or doing one of the 50,000,000,000,000* things that are more important, interesting and useful? :)

*yes, i have software to work that out

Pooky
03-11-2009, 21:35
I was writing a program such as that a few years ago. I gave up after about 6 months of work :( There are too many variables to consider and I had to make a lot of simplifications to make my program work with my current software engineering skills.

Nezalhualixtlan
03-11-2009, 22:16
I don't think on an army by army basis you'd ever be able to plug in points and come up with a solution that was meaningful and was able to compute in your lifetime.

As others have said, the best you can do is mathhammer the basics, and any calculator will do for this, but I regularly use Excel just to keep records of computations saved. Match up units to units, if you are interested in doing a point by point comparison try to roughly equate the amount of points spent on each unit, although other useful calculations can take into account the general average unit strength fielded (it can make sense to compare different point values to see how many points of a particular unit should be sufficient to take out a different value of another unit when you know point to point one unit is far superior in a particular type of match up). Take the hits fired or done by one, by the chance of hitting, multiply by the chance of wounding, then by the chance of bypassing the armor save (or any additional things like FNP or WBB) so you can get the average number of wounds caused by one unit to the other. You can iterate the numbers across the unit subtracting out the likely number of wounds caused, and running the numbers again, and again until you see the likely outcome if you are concerned with how many turns it might take, but generally what you really need to know becomes apparent after that first step, which unit stacks up well statistically against another one. You can do comparative calculations to see which of your potential units or weapons stack up best against a particular type of unit, or even a very general comparison of weapons based on S vs T and Sv.

Once you have a pretty general concept of what stacks up well against each other individually however, you need to spread into the realm of army vs army, and that is when tactics shine which really come down to understanding your army its strengths and weaknesses, and your opponents army's strengths and weaknesses and finding the most efficient allocation of your resources against his based on the general probabilities as you've come to understand them with the mathhammer above so as to stack the odds (legally rule wise in game) as far in your favor as possible before the random die start rolling. This is where it gets more complicated than any program is likely to be able to give you an answer to, there are just too many variables on a board that is completely open in an undefined 3-dimensional space.