PDA

View Full Version : Everything is playable or broken, depending on point cost .....



ArmyC
12-07-2015, 17:50
Here is a broad brush concept that I would like some feedback on.

There is a pre-painted miniatures game on the market right now that requires you to use an iPad, or similar for all game mechanics.

Golem Arcana

Every time anyone plays the game, the stats are collected, analyzed and point costs are adjusted.

Obviously, they gather enough data to justify a change, but once they are sure, they can change the point cost in the app, and it is a world wide change immediately.

Thus, your 150 point Monster that is wiping the field, could be adjusted and the next time you play, as you enter your list into the app, you notice that same monster now costs you 175 points.

Crazy talk?

The idea that such a concept could be implemented for Games Workshop may be pie in the sky, but wouldn't it be worth it?

I would certainly pay $10 a month for a subscription to such a service.

We already have data slates etc. It is not a long distance to what I am talking about.

If GW created a sanctioned tournament program with the goal of collecting this data, it could be used to adjust point values and cure the genetic disease that GW suffers from.

What disease? That of Game Designer Arrogance (GDA). The 'studio' at GW that thinks they can redesign a game, play a few rounds and publish a rule set.

The above may not be the case, but GW shows all the symptoms of GDA.

Torrent of Fire collects this data for the most part already. Would it be possible to adjust point costs simply by analyzing what is played and what armies win?

I think it would get us further down that road than we are at the moment.

SuperHappyTime
12-07-2015, 17:59
Agree, but first we would need the following:
1. A starting point system.
2. A tournament setting.

And I really think #1 is easy to do. Because of how everything has changed, an initial algorithm like the one that was displayed on an earlier thread can fantastically get near balancing everything.

Gorbad Ironclaw
12-07-2015, 18:08
You can't really compare GA to moat other games because if you are using the app to play (you don't have to) then you are likely playing or ptemade map and the app tracks who moves where, fights who, how that fight goes, etc. That is way more details than you could get from something as simple as a tournament tracking software. Even if battle mode in Warroom (PPs app) worked well and was used a lot it still wouldn't give you the kind a data you need to so that kind of thing. Not to mention, it creates huge logistical problems to change things that way unless, as with GA, the game is really just supposed to be a physical extension of the app.

Also, the premise is false. You can't balance everything with a points system, some things will just never be balanced.

Deadhorse
12-07-2015, 19:08
Summoning cannot be balanced by points.

For example, a chaos lord summons an infinite number of models on a 4+ each turn. How do you cost such an ability?
Units that summon units that can summon other units are also "infinite".

You cannot points cost infinte abilities. You'd have to errata these rules.

sunborn
13-07-2015, 01:35
You cannot points cost infinte abilities. You'd have to errata these rules.

I am afraid you can balance on infinite points. A Markov chain could do something like that, I imagine.

Since the mod killed the other thread, I will repost:

I have been thinking of a similar but better idea. Why don't we train a neural net to play games constantly in an effort to optimize the points of every single piece? In theory, only a few points could be static, like MEQ). This should be done before a release by the Gee Dub. But lets be honest, they are too ****ed. So if I had the programming ability, I would build such a thing and produce a side points list, backed up with hard data. As a side benefit, formations could be pointed out (like old apoc) as they add a certain benefit.

Wolf Lord Balrog
13-07-2015, 01:39
<snip>
Crazy talk?

The idea that such a concept could be implemented for Games Workshop may be pie in the sky ... <snip>

Here is your main problem right here. The current leadership of GW are too stupid to comprehend the concept, let alone know that its a good idea, or be able to execute it competently.

*repost from closed thread*

insectum7
13-07-2015, 03:26
I would certainly pay $10 a month for a subscription to such a service.
You lost me right there. For many reasons, no.

Also, there's a tremendous difference between the ability to collect data from an app, and the ability to collect data from a tabletop game where zero mechanics are run through one of them whiz-bang flashy number machines (computer).

(Also reposted from closed thread)

Edit:
Also, I'm not one of those guys that just demands everything be free, rulesets, codexes, dataslates, whatever. But at the moment I have a fixed cost of some boxes of models and a book every once and a while, possibly a magazine or dataslates. The idea of adding to that cost with an additional $120 a year is just ridiculous. Not only because I don't want to pay it, but it would immediately reduce the value of codexes. The point values in them would be null and void within weeks.

TBH I'm not against the idea of apps. Digital dataslates are ok by me. But a subscription for just points? Get out. I want content for my money, I'm not going to pay for a data collecting feedback program that tries to sell itself as a "service". This app would have to do a LOT more than be a popularity contest for point values. For a subscription fee of $10 I want at least a collection of updated FAQs, say 10 to 20 a month based on "liked" questions submitted by users, and ALL the codexes plus their updates if the design studio is updating them. A "prime membership" type option is the only thing I would consider to be of reasonable value for a subscription fee. Nor should all players have to use it. Adjusted point values based on feedback by paying members should be given away as free PDFs for non-subscribers.

Even then, I don't think the data collection for points adjustments could be handled correctly by what would most likely be "minimum development/program cost". I think doing it right would be difficult enough to preclude the sort of resources put towards it and sustained.

Baaltor
13-07-2015, 05:58
Also, I'm not one of those guys that just demands everything be free, rulesets, codexes, dataslates, whatever. But at the moment I have a fixed cost of some boxes of models and a book every once and a while, possibly a magazine or dataslates. The idea of adding to that cost with an additional $120 a year is just ridiculous. Not only because I don't want to pay it, but it would immediately reduce the value of codexes. The point values in them would be null and void within weeks.

TBH I'm not against the idea of apps. Digital dataslates are ok by me. But a subscription for just points? Get out. I want content for my money, I'm not going to pay for a data collecting feedback program that tries to sell itself as a "service". This app would have to do a LOT more than be a popularity contest for point values. For a subscription fee of $10 I want at least a collection of updated FAQs, say 10 to 20 a month based on "liked" questions submitted by users, and ALL the codexes plus their updates if the design studio is updating them. A "prime membership" type option is the only thing I would consider to be of reasonable value for a subscription fee.


I want that too, and it should have articles written by the game designers, and the ability to communicate with them too.

Lord Damocles
13-07-2015, 10:23
Copy/paste from the other closed thread:


How would you know if a particular unit was over/under performing for its cost?

What if people aren't taking a particular unit because the background/model is unpopular? Does the points cost get adjusted downwards?

How would it judge units which have little impact as far as killing is concerned, but are beneficial as force multipliers? In terms of 'earning their points back', my Tomb Spyders suck; but they're awesome.



Also, any plan which involves me a) having to use an app for anything, and/or b) having to pay a subscription, can go jump.

Denny
13-07-2015, 13:15
I also wonder about how someone would feel after having crafted their perfect 1500/2000/whatever value list to turn up for a game only to discover they are now 35 points over.

duffybear1988
13-07-2015, 13:47
Yeah that would suck a bit.

lanrak
13-07-2015, 18:34
Considering Jervis said that he thinks 40k points are within 20% of the actual worth of the models in game.
He has no way of actually knowing this as no one at GW actually bothers fully play testing everything any more.

At 2000 'GW points' your army could be actually 2400 actual points or 3600 actual points.

If the armies were within 2% accuracy, (at least ten times better than GW can get them.)I do not think anyone would be unhappy. with game balance...

totgeboren
13-07-2015, 19:30
If you know multivariate data analysis, this approach would definitely give you a superb means of fine-tuning the cost of different units, because you would have so many different battles to cross-analyse. The inclusion of a specific unit X could result in say a 5% increased chance of winning, and thus deserve a slight point increase. If you are good at programming, this could even be done online automatically as data comes in.

But what it cannot do is identify weakly written rules. Sometimes the fault isn't really with the points, it's that the rules for the model is written in a bad way. Like that new-ish Tyranid brainbug. It's just so incredibly weak in the role it is supposed to fill that you have no reason for using it. Any algorithm like this would just drop the cost way down until people start to use it (for target saturation or something), but it still wouldn't be able to do what it was intended to do.

Sir Didymus
14-07-2015, 15:08
GW lives in the Dark Ages, and sees the internet only as a medium to publish pretty pictures of dolls.

They also have no understanding of the player base need to be competitive.

and last but certainly not least, GW can't decide whether their rules are a commodity or a marketing tool.

-

Take newcomer Corvus Belli's Infinity in comparison, who give out their rules for free, support their game with army builders for free, and have designed their game around a CB tournament system to provide constant feedback to their game designers. All this for free, so players can concentrate about spending their money on CBs product: Miniatures. They have a growth of 400% pr. year and are struggling to produce enough minis to meet demands. While they've eliminated war gamers main source of complaint, spending money on bad products.

The_Real_Chris
15-07-2015, 10:14
In regards to the title, not quite. You can balance with points within a range set by the game system.

So for example, you can't realistically point up a one shot wonder weapon that on the first turn will fire and have a 50% chance of destroying the enemy army. So a 1 in 2 chance of winning. Should that be 50% of your points? 100%?

Whereas if playing heads or tails on a coin it is fair that you have a points cost of 100% and a 0-1 limit :)

Extreme, but it is to highlight that extremes on the armour/mobility/firepower can be hard for a system to deal with and subsequently cost.

MiyamatoMusashi
15-07-2015, 10:32
So for example, you can't realistically point up a one shot wonder weapon that on the first turn will fire and have a 50% chance of destroying the enemy army. So a 1 in 2 chance of winning. Should that be 50% of your points? 100%? ... Extreme, but it is to highlight that extremes on the armour/mobility/firepower can be hard for a system to deal with and subsequently cost.

Absolutely true, but the good news is, that one-shot wonder weapon would also be boring and hence should have no place in the game. Of course that was a deliberately extreme example, but other (though more realistic) extremes that are difficult to cost, are also usually boring and have no place in the game. So just leave them out, then you don't need to worry about their cost!

An example would be Titans in 40K. They look fantastic, they're incredible centrepieces for an army, but the entire game spins around whether they can be destroyed before they wipe out the enemy. If the enemy brings Strength D and AV (maybe their own Titan), it can be done. If not, it can't. That's as hard to cost as a 50/50 wonder weapon... and just as dull. Who has fun looking across the board and thinking "there is nothing I can do to win here"?

Titans therefore don't really work in 40K. They're not so bad in Apocalypse, where it's accepted that many of the normal rules in the game will break down and you're just playing for the spectacle; but in your standard match-up, leave the Titan at home.

Titans fit much better in Epic. That system was designed for large war machines to be dominant. 40K wasn't. If a unit or rule fits the system, a points cost can be found. If it wasn't, it won't matter that a points cost is difficult to find, because you should just leave it out. ;)

The_Real_Chris
15-07-2015, 11:04
I would argue that in general the 'kill it before it kills us' situation is always tricky. And it gets worse the higher the firepower and lower the survivability is. Epic has problems with units ported from 40k that have instant kill weapons but low armour. (Anyone bringing a Titan to a tourney had better have a plan to deal with Deathstrikes before they can fire!) Airpower can fall into this as well depending on system. It is certainly the least satisfying part of Epic and the part that in playtesting has most often made armies unbeatable or too brittle with regards to winning or losing.

Saying that a lot of 'kill it before it kills us' games can be good. Ogre is built around it. But plenty of other games have units like Tiger tanks that will dominate unless knocked out and that forms an important part of the narrative.

lanrak
15-07-2015, 19:53
Point values only can give accurate as possible value to in game worth.

This is only the first part of a two stage process though.

Play testing can determine synergistic anomalies and the range of units that should be included in the game, from a game play perspective.Is the second part of the process.

Games that have more tactical considerations can allow more powerful units to be taken, as there are lots of ways to deal with them.

Eg a Tiger tank can pick off your tanks at long range , before you get close enough to even damage it.If the game play is limited to 'just killing stuff' , this unit would be boring to play and play against...

However, a smoke bombardment from some attached 105mm howitzers and the Tiger is 'blind' , and has to move to find targets.
And as the Tiger has been forced to move it looses its accuracy.
And its at this point your ambushing Tank Destroyers take it out with a flank shot...

Greavous
16-07-2015, 08:24
my biggest problems with this system would be.
1. many 'pro' players using it and the ponits being adjusted for top end play, making the lower people suffer by some being under pointed and some being over, it would be like DOTA2 where everything is balanced just for the top 1% of people and they effectivly force you to play an army a certain way.
2. gangs of people all giving false info to bump up a specific army.
3. rewriting a list every game to meet 'new' point costs.
4. people who DONT use the app e.g. print off the origional points list and only ever use them. (like some players who only use a specific edition codex).

stroller
16-07-2015, 09:10
Crazy talk?

Yep.

To be honest, I have no interest in any gaming subscription service. As others have said, it would also be open to manipulation. I also like the concrete: I've bought the book. It's mine to do with as I please. I do play ipad games, but at home, when I'm being unsociable.

Herzlos
16-07-2015, 11:48
I'd much rather the games company (whoever I'm paying for the rules/army books) gave me a set of points values that was close enough for me to play a game. Within 10% is probably fine.

I'm too old and busy to playtest for paid-for materials.

Bloodknight
16-07-2015, 12:49
Thus, your 150 point Monster that is wiping the field, could be adjusted and the next time you play, as you enter your list into the app, you notice that same monster now costs you 175 points.

Crazy talk?

No. We had a similar thing on the Mekwars servers (a campaign/tournament setting for Megamek, a java based emulation of BattleTech) 8? years ago with floating point values. The data was collected (at the time the server hosted about 30000 games a year) and the underperforming units got cheaper. It worked. But:

It didn't prove very popular with the player base once the players found ways to game the system. Which they always do. So yeah, the Shadowhawk was quite popular for a while until some dicks started intentionally losing games with high-powered units to make them cheap and abusing those to rake in the big campaign benefits.

Denny
16-07-2015, 14:21
It didn't prove very popular with the player base once the players found ways to game the system. Which they always do. So yeah, the Shadowhawk was quite popular for a while until some dicks started intentionally losing games with high-powered units to make them cheap and abusing those to rake in the big campaign benefits.

Awesome. :)

I think this problem lies at the core of all discussions of balance. The question shouldn't be 'Will the system balance the game?' so much as 'Will this system balance the game when a significant minority of players are going to deliberately try and break the game balance by whatever means necessary?'