Math and Game Balance

[BitsHammer]

So I recently learned about a AMA on Reddit done by former-GW Game Dev James Hewitt, and ran across this little nugget:

 

 

Unpopular opinion time: maths does a pretty naff job of balancing a wargame.

 

We often used maths and spreadsheets and the like as a starting point for points values - I love designing a points calculator, me - but you'll never account for all the variables. In a game like 40k, with thousands of different units, each with their own unique rules, not to mention loads of different scenarios... what does "balance" even mean?

 

Like, take a Space Marine Devastator squad with four heavy bolters. If you're playing a game against an army of foot-slogging Orks, no vehicles in sight, and you're playing a scenario down the length of the table, with 30" between forces at the start, and minimal scenery... how many points is that Devastator Squad worth, compared to a Tactical Squad? What about if you're playing a Planetstrike game against a shooty Tau army in very dense terrain?

 

The only way to balance a game like this is to assign values, test and amend based on a broad consensus. Interestingly, with the General's Handbook / Chapter Approved, that's pretty much what's happening these days!

 

In the past I've argued that they should have some sort of formula and weighting system, but it seems that they did just that....and it didn't work as well as the idea sounds like it should.

[Rik Lightstar]

So I recently learned about a AMA on Reddit done by former-GW Game Dev James Hewitt, and ran across this little nugget:

 

 

Unpopular opinion time: maths does a pretty naff job of balancing a wargame.

 

We often used maths and spreadsheets and the like as a starting point for points values - I love designing a points calculator, me - but you'll never account for all the variables. In a game like 40k, with thousands of different units, each with their own unique rules, not to mention loads of different scenarios... what does "balance" even mean?

 

Like, take a Space Marine Devastator squad with four heavy bolters. If you're playing a game against an army of foot-slogging Orks, no vehicles in sight, and you're playing a scenario down the length of the table, with 30" between forces at the start, and minimal scenery... how many points is that Devastator Squad worth, compared to a Tactical Squad? What about if you're playing a Planetstrike game against a shooty Tau army in very dense terrain?

 

The only way to balance a game like this is to assign values, test and amend based on a broad consensus. Interestingly, with the General's Handbook / Chapter Approved, that's pretty much what's happening these days!

In the past I've argued that they should have some sort of formula and weighting system, but it seems that they did just that....and it didn't work as well as the idea sounds like it should.

Way back when in the days of Rogue Trader and 4th edition of WFB I think it was, they published the "formulae" they allegedly used to calculate the points for everything within the main rulebooks and encouraged you to make up your own weird and wonderful stuff.

 

In second edition I don't remember seeing an actual breakdown of how to assign a points value to something but the values allocated to things made it pretty clear that there was one.

 

I'm sure you could find access to these tables somewhere online and have some fun bringing it up to date. But as James Hewitt says, there are just too many variables for a simple formula to be effective.

 

Rik

[BitsHammer]

 

So I recently learned about a AMA on Reddit done by former-GW Game Dev James Hewitt, and ran across this little nugget:

Unpopular opinion time: maths does a pretty naff job of balancing a wargame.

 

We often used maths and spreadsheets and the like as a starting point for points values - I love designing a points calculator, me - but you'll never account for all the variables. In a game like 40k, with thousands of different units, each with their own unique rules, not to mention loads of different scenarios... what does "balance" even mean?

 

Like, take a Space Marine Devastator squad with four heavy bolters. If you're playing a game against an army of foot-slogging Orks, no vehicles in sight, and you're playing a scenario down the length of the table, with 30" between forces at the start, and minimal scenery... how many points is that Devastator Squad worth, compared to a Tactical Squad? What about if you're playing a Planetstrike game against a shooty Tau army in very dense terrain?

 

The only way to balance a game like this is to assign values, test and amend based on a broad consensus. Interestingly, with the General's Handbook / Chapter Approved, that's pretty much what's happening these days!

In the past I've argued that they should have some sort of formula and weighting system, but it seems that they did just that....and it didn't work as well as the idea sounds like it should.

Way back when in the days of Rogue Trader and 4th edition of WFB I think it was, they published the "formulae" they allegedly used to calculate the points for everything within the main rulebooks and encouraged you to make up your own weird and wonderful stuff.

 

In second edition I don't remember seeing an actual breakdown of how to assign a points value to something but the values allocated to things made it pretty clear that there was one.

 

I'm sure you could find access to these tables somewhere online and have some fun bringing it up to date. But as James Hewitt says, there are just too many variables for a simple formula to be effective.

 

Rik

 

I gave up on the idea some time ago, but it reinforces a point that came up on a recent Voxcast: balance isn't as clear cut as people think it is.

[BLACK BLΠFLY]

Having stratagems make properly calculating points for units even more difficult.

[jaxom]

Waaaaay back when at a Games Day Baltimore I was at a seminar and Gav Thorpe talked about points values as partially being internal references to an army. So 15 points of tactical marine was 15 points of value within a Space Marine codex and would not be the same points if you had a similar unit in a Guard codex.

[Brother Tyler]

I think that the quote from James Hewitt is spot on. The only way calculations and algorithms might work would be if the game was changed from a tabletop miniatures wargame to a boardgame with severe limitations such as those seen in Space Hulk, Lost Patrol, etc. There are so many variables beyond the basics of range, to hit, to wound, and save; and most of those variables are irregular/unregulated (e.g., different terrain, maneuver constructs, etc.). Math only provides a ballpark starting point, but a wargame as complex as Warhammer 40,000 involves just as much art in its creation/updating as it does science.

[chapter master 454]

So...as I have said once, twice and upon 40,000 times (there may be slight hyperbole there ;))

 

 

In Theory: Theory and Practice are the same

In Practice: Theory and Practice aren't the same

 

It really is difficult and even I had some deluded ideas in earlier discussions within and around 8th in regards to points cost (such as all weapons can have a set value where I was quickly shown that isn't possible. A Lascannon is far more potent with a marine than an ork. Hey, we all have bad ideas...Plato's Man anyone?).

[Marshal Rohr]

I’ve always leaned into armies should be balanced around the basic troops that will make the bulk of an army. A full strength Platoon of Guard should be roughly equivalent in cost to squad of tactical marines and that again to a mob of 30 boys and so on.

[Panzer]

I definitely agree with what James Hewitt wrote there, however sometimes GWs basis for things or their adjustments is so far off that I doubt they have a very good formular to start with and that it involves more winging/gutfeeling than it has to be.

[Exilyth]

That's why playtesting is so important - to find out how and where theory and practice diverge and to get a better idea about how to minimize that divergence.

[ERJAK]

I definitely agree with what James Hewitt wrote there, however sometimes GWs basis for things or their adjustments is so far off that I doubt they have a very good formular to start with and that it involves more winging/gutfeeling than it has to be.

Part of the issue is determining value. How do you account for the value of a wound, for example?

 

1 wound at T3 leadership 5-6 with a 6+ save is near worthless. Get 10 of those together and label them troops and now they're a useful cheap utility option. Get 40 of them together and make them immune to morale and suddenly they're the core of 80% of Chaos lists.

 

What about movement? Exactly how valuable is 7" vs 6" for a guardsman? An intercessor? A Berzerker?

Edited October 22, 2019 by ERJAK

[Lord_Caerolion]

I definitely agree with what James Hewitt wrote there, however sometimes GWs basis for things or their adjustments is so far off that I doubt they have a very good formular to start with and that it involves more winging/gutfeeling than it has to be.

 

Just see the previous editions of "no, plasma pistols cost the same as they always do, even if you can take a meltagun in addition to your existing weapons for 2/3 of the price!"

I also seem to recall this coming up in the Witch Hunters design interview, commenting about how it was hard to decide what points to give Arco-Flagellants, as they're a unit with no real save, that can turn into a unit of inconsistent blenders that you have no control over. What points do you give a squad that dies when someone looks at it funny, may roll really low for Attacks and do almost nothing, or roll ungodly high and annihilate everything?

There was also something along the lines of "ok, you've got Striking Scorpions, and they have their points value, but that's just to the Eldar. If you could take them as part of the Tau, who have no other real close combat like the Scorpions, their value to the army just increased significantly, so should they cost the same points?"

[MegaVolt87]

 

I definitely agree with what James Hewitt wrote there, however sometimes GWs basis for things or their adjustments is so far off that I doubt they have a very good formular to start with and that it involves more winging/gutfeeling than it has to be.

 

Just see the previous editions of "no, plasma pistols cost the same as they always do, even if you can take a meltagun in addition to your existing weapons for 2/3 of the price!"

I also seem to recall this coming up in the Witch Hunters design interview, commenting about how it was hard to decide what points to give Arco-Flagellants, as they're a unit with no real save, that can turn into a unit of inconsistent blenders that you have no control over. What points do you give a squad that dies when someone looks at it funny, may roll really low for Attacks and do almost nothing, or roll ungodly high and annihilate everything?

There was also something along the lines of "ok, you've got Striking Scorpions, and they have their points value, but that's just to the Eldar. If you could take them as part of the Tau, who have no other real close combat like the Scorpions, their value to the army just increased significantly, so should they cost the same points?"

 

Getting rid of random stats would be a start. New oblits are pointed at what can be assumed is their maximum dmg capability assuming max rolls. I would rather pay MORE points for higher guaranteed dmg than random potential max dmg. I think GW would benefit from hiring three mathematicians full time for the AoS, 40K and FW/ specialist games design teams. Plus they would make excellent guests as GW employees on vox cast etc and community interactions on social media. 

[Slasher956]

The only way(s) a maths based formula works is within a controlled environment.  When it becomes more, use what you have, then balance becomes harder.

 

Look at balanced wargames and you find that they either have generic stats (including weapons) and/or tight rules on terrain.  IE you must have 2 pieces of 3 x 4" LOS blocking & 4 3 x .5" linar terrain in a 2' square area of the board.

[Ishagu]

I think GW deserves far more credit than they get. Balancing a game like this is very difficult, and it has a massive number of units and factions that sets 40k apart from the competition. It's a unique beast.

 

40k 8th edition has been pretty darn good for the most part, and we've seen a large variety of units and factions used in all levels of the game. Play-testing is important, but GW gets more feedback done in 2 weeks of community play than any controlled play-testing can provide.

[MegaVolt87]

The only way(s) a maths based formula works is within a controlled environment.  When it becomes more, use what you have, then balance becomes harder.

 

Look at balanced wargames and you find that they either have generic stats (including weapons) and/or tight rules on terrain.  IE you must have 2 pieces of 3 x 4" LOS blocking & 4 3 x .5" linar terrain in a 2' square area of the board.

 

Old cityfight had recommended terrain densities for different points levels and table sizes. Older editions had superior terrain and more nauced LOS too. 

[BitsHammer]

 

The only way(s) a maths based formula works is within a controlled environment.  When it becomes more, use what you have, then balance becomes harder.

 

Look at balanced wargames and you find that they either have generic stats (including weapons) and/or tight rules on terrain.  IE you must have 2 pieces of 3 x 4" LOS blocking & 4 3 x .5" linar terrain in a 2' square area of the board.

 

Old cityfight had recommended terrain densities for different points levels and table sizes. Older editions had superior terrain and more nauced LOS too. 

 

I don't lament some of the things we've lost, but I miss older edition terrain rules.

[chapter master 454]

I do feel we could do with some updates to terrain really. Maybe a topic for a chapter approved to approach (and not being open play only).

[ERJAK]

The only way(s) a maths based formula works is within a controlled environment. When it becomes more, use what you have, then balance becomes harder.

 

Look at balanced wargames and you find that they either have generic stats (including weapons) and/or tight rules on terrain. IE you must have 2 pieces of 3 x 4" LOS blocking & 4 3 x .5" linar terrain in a 2' square area of the board.

There's no such thing as a balanced wargame. Even chess favors white.

[BitsHammer]

 

The only way(s) a maths based formula works is within a controlled environment. When it becomes more, use what you have, then balance becomes harder.

 

Look at balanced wargames and you find that they either have generic stats (including weapons) and/or tight rules on terrain. IE you must have 2 pieces of 3 x 4" LOS blocking & 4 3 x .5" linar terrain in a 2' square area of the board.

There's no such thing as a balanced wargame. Even chess favors white.

 

Very true, but there is room to give the second player something to balance it out.

 

I'd argue that if/when we finally get 9th (or even 8.5) the CP generation should go more of an AoS route where we get it on a turn by turn basis (1 CP a turn, +1 if you have a warlord on the table) with small bonuses (that cost points) for taking formations. Then give whomever is going second +1CP so they have more resources in exchange.

 

The whole battleline system would work well in 40k as well. But I'm biased about this.

[Vermintide]

There's something of a divide in mentality between the community and the designers, especially when it comes to online discussion. As I've noted before, the vast majority of online 40k content revolves around competitive tournament play. The community around that values mathammer disproportionately, and tends to focus around list-building in the context of "the meta".
 

As has been pointed out, mathematical balance only really means anything in a system with controlled variables. And that's the thing- When it comes to high-end meta competitive play, you know what you're going to be up against; like a game of chess opening with a classic old play, there's a familiar pattern to a lot of tournament games. You know exactly what your opponent is going to do with that blob of talos, and he knows exactly what you're going to try to do with those three smash captains. Within that almost scripted meta-context, mathhammer helps wring out efficiency and inform your optimal decisions.

Outside of that framework I think it's a lot less reliable. You don't know if you're going to walk into a game and meet some absolute madman who puts four hundred and twenty termagants (1680pts) down at the other side of the table. All your anti-tank is pointless and you just get swarmed. The points are really fairly arbitrary, in those terms.

[jaxom]

There's something of a divide in mentality between the community and the designers, especially when it comes to online discussion. As I've noted before, the vast majority of online 40k content revolves around competitive tournament play. The community around that values mathammer disproportionately, and tends to focus around list-building in the context of "the meta".

Personally, I'd say this seems to be the case for online discussion of rules and points. There's a lot of great online discussion of lore and hobby works by both competition and non-competition players. It is very hard to be on the other side of that fence as one who is happy regardless of "army efficiency" if I can sit down with a codex and come up with a few different lists that let me live the fantasy of playing that faction while maintaining a fun challenge for my opponent.

[Ciler]

There's something of a divide in mentality between the community and the designers, especially when it comes to online discussion. As I've noted before, the vast majority of online 40k content revolves around competitive tournament play. The community around that values mathammer disproportionately, and tends to focus around list-building in the context of "the meta".

 

As has been pointed out, mathematical balance only really means anything in a system with controlled variables. And that's the thing- When it comes to high-end meta competitive play, you know what you're going to be up against; like a game of chess opening with a classic old play, there's a familiar pattern to a lot of tournament games. You know exactly what your opponent is going to do with that blob of talos, and he knows exactly what you're going to try to do with those three smash captains. Within that almost scripted meta-context, mathhammer helps wring out efficiency and inform your optimal decisions.

 

Outside of that framework I think it's a lot less reliable. You don't know if you're going to walk into a game and meet some absolute madman who puts four hundred and twenty termagants (1680pts) down at the other side of the table. All your anti-tank is pointless and you just get swarmed. The points are really fairly arbitrary, in those terms.

A very good point. I also expect that, for the most part, rule devs never really consider what I would call "anti-fun" options that technically still fall within (as you pointed) the realm of the rules.

[BitsHammer]

 

There's something of a divide in mentality between the community and the designers, especially when it comes to online discussion. As I've noted before, the vast majority of online 40k content revolves around competitive tournament play. The community around that values mathammer disproportionately, and tends to focus around list-building in the context of "the meta".

 

As has been pointed out, mathematical balance only really means anything in a system with controlled variables. And that's the thing- When it comes to high-end meta competitive play, you know what you're going to be up against; like a game of chess opening with a classic old play, there's a familiar pattern to a lot of tournament games. You know exactly what your opponent is going to do with that blob of talos, and he knows exactly what you're going to try to do with those three smash captains. Within that almost scripted meta-context, mathhammer helps wring out efficiency and inform your optimal decisions.

 

Outside of that framework I think it's a lot less reliable. You don't know if you're going to walk into a game and meet some absolute madman who puts four hundred and twenty termagants (1680pts) down at the other side of the table. All your anti-tank is pointless and you just get swarmed. The points are really fairly arbitrary, in those terms.

A very good point. I also expect that, for the most part, rule devs never really consider what I would call "anti-fun" options that technically still fall within (as you pointed) the realm of the rules.

 

For some people those armies are fun.

 

I assume those people are probably accountants though, because even after so many years I can't find those things fun. 

[Vermintide]

I dunno, I'd probably find four hundred and twenty termagants fun. The only obstacle is, you know, Painting four hundred and twenty termagants.

But then again, I'm a Skaven at heart. So I would like that idea Just goes to show, though, that it's all really up in the air as far as balance is concerned.