# How to balance stats?

13 replies [Last post]
releppes
Offline
Joined: 09/17/2010

I'm working on a simple war game. In the game, units can be purchased. Each unit in the game has the following stats:

Strength
Armor
Movement
Range

What I'd like to do is assign a cost to each unit. Preferably something simple like add all the stats will give the cost of the unit. It sounds easy, but I think there's something more to it.

For example:

Say there's UnitA with S=4,A=4,M=3,R=1. Then I want to make a UnitB with M=4 and R=4. Now say I want both units to be the same cost. How would I pick appropriate values for UnitB strength and armor? Is there an algorithmic or is this something I need to figure via trial and error (ie: play testing)?

Maybe the question is more general. HowTo balance units when creating a war game? I figure there must be some way I can put a value on the importance of each attribute. In my game, strength and armor are near equal. I just can't see how to relate movement and range.

hulken
Offline
Joined: 04/18/2009
A easy way to do this is to

A easy way to do this is to give the units a "rock, paper and scissor" feeling to them. Then you can have the units costing the same to create. If one player builds onley "paper" units then the other can counter building the best units to counter this and so forth. Easy and self balancing. Also it works with more then three units types.

If you want to go mor oldscole you can use an exel spreadsheet to calculate al the values for eatch unit. If armour and attack is worth the same then you can calulate what the cost is per point, and the cost should be rufly the same for al the units, unless they have a special ability. Then you need to quantify that and ad that also.

I hope this is somewhat helpfull to you.

Relexx
Offline
Joined: 05/31/2010
My first question would be

My first question would be ...
Is S:4 A:4 M:3 R:1 more or less powerful a unit as S:4 A:4 M:1 R:3?

My guess is that there will need to be some form of scaling based upon movement/range for a units worth.

releppes
Offline
Joined: 09/17/2010
Relexx wrote:My first

Relexx wrote:
My first question would be ...
Is S:4 A:4 M:3 R:1 more or less powerful a unit as S:4 A:4 M:1 R:3?

My guess is that there will need to be some form of scaling based upon movement/range for a units worth.

This is exactly my question. How would I begin to quantify that?

At first I made the stipulation that Range and Movement would have the same cost. So the answer to the above question would be yes. Then I tried to relate Range to Strength by saying in an attack, Strength would taper off by 1 for each tile away from a target. I came up with saying Strength and Armor (being equal) would be a cost of 1 and Range and Movement (being equal) would have a cost of 1/2. This sounded good at the time, but then I realized I made some pretty big stipulations. There was no real basis for the comparisons I was making. How does one really say one unit is more valued on the battlefield?

There were other factors. For example: I was going to have terrain effect Movement, but not Range (keeping the game somewhat simple). This would definitely effect my assumption that Range and Movement were equal in value.

I had this naive thought that I was going to have all units be the same cost and vary the stats for each unit to provide variety. Now I realize it's not so simple. It's a big question, and I don't even know where to begin.

releppes
Offline
Joined: 09/17/2010
hulken wrote:A easy way to do

hulken wrote:
A easy way to do this is to give the units a "rock, paper and scissor" feeling to them. Then you can have the units costing the same to create. If one player builds onley "paper" units then the other can counter building the best units to counter this and so forth. Easy and self balancing. Also it works with more then three units types.

If you want to go mor oldscole you can use an exel spreadsheet to calculate al the values for eatch unit. If armour and attack is worth the same then you can calulate what the cost is per point, and the cost should be rufly the same for al the units, unless they have a special ability. Then you need to quantify that and ad that also.

I hope this is somewhat helpfull to you.

I think my strategy was your old school approach. Have a single cost for units, but each unit having varied stats. I got stuck on how to quantify units.

However, the game is a very simplified war game. The units don't even have health. It's a unit vs unit battle and the outcome is either (no damage, disabled, or dead). For the most part, battles are a luck fest with simple probabilities.

I like the idea of your "rock, paper and scissor" approach. How does one design such a system? Do you say unitA can kill unitB, but not unitC. Meaning, regardless how I define the stats for unitA, it's in the rule book that unitC is invulnerable to unitA's attack?

Or do I use my stats to say the probability of unitA killing unitB is very high, but very low if trying to kill unitC. Doing something like that sounds nice, but I don't know how I'd make it cyclic.

InvisibleJon
Offline
Joined: 07/27/2008
Figure out how useful each stat is.

Howdy,

Let's assume that Strength and Armor are, in fact, equally valuable in the game – That is: One point of Strength and one point of Armor "cancel each other out."

If that's true, then you just need to figure out the relationship between Movement and Range, then put them in terms of Strength and Armor.

I know virtually nothing about Movement and Range in your game, but it's likely that Movement should be more valuable than Range. Why? Range simply allows you to project and apply force over a longer distance, while Movement actually moves your unit, making it easier for you to claim positions, escape, etc.. In many games, Movement is used to determine how many actions you get to take, which is also very valuable.

Metaphysically, the values you give to the powers dictate how important you think they are in your game universe. Consider the universe created with the following values:

* Range = 1pt per
* Armor = 2 pts per
* Strength = 2 pts per
* Movement = 3 pts per

In this universe, it's very easy to build a long-range cannon that bombards targets from a distance, and it's hard to build a quick "assassin" unit. In this universe, things just don't move fast and range is so cheap that virtually every unit has a decent long-range attack.

Now consider this:

* Movement = 1pt per
* Armor = 2 pts per
* Strength = 2 pts per
* Range = 3 pts per

In this universe, it's easy to build quick units that zoom up to you and stab you at close range, and it's hard to build a long-range cannon. In this universe, everything is very mobile (perhaps they've mastered anti-gravity or hovercraft), but very few things can kill you at range.

The point system you choose will dictate how the game plays. Ask yourself how you want the game to play, then set your points accordingly.

Just a knee-jerk response, but I'd probably go with the following:

* Range = 3pt per
* Armor = 5 pts per
* Strength = 4 pts per
* Movement = 5 pts per

This point system:
* Slightly favors Strength over Armor, so it's a little easier to destroy things than it is to defend them. This means that the game will trend towards an end instead of "turtling," a stalemate, or just a random winner.
* Range is the least expensive, so you'll have a bit of shelling and bombardment before you get to close-range combat.
* Movement is as expensive as Armor, implying that the ability to move is as useful as the ability to prevent damage when you're hit. Because Movement is (relatively) expensive, you'll want to carefully consider how you position and move your forces. However, Movement isn't so expensive that you can't or won't purchase quick units.

...and that's where I'm at. I hope your game turns out the way you want it to!

hulken
Offline
Joined: 04/18/2009
You do not have to need to

You do not have to need to have a unit kill another. You onley need to give it proper advantage over the typ it is designed to kill. I do not know if you use dice to resolv combat, Give the unit a +2 if you use a t6 would have this effect.

If you have severa units it sould look somthing like this:
Swordman -> Maceman -> Spearman -> Horseman -> Artcher -> Swordman

This would be sthe simplest way to do this. If you want to make it a lilt more advanced you can make it like this:
Swordman -> Maceman, Spearman -> Horseman -> Artcher, Swordman, Maceman Artcher -> Swordman
Maceman -> Spearman
Spearman -> Horseman
Horsman -> Artcher, Swordman, Maceman
Artcher -> Swordman, Spearman
In this scenario the horceman will have a edge over every infantry exept the spearman. The artchers have advantage over al slower units.

How to quantify the range, if you want range and movement to be equal you have to have a 50/50 in the outcome in a fight betwen a unit with 3 movement and a unit with 3 range. (and the same attack, if the attack works with the range that is, if not then the moving unit need to have equal amount of attack as the range) in the fight have a 50/50 outcome then I would say range is worth the same as movement.

But I think range should be worth more. But I have to go to the bus now, mor on this later.

SiddGames
Offline
Joined: 08/02/2008
Playtest

I like Jon's analysis of how you want your game to feel to dictate relative costs.

An observation I'll repeat from someone else's post a long time ago on a similar topic -- you'll eventually have to start playtesting this. When you do, you may find that some units are very popular and some are never purchased. This may point to units that are too cheap or too expensive, which in turn may show you that you need to adjust the relative costs/values of the different stats.

In fact, I would say don't get bogged down RIGHT NOW in setting up all these elaborate costs among variable units. Give it your best guess and get to playtesting! One playtest with some poorly designed units will tell you a lot about their costs much faster than 6 hours of tedious spreadsheet manipulation to design "good" units that will still probably change after your first playtest, heh.

stubert
Offline
Joined: 01/26/2009
if there is a 1-1

if there is a 1-1 correspondence between strength and armor, then they should be "dollarized" the same.

i.e. - \$100 per unit of strength, and \$100 per unit of armor

Then, I believe that (especially if terrain is to be a factor) movement would weigh more heavily than range for three reasons:

1) you may be "obstructed" or out of range from another unit, but movement would allow you to either move from behind an obstruction, fire, then re-obscure yourself in a single move, or move within range, fire, then move back out of range (if movement rules allow),

2) you would be able to, as stated above, have much more freedom to establish guarded positions, claim territory, and maneuver strategically - which would potentially have a far greater impact on your victory than the range of your weapons,

3) if your movement values were higher, but enemy range was higher, you could move inside the range of enemy weapons more quickly than they could move away from you, closing the distance so that they were within range of YOUR weapons, effectively negating the benefit of the range of their weapons.

I would therefore place twice as much weight on movement than range (i.e.- \$200 per point of movement, and \$100 per unit of range)

I would also weight range and movement 1 1/2 as high as strength and armor, making the index

S - \$1
A - \$1
R - \$1.5
M - \$3

I would then go so far as to create a factoring index that would scale the costs in favor of certain combinations, such as:

A) High strength and high armor would be a nice combination, because in any battle with that unit, you would likely require multiple units to defeat it.

B) A unit with high movement and high defense would seemingly be unkillable, as you would not only have to be able to hit it (which presumably would be difficult with a high movement value), but you would have to do so multiple times, given its high armor value.

C) a unit with a high strength and a high range would be a very useful defensive stationary cannon, as you would be able to keep it near units you needed to keep safe so that it could fend off (or severely weaken) attacking enemy units before they could get within range of units you need to protect.

D) A unit with a high movement and a high range would be an excellent assassin unit, as it would be able to always stay outside of enemy ranges, but always keep them within its own - it would have to be called in to reinforce your units or else pinned down for the enemy to have any chance to destroy it.

For the record, any combinations of the stats would allow for a strategic advantage, and testplaying will reveal the merit of each, allowing for a reasonable set of scaling factors.

My scaling factors for the above would work out to something like:

A) 1.2
B) 1.5
C) 1.3
D) 1.8

This would mean that if you set standard unit value at 2, and high unit value at 4+, then the above units (A, B, C, D) would cost the following amounts (per dollar unit - if you use \$100 increments, then multiply these by 100):

STANDARD UNIT: \$13
S- 2 - \$2
A- 2 - \$2
R- 2 - \$3
M- 2 - \$6
scaling factor of 1
\$2+\$2+\$3+\$6 = \$13 x 1 = \$13

UNIT A - \$22.8
S- 5 - \$5
A- 5 - \$5
R- 2 - \$3
M- 2 - \$6
scaling factor of 1.2
\$5+\$5+\$3+\$6 = \$19 x 1.2 = \$22.8

UNIT B - \$37.5
S- 2 - \$2
A- 5 - \$5
R- 2 - \$3
M- 5 - \$15
scaling factor of 1.5
\$2+\$5+\$3+\$15 = \$25 x 1.5 = \$37.5

UNIT C - \$26.65
S- 5 - \$5
A- 2 - \$2
R- 5 - \$7.5
M- 2 - \$6
scaling factor of 1.3
\$5+\$2+\$7.5+\$6 = \$20.5 x 1.3 = \$26.65

UNIT D - \$47.7
S- 2 - \$2
A- 2 - \$2
R- 5 - \$7.5
M- 5 - \$15
scaling factor of 1.8
\$2+\$2+\$7.5+\$15 = \$26.5 x 1.8 = \$47.7

Even a standard unit with 3,3,3,3 would cost \$19.5, which is substantially enough lower than the value of your lowest valued special unit (and rounding off these costs could index them out at):

STANDARD UNIT (2-2-2-2) - \$125
ELITE UNIT (3-3-3-3) - \$200
BRUTE (UNIT A: 5-5-2-2) - \$250
GUARDIAN (UNIT C: 5-2-5-2) - \$275
CRUISER (UNIT B: 2-5-2-5) - \$375
ASSASSIN (UNIT D: 2-2-5-5) - \$475

This indexing allows you to establish a very nice baseline with scaling factors that make adjusting your math after testplay MUCH MUCH easier!!!!

If, say, you still thought that movement and range were more valuable than strength and armor, but the assassin just wasn't as valuable a unit as you originally thought, you could lower it's index and recalculate the cost using the same method, and the scaling would take care of itself.

In addition, if your testplayers suggest that you change multiple factors, you wouldn't need to do a lot of minor adjustments, just change the pertinent scaling factors (if you find that range and armor have the same value, and that guardian units are WAY better than you originally thought, then change the scaling factor for armor to \$1.5, and the scaling factor for UNIT C to 1.6).

This way, you don't have to retool the mathematical matrices that you use to come up with the per-unit-cost rating of the pieces, they will automatically scale themselves based on the net value that THE GAME PLACES UPON THEM.

Hope that helped...

releppes
Offline
Joined: 09/17/2010
SiddGames wrote:...In fact, I

SiddGames wrote:
...In fact, I would say don't get bogged down RIGHT NOW in setting up all these elaborate costs among variable units. Give it your best guess and get to playtesting! One playtest with some poorly designed units will tell you a lot about their costs much faster than 6 hours of tedious spreadsheet manipulation to design "good" units that will still probably change after your first playtest, heh.

I was thinking this was the way I'd have to go. I wanted to do some analysis up front. My game is pretty simple. I thought there'd be a simple algorithmic approach to doing unit cost and balance. In fact, that was one of my design goals. I intended the game to be a template one could easily expand into whatever they wanted. Use a simple formula to add new units, add a different theme, create new map scenarios and game objectives. It's more of a generic mechanics system. As such, I tried to keep everything very simple.

However, the whole topic of unit balance seems to throw the simple design philosophy out the window. Balance seems to involve so many variables. The comments on Movement vs Range for example. If terrain effects movement but not range, then the cost of either attribute might completely depend on how the map is constructed. And there's no real connection between strength and range. It's all, as pointed out, depends on how you want the game to play.

I was taking a cookie cutter approach to game design. Looks like I need to nail some things down and take a stab at what I really want the game to look like. It's hard to explain, but I was trying to avoid any personal style added to the game. I was really going for generics. The game itself is more of a prototype for making something else. At least that was what I was after.

It's like a Game SDK for computer games. Sure, you'll never make a great computer game from a prefab SDK, but it's a good starting point to learn. I was trying to make a generic template SDK for board gaming. I might use it as a starting point to make a two player war game or as an RPG kind of game. Neither game would be a block buster. The fun would be in making such games and learning from the experience.

Anyway, back to unit balancing. It looks like I have some work to do. Play testing, as you pointed out, is likely the best way to find out what's going to work and what's not going to work. As most game designers might know, it's not so easy to draft members into playing a new and untested game. Especially a horrible one ;)

releppes
Offline
Joined: 09/17/2010
scaling

The scaling system is a nice idea, because it takes care of other issues I was creating for myself.

I had this thought that some units might provide special abilities. Like a medic that can heal a disabled unit or a leader that would provide a +1 strength modifier to surrounding units. How does one even begin to put a costs on such units.

I think I got my answer. It's sounds like I pretty much take a stab in the dark. As part of the game, decide the types of units I want. If some units like Archers get +1 attributes for defense in the woods, then I just need to guess at the value of such abilities when assigning costs. If during play testing the unit is too powerful, then scale up the cost. It's the play testing I wanted to avoid, but I see it's the only way to really get a feel for the balance. There's just too many variables to try and take an algorithmic approach.

One of the other design goals I had for this "generic" game was that unit types could be designed on the fly during the game. It's why I was trying to find an algorithmic approach to calculating unit cost. I initially wanted it such that a player could create their own units. The stats of each unit would be it's cost. It was novel and cheap, but I see it's not so easy. One of the failures of such a system was tracking stats to each individual unit. Still, I thought it was an interesting approach to a generic game.

dekooning
Offline
Joined: 11/29/2010
help

how did you come up with the scaling factors for each unit

tridagam
Offline
Joined: 03/23/2009
Sounds like hulken nailed it to me

"rock, paper and scissor" is simplicity in its purist form. But its "perceived" complexity is endless. This will allow many hours counter tactical play without using time to make endless calculations and to stay true to a perceived reality. It allows what may be a complex thing to be simply manipulated. then if you want further control later on in development then overlay costs or recourses...or levels in tech (simple bow, compound bow, long bow, cross bow...)But all this rides quite elegantly on the "rock, paper and scissor" form.

stubert
Offline
Joined: 01/26/2009
Scaling factors explained

Well,

I started with the premise that strategic value should be higher than brute value.

that said, I ranked the "brute" unit (UNIT A - 5-5-2-2) the lowest, putting it just above the regular and elite units - it should be widely available for little cash, levelling the playing field for all players, as kind of a "super-elite" unit. (I gave it a 1.2, more than 1.1 so that I had room to scale it down if necessary, but not too much higher than an elite unit so that there was room for other scaling factors above it.)

I then thought that a stationary cannon (UNIT C - 5-2-5-2) would be another sort of "brute" unit (a "guardian") with more strategy than a regular "brute", but would have less strategic advantage than the other 2 units - so I ranked it next (closer to the brute than the others). (I gave this one a 1.3 - slightly better than a "brute")

Out of the other 2, I believed that the "assassin" unit (UNIT D - 2-2-5-5) should be about 1 1/2 times as powerful as the brutes, mainly because it would probably take 3 "brutes" to beat 2 "assassins" in a head-to-head fight. I ranked its strategic usefulness highest. (I gave this one a 1.8, which is 1.5x the value of the "brute" unit.)

Then, I didn't know how to rank the "cruiser" unit (UNIT B - 2-5-2-5), so I ranked it 2nd (better than the "guardian", but not as good as the "assassin"). Since I couldn't properly assess its usefulness, I just rated it as the average of the highest rated and lowest rated special units. (I gave this a 1.5 - the average of 1.2 and 1.8 - it can scale up or down later once its TRUE usefulness is determined through testplay).

I just reasoned it out using as simple a thought process as possible - the point is not to lock in the scaling factors, but rather to set up what you THINK should be the way the game should play out, and tailor the final game's stats to the testplay responses.

Just be careful:

DON'T LET YOURSELF GET LOCKED INTO A PARTICULAR MINDSET!! If you think the assassin units are great, but none of the testplayers can use them effectively, nor do they purchase them in any great amount, you will need to scale back their cost, even though you may not want to - the idea you've created is sound. Don't hold on to a particular detail you love if it will compromise the end result. - just listen to your testplayers - they may well become your target market if the game goes to production.