AoS Army Balancing - AoS Warscrolls

[pendrake]

Pendrake's Fighting Value Formula
Draft 1​

Step 1: Add all the favorable stat numbers.

Wounds+Brave+Range+Attacks+2Hit+2Wound+Rend number+Damage = A

On models with multiple lines use the longest range number.
On models with multiple lines add up the total number of attacks from all lines.
On models with multiple lines use the highest damage number.
For Any number written as N+ use the following:
6+ =1; 5+ =2; 4+ =3; 3+ =4; 2+ =5; 1+ =6 (7 minus N)
For the Rend value use the magnitude but ditch the - minus sign.

Step 2: Add Save + Move together = B

Step 3: Multiply the the results of the first two steps. A x B = nnn

Step 4: Divide nnn by 10.
_______________________________________________________

Short form:
( W+B+R+A+2H+2W+R+D ) x ( M+S ) = Fighting Value

Fighting Value / 10 = Points Per Model ( but round up if there is any decimal )

 

[pendrake]

Point Values using Draft 1

14 Jungle Swarms [raw value 140]
38 Kroxigors [raw value 372]
51 Bastilodon [raw value 504]
31 Skinks
20 Temple Guard
28 Slann [Red Devils calculation]

 

[pendrake]

Pendrake's Fighting Value Formula
Draft 2​

Step 1: Add all the favorable stat numbers.

Brave+Range+Attacks+2Hit+2Wound+Rend number+Damage+Abs+Opt = A

On models with multiple lines use the longest range number.
On models with multiple lines add up the maximum possible number of attacks the model could legally use.
On models with multiple lines use the highest damage number.
For Any number written as N+ use the following:
6+ =1; 5+ =2; 4+ =3; 3+ =4; 2+ =5; 1+ =6 (7 minus N)
For the Rend value use the magnitude but ditch the - minus sign.
Add 1-5 points for special Abilities
Add 1-5 points for helpful Options

Step 2: Add Wounds + Save + Move together = B

Step 3: Multiply the the results of the first two steps. A x B = nnn

Step 4: Divide nnn by 10.
_______________________________________________________

Short form:
( B+R+A+2H+2W+R+D+Ab+Op ) x ( W + M+S ) = Fighting Value

Fighting Value / 10 = Points Per Model ( but round up if there is any decimal )


Sample Point Values Calculated
___________________

26 Temple Guard
38 Skinks

 

[The Red Devil]

Short form:
( W+B+R+A+2H+2W+R+D ) x ( M+S ) = Fighting Value

Would not this generate wrong in regards to the Rend? After looking at the warscrolls as far as I can see, if a unit has this it will only be negative? I.e. in your algorithm the rend will reduce the fighting value instead of increasing it.

Think it would be something like this? Though I am no math major either, so I'm certain its flawed
( W+B+R+A+2H+2W-(-R)+D ) x ( M+S ) = Fighting Value

 

[pendrake]

Would not this generate wrong in regards to the Rend? After looking at the warscrolls as far as I can see, if a unit has this it will only be negative? I.e. in your algorithm the rend will reduce the fighting value instead of increasing it.

Think it would be something like this? Though I am no math major either, so I'm certain its flawed
( W+B+R+A+2H+2W-(-R)+D ) x ( M+S ) = Fighting Value

I may well be confused about what Rend is. The larger the magnitude the better it is? I knocked off the minus sign before plugging in the value. And I think that is what you are doing with the double minus sign? e.g. "-(-R)" I think math majors have got special notion for that but I don't remember what it is either.

I went and added a notation to draft 1 and 2.

 

[The Red Devil]

I think math majors have got special notion for that but I don't remember what it is either.

Yea, abs (absolute value). The math indicator I think is |x| though while I use this a lot in programming it is only through the built in language construct for it, not by the indicator itself.

From what I understand the Rend, is used on the enemy units "save rolls". To negate them down. So if I have a save of four and I throw a four, but due to the attacking model has a Rend on -2 my four is actually a 2. or in reality the save of 4 became a 6 instead as we use in Ed. 8

Guess they worded it different to make it easier to understand for younger players.

 

[pendrake]

It is basically a high ST attack, ruining an Armour save? by another name?

 

[The Red Devil]

The first algorithm is seriously flawed on ranged units.

When you use it a Skink will cost 31 points, while a Temple Guard will cost 20 points each.

Formula for Skink:
(1+10 + 16 + 1 + 2 + 3 + |0| + 1) * (8 + 1) = 306 / 10 = 31 points (rounded up)

Formula for Temple Guard:
(1 + 10 + 1 + 2 + 4 + 4 + |-1| + 1) * (5 + 3) = 192 / 10 = 20 points (rounded up)

 

[The Red Devil]

The second is equally flawed, note on closer look both seems to be flawed on both ranged units and fast movement units (i.e. units with high movement).

By using the second formula the problem is how to decide Abs and Opt. I.e. who will decide what is the correct value to use here? The problem I see here, is that people will be biased and argue either for a lower or higher value all depending on what side of the table they are on.

However if I try to set a value on the two by just counting the abilities/values this is what I get.

A Skink will cost 38 points, while a Temple Guard will cost 38 points each.

Formula for Skink:
(10 + 16 + 1 + 2 + 3 + |0| + 1 + 3 + 2) * (1 + 1 + 8) = 380 = 38 points

Formula for Temple Guard:
(10 + 1 + 2 + 4 + 4 + |-1| + 1 + 2 + 3) * (1 + 3 + 5) = 252 = 26 points

 

[The Red Devil]

It is basically a high ST attack, ruining an Armour save? by another name?

Yes, if we compare to 8 Ed. I guess this is the best way of explaining it. Only difference you are told at once how much to ruin the Armour save with.

 

[The Red Devil]

After running these two tests against the two initial algorithms, it seems we have a few considerations we need to add, which unfortunately add a lot more complexity to the system.

1. There is a difference between short range and long range units. How can we make certain the algorithm does not favor one of them.
2. There is a difference between slow and fast units. How can we make certain the algorithm does not favor one of them.
3. How to calculate the difference of a warscroll that contain a unit vs. one that contain a hero? I.e. with the current algorithms the smallest warscroll of Skinks cost 310 points or 380 points, while a Slann costs 28 points or 47 points.

By an initial look at the warscroll stats, I think the only way we could possibly make an algorithm like this work, is if we change it completely and divide it into sections. Separated by type, i.e. Attack, Save etc. and multiply, add, and even subtract or divide on some, depending on unit type.

With other words, on a normal range unit we use the attach points once in the algorithm, while on a range unit, we need to include it in a different part of the algorithm to reduce the number slightly.

The problem with this is that it would be a mighty task actually creating an algorithm that work, and that will work equally across all races.

 

[Ixt]

I think that both players agreeing to:

'x battalions, x warscrolls, x wounds per unit max, x wounds per army max'

is a good place to start.

"2 Battalions, 5 warscrolls, 50 wounds per unit max, 350 wounds per army max," or whatever. Nuances will come with playtesting, and adjusted accordingly.

 

[The Red Devil]

'x battalions, x warscrolls, x wounds per unit max, x wounds per army max'

Yes, this approach sounds like it could work. The problem is how to count the special warscrolls, i.e. where the wounds are unlimited like with Kroak.

 

[Ixt]

"No special warscrolls," "1 special warscroll per army," "special warscrolls count as two/three warscrolls"

Ultimately, upon inplementing that rudimentary balancing mechanism, the rest of the game can be balanced on a unit-by-unit basis (just like 8th), or by simply talking with your opponent.

 

[The Red Devil]

I have tried a few different algorithm to automatically calculate points, and I do not believe it is really feasible to create this. Mainly due to the difficulty of actually creating accurate point values across all the warscrolls.

Ultimately, upon inplementing that rudimentary balancing mechanism, the rest of the game can be balanced on a unit-by-unit basis (just like 8th), or by simply talking with your opponent.

While what you suggest looks better, I feel it is still flawed to go by wounds, simply since the wounds on some of the models are "better" than other wounds.

I.e. if we go by wounds, it reduces the chance someone bring less efficient models, i.e. considering how efficient a model is vs its wound count. and the potential damage output it can have.


I guess the biggest gripe I have with the "no rules regarding balance" is that it feels wrong that we as the users/customers of the product need to invent the rules on how to balance the game so it will be fair for everyone.

 

[Ixt]

I agree. Points just won't cut it here.

I.e. if we go by wounds, it reduces the chance someone bring less efficient models, i.e. considering how efficient a model is vs its wound count. and the potential damage output it can have.

Yet, that's the same as it's always been with a point-based currency. "Why bring Saurus Warriors, when you could bring Temple Guard?" TG are way more efficient than Saurus... but they have different roles. That's why you bring Saurus. "Why bring x, when you could have x?" The answer is, 'because.'

So, has anything changed in that regard? To me, it seems not... especially with all of the dynamic new abilities and support options. Seriously, if you substitute the word 'wounds' for 'points' in your post, you'll see that it rings true for 8th, which was [mostly] a fine ruleset.

 

[The Red Devil]

So, has anything changed in that regard? To me, it seems not... especially with all of the dynamic new abilities and support options. Seriously, if you substitute the word 'wounds' for 'points' in your post, you'll see that it rings true for 8th, which was [mostly] a fine ruleset.

True, but there is a subtle difference between the points in Ed. 8 and wounds in AoS.

Now both a Skink and a Temple Guard has 1 wound. While in Ed. 8 you could almost get three skinks for the price of one Temple Guard. I can mention other similar cases as well, but I am certain the point has been made.

Though it will be interesting to see what kind of a system that becomes the norm, and if it will be fair enough to allow tournaments.

 

[pendrake]

First off thanks for giving it a serious look.

The first algorithm [...] a Skink will cost 31 points, while a Temple Guard will cost 20 points each.

[2nd Algorithim]... A Skink will cost 38 points, while a Temple Guard will cost 26 points each.

Intuitively, It does look like I have overrated shooting units. But have I? In these new rules they shoot after they have been engaged in close combat. The entire unit that survived the first round of CC shoots back on the next shooting phase? and then fights again in the next CC?
[[.. looks at Ruglud and his xbow Orcs ..]]

By an initial look at the warscroll stats, I think the only way we could possibly make an algorithm like this work, is if we change it completely

This is why I labeled them draft 1 and 2.

... on a normal range unit we use the attac[k?] points once in the algorithm, while on a ranged unit, we need to include it in a different part of the algorithm to reduce the number slightly ...

There is a fresh idea in here somewhere... ...specifics?

The problem with this is that it would be a mighty task actually creating an algorithm that work, and that will work equally across all races.

Which begs the questions: has there ever been such a formula? and if not, what were the point values pulled out of in all the previous editions?

Given that it is such a mighty task, maybe it is taking longer than they wanted...

I guess the biggest gripe I have with the "no rules regarding balance" is that it feels wrong that we as the users/customers of the product need to invent the rules on how to balance the game so it will be fair for everyone.

This is because a certain UK enterprise has habituated its customers to expect this.

 

[The Red Devil]

There is a fresh idea in here somewhere... ...specifics?

I spend a hour or two yesterday trying to make a different algorithm, but the problem was that no matter how I attacked it, some of the models kept getting either too low or too high score than the others. Though a key element if an algorithm like this is used, will be to divide the total on the core number of models in the warscroll. I.e. for Saurius and Skinks that is 10, and for temple guard it is 5. This would also solve the Slann vs. Skink issue since the Slann would cost the initial points.

ch begs the questions: has there ever been such a formula? and if not, what were the point values pulled out of in all the previous editions?

Normally balancing like this is done by using an algorithm for the initial numbers. But then after that the hard work start, with play testing and then adjusting the numbers until they feel right.

That is the problem I see here, i.e. that I do not believe the community will be able to agree on the "adjusted" number. Since me and you might have different thoughts about how efficient a warscroll is, due to we use different strategies when playing.

 

[pendrake]

That is the problem I see here, i.e. that I do not believe the community will be able to agree on the "adjusted" number.

I have been paying attention to Warhammer since 5e. I would observe that the community has never ever agreed on the 'adjusted' numbers; nevermind that those numbers were provided by "The Citadel".

I spend a hour or two yesterday trying to make a different algorithm, but the problem was that no matter how I attacked it, some of the models kept getting either too low or too high score than the others.

Too low or high defined by your... intuition... ?

Though a key element if an algorithm like this is used, will be to divide the total on the core number of models in the warscroll. I.e. for Saurius and Skinks that is 10, and for temple guard it is 5. This would also solve the Slann vs. Skink issue since the Slann would cost the initial points.

So, maybe where I was dividing by 10, divide by 5 or sometimes 1?