Forum adverts like this one are shown to any user who is not logged in. Join us by filling out a tiny 3 field form and you will get your own, free, dakka user account which gives a good range of benefits to you: No adverts like this in the forums anymore. Times and dates in your local timezone. Full tracking of what you have read so you can skip to your first unread post, easily see what has changed since you last logged in, and easily see what is new at a glance. Email notifications for threads you want to watch closely. Being a part of the oldest wargaming community on the net. If you are already a member then feel free to login now.

 Marmatag wrote:
I'm laughing at people who think Kabalite Warriors should be 8ppm.

For the first time in i can't remember when, Dark Eldar are actually playable. Leave Britney alone.

They were pretty OP in 7th before jink got nerfed to make passengers hit on 6's during jink.

Kdash wrote:

 Mr Morden wrote:

gendoikari87 wrote:
Look people i have the calculator if anyone wants it but heres what straight math says a few units should be:

Marines: 9.5
Sisters: 9
Guard: 5.5
Terminator(pf sb): 39.3
Gk terminator: 39.2
Gk marine: 20.1

Ok how eactly does the Marine pay nothing but 0.5 pts for the following bonuses:

+1 WS, S, T, Chapter Tactics

You can't include Chapter Tactics in that, because we all know that SoB are getting their own versions pretty soon. So it is a mute point.

I do agree with the rest though, that any statline increases should be reflected by more than a 0.5 point increase.

I mean, should it actually? This is kind of a general marine problem, paying for stats that don't actually come into effect. A Devastator squad really does not care that they have WS3+, nor does a Khorne Berzerker care that he has BS3+. It's even more egregious with HQ's. WS2+ reroll 1's is a pretty insane melee statline for a Captain... but if he's just holding a chainsword and a bolt pistol those stats are pretty irrelevant. Far less so if you give him a Thunder Hammer instead.

So Kabs were OP because their transports were undercosted for their protective ability. But Tacs were crap because you needed undercosted-to-zero transports for them to be OP? Double standard much?

Bharring wrote:
16% is actually the lower-bound *limit* that a stat change can impact short of having no impact at all.

The mimimum non-zero impact a +1T gives is 20%: going from a 2+ to a 3+ to be wounded yields a 20% increase in durability. But 3+ to 4+ is a 25% increase, 4+ to a 5+ is a 33% increase, and 5+ to a 6+ is a 50% increase.

Now, 1pt of T often has a 0% increase. So the average might be lower than the above might suggest. For more detail, here's each step vs a couple common strengths:

Step : S3 : S4 : S5
1=>2 : 20% : 0 % : 0%
2=>3 : 25% : 20% : 20%
3=>4 : 33% : 25% : 0%
4=>5 : 0% : 33% : 25%
5=>6 : 50% : 0% : 33%

Note how going form 1 to 2 has barely any effect, whereas 2=>3, 4, 5, and 6 have a much larger effect. You can't fairly point +1T because it's not a consistent value.

*sigh* let me explain how my calculator works. It has 9 sections armiger ogryn and sister. It compares how the unit being measured compares in melee, ranged and defense against the stated units so a tabfor armiger melee defense and ranged. The same for ogryns and sisters it then averages all these factors together with an algorithm that weights shooting and defense the same and melee for half. Theres also an output that gives just the base average of all 9 sections

It is not as simple as +1 T is half a point

Bharring wrote:
Can all (or at least mostly) agree that Kabs being 6 and Guardians being 8 is bad?

Yeah theres a problem, but the problem is with the guardians.

 Xenomancers wrote:
They were pretty OP in 7th before jink got nerfed to make passengers hit on 6's during jink.

No. They were crap.
Everything but reaver spamming was crap unless you were playing an equally screwed bottom tier faction. (So... Orks. Tyranids without Flyrant Spam. Genuine CSM).

For that to work, all underlying assumptions for the algorithm have to be true a vast majority of the time. That's incredibly unlikely.

Dandelion wrote:

Uh, vets don't have better WS than Guardsmen. Which puts your +1 BS at 2 points.

Ah, my bad. Vets are overcosted by 1 then.

Dandelion wrote:

Of course, if you look at current prices conscripts and infantry are the same points. Which means that +1 WS, +1 BS and +3 Ld are free...

Conscripts were nerfed to remove them from the game, not to make them appropriately priced.
When they were 3 points, there was a very clear floor for what the minimum a unit could cost was.

Dandelion wrote:

Also, did you just say that Skitarii rangers are overcosted? If anything they're undercosted, so your math can't be right. Can you imagine 6 pt Rangers?

Read the whole post. Rangers are correctly priced at 7 points because they pay for a bolter-equivalent upgrade, just the same as a Guardsman would have to. This is one of the reasons we can tell that units stem off of a Guardsman as the basic unit, and not off of a Marine.

To the people complaining that all stats aren't worth the same in practical application: And? That doesn't change the formula GW used to price things, and we can see where they've stuck with that model based on the point increase to Boys when they received exploding 6s at range (effectively +1 BS).

Spoletta wrote:

Bharring wrote:
Can all (or at least mostly) agree that Kabs being 6 and Guardians being 8 is bad?

Yeah theres a problem, but the problem is with the guardians.

This.

gendoikari87 wrote:

Bharring wrote:
16% is actually the lower-bound *limit* that a stat change can impact short of having no impact at all.

The mimimum non-zero impact a +1T gives is 20%: going from a 2+ to a 3+ to be wounded yields a 20% increase in durability. But 3+ to 4+ is a 25% increase, 4+ to a 5+ is a 33% increase, and 5+ to a 6+ is a 50% increase.

Now, 1pt of T often has a 0% increase. So the average might be lower than the above might suggest. For more detail, here's each step vs a couple common strengths:

Step : S3 : S4 : S5
1=>2 : 20% : 0 % : 0%
2=>3 : 25% : 20% : 20%
3=>4 : 33% : 25% : 0%
4=>5 : 0% : 33% : 25%
5=>6 : 50% : 0% : 33%

Note how going form 1 to 2 has barely any effect, whereas 2=>3, 4, 5, and 6 have a much larger effect. You can't fairly point +1T because it's not a consistent value.

*sigh* let me explain how my calculator works. It has 9 sections armiger ogryn and sister. It compares how the unit being measured compares in melee, ranged and defense against the stated units so a tabfor armiger melee defense and ranged. The same for ogryns and sisters it then averages all these factors together with an algorithm that weights shooting and defense the same and melee for half. Theres also an output that gives just the base average of all 9 sections

It is not as simple as +1 T is half a point

I'll admit that this is the best attempt i have seen on this forum to math stuff out (surely better than pitting models one against the other ), yet that mathematical model is still awfully incomplete to math out something of this complexity.
Assigning points through a formula is absolutely impossible, that's the sad reality.

So rather than laughing and assuming that the infantry paradigm you support is the cheap one, you could mention that. Because using guardian/marine/necron warrior/dire avenger paradigm, kabalites are nuts.

Gendo,
In other words, you overfit an equation based on maybe 18 data points (9 sections, across armiger, ogryn, and sisters), combined with vague assumptions (shooting = defense = 2x melee) and a very shortsighted postulate (unit is the sum of values, not the product) that various stats that actually do interplay are in fact independent.

Marines would be interpolation, I get that. But I would not trust that equation to be accurate.

I'm pretty sure GW doesn't use a strict formula, and if they do then doing so is misguided. Any formula would require too many false assumptions and fail to account for how stats actually interact with the game.

Martel732 wrote:
We can use the canonical sister vs marine comparison to examine this issue.

S2 hit AP0:
Sister: 0.33*0.33= 0.11 wounds
Marine: 0.166*0.33= 0.055 wounds
Marine takes half the wounds, or 50% less.

S3 hit AP0:
Sister: 0.5*0.33= 0.17 wounds
Marine=0.33*0.33=0.11 wounds
Marine takes 35% less wounds

S4 hit AP0
Sister: 0.66*0.33=0.22 wounds
Marine: 0.5*0.33=0.17 wounds
Marine takes 23% less wounds

S4 hit AP -1:
Sister: 0.66*0.5=0.33 wounds
Marine: 0.5*0.5=0.25 wounds
Marine takes 24% less wounds

S5 hit AP 0:
Sister: 0.66*0.33=0.22 wounds
Marine: 0.66*0.33=0.22 wounds
Marine takes identical wounds

S5 hit AP -1:
Sister: 0.66*0.5=0.33 wounds
Marine: 0.66*0.5=0.33 wounds
Marine takes identical wounds


S6 hit AP0:
SIster: 0.83*0.33=0.27 wounds
Marines: 0.66*0.33=0.22 wounds
Marine takes 19% less wounds

S6 hit AP-1:
Sister: 0.83*0.5=0.42 wounds
Marine: 0.66*0.5= 0.33 wounds
Marine takes 21% fewer wounds

S6 hit AP-2:
Sister: 0.833*0.66= 0.55 wounds
Marine: 0.66*0.66= 0.44 wounds
Marine takes 20% fewer wounds

This is just a sampling of weapon types. 20% pops up a lot in this analysis, but there is also a 0%. Enemy weapon distribution makes a huge difference. Obviously, S5 is the shaft for marines. How do you get a 16% average out without knowing enemy weapon distribution?

The analysis completely changes for boyz vs sisters. I'm a science guy, and I'm telling you the math is not simple on this at all.

your calculations are all wrong here. rather they are looking at the wrong perspective. The increase or decrease will always be 16.666% because thats 1/6

PuppetSoul wrote:

Dandelion wrote:

Uh, vets don't have better WS than Guardsmen. Which puts your +1 BS at 2 points.

Ah, my bad. Vets are overcosted by 1 then.

Dandelion wrote:

Of course, if you look at current prices conscripts and infantry are the same points. Which means that +1 WS, +1 BS and +3 Ld are free...

Conscripts were nerfed to remove them from the game, not to make them appropriately priced.
When they were 3 points, there was a very clear floor for what the minimum a unit could cost was.

Dandelion wrote:

Also, did you just say that Skitarii rangers are overcosted? If anything they're undercosted, so your math can't be right. Can you imagine 6 pt Rangers?

Read the whole post. Rangers are correctly priced at 7 points because they pay for a bolter-equivalent upgrade, just the same as a Guardsman would have to. This is one of the reasons we can tell that units stem off of a Guardsman as the basic unit, and not off of a Marine.

To the people complaining that all stats aren't worth the same in practical application: And? That doesn't change the formula GW used to price things, and we can see where they've stuck with that model based on the point increase to Boys when they received exploding 6s at range (effectively +1 BS).

Exploding 6's equals +0,33 BS.

"your calculations are all wrong here. rather they are looking at the wrong perspective. The increase or decrease will always be 16.666% because thats 1/6"

That's not even remotely correct. My calculations show the exact pertinent result we need to look at.

Lmao at the notion that exploding 6s is equivalent to +1BS.

 Arachnofiend wrote:
I'm pretty sure GW doesn't use a strict formula, and if they do then doing so is misguided. Any formula would require too many false assumptions and fail to account for how stats actually interact with the game.

This. False assumptions will kill you.


Automatically Appended Next Post:

Bremon wrote:
Lmao at the notion that exploding 6s is equivalent to +1BS.

So many people who play this are so bad at math.

Spoletta wrote:

gendoikari87 wrote:

Bharring wrote:
16% is actually the lower-bound *limit* that a stat change can impact short of having no impact at all.

The mimimum non-zero impact a +1T gives is 20%: going from a 2+ to a 3+ to be wounded yields a 20% increase in durability. But 3+ to 4+ is a 25% increase, 4+ to a 5+ is a 33% increase, and 5+ to a 6+ is a 50% increase.

Now, 1pt of T often has a 0% increase. So the average might be lower than the above might suggest. For more detail, here's each step vs a couple common strengths:

Step : S3 : S4 : S5
1=>2 : 20% : 0 % : 0%
2=>3 : 25% : 20% : 20%
3=>4 : 33% : 25% : 0%
4=>5 : 0% : 33% : 25%
5=>6 : 50% : 0% : 33%

Note how going form 1 to 2 has barely any effect, whereas 2=>3, 4, 5, and 6 have a much larger effect. You can't fairly point +1T because it's not a consistent value.

*sigh* let me explain how my calculator works. It has 9 sections armiger ogryn and sister. It compares how the unit being measured compares in melee, ranged and defense against the stated units so a tabfor armiger melee defense and ranged. The same for ogryns and sisters it then averages all these factors together with an algorithm that weights shooting and defense the same and melee for half. Theres also an output that gives just the base average of all 9 sections

It is not as simple as +1 T is half a point

I'll admit that this is the best attempt i have seen on this forum to math stuff out (surely better than pitting models one against the other ), yet that mathematical model is still awfully incomplete to math out something of this complexity.
Assigning points through a formula is absolutely impossible, that's the sad reality.

it gets close enough. 100% isimpossible 99.9% is possible if you do this for every unit in the game but you can get a good 90% close withthis, certainly better than gws gut feeling approach

So you claim. How do you know it's 90%?

Bremon wrote:
Lmao at the notion that exploding 6s is equivalent to +1BS.

depends on what kind of exploding 6s


Automatically Appended Next Post:

Martel732 wrote:
So you claim. How do you know it's 90%?

since your criticizing it I assume you have a better model? Id like to see it

I am starting to see how intelligent design endures in certain parts of the world.

The last time GW pointed units on this basis was maybe 2nd edition and 4th edition WHFB (if anyone can remember that far back).

gendoikari87 wrote:

Spoletta wrote:

gendoikari87 wrote:

Bharring wrote:
16% is actually the lower-bound *limit* that a stat change can impact short of having no impact at all.

The mimimum non-zero impact a +1T gives is 20%: going from a 2+ to a 3+ to be wounded yields a 20% increase in durability. But 3+ to 4+ is a 25% increase, 4+ to a 5+ is a 33% increase, and 5+ to a 6+ is a 50% increase.

Now, 1pt of T often has a 0% increase. So the average might be lower than the above might suggest. For more detail, here's each step vs a couple common strengths:

Step : S3 : S4 : S5
1=>2 : 20% : 0 % : 0%
2=>3 : 25% : 20% : 20%
3=>4 : 33% : 25% : 0%
4=>5 : 0% : 33% : 25%
5=>6 : 50% : 0% : 33%

Note how going form 1 to 2 has barely any effect, whereas 2=>3, 4, 5, and 6 have a much larger effect. You can't fairly point +1T because it's not a consistent value.

*sigh* let me explain how my calculator works. It has 9 sections armiger ogryn and sister. It compares how the unit being measured compares in melee, ranged and defense against the stated units so a tabfor armiger melee defense and ranged. The same for ogryns and sisters it then averages all these factors together with an algorithm that weights shooting and defense the same and melee for half. Theres also an output that gives just the base average of all 9 sections

It is not as simple as +1 T is half a point

I'll admit that this is the best attempt i have seen on this forum to math stuff out (surely better than pitting models one against the other ), yet that mathematical model is still awfully incomplete to math out something of this complexity.
Assigning points through a formula is absolutely impossible, that's the sad reality.

it gets close enough. 100% isimpossible 99.9% is possible if you do this for every unit in the game but you can get a good 90% close withthis, certainly better than gws gut feeling approach

"Gut feeling" is 99% of the times more accurate than the corresponding algurithm when done by someone with enough experience. I know that at least for engineering that's how the world goes.
I would be ready to bet that if we payed a pro to create an algorithm that calculates the point and apply it to the game as is, the result would be more unbalanced than the actual state created with "Gut feelings".

Bharring wrote:
So Kabs were OP because their transports were undercosted for their protective ability. But Tacs were crap because you needed undercosted-to-zero transports for them to be OP? Double standard much?

These things never existed at the same time. DE was pretty far gone once 7.5 codex came out. 7.5 was basically another edition that only a few armies got codex for.

"since your criticizing it I assume you have a better model? Id like to see it"

It can't be modeled accurately. You are the ones making the claims of accuracy.

"your calculations are all wrong here. rather they are looking at the wrong perspective. The increase or decrease will always be 16.666% because thats 1/6"

This is a common mistake, similar to the "odds of 6 dice all showing the same number) being 1 in 6 to the 5th not the 6th. The intuitive answer is so appealing, but it misses the critical.

If an event is 16% likely to happen, and the odds change such that 16% of the time that it would have not happened, now it will happen, the event now has a 32% chance of happening. The odds of it occuring have increasd by 16% *of the attempts*. The odds of it occuring have increased by 100% *of it's original chance*.

The odds of succeeding on a 2+ are 83%. If you change that to a 3+ (AP-1, for instance), the odds go down by 16% of the total possibilities. So 16% of the time, you now fail when you would have succeeded. But the odds go down by 20% of the chance of success: for every 5 times you would have succeeded, now you only pass 4 of them (out of 5). Conversely, your failure rate has doubled: for every 1 failure you would have had, now you have 2. So, when discussing the impact of a -1 on a 2+ save, if it took 6 wounds to kill you, it now takes 3. You have halved your durability.

The difference is because a saved would may as well not have happened. So you have a 1/6 chance of dying to the first wound. Assuming you pass that, you have a 1/6 chance of dying on the next wound. And again. And again.

gendoikari87 wrote:

Bremon wrote:
Lmao at the notion that exploding 6s is equivalent to +1BS.

depends on what kind of exploding 6s


Automatically Appended Next Post:

Martel732 wrote:
So you claim. How do you know it's 90%?

since your criticizing it I assume you have a better model? Id like to see it

Indeed the problem here is a matter of terms. "Exploding dice" is tipically referred to something like the Tesla rule, where a 6 generates 2 hits. In that case, it is indeed an additional point of BS.
Dakkadakkadakka is not truly an "Exploding dice", it's less powerfull and the effect depends on the BS of the model.

Writing the equations to properly balance this game would likely be categorically more difficult and complicated than writing the actual values that balance this game.

Spoletta wrote:

gendoikari87 wrote:

Spoletta wrote:

gendoikari87 wrote:

Bharring wrote:
16% is actually the lower-bound *limit* that a stat change can impact short of having no impact at all.

The mimimum non-zero impact a +1T gives is 20%: going from a 2+ to a 3+ to be wounded yields a 20% increase in durability. But 3+ to 4+ is a 25% increase, 4+ to a 5+ is a 33% increase, and 5+ to a 6+ is a 50% increase.

Now, 1pt of T often has a 0% increase. So the average might be lower than the above might suggest. For more detail, here's each step vs a couple common strengths:

Step : S3 : S4 : S5
1=>2 : 20% : 0 % : 0%
2=>3 : 25% : 20% : 20%
3=>4 : 33% : 25% : 0%
4=>5 : 0% : 33% : 25%
5=>6 : 50% : 0% : 33%

Note how going form 1 to 2 has barely any effect, whereas 2=>3, 4, 5, and 6 have a much larger effect. You can't fairly point +1T because it's not a consistent value.

*sigh* let me explain how my calculator works. It has 9 sections armiger ogryn and sister. It compares how the unit being measured compares in melee, ranged and defense against the stated units so a tabfor armiger melee defense and ranged. The same for ogryns and sisters it then averages all these factors together with an algorithm that weights shooting and defense the same and melee for half. Theres also an output that gives just the base average of all 9 sections

It is not as simple as +1 T is half a point

I'll admit that this is the best attempt i have seen on this forum to math stuff out (surely better than pitting models one against the other ), yet that mathematical model is still awfully incomplete to math out something of this complexity.
Assigning points through a formula is absolutely impossible, that's the sad reality.

it gets close enough. 100% isimpossible 99.9% is possible if you do this for every unit in the game but you can get a good 90% close withthis, certainly better than gws gut feeling approach

"Gut feeling" is 99% of the times more accurate than the corresponding algurithm when done by someone with enough experience. I know that at least for engineering that's how the world goes.
I would be ready to bet that if we payed a pro to create an algorithm that calculates the point and apply it to the game as is, the result would be more unbalanced than the actual state created with "Gut feelings".

Dude...are centurians and tactical marines part of this gut feeling? Gut feeling is worthless. Formula like this should have been the basis for every point cost in the index when play testing started. Then obvious discrepancies get moved up on down based on experience. If that were done in the beginning. We'd basically have a balanced game.

It's not gut feeling. It's probably anecdotal testing.