Random Picks and Modding...

[Agrajag]

Just a tiny little thing I wish to say, I think you should make all calculation by multiplying.
That is, for a fighting commander, 100 HP and 20 Attack will be much more valuable than two commanaders, one with 20 attack and one with 100 HP.
So the higher your stats, the better your other stats become, because with the improvement of one stat you increase the usability of the other (increasing HP to 100 as an example, will mean that that commander will live ~10 times more than an ordinary commander, thus making all of his abilities much more useful).

[Ighalli]

I strongly agree with Agrajag about multiplicitive values. The problem with that is weighing the various stats (not to mention abilities) and then normalizing them (to go from the product to the price). The additive math is definately easier, but it should be obvious that a mage assassin is worth more than an assassin and a mage.

[Chazar]

On the other hand, why should a higher attack value make a fragile seraph (having a high cost by these calculations) even more expensive? I doubt that the attack value can be ever high enough to be of a use for most mages...

[Agrajag]

Chazar said:
On the other hand, why should a higher attack value make a fragile seraph (having a high cost by these calculations) even more expensive? I doubt that the attack value can be ever high enough to be of a use for most mages...

This would be a problem in the additive as well as the multiplicitive system which can easily be solved by applying "factors" for what is important for what each kind of unit (so you divide strength by 10 for a mage because it is almost completely useless for him, as an example).
The factors will be difficult to calculate though and increase the effort required by quite a bit (and or opposite to the original idea which was an absolute value for stats rather than one dependent on unit type.

[Scott Hebert]

Ighalli said:
I strongly agree with Agrajag about multiplicitive values. The problem with that is weighing the various stats (not to mention abilities) and then normalizing them (to go from the product to the price). The additive math is definately easier, but it should be obvious that a mage assassin is worth more than an assassin and a mage.

I disagree, and for a number of reasons. First, on the 'mage assassin' > 'mage' plus 'assassin'... why is that true? Yes, a mage assassin has more options than either a mage or an assassin (and thus should cost more), but if I have a mage and an assassin, and you have a mage assassin, you can only assassinate OR research/forge/ritualize/do magey stuff. I can do both (if not as well as the assassin mage).

For me, the 'opportunity cost' of having multiple commanders has to be worth something. (It's sort of an application of the chess idea of the 'overused piece'.)

Now, should they be equal? IOW, would the mage+assassin's cost equal the assassin mage's cost? Probably not. Even if the mage and the assassin mage had the same magic skills, and the assassin and the assassin mage had the same combat skills, the assassin mage most likely would not cost as much as the assassin and the mage combined, due to some overlap in other abilities. For example:

Assassin: 30g
Mage: 60g
Assassin/Mage: 80g

(These are 'off the top of my head', but the Void Child of R'lyeh, compared with a regular assassin or a 1-path mage, seems to bear this out.)

And yes, it's true that I can only buy one commander a turn, but look at the difference in upkeep should I buy one assassin and one mage, vs. two assassin/mages.

There are all kinds of permutations, but I don't see where a multiplicative system, as opposed to an additive system, is any more appropriate. Perhaps you can give examples of certain abilities that are better on 'better' commanders?

[Scott Hebert]

Agrajag said:

This would be a problem in the additive as well as the multiplicitive system which can easily be solved by applying "factors" for what is important for what each kind of unit (so you divide strength by 10 for a mage because it is almost completely useless for him, as an example).
The factors will be difficult to calculate though and increase the effort required by quite a bit (and or opposite to the original idea which was an absolute value for stats rather than one dependent on unit type.

This is more or less correct. A 'pure mage's' attack and strength stats are basically worthless. Unfortunately, that's not really inside my purview. (I.e., it's not my fault that attack and strength factor so little into magic.) Besides, if it were abnormally high, it would stand to reason that you could, theoretically, make some sort of combat commander out of them.

However, this is more or less a moot point, because 'pure mages' generally have low (normal) combat stats.

Perhaps I need to be clearer in what I intend to do. I do not plan to create a system to value commanders, and then 'pigeonhole' all the commanders into that system. I intend to attempt to discover a system that gives the large majority of commanders an accurate gold cost, and then apply that, to see which commanders are over/undercosted.

Any attempt to apply different standards to different commanders will require a subjective apportioning of the commanders into different roles that I feel could jeopardize the larger work.

[Agrajag]

Scott Hebert said:
Perhaps you can give examples of certain abilities that are better on 'better' commanders?

Just as a simple example, take this none-existing unit:
HP 1
Attack 30
Defence 1
Strength 30
Protection 1
AP 30
and this unit:
HP 16
Attack 16
Defence 16
Strength 16
AP 16

According to the additive system, both are exactly the same, while it is obvious that the second is a much better unit.
as another example you can take the first unit and up its HP, Defence and Protection to 30, according to the additive system that unit is now almost twice as strong as before, but according to the multiplicitive system it is 27000 times better, which IMO is a better represntation. (Okay, I realize 27000 is a bit too high, but you can scale down the results to recieve more normal values.)

[Scott Hebert]

I see where you're coming from, but (IMO) I don't think I have to worry about it. After all, I doubt anyone is going to try to make the first unit, right?

And then again, there's always extenuating circumstances. Give that 30 Att, 30 Str guy Etherealness and Luck, and... well, his low def/prot/hp was just mitigated to a large degree.

I'm not saying your example isn't instructive, but do you have an example from the game that is this distorted?

[Endoperez]

Would these values be counted before or after adjusting the equipment? As an example, Ulmish Lord Guardian will get prot 20 but -8 or so def with his equipment.

[Oversway]

One commander factor that I don't see mentioned (but may have missed) is slots. Maybe it doesn't matter so much since most commanders have either humanoid or horseman slots.

Does the mounted tag seem to be factored into the cost in the regular game? Perhaps it is that mounted commanders also have other good stats, but they always seem rather expensive.