Character creation: maximizing damage per turn

[phiwum]

I'm afraid I don't know what you mean, but I was more interested in what counts as an "optimal starting human character" anyway. How do you define that term?

[hcobb]

Originally Posted by phiwum

I'm afraid I don't know what you mean, but I was more interested in what counts as an "optimal starting human character" anyway. How do you define that term?

It's whatever works for the monster in question. A human wizard vs a wizard wrath, crossbow and spear types for dragons, and so on.

[phiwum]

So is it a matter of your judgment or a real combinatorial exercise to determine what's ideal? Did you prove mathematically what the ideal starting character is? What are the ideal wizard spells or did you just go with Staff III or something?

Sorry if my questions sound too aggressive. I'm just trying to figure out how you calculated your numbers. Thanks.

[Nils_Lindeberg]

As an answer to the initial topic. There has been several simulation programs written, where you match basic starting characters and simulate 1 on 1 fights between them, thousands of times.

There was no clear winner, since the basic rule is that 1 DX higher characters usually beats the slower characters due to striking first. But when DX difference becomes too big due to high ST or the slower character is wearing armor, it goes the other way. Rock-paper-scissors comes to mind.

What my takeaway was, after writing one such monte carlo simulation was that the system is beautifully balanced. DX is not a dominating attribute like many people seem to believe. You could find both 9/15 and 15/9 builds as well as javelin experts in the 32p startup characters winners circle. But overall 11/13, 12/12 and 13/11 performed the best.

Ideally, you want 1 higher DX than your opponent, not more. If you have more, versions of your build with more armor will do better, so it is brought down to 1 higher. adjDX higher than 13 usually didn't do well, just as adjDX 8 or lower didn't do well either. High DX due to expertise was balanced, especially at higher attribute levels than 32. Even mastery was balanced once you reached 40 point characters.

Armor is better than people think, even in a one-on-one.

But basically, you can't go wrong with adj DX 9-12 and as much ST as your IQ allows. If you know your opponent adjDX try to adjust your adjDx to be one higher.

My simulation and the others didn't take HtH, ranged or special weapons like Bola into account. It would have required too many assumptions on the arena set-up. I took pole weapons into account and even that was a headache.

[phiwum]

Your analysis, Nils, is much more complicated than mine. I was only wondering where the sweet spot was for maximum damage per turn, which is easy to analyze mathematically. If you want to compare who does better in a fight, the analysis becomes quite difficult and so simulations are a better way to go.

Interesting observations.

[larsdangly]

Everyone who focuses on close-combat specialists (a small subset of all characters, but something we end up discussing a lot!) should be aware of the average damage output per turn, as in the OP. I'm sure everyone is aware that the relative value of characters depends a lot on situational variables (e.g., if everyone gets a few turns to get ready before closing, an arquebuser will wreck most people most of the time).

But another, more generic thing to consider in matchups is the effect of armor protection on a foe's effective offensive firepower. E.g., taken at face value, a horsebow specialist (10/13(16)/9 MW III)and a heavy crossbow specialist (15/8(11)/9 MW III) deliver similar damage-on-target in one turn of close range prepared fire (roughly 6.6, expected value). But when firing on a foe in heavy armor (say, chain+large shield) the horsebow's expected value drops to 0.3 (each hit has a 1 in 6 chance of scoring 1 pt of damage) whereas the heavy crossbow drops to about 3.5 (this is approximate; I just subtracted the armor from the mean damage in this case because I couldn't be bothered to calculate the probabilities of all the various outcomes; also, I've ignored double damage outcome, etc.).

The point is, your power ranking is very, very sensitive to your matchup with armor. I don't think it's really worth talking about which 'build' is best, even in brute force rankings like this, unless you are clear about how hard or soft your target is.

Also, this feature of the game leads to a surprisingly realistic arms race between weapon types and armor types - better than occurs in most roleplaying games (certainly way better than 5E D&D).