[GM] Individual Melee Combat Balance

[Ts_]

Originally Posted by Nymdok

I take as a given that all the attack manuvers are balanced already. I have NO statistical proof of this, but I assume that any glaring errors would have come up by now.

Well, they are "balanced" in a way that all maneuvers have serious drawbacks in a 1-on-1 ("skip" a turn, no defense). However, consider a brute with skill 10 and DR 20 vs a fencer with skill 20 but low DR and not enough strength to get through DR 20. Evaluate, feint, all-out-attack etc. suddenly make sense. Similarly, Aim seems like an essential maneuver for ranged combat.

Originally Posted by Nymdok

I specifically use deceptive attack (Even though it has its peculiarites) because its the only manuver that can reduce the Active defense of an opponent on the SAME turn its applied.

All-out-attack(Feint)? Or maybe a possible strategy is to Feint until you succeed and then follow up with a normal attack (deceptive, of course). Attacking every third turn (you win a quick contest about 50% of the time with equal skills, so after two expected rounds of feinting) is certainly bad, but the penalty to defense might make up for it (quite doubtful in most situations, but maybe I can figure it out precisely).

Originally Posted by Nymdok

NOTE: One fo the limitations of this model so far is that it assumes you dont Deceptive Attack your skill to below 16. I made this decision because below that, deceptive Attack can actually REDUCE your odds of landing a blow.

Umm, that would not be a limitation in your model if it was always like this. But: Skill 20 vs 20, thats attack 20 vs. defense 13 or defense 16 with a +2 shield and combat reflexes.
Assuming my math is correct, the chance to hit with 20 vs 13 reduced to 16 vs 11 is ... 42.6%, including automatical hits on a crit of <= 6.
If you further reduce to 14 vs 10 the chance to hit is 46.3%, but you have less good crits and risk more crit fails.
It's more pronounced with attack 20 vs. defense 16. Reduced to 16 vs 13 that's a 23.7% chance to hit, 14 vs. 12 is 24.9% and 12 vs 11 is 28.9%.

Anyway, I can see why 16 is a good point to land on, but it's not always the sweet spot for the chance to hit. I guess 75% less critical failures make up for 20% less hits, but I'm not entirely sure.

Regards, Ts

[Bruno]

One place where the "general rule" about stopping at 16 breaks down is if your opponant has defenses so high that you need to take a greater penalty to just bring his defense down TO 16 and actually start impacting his chance of success/critical failure at all.

Although at that point it's usually worth more to just take a Rapid Strike and double your chances of getting a critical success and skipping his defenses altogether.

But frankly, this sort of tactical simulation is NOT the point. This isn't a systematic attempt to replace actually playing the game - it's just meant to be a tool to try and make sure you're aiming in the right area.

[Nymdok]

Originally Posted by Ts_

Well, they are "balanced" in a way that all maneuvers have serious drawbacks...

All-out-attack(Feint)? ....

Umm, that would not be a limitation in your model if it was always like this. But: Skill 20 vs 20, thats attack 20 vs. defense 13 or defense 16 with a +2 shield and combat reflexes.
Assuming my math is correct, the chance to hit with 20 vs 13 reduced to 16 vs 11 is ... 42.6%, including automatical hits on a crit of <= 6.
If you further reduce to 14 vs 10 the chance to hit is 46.3%, but you have less good crits and risk more crit fails.
It's more pronounced with attack 20 vs. defense 16. Reduced to 16 vs 13 that's a 23.7% chance to hit, 14 vs. 12 is 24.9% and 12 vs 11 is 28.9%.

Anyway, I can see why 16 is a good point to land on, but it's not always the sweet spot for the chance to hit. I guess 75% less critical failures make up for 20% less hits, but I'm not entirely sure.

Regards, Ts

Im assumign that AOA (Feint, Attack) is balanced by the fact that you get no defense. This is, of course, more true in the general case than for a specific case in which it may not be true at all. For example if you have DR over your ENTIRE body which is more than the Damage the opponent can deliver on every turn, and the opponent has a High Active Defense, then Swing Away!

Your math IS right. Or at the very least, my chart agrees with you. :)

The trick is that for 20 Attack vs 16 AD we see that
16 v 14 = 17.5
but
14 v 13 = 16.3

That extra -2:-1 step DECREASES your odds of hitting by 1.2%, not alot but it gets more severe (as you might expect).


For 20 v 13 we have

20 v 13 = 23.7%
12 v 9 = 47.0%
10 v 8 = 37.5%

Which drops your odds by 12.5% !

which again seems reasonable when considering that past 12, were 'giving up' more area under the curve to hit and not gaining enough on the 'fails his defense' side.

Nymdok

[Nymdok]

Originally Posted by Bruno

...But frankly, this sort of tactical simulation is NOT the point. This isn't a systematic attempt to replace actually playing the game - it's just meant to be a tool to try and make sure you're aiming in the right area.

Precisely!

Nymdok

[Ts_]

"But numbers are ... you must ... how can you without them?"

Anyway. This was not meant to replace the game (well, not entirely), but to give the GM more understanding of the stats involved. I thought doing so was in line with the OP, and that I just tried to improve the technical quality, but maybe not.

I'll be watching this thread anyway for good advice!

Regards
Ts

[Nymdok]

Originally Posted by Ts_

"But numbers are ... you must ... how can you without them?"

Anyway. This was not meant to replace the game (well, not entirely), but to give the GM more understanding of the stats involved. I thought doing so was in line with the OP, and that I just tried to improve the technical quality, but maybe not.

Ts, I appreciate the input!

Im not trying to dimiss your arguments, Im simply trying to show why I made the approximations I did.

Nymdok

p.s. I used these numbers to run up a little example here
http://forums.sjgames.com/showthread...569#post876569

Id love to hear your thoughts in that thread or this one!