MECHANICS: Combat skills formulas

The place with help files for new players, weapon and equipment stats and other pertinent information for players of all levels and experience.
Adael
Posts: 1081
Joined: Sun Mar 22, 2015 12:34 am

Re: Combat skills formulas

Post by Adael » Tue Jan 22, 2019 3:01 pm

Skadi wrote:@Stark - is there any integer truncation within a step - that is, are 16/5 and 18/5 considered the same within a calculation (both becoming 3) or are they 3.2 and 3.6 respectively?
Things get truncated at the end of a step, so they're 3.2 and 3.6 within the step.
Skadi wrote:@Fermin 10 DB is pretty huge at the top end, but honestly I think master/zerk is really underlooked. For an 18 lb weapon, +10 OB is (18/5) * (10/3) = +12% bash chance. That's a TON when a nonmaster has something like 10% chance otherwise vs 150 db. Zerk offensive and the numbers get even sillier.
Zerk/offensive master is still just a 12% increase over nonmaster being zerk. The amount of OB postures gives you doesn't change with mood.
But I suppose the fractional increase masters get is what people notice more. Masters might have a 72% chance to bash someone abs/combo instead of 60% for nonmaster, which feels like less noticeable of a difference than, say, the 20% chance to bash a dodger when non-master has an 8% (pulling numbers out of my ass but the point gets across I think :P).
Reyne wrote:Assuming Stark's tool is correct (which I am), dodge gets seriously diminished returns against clubs past ~135 DB due to the clubs bonus always giving at least 9% chance to bash.
That's not necessarily a bad thing. The point of clubs is to make things sit, and 9% isn't that high imo.

Clubs aside, every 10 DB is a decrease in bash chance by 15%.

Coding, but lowering how much the PB of a character affects bash would be interesting. It wouldn't affect absers, but the combo bashers would be less powerful.

Monsoon
Posts: 105
Joined: Wed Apr 19, 2017 7:56 pm

Re: Combat skills formulas

Post by Monsoon » Tue Jan 22, 2019 3:43 pm

as far as I knew pb wasn’t supposed to effect bash chance. Ghost code :shock:

Fermin
Posts: 453
Joined: Sat Sep 10, 2016 12:16 pm

Re: Combat skills formulas

Post by Fermin » Tue Jan 22, 2019 4:43 pm

I certainly like dodge, a lot of fun aspects to it and like I said, looking over the numbers and formulas everything looks really very balanced. Guess I'm just fighting too many berserk fades!

What would be chance to bash 160 dodge with a berserk fade? Thought it capped out at like 60 but maybe i'm missing someting...

stark
Posts: 56
Joined: Sat Aug 26, 2017 4:03 pm

Re: Combat skills formulas

Post by stark » Tue Jan 22, 2019 5:32 pm

Skadi wrote:@Stark - is there any integer truncation within a step - that is, are 16/5 and 18/5 considered the same within a calculation (both becoming 3) or are they 3.2 and 3.6 respectively?
Yeah so the probability variable is an integer, which means any time it gets altered at all it will be truncated, it's how C does things by default.
Fermin wrote:I certainly like dodge, a lot of fun aspects to it and like I said, looking over the numbers and formulas everything looks really very balanced. Guess I'm just fighting too many berserk fades!

What would be chance to bash 160 dodge with a berserk fade? Thought it capped out at like 60 but maybe i'm missing someting...
It would cap at 65% because the berserk fade is usually mounted. If they're wielding a club 74%. But against 160db, a 18lbs non-club weapon at 210 ob mounted bashes it 47% of the time. So really, we're saying absolute best basher bashes absolute best db 1/2 of the time. That does feel quite balanced.

Skadi
Posts: 32
Joined: Sat Jul 01, 2017 1:10 pm

Re: Combat skills formulas

Post by Skadi » Tue Jan 22, 2019 8:27 pm

Adael wrote:Zerk/offensive master is still just a 12% increase over nonmaster being zerk. The amount of OB postures gives you doesn't change with mood.
But I suppose the fractional increase masters get is what people notice more. Masters might have a 72% chance to bash someone abs/combo instead of 60% for nonmaster, which feels like less noticeable of a difference than, say, the 20% chance to bash a dodger when non-master has an 8% (pulling numbers out of my ass but the point gets across I think :P).
Yeah, but imo that's a huge increase (2.5x!). Going with your numbers, that's going from a 1/12 chance to a 1/5 chance - with dodgers, most of the time you don't get too many chances to land an OUCH bash, so you need to make the few you get in the right situation count.

In an imaginary scenario where you have 3 people on a dodger and attempting a bash:

Chance at least 1 of them land a bash (20% chance, all masters): ~ 48.8%
Chance at least 1 of them land a bash (8% chance, all nonmasters): ~ 22.1%

I think the other thing is that DS seems to 1) zerk more readily and 2) have more characters with attack than LS, so it's more common for LS dodgers to get destroyed in a bash than DS dodgers (bonus points for FCs, who are probably riding, have max 18 dex, and may be weaving and taking the -10 db malus in addition to having lower HPs). You get a lot of hunters on LS that aren't contributing the same firepower even if they manage to land a bash.

Skadi
Posts: 32
Joined: Sat Jul 01, 2017 1:10 pm

Re: Combat skills formulas

Post by Skadi » Tue Jan 22, 2019 8:29 pm

stark wrote:
Skadi wrote:@Stark - is there any integer truncation within a step - that is, are 16/5 and 18/5 considered the same within a calculation (both becoming 3) or are they 3.2 and 3.6 respectively?
Yeah so the probability variable is an integer, which means any time it gets altered at all it will be truncated, it's how C does things by default.
Fermin wrote:I certainly like dodge, a lot of fun aspects to it and like I said, looking over the numbers and formulas everything looks really very balanced. Guess I'm just fighting too many berserk fades!

What would be chance to bash 160 dodge with a berserk fade? Thought it capped out at like 60 but maybe i'm missing someting...
It would cap at 65% because the berserk fade is usually mounted. If they're wielding a club 74%. But against 160db, a 18lbs non-club weapon at 210 ob mounted bashes it 47% of the time. So really, we're saying absolute best basher bashes absolute best db 1/2 of the time. That does feel quite balanced.
Thanks - I'm coming from Python so seeing X/5 as part of a step (rather than X/5.0 to make it a float) scared me and I wanted to check.

Adael
Posts: 1081
Joined: Sun Mar 22, 2015 12:34 am

Re: Combat skills formulas

Post by Adael » Mon Feb 04, 2019 11:46 am

stark wrote:Stab Chance

1) start with a 50% probability
2) add 2x level difference to the probability
3) multiply probability by stab skill / 99
4) If weapon skill is less than 80% the probability gets multiplied by weapon_skill / 99
5) If encumbrance is over 50%, the probability is reduced by 10 * ((carrying + equipped) * 10 / (can carry + can equip) - 5)
6) If the victim has notice on, the probability is reduced by notice / 5
7) If the victim is sleeping, the probability gets doubled
8) If the attacker has sneak on (whether or not the sneak was successful in hiding the attacker's entrance), the probability gets increased by: (sneak_skill - 1d101) / 10. This is where no-hide stabs come from. If a stabber doesn't prac sneak, they will never land a no-hide stab.
The bolded line has me confused. Nowhere does the code seem to take into account whether a stabber is actually hidden, or what their hide % is, for a successful stab to land. It seems to me you could still land in a no-hide even without sneak pracced?

stark
Posts: 56
Joined: Sat Aug 26, 2017 4:03 pm

Re: Combat skills formulas

Post by stark » Mon Feb 04, 2019 1:43 pm

Adael wrote:The bolded line has me confused. Nowhere does the code seem to take into account whether a stabber is actually hidden, or what their hide % is, for a successful stab to land. It seems to me you could still land in a no-hide even without sneak pracced?
It does though. There is a check after the stab / notice etc calculations that goes:

* If you're hidden, remove the hide bit (so that after the stab, which is being executed, you're standing normally in the room)
* Else, set the probability to 0

And after that, the sneak portion is added. So if you're not hidden, the only way to get any positive points on that probability integer after the zeroing out is via what sneak provides.

EDIT: I edited my original post to add a step 8) to the stab chance sequence to make it extra clear:

7) If the victim is sleeping, the probability gets doubled
8) If the attacker is not hidden, the probability is set to 0.
9) If the attacker has sneak on (whether or not the sneak was successful in hiding the attacker's entrance), the probability gets increased by: (sneak_skill - 1d101) / 10. This is where no-hide stabs come from. If a stabber doesn't prac sneak, they will never land a no-hide stab.

Elyse
Posts: 330
Joined: Wed Feb 18, 2015 1:47 pm

Re: Combat skills formulas

Post by Elyse » Mon Feb 04, 2019 2:18 pm

I could have sworn there were records of *mounted* stabs by non-fades (e.g. sneak not on; clearly not hidden) that on very rare occasions succeeded against low level targets. This is such a corner case that it really doesn't matter, but I do wonder where the (tiny) success percentage came from, if sneak is the only source of successful no-hide stabs. Of course, I could just be remembering wrong.

Adael
Posts: 1081
Joined: Sun Mar 22, 2015 12:34 am

Re: Combat skills formulas

Post by Adael » Mon Feb 04, 2019 3:31 pm

@stark Aaah thanks, that makes it more clear.

Post Reply