Talk:Hit Chance

From TheKolWiki
Jump to: navigation, search

Some experience fighting tough monsters tells me that the hit and critical rolls are both the same roll, i.e. the algorithm is, in pseudocode:

roll for fumble
if no fumble:
{
  generate random double
  if roll > hit% then miss
  else if roll < critical% then critical hit
  else "normal" hit
}

Basically, critical hits are a fixed percentage of all non-fumble attack attempts, not a percentage of hits. I'll edit this into the article, please correct me if I'm proven wrong. --Poochy 05:03, 26 December 2006 (CST)

  • pcentella has done a lot of work on this as well -- can you post your results in the simulator of loathing thread? I think I see what you and he are saying now, but it would be nice to get confirmation. --DirkDiggler 05:11, 26 December 2006 (CST)
    • Err, I kinda don't have my exact numbers anymore. I originally came up with this hypothesis because I got bored while plinking away at some topiary golems (before the update that caused some monster stat variance), so I started counting my critical hits versus normal hits. Eventually, I formed the above hypothesis out of a couple more tests done by varying my Muscle a bit. Unfortunately, I didn't think the raw data itself was going to be of much importance, and didn't bother saving it. D'oh! --Poochy 06:26, 26 December 2006 (CST)
  • If fumble is a normal miss and crit a normal hit and they're equal for a monster hit chance, I don't see how they're relevant, since they cancel each other out. Either way, base * (1-(fumb+crit)) assumes crit is a miss, since it decreases the base chance and that's wrong imo. If you factor them in, they should be base * (1+crit) * (1-fumb). --MoreCurious 14:59, 30 March 2007 (CDT)
  • Kind of a random question, but which stat is used for determining hit chances for each character type if no weapon is equipped? Lostcalpolydude 04:36, 3 April 2007 (CDT)
  • After doing some spading, I think it is linear, and that the formula is basically correct, except the denominator is 11 instead of 10.5. Giving this a day or two before putting it in.--Top1214 16:03, 29 September 2010 (UTC)
    • Hmm.. That seems surprising. I would think that it was more like the Monster Hit Chance stuff, in that it wasn't linear, but looked a lot like it if you weren't looking very closely. (As in, it's a CDF of a Triangular Distribution). --RoyalTonberry 20:29, 29 September 2010 (UTC)
      • If you go to the kolspading page where I was working on this, I did over 700 rounds at muscle +5 and never missed and 1000 rounds at muscle -6 and never hit.--Top1214 19:50, 9 October 2010 (UTC)
        • Which would make perfect sense if the formula were similar to the monster hit chance calculation. The interesting parts are where you are at muscle-6 and hit some percent of the time, and muscle+4 and hit some percent of the time. It makes sense if you were to model the hit chance as:
base = player_muscle - monster_defense
will_hit = base + rand(0,5) - rand(0,5) >= 0

Which is triangular, and you only hit if your base number modified by those rolls is greater than or equal to 0. So if your muscle - monster defense is -6 (the monster has 6 more defense than your muscle), you will never hit. If your muscle - monster defense is -5 (the monster has 5 more defense than your muscle), you will hit if you get a 5 on the adding roll, and a 0 on the subtracting roll (which will happen about 1/12 of the time, or 8.33% (not counting crits). This is what the Monster Hit Chance looks like. It would make absolutely perfect sense that Player Hit Chance looks like this as well. Maybe I just need to gather the data and look at it. --RoyalTonberry 20:17, 9 October 2010 (UTC)

        • True it could be, but if it is, b/c of the small random number, it approximates linear quite well. The tails of Monster hit chance made it so that there were values that were expected to be 0, and weren't. This was cool and interesting to note. It made a difference at certain values. Here, though, the difference between a CDF and linear is going to be what, a 4% chance of hit probability? I know there's satisfaction of getting it exactly right, but getting it exactly right doesn't change a whole lot. I don't think 1 extra regular hit over 25 rounds will matter in any practical sense. To me, burning more turns (or time) on it just doesn't seem that valuable. I hesitated even changing the page at all, b/c again, the change would just affect percentages a little bit.--Top1214 20:49, 9 October 2010 (UTC)
  • The page says
normal hit 	= 	(1-fumbrate)*(baserate-critrate)
critical hit 	= 	(baserate) or (critrate), whichever is lesser
miss 	= 	(1-fumbrate)*(1-baserate)

which basically means that if you can't hit without a critical (baserate=0), you will never get a critical (critical hit = 0 or critrate, whichever is lesser = 0). That's almost certainly wrong, and in direct contradiction of "critical hits are a certain percentage of all non-fumble attack attempts". Shouldn't it be

critical hit 	= (1-fumbrate) * critrate
normal hit 	= (1-fumbrate) * (1-critrate) * baserate
miss 	= 	1 - critical hit - normal hit = 1 - (1-fumbrate) * (critrate + (1-critrate)*baserate)

? --XKiv 16:04, 10 June 2011 (UTC)


Moxie classes

Moxie classes seem to be having a bonus to ranged hit - for example, with 210 moxie I was always hitting at the post-war hippy camp, and with equal moxie at the goatlet (68) I was always hitting. This seems to be a change that happened in NS-13, but isn't an automatic 100% hit chance - was missing ~50% of the time with a stolen accordion at the Spooky forest having just hit level 2 as an AT. Tested the +hit bonus on both AT and DB, and with different signs, and my clean AT is having a much easier time hitting than he should be - at 32 moxie I wasn't missing against the Boss bat. Will investigate this more when my next multi gets his activation email sent, need the current one for spading damage. Draco Cracona 13:24, 19 October 2007 (CDT)

Mysticality Weapons

Who determined that with Myst weapons you use your mysticality for the chanche to hit? I have Mys: 21 (12), Mus 8 (6), equipping a pasta of Peril, and I have a really hard time hitting monster in the Degrassi Knoll. The only effect I have is Missing Kidney... --Luos 03:56, 25 October 2007 (CDT)

  • It looks like a case of a naïve edit creating a misleading comment after the NS13 nerfs for Spirit/glove. I believe Muscle was always used as to-hit for melee combat (and have edited the article to reflect this), but have no proof of it. --Bagatelle 22:07, 25 October 2007 (CDT)
  • I was under the impression that Myst weapons did use Myst to determine To-Hit chances since NS13. I swear I've hit things like rats and bats with a 20-30 Myst, a saucepan, and a significantly lower Muscle. I'll do some difinitive spading on this at the end of the current run (adventure with a Muscle at least 7 below the safe adventuring level of a certain area, but myst at leat 7 above it and see if I miss at all, then try the reverse). --Wildfire393 16:49, 30 October 2007 (CDT)
  • Nevermind end of this run, I forgot I had 30 extra turns I needed something to do with. Muscle of 100, Mysticality of 140. Using Knob Goblin Tongs (mysticality weapon). Adventuring in Black Forest (creatures require 140 stat to be hit, can't be hit below 123). Missed every round. Ok. Nevermind. I thought there was a correlation, apparently not. --Wildfire393 16:56, 30 October 2007 (CDT)

No More Misses

No more misses on regular attacks! Gotta update this and that. Did some experimenting. With 199 Muscle and Clumsy (Always Fumble), fighting a ML ~70 goat, tried a Shieldbutting, ended up getting a glancing blow, with a "you're out of your league" comment but no extra shieldbutt damage or delevel. Same results with every other -butt and -butt combo I could think of. Used a caret, next attack crit. Don't want to forgot, putting it here. --Johnny Treehugger 04:34, 6 September 2012 (CEST)

Even if you consider every glancing blow a miss, the page is very, very wrong. It says you can't miss if your attack is 5 higher than the monster's defense, but I get glancing blows at 1130 muscle vs. a monster with 1104 defense (Manuel's number). --Greycat 22:41, 21 January 2013 (CET)

I think that's a high-ML specialthing. --Raijinili (talk) 18:37, 8 May 2013 (UTC)