Ironbound

Announcement

Collapse
No announcement yet.

Crit %`s over time

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • Jackal31
    replied
    Originally posted by Earlomorton View Post
    I think you also have to count misses from terror. Based on your rough critical hit % you would have had to miss, be evaded, or be parried on ~96 attacks or about 15% total attack volume. This is in the realm of plausible if you played against a good number of Sins.
    I thought about that too. Not just terror, but there is a pirate token (rum?) as well. I might track my next game for misses in comparison to recorded attempts. I'm not going too much more in depth.

    Leave a comment:


  • Earlomorton
    replied
    I think you also have to count misses from terror. Based on your rough critical hit % you would have had to miss, be evaded, or be parried on ~96 attacks or about 15% total attack volume. This is in the realm of plausible if you played against a good number of Sins.

    Leave a comment:


  • Jackal31
    replied
    That makes sense. If the crit is calculated from strike, then all attacks counted but resulted in parry or evade should be reduced. Makes me wonder how many there are in that collection.

    Leave a comment:


  • Earlomorton
    replied
    Originally posted by Jackal31 View Post
    we should reduce the amount of attacks that were evaded or parried when completing the calculations.
    This. If you attack 20 times and they Parry or Evade 3, you should use 17 attacks in your % critical calculations. Ignore doing the same for blocks because you can still critical hit on those.

    Leave a comment:


  • Jackal31
    replied
    You mean if I crit and they evade or parry?. I am using the stat at the end of the game. So to try and calculate overall damage, that would be a different approach.

    as I think about this more, an outcome of evade would still count as an attack, so any successful strikes have a possible chance of crit....is that the question? It would be fair to suggest that measure, but I did not take that into account.

    IF there were an average of 1 evade/parry per game, that would make sense. That means, we should reduce the amount of attacks that were evaded or parried when completing the calculations.

    Leave a comment:


  • Earlomorton
    replied
    Just curious, did you factor in your opponent's evades and parries in your calculation?

    Leave a comment:


  • Jackal31
    replied
    Originally posted by SteveJones View Post
    it's all over the place, really. i have monitored my crit percentage over many games and it's not consistent, but the usual response from devs or dev apologists is that the RNG is "true" and would calculate "correctly" (or according to percentage) over infinite sessions.

    argument against "capped rng" is that the game would become a fiesta of precise calculations where only the most efficient builds would persist in the meta.
    So, I monitored my last 31 matches and here's what I found...

    I should note, I'm playing assassin with emerald sword, efreet nail, dire charm, and ring of shadows. It's a standard 35 main and 30 off.

    In the 31 games, I had 168 of 629 strikes as critical supporting a 26.7%. I also went from Creulty IV to Creulty V and in those 15 games I had a 27%. So even if the weapon strike was balanced, the average across all games should be closer to 31.5 overall. My crit success is 5% lower. That's a 1 in 20, or 31 total strikes not collected.

    Ironically, I didn't notice at first, but once I got 100 wins, my matches got tougher. I've only won 5 of 15 games since (105-40 overall). Not sure if there's a design for this or not.

    Anyway, my build supports getting the crit. I'm thinking of changing the build so it's not dependent on crit since the accuracy is not consistent. The same arguments can be made about parry and other characteristics too (but the stats would be more difficult to collect and support).

    The only other thought I have is the consistency for all players. If everyone is getting the same results, then even if the function is broken, it's fair across the board. It's just deceptive to the user when the stats appear which builds the negative reactions.

    Leave a comment:


  • Earlomorton
    replied
    You'd think limiting it to 8 or 6 sigma deviations wouldn't do too much to a particular matches randomness. But I'm not a statistics or probability expert.

    Leave a comment:


  • SteveJones
    replied
    it's all over the place, really. i have monitored my crit percentage over many games and it's not consistent, but the usual response from devs or dev apologists is that the RNG is "true" and would calculate "correctly" (or according to percentage) over infinite sessions.

    argument against "capped rng" is that the game would become a fiesta of precise calculations where only the most efficient builds would persist in the meta.

    Leave a comment:


  • Jackal31
    replied
    A "true rng" will still support accurate averages though. Just like coin flipping (either electronically or not). It'll still be close to 50% with more and more trials. An electronic rng will still support accurate averages if set up correctly for favorable outcomes. I have had only a couple, that I can remember, where my crit % is near 40. So I know it's possible, but it doesn't appear to be consistent to balance.

    I've played games to support these percentages in my profile, and normally don't complain. This routine outcome is why I brought it up though. Based off my recent activity, I should have my next 6 games at 50% to balance this, or my next 15 games at 40%.... which neither makes sense (though theoretically possible).

    I might monitor it over more games...has anyone on discord done this already?

    Leave a comment:


  • SteveJones
    replied
    it's called "true rng" - meaning, the RNG is true across infinite variations (whihc nobody plays) but generally not accurate within limited sample sizes (which is the game we all play).

    everyone complains about it on the discord.

    Leave a comment:


  • Jackal31
    started a topic Crit %`s over time

    Crit %`s over time

    Hey everyone,

    recently joined over past few weeks and read though some of the forums looking for different things...that said, if I overlooked this, my apologies.

    I have been monitoring the crit success for myself and my opponents over some games. My current set up has main at 33% and off at 28%. My attacks are usually between 20-28 strikes per game. That should support between 7-9 crit strikes, however I am only getting 6-7 regularly and over my last few games it has been closer to 21% success (I've recorded 27/132 total for approximately 20.5%).

    I don't understand how this function is reflective of opportunity. I have my character set up for this, so when I collect a game where I have 3/28 success and lose by less than 10hp, it becomes frustrating. I also have the Greater Dire Charm equipped, so it really costs me 25 more hp when success is expected to be 30%.

    My second thought was it'll level out over time, but with only 6 games to compare and a 21% success....it doesn't appear to be true.

    Thoughts and feedback are appreciated.
Working...
X