+1% Crit Arrow Myth
Comments
-
even with all those calculations, and oh by god that was some very scary math jargon, you can always go wrong just cause...your having a bad day
even a lvl1 barb will score a few misses on a lvl 1 mob(with a accuracy% of 99), what if the miss you scored was supposed to be your called shot? cause crit% is independant of accuracy, so infact even if you do calculate with something like 10^10 shots(hits+misses both) then you can still wont get a accurate value since hit is not at 100%. and if you scored a crit at that exact shot you missed then no one but the game engine will know what really happened back there.
now if a lvl 150 archer attacks a lvl 1 mob...i dont know if at lvl cap with end game weapons(most importantly highest accuracy ever) evasion:accuracy ratio, ie hit%, will be on 100 or still on 99. but at that case the mob is bound to die out on the first or second hit..even if the archer is using his bare hands armed with a crit arrow.
and i`d like to believe that the trigger resets with every lock on/killed mob. if the attack fixtures did NOT reset then it would be impossible for me to score even 5 misses for 100x2shots fired for a mob on whom i got a hit(accuracy)% of 98.
even taking the worst case scenario of having 2 misses at the ending and 2 misses at the starting of a 100x2 attack spectrum,
ie, [-,-...-,-,-,-,-,-,-,I,I] [I,I,-,-,-,-... (-= hit ; I=miss ; [x]=100shots)
then a 5th miss wouldnt be possible in another 98hits, but this clearly is NOT the case since i do end up missing before the string of 98hits completes.
so unless you get a damage immune mob on whom you got a hit% of 100 as your punching-bag-guinea-pig, you can go calculate while i go play a calming round of Tower of Hannoi, and after im finished around the turn of the millenium you will still be wrong.
only other way to get a perfect test dummie would be to attack a mob(on whom you got 100% hit rate) while out of its range, ie using a ranged weapon and checking for the number of crits you score per 100shots fired with +crit ammo. this would be the only possible way of solving the myth, but thats only if you get a mob on whom you got a hit rate of 100%.
@Psynopsis: no other factor would matter since the controlled environment is already flawed, cause there is a chance to score a miss and upset the crit rolls. so even if he got a sudden boost of +1000accuracy*(equips/buff/herbs/level ups/etc) and shifted his hit% from 98 to 99, EVEN then he wont get a true score since there is a 1% scope of failure on calculations.
at the end of the day, not like +1% crit will Really make a difference for any class who doesnt use 100% magic based attacks(tho it will make a small difference to archers, since it will addon to the base crit%) since with 1% crit(for any base lvl 1 char) or 2% crit(with +crit arrow), if you score a miss with the physical attack and if the engine makes that your crit roll, then thats when your having a bad day 0 -
No, sorry, your "scary math jargon" is wrong. b:laugheven with all those calculations, and oh by god that was some very scary math jargon, you can always go wrong just cause...your having a bad day
even a lvl1 barb will score a few misses on a lvl 1 mob(with a accuracy% of 99), what if the miss you scored was supposed to be your called shot? cause crit% is independant of accuracy, so infact even if you do calculate with something like 10^10 shots(hits+misses both) then you can still wont get a accurate value since hit is not at 100%. and if you scored a crit at that exact shot you missed then no one but the game engine will know what really happened back there.
As you correctly point out, hits and crits are independent of each other. Missing just reduces your sample size, it has no effect on increasing or decreasing the chances that your hits will results in crits or not.[SIGPIC][/SIGPIC]
PWI Merchanting Guides: warrenwolfy.wordpress.com0 -
even with all those calculations, and oh by god that was some very scary math jargon, you can always go wrong just cause...your having a bad day
even a lvl1 barb will score a few misses on a lvl 1 mob(with a accuracy% of 99), what if the miss you scored was supposed to be your called shot? cause crit% is independant of accuracy, so infact even if you do calculate with something like 10^10 shots(hits+misses both) then you can still wont get a accurate value since hit is not at 100%. and if you scored a crit at that exact shot you missed then no one but the game engine will know what really happened back there.
now if a lvl 150 archer attacks a lvl 1 mob...i dont know if at lvl cap with end game weapons(most importantly highest accuracy ever) evasion:accuracy ratio, ie hit%, will be on 100 or still on 99. but at that case the mob is bound to die out on the first or second hit..even if the archer is using his bare hands armed with a crit arrow.
and i`d like to believe that the trigger resets with every lock on/killed mob. if the attack fixtures did NOT reset then it would be impossible for me to score even 5 misses for 100x2shots fired for a mob on whom i got a hit(accuracy)% of 98.
even taking the worst case scenario of having 2 misses at the ending and 2 misses at the starting of a 100x2 attack spectrum,
ie, [-,-...-,-,-,-,-,-,-,I,I] [I,I,-,-,-,-... (-= hit ; I=miss ; [x]=100shots)
then a 5th miss wouldnt be possible in another 98hits, but this clearly is NOT the case since i do end up missing before the string of 98hits completes.
so unless you get a damage immune mob on whom you got a hit% of 100 as your punching-bag-guinea-pig, you can go calculate while i go play a calming round of Tower of Hannoi, and after im finished around the turn of the millenium you will still be wrong.
only other way to get a perfect test dummie would be to attack a mob(on whom you got 100% hit rate) while out of its range, ie using a ranged weapon and checking for the number of crits you score per 100shots fired with +crit ammo. this would be the only possible way of solving the myth, but thats only if you get a mob on whom you got a hit rate of 100%.
@Psynopsis: no other factor would matter since the controlled environment is already flawed, cause there is a chance to score a miss and upset the crit rolls. so even if he got a sudden boost of +1000accuracy*(equips/buff/herbs/level ups/etc) and shifted his hit% from 98 to 99, EVEN then he wont get a true score since there is a 1% scope of failure on calculations.
at the end of the day, not like +1% crit will Really make a difference for any class who doesnt use 100% magic based attacks(tho it will make a small difference to archers, since it will addon to the base crit%) since with 1% crit(for any base lvl 1 char) or 2% crit(with +crit arrow), if you score a miss with the physical attack and if the engine makes that your crit roll, then thats when your having a bad day
or u could hire a mage
we have 100/% hit rates0 -
too bad the arena is broken. Is there any inanimate objects that can be attacked like the stakes in TW?
If there is such a thing avail for attack, then equip crit arrow and use fists and count the amnt of crits u get in a 15min period for example. Do same without arrow and compare.0 -
Psynopsis - Heavens Tear wrote: »OK, yes, the sample size is a little small. But that would depend on the mean, standard deviation, and median of the data, none of which were listed in your posts.
The data are binary - did it critical or did it not? So there is no mean, sdev, nor median.0 -
Solandri - Heavens Tear wrote: »The data are binary - did it critical or did it not? So there is no mean, sdev, nor median.
Or, rather, there are but they can all be computed from the counts of successes and failures which the OP gave if the data is binary. However, I'm still curious about the count of misses (or the game's stated miss rate if the count is unavailable) since correcting the standard deviation for the possibility of observational correlation is the last piece to answering whether we have enough observations for a reasonable conclusion.0 -
While it's great that you know some words used in statistics, please don't contribute unless you can contribute something more than just random words you heard in your year 9 maths class.Psynopsis - Heavens Tear wrote: »OK, yes, the sample size is a little small. But that would depend on the mean, standard deviation, and median of the data, none of which were listed in your posts.
For starters, you are given all the information to calculate the mean, median and standard deviation.
Secondly, they're fairly meaningless for what we're doing here.
Except he's testing the crit rate, so unless one of them changed his crit rate, which we can fairly safely assume didn't happen because he'd know it'd invalidate the experiment, they have no effect. We don't care about the damage, so the damage buff didn't matter.In addition, there are a lot of factors that could have screwed it up. Did you ever change equipment? Did you ever level in between hits? Did you get buffs from a passing player, or did they disappear during that time? If the answer was yes to any of those questions, the statistics you listed are 100% invalid, because we cannot tell which data are accurate.even with all those calculations, and oh by god that was some very scary math jargon, you can always go wrong just cause...your having a bad day
even a lvl1 barb will score a few misses on a lvl 1 mob(with a accuracy% of 99), what if the miss you scored was supposed to be your called shot? cause crit% is independant of accuracy, so infact even if you do calculate with something like 10^10 shots(hits+misses both) then you can still wont get a accurate value since hit is not at 100%. and if you scored a crit at that exact shot you missed then no one but the game engine will know what really happened back there.
But the odds to miss a crit and a non crit balance out. Let's take an easy example of a 10% crit chance, and a 20% miss chance.
You attack 200 times. You'd expect 20 crits out of those.
You have a 20% miss chance, so you'd miss 40 hits, of which 4 of them (20%*20 crits) are crits.
You end up hitting 160 times, with 16 crits. Still the same crit rate.0 -
Eh, you need the hit rate if the figure given is the number of attacks, since we need to know the number of hits, but if we're given the actual number of hits then it doesn't matter, since that's the value we're using.Bobncut - Sanctuary wrote: »Or, rather, there are but they can all be computed from the counts of successes and failures which the OP gave if the data is binary. However, I'm still curious about the count of misses (or the game's stated miss rate if the count is unavailable) since correcting the standard deviation for the possibility of observational correlation is the last piece to answering whether we have enough observations for a reasonable conclusion.0 -
If you look in overall the +1% critical arrow give only+1% dmg to all damage which is relatively low. And for the percent, it adds an extra percent in your character information stats and therefore you have an extra percent of critical hit, you don't need to do math, you only need to read.~~~ My Guides ~~~
Mystical Tome (h ttp://pwi-forum.perfectworld.com/showthread.php?t=250862)
AOE Grinding with Venomancer (h ttp://pwi-wiki.perfectworld.com/index.php/Guide_to_AOE_Grinding_With_Venomancer)0 -
TameThat - Sanctuary wrote: »If you look in overall the +1% critical arrow give only+1% dmg to all damage which is relatively low. And for the percent, it adds an extra percent in your character information stats and therefore you have an extra percent of critical hit, you don't need to do math, you only need to read.
What they are saying is that, yes it does display +1% more crit on your character stats, however it may not actually apply.Main characters
Celestial Sage Venomancer Zoe - 100
Sage Barbarian Malego - 910 -
While it's great that you know some words used in statistics, please don't contribute unless you can contribute something more than just random words you heard in your year 9 maths class.
For starters, you are given all the information to calculate the mean, median and standard deviation.
Secondly, they're fairly meaningless for what we're doing here.
Except he's testing the crit rate, so unless one of them changed his crit rate, which we can fairly safely assume didn't happen because he'd know it'd invalidate the experiment, they have no effect. We don't care about the damage, so the damage buff didn't matter.
But the odds to miss a crit and a non crit balance out. Let's take an easy example of a 10% crit chance, and a 20% miss chance.
You attack 200 times. You'd expect 20 crits out of those.
You have a 20% miss chance, so you'd miss 40 hits, of which 4 of them (20%*20 crits) are crits.
You end up hitting 160 times, with 16 crits. Still the same crit rate.
I'd like to point out that the basic assumption is flawed.
The critical percent is not the portion of shots that will critical, its the chance that it will critical.
Therefore all the math being posted and debated isn't really a true reflection.
Perhaps you should calculated the number of possible hit combinations based on the probability of 2% - solve for "r" in a combinatoric.0 -
Uh...wha? Care to explain a bit more?angellicdeity wrote: »I'd like to point out that the basic assumption is flawed.
The critical percent is not the portion of shots that will critical, its the chance that it will critical.
Therefore all the math being posted and debated isn't really a true reflection.
Perhaps you should calculated the number of possible hit combinations based on the probability of 2% - solve for "r" in a combinatoric.
By "solve for r in a combinatoric", do you mean solving nCr = n!/(r!(n-r)!)? How does that help?
And the basic assumption is not flawed. The basic assumption is that for a sufficiently large number of trials E(X) is equal to the mean. If it is flawed, kindly tell us how it is flawed. And in the meantime, I would direct you to read up on probability.0 -
Uh...wha? Care to explain a bit more?
By "solve for r in a combinatoric", do you mean solving rCk = r!/(k!(r-k)!)? How does that help?
And the basic assumption is not flawed. The basic assumption is that for a sufficiently large number of trials E(X) is equal to the mean. If it is flawed, kindly tell us how it is flawed. And in the meantime, I would direct you to read up on probability.
Sure it is.
You are assuming that chance is directly proportional to the portion.
I'd advise you to read up on a bit of probability here, as anyone sufficiently versed in it would understand where the use of combinatorics would follow suit.
Furthermore, anyone versed in the study of probability wouldn't have made the fundamental mistake of saying that any of the calculations presented in this thread reflected the "chance" of criticalling in any sort of way.0 -
I've taken college-level statistics and got an A in the course. Most of the math here is fine (there's obviously been some disagreement), but the point is, %chance affects proportion over a large enough sample size. While we may debate on the necessary sample size, based on the calculations above, I think it's fair to say the n=6700 is well over the requirement for a standard 95% confidence test.0
-
Like I said earlier, his data seems fairly accurate.
The biggest question I have now is, does it even work AT ALL(maybe it's not even working on bows!? maybe the arrow itself is broken)0 -
I'm not assuming it's proportional at all. I'm merely assuming that I can find a meaningful confidence interval at a chosen confidence level given the number of trials, and the number of successes.angellicdeity wrote: »Sure it is.
You are assuming that chance is directly proportional to the portion.
I'd advise you to read up on a bit of probability here, as anyone sufficiently versed in it would understand where the use of combinatorics would follow suit.
Furthermore, anyone versed in the study of probability wouldn't have made the fundamental mistake of saying that any of the calculations presented in this thread reflected the "chance" of criticalling in any sort of way.
Nothing you have said (I'm not counting just going "You're wrong", and nothing I have found when searching has suggested to me that I am unable to do so.
From reading your posts it seems to me that you feel that after flipping a coin a billion times, and recording whether it's heads or tails, we are unable to say anything about the coin, or the chance of it showing heads or tails. Which of course begs the question
I get that you have a combinatoric in the binomial distribution, I just can't see how solving for r, as you suggested we do, is at all helpful in determining a value for p.
Unless I'm misunderstanding you, r, which we're solving for, is the number of successes.
P(X=r)=nCr p^r (p-1)^(n-r).
We can measure n and r, we're trying to find p.
It would be far more helpful for you to actually post something up, like a useful link, a worked example, something like that instead of just telling me I'm wrong, and not backing it up.0 -
Well what is the actual crit rate? Is it actually 1 % without the charm? Could it be 1.4?
Is the crit rate calcluated after hit or before?
Binomial distribution, with n and p.
std dev = sqrt (np(1-p))
Maybe this is just a huge plot to get us to sell our 1% arrows.
0 -
Now please join the lost city server so i can slaughter you for your stupidity.OMGLAZERZ - Heavens Tear wrote: »I was curious to see if the crit arrow worked, so I went and tested the theory that it works. I mean, everyone says it does, so it has to, right?
So, I set out to verify what seemed to be common-sense logic : if it says it's giving you +1% crit that it is giving you just that: +1% crit.
Common sense, of course.
So, I started my testing, and then they patched PWI, so I started my testing over again which was done in both tiger form and human form on my barbarian.
Over about 6,760 hits I had exactly 138 crits which comes out to a 2.04% crit rate with the arrow equipped, and a base crit rate of 2%. Having the arrow equipped should have boosted that to a 3% crit rate. I also tested the Christmas Blessing against it and over 1020 hits had 29 crits which comes out to about a 3% crit rate.
As a baseline to compare it to, I unequipped the arrow and did a run of 1025 hits where I ended up getting 22 crits, or about a 2.15% crit rate which was the same with the arrow equipped.
What does this tell us? That the Crit arrow doesn't actually give you +1% crit rate with a non-bow equipped.
IT'S.
AN.
ARROW!!!!
Omg I'm so tired of stupid people... Of course it's not going to give extra crit rate if you're not using a bow! What made you think it would!? Omg. I would take my cleric and tempest your stupid ****!0 -
This is when I come in and say the following:
You all honestly just wasted your time/hours testing this out? Wow.
Heres some stuff I added
Crit arrow = 2,500
Bow = 3,500 (roughly)
Finding out how much time you wasted on something useless = Priceless b:pleased0 -
angellicdeity wrote: »I'd like to point out that the basic assumption is flawed.
The critical percent is not the portion of shots that will critical, its the chance that it will critical.
I thought that was only a factor if an event can happen multiple times in a trial? e.g. mob drops - when you kill a mob it can drop more than one item. In that case, you can't just count the number of a specific item you get vs. the number of mobs you kill. You'd have to factor for the number of drops.
In this case, a single hit can't critical twice. It can either critical or not critical. So the portion of shots that critical gives you the chance it will critical.
Edit: Never mind, I figured out what you are saying. Yes we know that the portion of shots that critical is not the chance that it will critical. The statistical methods that have been given are methods of determining from a sample the likelihood that a hypothesis (e.g. "The chance to critical is 3%") is correct.
Basically you ask, "If the chance to critical is really 3%, what is the odds of me taking 6700 shots and only getting 2%?" If the odds are less than 5%, then you can say with 95% confidence that the chance to critical is not really 3%. Standard stuff. (You can also ask the inverse. "If I take 6700 shots and only get 2% criticals, what are the odds that the chance to critical is really 3%." It'll be subtly different - exactly how is left as an exercise for the reader. LOL)
By the way, I'm curious how this data was collected. Is there a logging feature somewhere that I don't know about?0 -
i am not good at quoting multiple messages, and most of the times end up messing up the format and having to recheck->edit the post to make it more presentable. so i`ll just skip quoting and just do a bulk reply
yea, i did say unless you do 100% magic attack the data the OP gave is utterly flawed. since the OP did not mention what character he used to get the data, im taking a worst case scenario, ie physical attacks.
and no, the chances to crit:hit does NOT balance out. i mean jeez man, think about it for a minute. on one side you got a chance to hit and do more damage After landing a successful hit, on the other you got a chance to get lucky and only land a normal hit and those two are NOT a related possibility.
accorind to Your 'easy example' what if you miss on all the 20 shots that were Supposed to be crits? or what if you miss Only the normal hits and land All critical hits? its all a matter of luck(game mechanics) if your crit rolls when you miss or not. and you cant go and put a formula when it comes down to luck really.
so no matter what theory of co-relation you come up with for accuracy:crit, unless the game designers go back and reconfigure the Entire attack roll system of PWi, you will and i mean Will be wrong about that theory.
price of crit arrow:2k~10k each.
bow(NPC brand):3,500(roughly....right)
endgame bow(molded, TT, warsoul, rank equip, ripoff, etc):way too much
[bound]endgame bow(molded, TT, warsoul, rank equip, ripoff, etc):0
you trying to pull the mastercard(tm) joke : cliche = /
and what about the possibility that you crit AFTER you stopped collecting data? i mean the game logic was gonna send a crit your way, even the nasty game engine who always spawns mobs at you agreed to present you a crit for your efforts, even the mob who was supposed to be next on your hit list came to terms that it was gonna suffer a fatal crit but....you just stopped 'collecting data'.
when your talking bout such a huge difference in crit:normal hit, even one crit will do a substantial difference in determing the final answer.
same goes for what if you crit BEFORE started collecting data?
[when there is a chance of possibility, you can never get a definite answer. there will always be a error of atleast +/- 1point.]
@angry kiddo from lost city server:grow some nuts instead of making a new account Just to post here and flame everyone who responded to the thread. your just another cowardly hypocritic moron(no flaming intended, im just stating facts) who cant use his main account to make a point. putting that aside congrats on your first post count now please feel free to reply back with your main account. 0 -
Eh, you need the hit rate if the figure given is the number of attacks, since we need to know the number of hits, but if we're given the actual number of hits then it doesn't matter, since that's the value we're using.
I am assuming that when he says he gave us the hits he means that his numbers are all net of misses. Since he did the experiment with a barb and may have had a particularly high miss percentage, I am merely raising the point that in the observations we may have correlation between misses and crits. Since I am also willing to assume the game developers used the simplest possible way to imlement probabilities here (which would have the chances of crits and misses independent of each other), we can assume that the truncated sample is giving consistent estimates. However, since the question of real interest is whether the sample size is large enough, we would not make the simplifying assumption that the sample has the same properties as the underlying distribution (consistency is about what is true once you have enough observations, so assuming its implications in determining if we had enough would be circular). Intuitively, what I want to know is if we can rule out to a reasonable certainty the possibility that about 10 extra crits (i.e., 10 more than we would expect given the number of misses) went unobserved because they happened on misses. That is an unlikely possibility if the proportion of misses is low and we can possibly ignore it completely. But, if the chance is fairly high (and 20% would not be out of the question for a barb), then we should have a wider confidence interval than people have been calculating by assuming observational independence. And, the width of the confidence interval relative to the hypotheses we want to test is what tells us if we have enough observations.0 -
Solandri - Heavens Tear wrote: »The data are binary - did it critical or did it not? So there is no mean, sdev, nor median.
Binary distributions can still behave and function like normal distributions.While it's great that you know some words used in statistics, please don't contribute unless you can contribute something more than just random words you heard in your year 9 maths class.
For starters, you are given all the information to calculate the mean, median and standard deviation.
You wish it was only a 9th grade math class. To give you a little background, I took both AP Calc AB and AP Calc BC and got 5's on both, and now I'm taking AP Statistics. So I know what I'm talking about.
And if you don't know what AP tests are, here's a little background information:
AP tests are held in May, after the completion of an AP course taken throughout the high school year. People who take the tests are then given a score of 1-5, 1 being the lowest, and 5 being the highest. These AP tests are similar to college-level courses, and if you pass an AP test (score of 3 or higher) not in the area of your major, you can waive college credits (as long as the college is in the US). Long story short, those 5's I got on the AP tests are the equivalent of an A in a first-year college-level Calculus course, both semesters.
And I forgot to add one thing: conducting only one sample will not necessarily give you the true mean value of the crit rate. Sometimes, you have good days, when you have lots of crits, and sometimes you have bad days, when you have almost no crits. Therefore, you should follow the Law of Large Numbers (aka when sample size increases, sample mean approaches population mean aka true mean) and conduct multiple samples. I would suggest 10 samples of 1000 hits each, for starters.0 -
Psynopsis - Heavens Tear wrote: »You wish it was only a 9th grade math class. To give you a little background, I took both AP Calc AB and AP Calc BC and got 5's on both, and now I'm taking AP Statistics. So I know what I'm talking about.
Look, I certainly don't want to discourage you. Someone who's able to get 5's in those AP courses has a bright future ahead. So don't take our criticism too harshly. But you still have a lot to learn about statistics. Take it from someone who's had grad school courses in it at MIT.
It's a very fun and very applicable topic. But a lot of the stuff you learn early on is simplified or limited in scope to make the math easier to deal with.
Normal (gaussian) distributions are just the tip of the iceberg. There are lots of other probability distribution functions. When you learn about erf, then the universe will seem on the verge of making complete sense. Until it all comes crashing down when you learn that erf can't be solved (so far), only approximated or solved numerically with a computer.
Since the crit/no crit values are binary, the relevant probability distribution in this case is a binomial distribution, which is what jemima used.And I forgot to add one thing: conducting only one sample will not necessarily give you the true mean value of the crit rate. Sometimes, you have good days, when you have lots of crits, and sometimes you have bad days, when you have almost no crits. Therefore, you should follow the Law of Large Numbers (aka when sample size increases, sample mean approaches population mean aka true mean) and conduct multiple samples. I would suggest 10 samples of 1000 hits each, for starters.
That was the whole point of everything Jemima and I posted. The numbers we gave were calculations of how likely it was that the observed critical rate corresponded to the actual critical rate. Since the calculated margin of error was less than half a percent for the 6700 sample case, we deemed it reliable (for a 95% confidence). The ~1000 sample cases are very close to reliable. You should learn how to calculate these in your college-level statistics courses (the grad-level courses will teach you how to derive the formulae used to calculate them).
There's no shame in not knowing this stuff yet. We all had to learn it at some point (well, those of us crazy enough to take it). Like I said, you have a bright future ahead. Just understand that there's a lot more out there than what you've learned so far.0 -
I'm not here to rain on anyone's parade, but a 3% crit rate is still supposed to be a 3% crit rate. Period.
On my venomancer, a mage veno at that, with a 3% crit rate with the Christmas Blessing and a base 2% crit chance, she did the following
Crits : 263
Non-Crits 8568
Total 8831
Act Crit % 0.0298
Ext Crit % 0.03
Expected 264
Diff 1
Spells don't miss.
With the crit arrow equipped and the Christmas Blessing equipped there should be an expected 4% crit rate showing, and so far, there isn't. I'm still working on it, but there are the numbers so far, and the crit arrow isn't doing anything. You can tell because the crit rate isn't going above 3% at all when it should be closer to 4%, not down between 2.5% and 3%.
Crits :: 87
Non-Crits 3198
Total 3285
Act Crit % 0.0265
Ext Crit % 0.04
Expected 131
Diff 44
You may argue statistics, probability and other mathematics all you want, but this isn't math here, this is hard numbers.[SIGPIC][/SIGPIC]0 -
Solandri - Heavens Tear wrote: »By the way, I'm curious how this data was collected. Is there a logging feature somewhere that I don't know about?
The client tells you whether your damage was a critical or not, and it does log the damage done in a chat window if you have it set to do so via the game settings. Beyond that I simply logged the data itself by hand, put it into a spreadsheet, and came up with the results doing simple sums and averages.
To be honest, I don't know a lot about statistics or probability outside of being able to estimate how many times a certain event will happen in x trials, but I'm not doing probability or statistical estimation. I'm simply collecting the data and analyzing it, and taking a determination that what is expected to happen simply isn't happening.
If you do not believe me, don't agree with my numbers, or even think I have a clue about anything (I wouldn't blame you in the least), then I challenge you to verify with your own tests whether or not my theory that the crit arrow doesn't do anything without a bow equipped yourself.
That being said, it's my advice to everyone to stop buying crit arrows as it's a waste of coin.[SIGPIC][/SIGPIC]0 -
can i buy your x-mas blessing? pm me in game b:pleased[SIGPIC][/SIGPIC]
6X Sin on RT0 -
LOL, I finally found this tread.
I read it entirely but still I do not think that we can do a very acurate experiment to determine 1% difference.
It's just to small.
I wonder if any of the people arguing above have access to the game formulas to determine how the game determine when to critical or how the game determines a random number?
The formula could be made in a way to make critical hits happen more often under certain conditions (mobs, days, hours, level,weapons,quests, grinding etc.) and less in other and still keep the percentage legit.0 -
Kitsuneaki - Sanctuary wrote: »can i buy your x-mas blessing? pm me in game b:pleased
Um, no, my veno needs it, and I play on Heaven's Tear, but the answer is still no considering the eventuality is she is going to have a herc.Granrey - Sanctuary wrote: »LOL, I finally found this tread.
I read it entirely but still I do not think that we can do a very acurate experiment to determine 1% difference.
It's just to small.
I wonder if any of the people arguing above have access to the game formulas to determine how the game determine when to critical or how the game determines a random number?
The formula could be made in a way to make critical hits happen more often under certain conditions (mobs, days, hours, level,weapons,quests, grinding etc.) and less in other and still keep the percentage legit.
No, it's not like accuracy, evasion or armor which are numbers and then a percentage is generated based on the mob. It's a set percentage that doesn't change unless you equip something that increases it, or gain enough dex to increase it.
And yes, you can determine a difference between percentages given large enough samples.
In fact, I challenge you to prove that you cannot see a difference between a 2% crit rate and a 3% crit rate over, say, 10000 data points.[SIGPIC][/SIGPIC]0 -
OMGLAZERZ - Heavens Tear wrote: »It's a set percentage that doesn't change unless you equip something that increases it, or gain enough dex to increase it.
.
Sorry, how do we know that? Do you have some info to validate this?0
This discussion has been closed.
Categories
- All Categories
- 181.9K PWI
- 699 Official Announcements
- 2 Rules of Conduct
- 264 Cabbage Patch Notes
- 61K General Discussion
- 1.5K Quality Corner
- 11.1K Suggestion Box
- 77.4K Archosaur City
- 3.5K Cash Shop Huddle
- 14.3K Server Symposium
- 18.1K Dungeons & Tactics
- 2K The Crafting Nook
- 4.9K Guild Banter
- 6.6K The Trading Post
- 28K Class Discussion
- 1.9K Arigora Colosseum
- 78 TW & Cross Server Battles
- 337 Nation Wars
- 8.2K Off-Topic Discussion
- 3.7K The Fanatics Forum
- 207 Screenshots and Videos
- 22.8K Support Desk