I don't call it bad because it's luck-based. I call it bad because what it contributes is insignificant compared to all the other options and doesn't really excel at anything except niche crit-based builds that still fall short of the alternatives.
But if the Crit can happen often enough for a particular person and overcome the 16 average damage per rest of the battle master. Then relying on that luck becomes worth it. A Crit based build driven by luck in the hands of some one like Tallesin who has gotten an inordinate number of critical successes is going to seem a lot more viable than it would if say... Will Wheaton were trying to play that same build.
No one has ever beaten the Law of Large Numbers; there's a reason casinos and lotteries are still in business. Play long enough and whatever luck you imagine you have will regress back to the mean.
No. Casino's and Lotteries are still in business because the games are at least partially rigged. It's not just the average mean. They know for a fact that some people will get away with more money than they lose. They are set up so that more people lose more money than they win and they let the "lucky ones" win and even win big because they are the minority group.
It's a matter of acceptable losses that also work as advertising to draw others in with the hopes of being those lucky few despite the fact that the odds are actually stacked against the average person. Usually by an average of about 20% when it comes to the Casino business though the more unscrupulous Casino's it's even higher. I don't remember what it is for Lotteries. But this is the reality. This is why there is a long standing saying of "the House Always Wins" Because it's not even odds to start with.
If by 'rigged' you mean 'the odds are never 50/50' then you'd be right. Roulette is the closest game I've found and the odds are STILL tilted in favor of the House by the green 00. You can bet Red/Black (as close to 50/50 as it gets) for a solid year and unless you hit a big win and walk away, the House still takes all your money. They do that because for every player who CAN walk away with a good stack of chips, there will be more than one who will get caught up in the excitement and roll the dice, or spin the wheel, or play one more hand.
Unless you are telekinetic the dice rolls will always even out to the rules of probability. If your building because you are “lucky” your lying to yourself. Can crazy lucky runs occur? Sure I saw a DM roll 10 crit fails in a row once (and it saved my character) but that was a run not a long term average. This is one of the hardest things for folks to accept. You crit 1 in 20 rolls on average is there any build that gives you 20x the damage of a normal roll? What do you actually get? 2x the damage that is all, yes if you get to use extra dice you get a larger hit but your still only getting 2x the damage for 1/20 chance. Oh you have an expanded crit range? Is it expanded enough that half your rolls are crits? If not focusing on crit damage is a waste of time and effort. Sadly math doesn’t lie but our perceptions often do.
I don't call it bad because it's luck-based. I call it bad because what it contributes is insignificant compared to all the other options and doesn't really excel at anything except niche crit-based builds that still fall short of the alternatives.
But if the Crit can happen often enough for a particular person and overcome the 16 average damage per rest of the battle master. Then relying on that luck becomes worth it. A Crit based build driven by luck in the hands of some one like Tallesin who has gotten an inordinate number of critical successes is going to seem a lot more viable than it would if say... Will Wheaton were trying to play that same build.
No one has ever beaten the Law of Large Numbers; there's a reason casinos and lotteries are still in business. Play long enough and whatever luck you imagine you have will regress back to the mean.
No. Casino's and Lotteries are still in business because the games are at least partially rigged. It's not just the average mean. They know for a fact that some people will get away with more money than they lose. They are set up so that more people lose more money than they win and they let the "lucky ones" win and even win big because they are the minority group.
It's a matter of acceptable losses that also work as advertising to draw others in with the hopes of being those lucky few despite the fact that the odds are actually stacked against the average person. Usually by an average of about 20% when it comes to the Casino business though the more unscrupulous Casino's it's even higher. I don't remember what it is for Lotteries. But this is the reality. This is why there is a long standing saying of "the House Always Wins" Because it's not even odds to start with.
If by 'rigged' you mean 'the odds are never 50/50' then you'd be right. Roulette is the closest game I've found and the odds are STILL tilted in favor of the House by the green 00. You can bet Red/Black (as close to 50/50 as it gets) for a solid year and unless you hit a big win and walk away, the House still takes all your money. They do that because for every player who CAN walk away with a good stack of chips, there will be more than one who will get caught up in the excitement and roll the dice, or spin the wheel, or play one more hand.
The probabilities of Roulette are actually much more complicated than that. And nothing in it is 50/50. It does a good job into faking people into thinking it is 50/50 but it's not Roulette will take your money because of these shifted odds that aren't quite what they appear.
But your best odds are actually in the card games. Even though they are done with multiple decks in casino's the number and type of cards are always the same and always equal out overall. The issue is that most people can't keep track of that many cards in their head and some places do their best not to advertise the number of decks that they are using exactly. though most of them tend to use at least 6 and prefer at least 8. The Skill Factor is what tends to set your ability to win or lose over the luck factor in those games. This is the reason for the fact that Card Counting while not outright banned in most places, will still get the casino to exercise their right to ask you to leave. it shifts the skill curve when your capable of doing it accurately.
Unless you are telekinetic the dice rolls will always even out to the rules of probability. If your building because you are “lucky” your lying to yourself. Can crazy lucky runs occur? Sure I saw a DM roll 10 crit fails in a row once (and it saved my character) but that was a run not a long term average. This is one of the hardest things for folks to accept. You crit 1 in 20 rolls on average is there any build that gives you 20x the damage of a normal roll? What do you actually get? 2x the damage that is all, yes if you get to use extra dice you get a larger hit but your still only getting 2x the damage for 1/20 chance. Oh you have an expanded crit range? Is it expanded enough that half your rolls are crits? If not focusing on crit damage is a waste of time and effort. Sadly math doesn’t lie but our perceptions often do.
My Rolls haven't averaged out in years... My dice rolls pretty much always add up on the low side. I get far more ones than I get 20's... and I usually get far more low rolls than I get high rolls. This is why in 3.x/PF i either stack modifiers to the point that rolling almost doesn't matter... And then on top of it i do things like take increased crit ranges so I can get more out of the few good rolls that I have while making my low rolls basically guaranteed if I can. It's also why I tend to either point buy or Array over rolled stats.
yes. Statistically over many many rolls it should average out. Maybe I was much luckier in my younger years and i just don't remember it. But that is not how things work out for me now.
Also. There's a thing about statistical math. It explains the majority. There are usually outliers when it comes to statistics. Anomolies that just don't fit what the math says that everything should fit by. It accepts that somewhere there is likely an exception to things.
Keep in mind. This does not mean that I don't occasionally pop off in play. Even at some key moments where the party really needed my character to do so. But that on session where I do that is usually offset by the number of sessions where I don't even come close.
It's actually one of the reasons that I play 5e. The thresh-holds are so much lower and they don't move nearly as much as in those editions.
Fateless, I feel your pain; but I challenge you to actually record all (or at least your next 500-1000) d20 & d100 rolls and see if reality and your perceptions are actually in line with each other. I expect you will be surprised. The likelihood of the 10 crits in a row I saw is insanely low - but it has a likelihood and such things do happen and are allowed for in probability. With the exception of the increased crit range abilities everything you are talking about doing actually improves general damage probably more than they have impact on crits. That was my my point about crit damage vs crit likelihood. Your go play roulette and pick a number you have a 1 in 38 chance of it coming up, if it does you get paid 35x your bet. Not quite an even return but close. In D&D your typical crit chance is 1 in 20 but your return is not 20x normal damage it’s only 2x, even if you have all possible increases in crit range that only brings the chance down to roughly 1 in 7 but your actual return is still only 2x not 7x. The return is not worth the costs of everything you put into it. This what was meant about crit fishing as a strategy being bad - focus on getting the most out of your regular hits and you will get the most out of your crits for free😁
"The probabilities of Roulette are actually much more complicated than that. And nothing in it is 50/50. It does a good job of faking people into thinking it is 50/50 but it's not Roulette will take your money because of these shifted odds that aren't quite what they appear."
As I said, it's as close as 50/50 as you're going to get.
36 numbers...18 red and 18 black. You could call that 50/50. However, there are also two green numbers...0 and 00. So the odds of hit a specific number of a specific color is about 2.3 something percent. Multiply that by 18 and you get 47.3 something percent. So a red/black bet has a 47.3% chance of hitting. Sure, you can drop a twenty on red, have it hit, and make double your money. You can even cover BOTH colors and win if either one hits...but that's dumb because not only do you automatically LOSE on the other bet, if either of the green numbers comes up you lose it all.
The House NEVER has 50/50 odds...ever. If they did everyone would flock to that game and over the course of the year, they would likely break even. Breaking even doesn't pay the rent.
Fateless, I feel your pain; but I challenge you to actually record all (or at least your next 500-1000) d20 & d100 rolls and see if reality and your perceptions are actually in line with each other. I expect you will be surprised. The likelihood of the 10 crits in a row I saw is insanely low - but it has a likelihood and such things do happen and are allowed for in probability. With the exception of the increased crit range abilities everything you are talking about doing actually improves general damage probably more than they have impact on crits. That was my my point about crit damage vs crit likelihood. Your go play roulette and pick a number you have a 1 in 38 chance of it coming up, if it does you get paid 35x your bet. Not quite an even return but close. In D&D your typical crit chance is 1 in 20 but your return is not 20x normal damage it’s only 2x, even if you have all possible increases in crit range that only brings the chance down to roughly 1 in 7 but your actual return is still only 2x not 7x. The return is not worth the costs of everything you put into it. This what was meant about crit fishing as a strategy being bad - focus on getting the most out of your regular hits and you will get the most out of your crits for free😁
i actually did this a few years ago. But you have to remember. the statistical average is over all rolls being done by everybody. not just me. My Rolls averaged out to about an 8.7 roughly. If Somebody is going to roll 4 ones in a game... Or even 4 runs in a row in my groups it's usually me. I'm also the one that they can reasonable expect to fail a DC 25 roll with a 23 modifier in my group. Something I've done several time though statistically that should be a low possibility of happening as well. But I probably should record them again just to see if it's changed any.
Also In PF and 3.x Where I take the various crit modifiers. The Maximum crit range is about 1 in 4. Because you can set things up so that you crit on anything that is a 15 and above. I still don't crit that often. Even though that realistically should put my crit chances at about 30%. There is nothing like that in 5e and I relly wish that there was. I could really use it.
Interestingly if you want to see people shaping up on rolls. For people that really have made 100's if not thousands of recorded rolls. We have Critical Role.
The wildemount one in particular is interesting because it's the whole campaign and over 15k rolls in total. There are some unknowns but the known rolls are still in the thousands. and since they tend to pretty well mention their nat 1's and their nat 20's. Even if they are the unused die on advantage or disadvantage. It's not a complete picture but it is a good basis to study statistically.
Nott and Yasha in that campaign were the closest to parity between their 1's and their 20's. Tallesin is the one where they are the most unbalanced by their stats.
Those rankings are interesting, on several accounts, but looking at the DM and the top 3-4 players I start thinking about subconscious telekinesis! I started to consider that there was something lopsided about the dice but if you look below the top few and DM the rolls are pretty close to even on 20’s and 1’s. If everyone was really lopsided in ratio I would think about the dice being off. While it would be a pain in the tail to do it might be interesting to have everyone at the table track their 20’s and 1’s and compare after a few sessions to see if you are off kilter and if anyone else is too. I did a study years ago on astrology like that and found that if they didn’t know at the start what the horoscope said for the day the matches they got in the evening were just random chance - pretty much in the middle of the chance range. If they read the horoscopes in the morning the results showed a different pattern: a) those that were neutral in belief were still down the middle of the chance range b) those that denied astrology were skewed to the negative but still in the chance range c) those that believed were also skewed but to the positive side but still in the chance range
seems our beliefs in something can have an effect and I’ve seen it show up in dice rolls as well. way back in college we were having a fair and my dorm put together a craps table, we made a ton of money with me rolling the dice as I seemed to be able to alter the chance of a number by concentrating on it to get a streak going then when folks started following my bets switching the number in my mind so sthey would lose. Yes I was a nasty guy back then but it was for a good cause 😳😁.
some folks do seem to have a natural subconscious telekinetic ability but it’s very hard to detect without a lot of rolls and careful analysis so you could be accurate in your belief but that also means you should be able to change those results with work.
A former DM who works in computer programming said there is a difference between truly random rolls of the dice and what you get from online gaming. That number seeds like pc gaming uses, are never truly random and you will tend to get more Nat 20s and Nat 1s, than you would get if you were at a gaming table with dice. This always felt intuitively true, although never tried to dig more into it. It seems like my party does have a higher rate of both extremes, than we should be seeing rolling a d20.
A former DM who works in computer programming said there is a difference between truly random rolls of the dice and what you get from online gaming. That number seeds like pc gaming uses, are never truly random and you will tend to get more Nat 20s and Nat 1s, than you would get if you were at a gaming table with dice. This always felt intuitively true, although never tried to dig more into it. It seems like my party does have a higher rate of both extremes, than we should be seeing rolling a d20.
This would not surprise me at all. I don't know about 1s and 20s but all of my friends who code say that there is no such thing as a truly random computer-generated number which is why many programs include something called a 'streak-breaker' where if you hit or miss too many times in a row the breaker kicks in to flip the result. It helps prevent the RNG from getting stuck in a rut.
This is true, computers use what is known as A “semi-random” number generator. However the examples I was presenting (the 10 crit fail sequence, the dice table in college) were rolled with regular dice. I don’t know what the story is with the critical roll data, never watched and not interested enough to go look. What was most interesting to me was the ratio of 20s to 1’s being close to 2:1. Even with a semi random generator I wouldn’t expect that, more likely that the occurrence of the is more like 2/3 out of 20 for each end than that one end is twice as likely as the other.
i actually did this a few years ago. But you have to remember. the statistical average is over all rolls being done by everybody. not just me.
While you can include everyone at the table for the statistics it is just as legitimate to only look at a subset (your rolls) but you will want to record as many total rolls to get the same statistical validity of the results. Since it is your rolls in particular you are interested in that is the way to do it.
This is true, computers use what is known as A “semi-random” number generator. However the examples I was presenting (the 10 crit fail sequence, the dice table in college) were rolled with regular dice. I don’t know what the story is with the critical roll data, never watched and not interested enough to go look. What was most interesting to me was the ratio of 20s to 1’s being close to 2:1. Even with a semi random generator I wouldn’t expect that, more likely that the occurrence of the is more like 2/3 out of 20 for each end than that one end is twice as likely as the other.
I noticed that too, with more 20s than 1s. But it's also such a small sample size, that things like that can happen. I have told my friends, the larger the sample size, the closer to the average the numbers should be. With only 100 or 200 rolls, that's a small sample size. Betting those percentages balance out if we were looking at 100,000 rolls.
that is quite likely but keep in mind that the roughly 300 critical rolls (20's + 1's) actually represents around 6000 total rolls for each character and by then the results should be closer to statistical averages. what I didn't see was the total numbers of rolls (it may be there and I missed it) to see if the numbers of 20's and 1's are running about 5% each or are they higher or lower.
Just went back and looked at the critical roll stats again. It’s difficult to tell for sure because the stats are by character not player but most are either too small a sample size to say much about or actually not too far off 50-50 when you look at 20’s/total or 1’s/total. You can’t combine anything except the DM stats (I’m assuming it’s the same DM all the time let me know if I’m wrong) which if combined show that over 75% of crits from the DM are 20’s - that’s one nasty DM to go up against. (That is after 465 crits representing around 4600 total D20 rolls, you would not expect it to be that far off the mean with that sample size.
Just went back and looked at the critical roll stats again. It’s difficult to tell for sure because the stats are by character not player but most are either too small a sample size to say much about or actually not too far off 50-50 when you look at 20’s/total or 1’s/total. You can’t combine anything except the DM stats (I’m assuming it’s the same DM all the time let me know if I’m wrong) which if combined show that over 75% of crits from the DM are 20’s - that’s one nasty DM to go up against. (That is after 465 crits representing around 4600 total D20 rolls, you would not expect it to be that far off the mean with that sample size.
They should be relatively close to the 5% number. But clearly, with the smaller sample size, some people have been more lucky, than unlucky. ;)
This right here is the document recording all of the rolls made for Campaign 2 of Critical Role.
The Primary cast members
Are the first 8 spots On the previous link that I linked for the campaign. with 7 and 8 being the same player (Caduceus was made after Molly died). If you don't know which player is which I won't necessarily spoil that because it let's you look at it in more of a blind fashion to not actually know by player name.
Every other Character beyond that is usually some kind of guest character or notable npc's. this is why their rolls are low.
But if your wanting to understand sample sizes for the top 7 people that were the main cast in most episodes. in alphabetical order because it's easiest.
Beau did 3075 rolls in total Caduceus Rolled 1325 rolls and Molly 447 for a total of 1772 Caleb did 1736 rolls, Fjord did 2127 Jester did 2055 Nott/Veth did 1796 as Nott and 587 as Veth for a total of 2383 rolls. Yasha did 1610 rolls.
Matt The DM's rolls sadly aren't in the sheet for the most part. Except for certain notable NPC's.
It should be noted that these do include rolls other than d20's because there are damage rolls mixed in as well. Though even if we go overly conservative and say that only one half of their rolls were d20's despite the fact that Damage is the only place smaller dice are used with any consistancy.
This means that Beau rolled 1532 D20 rolls as conservative estimate. This means that 9% of her rolls were natural 20's and 6.9% of her rolls were ones. This could be from Electronic Bias. I really don't remember. But it does put The extreme's within 2% of each other basically and allowing for error of incomplete numbers this does suggest that they are in parity for this specific person yes. and them both being on the high side.
Noth/Veth, and Yasha are also in very obvious parity by the numbers that we are given. Since they are almost dead on in our sample size on both sides.
Vex at first looks like she might be a little below the average mark with more 1's than 20's. But then if you look at the other Campaign she is slightly to the 20's by a not to disimilar margin so she's clearly in parity herself. With a rather large sample size between the two campaigns even if we go conservative on it to try and account for just d20's.
So let's look at one that seems to be much more out of whack. That of Caduceus combined with Molly.
Being conservative we are considering 886 D20 rolls for our sample size specific to that one person. With a Recording of 68 20's and and 40 1's. this is 7.6% of 20's and 4.5% of the rolls were 1's... This is a 3.1% difference in our sample size. This could certainly be that the sample size just isn't big enough.
So let's grab more dice rolls for this same person. In the Previous campaign he played Percy which gives us another 2082 total Rolls. Going overly conservative we're looking a 1041 rolls with 104 20's and 49 1's. Automatically from the numbers you can tell that the 20's were about 10% of the rolls and 4.7% of the 1's. He's skewing High and all he rolled in this campaign were actual dice.
This could be blamed on something called the Golden Snitch in campaign 1. So it's possible that the die has a weighting issue and it was one he rolled often. However he gave up that die to Matt the DM at the end of the Campaign so it was not available for Campaign 2. So while it's a bit more acceptable this particular individual is still skewing high on the rolls. Blended together we are looking at conservative estimates of 1972 Conservatively acceptable D20 rolls with 172 20's rolled and 89 1's rolled. This gives us an 8.7% on 20's rolled and a 4.6% on the 1's, which means were still looking at a difference of 4.1%. Unless we completely fine tooth the data like analysts however this is a pretty significant difference and we can't know if going into such detail would bring parity or not. If somebody wants to try I'm all for hearing the result. But just going by what we have. it's something rather notable.
Now it's possible that he's just going through a long (about 6 years) streak of skewing high and overall his numbers will even out. But at this point. He's a bit above the average on his rolls. By about a full number.
Interestingly another person at the table also Skew's Slightly High. But it seems to be at more acceptable levels. So let's look at him a little closer as well. In Campaign 2 going quite conservative again that would be 1063 Rolls. 99 being 20's and 63 being 1's. This gives us 9.3% as 20's and 5.9% as 1's. It again could be leaning to extreme's of 1's and 20's because of digital rollers. We'd have to go through days of footage and check. But I'm not going to argue that case. But he's still got a 3.4% difference between his high extreme and his low extreme. it's also higher than what Cad/Molly/Percy's player has for Campaign 2.
So how about we Expand his sample size as well. The Character he played in Campaign 1 is named Grog. There are a total of 2449 rolls for Grog recorded in Campaign 1, so our conservative number is 1224. He's recorded as having 86 20's and 58 1's. This gives us 7% and 4.7% respectively. And we know this game was played with Actual Dice so it's not skewed by electronic shenanigans. He's still got a 3.3% difference leaning towards the high side of the dice.
Even when we combine the numbers from the 2 sample sizes. We're looking at the conservative sample size of 2287, with 185 recorded 20's and 121 recorded 1's. Giving us 8% for the 20's to be rolled and 5.2% on the 1's being rolled over all of this. Granted some of these may be electronic and so extreme's may be slightly higher. But we're still looking at a 2.8% difference in favor of the high rolls. Overall he doesn't skew quite as high and again in the long run over a lot more rolls. He may well even out. But with our current data set. he's on a pretty good multi-year streak of playing weekly of getting a bit better than average. And he certainly isn't as skewed as the Cad/Molly/Percy player is over both campaigns.
And Again. i'm aware we aren't working with complete data. And some of my numbers are just conservative estimates that can adjust some things. The Second Campaign may even be suffereing a bit from psuedo random bias to Explain in that campaign why w'ere seeing numbers greater than 5% on both ends and the overly conservative estimate likely contributed to them being higher as well. And I totally accept that. But Not all of it is Perceptual Bias. For at least some of this there has to be some other factor in play. Whether it's particular long runs of slightly skewed luck. Or just the fact that certain of us are running slightly below average and others are running slightly above. it's hard to say. Because we have two overlapping situations. The Average of an Individual and the true objective average of the numbers that actually applies to all dice. Ideally math would like both to end up in the same place..
But it does stand to reason that if you can identify that your on one of these long higher swings that can span campaigns, And perhaps make the bias towards more extreme numbers like 1's and 20's in your favor. You can actually base builds around taking advantage of that. Though at least one of those things is quite difficult to do until you've basically already gotten through it.
(As a little side note that might interest some. Beau made so many rolls during the campaign that are so much higher than others, except perhaps for the DM Matt because she played a Monk. Most monks naturally make a lot of attack rolls per turn and then from that a lot of damage rolls. Grog in the First Campaign worked under the same principle but in a different way. he was a Berserker Barbarian. So he rolled quite regularly with advantage and he made decent amount of use of the berserk frenzy's bonus action attacks at key points in things.)
It looks like Beau was making a LOT of rolls compared to others. Was this because of the Monk Flurry or Blows and Stunning Strike (which can turn a single roll into several)?
For bad luck in rolls, Will Wheaton, on Critical Role, has to "take the cake!"
Out of 54 total rolls with a d20 (a very, very small sample size), he rolled a 1 or 2, 17 times! He rolled a 4 or lower, 25 out of 54 times! He rolled a 6 or lower, 31 times! He rolled a 9 or less, on 39 out of 54 rolls! That is mind bogglingly bad luck! Wow!
For bad luck in rolls, Will Wheaton, on Critical Role, has to "take the cake!"
Out of 54 total rolls with a d20 (a very, very small sample size), he rolled a 1 or 2, 17 times! He rolled a 4 or lower, 25 out of 54 times! He rolled a 6 or lower, 31 times! He rolled a 9 or less, on 39 out of 54 rolls! That is mind bogglingly bad luck! Wow!
If by 'rigged' you mean 'the odds are never 50/50' then you'd be right. Roulette is the closest game I've found and the odds are STILL tilted in favor of the House by the green 00. You can bet Red/Black (as close to 50/50 as it gets) for a solid year and unless you hit a big win and walk away, the House still takes all your money. They do that because for every player who CAN walk away with a good stack of chips, there will be more than one who will get caught up in the excitement and roll the dice, or spin the wheel, or play one more hand.
Unless you are telekinetic the dice rolls will always even out to the rules of probability. If your building because you are “lucky” your lying to yourself. Can crazy lucky runs occur? Sure I saw a DM roll 10 crit fails in a row once (and it saved my character) but that was a run not a long term average. This is one of the hardest things for folks to accept. You crit 1 in 20 rolls on average is there any build that gives you 20x the damage of a normal roll? What do you actually get? 2x the damage that is all, yes if you get to use extra dice you get a larger hit but your still only getting 2x the damage for 1/20 chance. Oh you have an expanded crit range? Is it expanded enough that half your rolls are crits? If not focusing on crit damage is a waste of time and effort. Sadly math doesn’t lie but our perceptions often do.
Wisea$$ DM and Player since 1979.
The probabilities of Roulette are actually much more complicated than that. And nothing in it is 50/50. It does a good job into faking people into thinking it is 50/50 but it's not Roulette will take your money because of these shifted odds that aren't quite what they appear.
But your best odds are actually in the card games. Even though they are done with multiple decks in casino's the number and type of cards are always the same and always equal out overall. The issue is that most people can't keep track of that many cards in their head and some places do their best not to advertise the number of decks that they are using exactly. though most of them tend to use at least 6 and prefer at least 8. The Skill Factor is what tends to set your ability to win or lose over the luck factor in those games. This is the reason for the fact that Card Counting while not outright banned in most places, will still get the casino to exercise their right to ask you to leave. it shifts the skill curve when your capable of doing it accurately.
My Rolls haven't averaged out in years... My dice rolls pretty much always add up on the low side. I get far more ones than I get 20's... and I usually get far more low rolls than I get high rolls. This is why in 3.x/PF i either stack modifiers to the point that rolling almost doesn't matter... And then on top of it i do things like take increased crit ranges so I can get more out of the few good rolls that I have while making my low rolls basically guaranteed if I can. It's also why I tend to either point buy or Array over rolled stats.
yes. Statistically over many many rolls it should average out. Maybe I was much luckier in my younger years and i just don't remember it. But that is not how things work out for me now.
Also. There's a thing about statistical math. It explains the majority. There are usually outliers when it comes to statistics. Anomolies that just don't fit what the math says that everything should fit by. It accepts that somewhere there is likely an exception to things.
Keep in mind. This does not mean that I don't occasionally pop off in play. Even at some key moments where the party really needed my character to do so. But that on session where I do that is usually offset by the number of sessions where I don't even come close.
It's actually one of the reasons that I play 5e. The thresh-holds are so much lower and they don't move nearly as much as in those editions.
Fateless, I feel your pain; but I challenge you to actually record all (or at least your next 500-1000) d20 & d100 rolls and see if reality and your perceptions are actually in line with each other. I expect you will be surprised. The likelihood of the 10 crits in a row I saw is insanely low - but it has a likelihood and such things do happen and are allowed for in probability. With the exception of the increased crit range abilities everything you are talking about doing actually improves general damage probably more than they have impact on crits. That was my my point about crit damage vs crit likelihood. Your go play roulette and pick a number you have a 1 in 38 chance of it coming up, if it does you get paid 35x your bet. Not quite an even return but close. In D&D your typical crit chance is 1 in 20 but your return is not 20x normal damage it’s only 2x, even if you have all possible increases in crit range that only brings the chance down to roughly 1 in 7 but your actual return is still only 2x not 7x. The return is not worth the costs of everything you put into it. This what was meant about crit fishing as a strategy being bad - focus on getting the most out of your regular hits and you will get the most out of your crits for free😁
Wisea$$ DM and Player since 1979.
"The probabilities of Roulette are actually much more complicated than that. And nothing in it is 50/50. It does a good job of faking people into thinking it is 50/50 but it's not Roulette will take your money because of these shifted odds that aren't quite what they appear."
As I said, it's as close as 50/50 as you're going to get.
36 numbers...18 red and 18 black. You could call that 50/50. However, there are also two green numbers...0 and 00. So the odds of hit a specific number of a specific color is about 2.3 something percent. Multiply that by 18 and you get 47.3 something percent. So a red/black bet has a 47.3% chance of hitting. Sure, you can drop a twenty on red, have it hit, and make double your money. You can even cover BOTH colors and win if either one hits...but that's dumb because not only do you automatically LOSE on the other bet, if either of the green numbers comes up you lose it all.
The House NEVER has 50/50 odds...ever. If they did everyone would flock to that game and over the course of the year, they would likely break even. Breaking even doesn't pay the rent.
i actually did this a few years ago. But you have to remember. the statistical average is over all rolls being done by everybody. not just me. My Rolls averaged out to about an 8.7 roughly. If Somebody is going to roll 4 ones in a game... Or even 4 runs in a row in my groups it's usually me. I'm also the one that they can reasonable expect to fail a DC 25 roll with a 23 modifier in my group. Something I've done several time though statistically that should be a low possibility of happening as well. But I probably should record them again just to see if it's changed any.
Also In PF and 3.x Where I take the various crit modifiers. The Maximum crit range is about 1 in 4. Because you can set things up so that you crit on anything that is a 15 and above. I still don't crit that often. Even though that realistically should put my crit chances at about 30%. There is nothing like that in 5e and I relly wish that there was. I could really use it.
Interestingly if you want to see people shaping up on rolls. For people that really have made 100's if not thousands of recorded rolls. We have Critical Role.
Rankings - Vox Machina — CritRoleStats
Wildemount Character Rankings — CritRoleStats
The wildemount one in particular is interesting because it's the whole campaign and over 15k rolls in total. There are some unknowns but the known rolls are still in the thousands. and since they tend to pretty well mention their nat 1's and their nat 20's. Even if they are the unused die on advantage or disadvantage. It's not a complete picture but it is a good basis to study statistically.
Nott and Yasha in that campaign were the closest to parity between their 1's and their 20's. Tallesin is the one where they are the most unbalanced by their stats.
But it shows something. People talk about
Those rankings are interesting, on several accounts, but looking at the DM and the top 3-4 players I start thinking about subconscious telekinesis! I started to consider that there was something lopsided about the dice but if you look below the top few and DM the rolls are pretty close to even on 20’s and 1’s. If everyone was really lopsided in ratio I would think about the dice being off. While it would be a pain in the tail to do it might be interesting to have everyone at the table track their 20’s and 1’s and compare after a few sessions to see if you are off kilter and if anyone else is too. I did a study years ago on astrology like that and found that if they didn’t know at the start what the horoscope said for the day the matches they got in the evening were just random chance - pretty much in the middle of the chance range. If they read the horoscopes in the morning the results showed a different pattern:
a) those that were neutral in belief were still down the middle of the chance range
b) those that denied astrology were skewed to the negative but still in the chance range
c) those that believed were also skewed but to the positive side but still in the chance range
seems our beliefs in something can have an effect and I’ve seen it show up in dice rolls as well.
way back in college we were having a fair and my dorm put together a craps table, we made a ton of money with me rolling the dice as I seemed to be able to alter the chance of a number by concentrating on it to get a streak going then when folks started following my bets switching the number in my mind so sthey would lose. Yes I was a nasty guy back then but it was for a good cause 😳😁.
some folks do seem to have a natural subconscious telekinetic ability but it’s very hard to detect without a lot of rolls and careful analysis so you could be accurate in your belief but that also means you should be able to change those results with work.
Wisea$$ DM and Player since 1979.
A former DM who works in computer programming said there is a difference between truly random rolls of the dice and what you get from online gaming. That number seeds like pc gaming uses, are never truly random and you will tend to get more Nat 20s and Nat 1s, than you would get if you were at a gaming table with dice. This always felt intuitively true, although never tried to dig more into it. It seems like my party does have a higher rate of both extremes, than we should be seeing rolling a d20.
This would not surprise me at all. I don't know about 1s and 20s but all of my friends who code say that there is no such thing as a truly random computer-generated number which is why many programs include something called a 'streak-breaker' where if you hit or miss too many times in a row the breaker kicks in to flip the result. It helps prevent the RNG from getting stuck in a rut.
This is true, computers use what is known as A “semi-random” number generator. However the examples I was presenting (the 10 crit fail sequence, the dice table in college) were rolled with regular dice. I don’t know what the story is with the critical roll data, never watched and not interested enough to go look. What was most interesting to me was the ratio of 20s to 1’s being close to 2:1. Even with a semi random generator I wouldn’t expect that, more likely that the occurrence of the is more like 2/3 out of 20 for each end than that one end is twice as likely as the other.
Wisea$$ DM and Player since 1979.
While you can include everyone at the table for the statistics it is just as legitimate to only look at a subset (your rolls) but you will want to record as many total rolls to get the same statistical validity of the results. Since it is your rolls in particular you are interested in that is the way to do it.
Wisea$$ DM and Player since 1979.
I noticed that too, with more 20s than 1s. But it's also such a small sample size, that things like that can happen. I have told my friends, the larger the sample size, the closer to the average the numbers should be. With only 100 or 200 rolls, that's a small sample size. Betting those percentages balance out if we were looking at 100,000 rolls.
that is quite likely but keep in mind that the roughly 300 critical rolls (20's + 1's) actually represents around 6000 total rolls for each character and by then the results should be closer to statistical averages. what I didn't see was the total numbers of rolls (it may be there and I missed it) to see if the numbers of 20's and 1's are running about 5% each or are they higher or lower.
Wisea$$ DM and Player since 1979.
Just went back and looked at the critical roll stats again. It’s difficult to tell for sure because the stats are by character not player but most are either too small a sample size to say much about or actually not too far off 50-50 when you look at 20’s/total or 1’s/total. You can’t combine anything except the DM stats (I’m assuming it’s the same DM all the time let me know if I’m wrong) which if combined show that over 75% of crits from the DM are 20’s - that’s one nasty DM to go up against. (That is after 465 crits representing around 4600 total D20 rolls, you would not expect it to be that far off the mean with that sample size.
Wisea$$ DM and Player since 1979.
They should be relatively close to the 5% number. But clearly, with the smaller sample size, some people have been more lucky, than unlucky. ;)
All Rolls - Wildemount - Google Sheets
This right here is the document recording all of the rolls made for Campaign 2 of Critical Role.
The Primary cast members
Are the first 8 spots On the previous link that I linked for the campaign. with 7 and 8 being the same player (Caduceus was made after Molly died). If you don't know which player is which I won't necessarily spoil that because it let's you look at it in more of a blind fashion to not actually know by player name.
Every other Character beyond that is usually some kind of guest character or notable npc's. this is why their rolls are low.
But if your wanting to understand sample sizes for the top 7 people that were the main cast in most episodes. in alphabetical order because it's easiest.
Beau did 3075 rolls in total
Caduceus Rolled 1325 rolls and Molly 447 for a total of 1772
Caleb did 1736 rolls,
Fjord did 2127
Jester did 2055
Nott/Veth did 1796 as Nott and 587 as Veth for a total of 2383 rolls.
Yasha did 1610 rolls.
Matt The DM's rolls sadly aren't in the sheet for the most part. Except for certain notable NPC's.
It should be noted that these do include rolls other than d20's because there are damage rolls mixed in as well. Though even if we go overly conservative and say that only one half of their rolls were d20's despite the fact that Damage is the only place smaller dice are used with any consistancy.
This means that Beau rolled 1532 D20 rolls as conservative estimate. This means that 9% of her rolls were natural 20's and 6.9% of her rolls were ones. This could be from Electronic Bias. I really don't remember. But it does put The extreme's within 2% of each other basically and allowing for error of incomplete numbers this does suggest that they are in parity for this specific person yes. and them both being on the high side.
Noth/Veth, and Yasha are also in very obvious parity by the numbers that we are given. Since they are almost dead on in our sample size on both sides.
Vex at first looks like she might be a little below the average mark with more 1's than 20's. But then if you look at the other Campaign she is slightly to the 20's by a not to disimilar margin so she's clearly in parity herself. With a rather large sample size between the two campaigns even if we go conservative on it to try and account for just d20's.
So let's look at one that seems to be much more out of whack. That of Caduceus combined with Molly.
Being conservative we are considering 886 D20 rolls for our sample size specific to that one person. With a Recording of 68 20's and and 40 1's. this is 7.6% of 20's and 4.5% of the rolls were 1's... This is a 3.1% difference in our sample size. This could certainly be that the sample size just isn't big enough.
So let's grab more dice rolls for this same person. In the Previous campaign he played Percy which gives us another 2082 total Rolls. Going overly conservative we're looking a 1041 rolls with 104 20's and 49 1's. Automatically from the numbers you can tell that the 20's were about 10% of the rolls and 4.7% of the 1's. He's skewing High and all he rolled in this campaign were actual dice.
This could be blamed on something called the Golden Snitch in campaign 1. So it's possible that the die has a weighting issue and it was one he rolled often. However he gave up that die to Matt the DM at the end of the Campaign so it was not available for Campaign 2. So while it's a bit more acceptable this particular individual is still skewing high on the rolls. Blended together we are looking at conservative estimates of 1972 Conservatively acceptable D20 rolls with 172 20's rolled and 89 1's rolled. This gives us an 8.7% on 20's rolled and a 4.6% on the 1's, which means were still looking at a difference of 4.1%. Unless we completely fine tooth the data like analysts however this is a pretty significant difference and we can't know if going into such detail would bring parity or not. If somebody wants to try I'm all for hearing the result. But just going by what we have. it's something rather notable.
Now it's possible that he's just going through a long (about 6 years) streak of skewing high and overall his numbers will even out. But at this point. He's a bit above the average on his rolls. By about a full number.
Interestingly another person at the table also Skew's Slightly High. But it seems to be at more acceptable levels. So let's look at him a little closer as well. In Campaign 2 going quite conservative again that would be 1063 Rolls. 99 being 20's and 63 being 1's. This gives us 9.3% as 20's and 5.9% as 1's. It again could be leaning to extreme's of 1's and 20's because of digital rollers. We'd have to go through days of footage and check. But I'm not going to argue that case. But he's still got a 3.4% difference between his high extreme and his low extreme. it's also higher than what Cad/Molly/Percy's player has for Campaign 2.
So how about we Expand his sample size as well. The Character he played in Campaign 1 is named Grog. There are a total of 2449 rolls for Grog recorded in Campaign 1, so our conservative number is 1224. He's recorded as having 86 20's and 58 1's. This gives us 7% and 4.7% respectively. And we know this game was played with Actual Dice so it's not skewed by electronic shenanigans. He's still got a 3.3% difference leaning towards the high side of the dice.
Even when we combine the numbers from the 2 sample sizes. We're looking at the conservative sample size of 2287, with 185 recorded 20's and 121 recorded 1's. Giving us 8% for the 20's to be rolled and 5.2% on the 1's being rolled over all of this. Granted some of these may be electronic and so extreme's may be slightly higher. But we're still looking at a 2.8% difference in favor of the high rolls. Overall he doesn't skew quite as high and again in the long run over a lot more rolls. He may well even out. But with our current data set. he's on a pretty good multi-year streak of playing weekly of getting a bit better than average. And he certainly isn't as skewed as the Cad/Molly/Percy player is over both campaigns.
And Again. i'm aware we aren't working with complete data. And some of my numbers are just conservative estimates that can adjust some things. The Second Campaign may even be suffereing a bit from psuedo random bias to Explain in that campaign why w'ere seeing numbers greater than 5% on both ends and the overly conservative estimate likely contributed to them being higher as well. And I totally accept that. But Not all of it is Perceptual Bias. For at least some of this there has to be some other factor in play. Whether it's particular long runs of slightly skewed luck. Or just the fact that certain of us are running slightly below average and others are running slightly above. it's hard to say. Because we have two overlapping situations. The Average of an Individual and the true objective average of the numbers that actually applies to all dice. Ideally math would like both to end up in the same place..
But it does stand to reason that if you can identify that your on one of these long higher swings that can span campaigns, And perhaps make the bias towards more extreme numbers like 1's and 20's in your favor. You can actually base builds around taking advantage of that. Though at least one of those things is quite difficult to do until you've basically already gotten through it.
(As a little side note that might interest some. Beau made so many rolls during the campaign that are so much higher than others, except perhaps for the DM Matt because she played a Monk. Most monks naturally make a lot of attack rolls per turn and then from that a lot of damage rolls. Grog in the First Campaign worked under the same principle but in a different way. he was a Berserker Barbarian. So he rolled quite regularly with advantage and he made decent amount of use of the berserk frenzy's bonus action attacks at key points in things.)
It looks like Beau was making a LOT of rolls compared to others. Was this because of the Monk Flurry or Blows and Stunning Strike (which can turn a single roll into several)?
For bad luck in rolls, Will Wheaton, on Critical Role, has to "take the cake!"
Out of 54 total rolls with a d20 (a very, very small sample size), he rolled a 1 or 2, 17 times! He rolled a 4 or lower, 25 out of 54 times! He rolled a 6 or lower, 31 times! He rolled a 9 or less, on 39 out of 54 rolls! That is mind bogglingly bad luck! Wow!
p.s. he only rolled one Nat 20.
A Distant Chime: Intro Statistics for RPGs: The Wheaton Dice Curse (espharel.blogspot.com)
They've had interviews about Will's terrible dice luck. it's like he's the Karma balancer for Tallison Jaffee or something...