Kramnik doesn't understand statistics. The chance of the streak is a little under 1 percent with that ELO difference, but that is if Hikaru had one attempt. These aren't the first 46 games Hikaru played online.
I think in general none of us really understand statistics when it can be so complex. Chess.com's analysis' in situations like these are usually pretty spot on and I wouldn't be surprised if they release some sort of statement backing what is virtually their #1 streamer. Its quite likely that Hikaru achieved his streak legitimately, but its also not impossible he had engine assistance - just improbable.
It's more that given the ratings of the players, the correct statistical analysis is the confidence interval of beating a similar group of players (ie not a single variable statistic analysis such as a win %). Even a 1% confidence interval would not be unusual if a 50-game session occurs every 3rd day of the year. But this is all speculation, as Kramnik didn't try to learn anything about statistics or epistemology. What was the accuracy rate across all games? That's an actual metric. @@jacksarkisian
@@jacksarkisianlol of course some people understand statistics. You just need to take a statistics course. For complex statistics needed for machine learning or data science you might need an advanced degree, but you do not need an advanced degree to realize statistical distributions have outliers. Moreover, people need to understand where the burden of proof lies in accusations, a fallacious argument does not constitute a serious accusation especially given all we know of Hikaru’s capabilities at blitz since he was a teenager and considering his opponents were FMs and IMs. The burden of proof always lies with the accuser since otherwise anyone could accuse anyone of anything on a whim and it is not serious to take accusations at face value even if a famous person makes them. It doesn’t matter who Kramnik is, his chess expertise as former champion is completely irrelevant to his argument here, invoking his authority in chess is just a red herring as the argument is not based in specific chess details but on the supposed atypicality of Hikaru’s run. Be serious. In the past people like Fischer accused Karpov and Kasparov of arranging their matches. That’s just how chess players are, generally paranoid.
@@ormerodp Hikaru covered the average accuracy across all 46 games in his most recent video: ruclips.net/video/Zm2R--_fBj0/видео.html I do think its highly unlikely there was any kind of foul play - I think the odds of Hikaru cheating are closer to 0.001%. But if you told me in 2021 that Ding Liren would be the next World Champion after Magnus, I'd also have told you it was impossible. Stranger things have happened.
A signifficantly large streak is the kind of thing that warrants a closer look with real statistics, but it is not evidence of anything by itself. Especially because winning beggets winning at least psychologically. Most players are going to second-guess themselves more against a player that has a reputation for being strong which gives the player with the reputation the opportunity to throw bad looking moves to make their opponents have to waste precious mental energy and time. Few players can consistently get away with that, especially at the highest level, but Hikaru (magnus being the other) can and does do that very often, from what I have seen. I would attribute the ability to do that to the size of his streaks more than playing suspiciously good at key times and positions. Though I would be interested in seeing actual statistics about that. And how would you create a tool that helps you with that? I mean, I suppose you could, but I don't think that Hikaru would be able to afford the cost of developing such a tool considering current technology and costs. And that doesn't even take into consideration that, if I understand things correctly, the accusation comes for a time where he was live-streaming. So you mean to tell me that he was reacting to chat, and looking at the tool, AND playing at a high level at the same time? Because a tool can only give you hints or confirm a suspicion without actually having to think it through, but you can't use a tool for every single move or you are going to get caught immediately. Some people underestimate the mental tax of livestreaming. Being able to live stream and do a single thing well is impressive enough. Being able to do two other things well is even more unbelievable to me than the size of his streak.
Chess.com has these tools at their disposal to investigate such matters. But a major issue Chess.com face is the fact they could never share how they come to such conclusions. Explicitly explaining the calculations and methodology used to draw conclusions on whether a player is cheating or not indirectly makes it a whole lot easier for people to cheat in a future scenario. If you know exactly how the fairplay process works on Chess.com, cheaters will actively avoid the invisible tripwires to avoid detection. Also if you're curious, Hikaru uploaded the stream in which he achieved this incredible streak on RUclips. You can find it here: ruclips.net/video/B4Bs7dPUbyI/видео.html
"You have not been clickbaited"... meanwhile thumbnail shows Hikaru's profile showing "Closed: fair play" and showing 4150 rating for Hikaru... If this isn't clickbait, what is???
I think there is no shot of him cheating. I just do see it. 400 fide is a massive difference and also as have seen false accusations by gms before - I’m referencing the accusation on witty
We can use an average -350 Elo, but we have to create a table of trials to get a feeling as to whether the outcome is likely, but likely in pure statistical terms. Using an average is OK since it would be easy to beat a lower-ranked player in the group, but easier to tie with the closer-ranked players. There is a very high probability of a tie with close Elos. NOTE 1: I could be wrong about this, but from what I have gathered, the win rate for a +350 Elo is not 88%, but 78.7% because we are talking about wins and not wins + ½ ties. There appears to be some confusion over what "win rate" means, and in this case a 350 Elo gap means: WIn: 78.7% Loss: 0.8% Tie: 20.5% NOTE 2: The Elo ratings, thus p values, at this range are very sensitive. If you want to assume Hikaru's wins are 'sus' (as Hikaru called Hans' games) then use the known values. But if you want to claim the results are wrong, then assume the Elo ratings are wrong, thus the win% should be higher. If we assume an average Elo difference of 450 then Hikaru's streak is not very 'sus' as p goes from .78 to .88 NOTE 3: This is blitz chess, and it is quite possible that ties are far less common because of time constraints. The typical chess definition of win adds ½ of the tie value to the win column. I am using stats on the odds of a streak of 45 wins within a range of 46 games. Elo difference: 350 p result: 0.787 win based on Elo math Desired Outcome: 45.5/46, but we look at 45/45 as well Trials: Selections 1 to 10,000 45.5/46 One trial (start now, end after 46 games): Probability ≈ 0.00024 or 1 in 4,000 (not 1%) 2,600 trials (play 10x a day, every weekday, for a year): 9.5% 10,000 trials: 33% 45/45 One trial: 0.00002 2,600 trials: 1.1% 10,000 trials: 4.3%
I follow hikaru he uploaded the games he played against said opponents and you can clearly see he makes errors and he takes you through his thought process on every move so unless kramnik can come up with concrete evidence that hikaru is cheating im just gna assun5e his a bitter old man that is having a hard time accepting his past his prime
I don't think Hikaru cheated, but note that when Hans played a game (even OTB) he clearly makes errors. But people chimed in with the absurd "when you cheat, you cheat at the right time and fool people that way). This was a very common and absurd logical fallacy.
Let's see. Hikaru beats 45,5-0,5 players that are on average ~400 points lower than him. A bit unusual, but not completely unplausible. Knowing Hikaru and how much he plays, such insane runs bound to happen eventually. If he scored 45,5-0,5 against 3300 rated opponents, then I would be very suspicious
Actually a fair point I thought of while making this video, but the simple explanation is that Elo on Chess.com (especially Blitz) can be massively inflated. Kramnik's statistical analysis drawing the "3600+" figure is also questionable at best.
Hikaru unfairly accused Hans OVER and OVER for weeks with unfair comments like, "A 2700 does not analyze like that". Constant. Barely indirect. More direct than what Kramnik did. What goes around comes around. Hikaru is a terrible human being and deserves what he gets.
Thanks for noticing, although I'm not quite a genius, but close. You on the other hand are extremely limited in cognitive ability. Hans was not embroiled in the controversy of online play. He was being accused of cheating OTB. Perhaps you should learn to read and discern before being confrontational. And btw, I should note that I am not even a Hans fan. Just a fan of truth and justice. As far as brain cells, anytime any venue you want to match wits with me on any subject at all I'd be happy to hand you your a**. @@theworldofwoo8320
Even he is a terrible person that doesn't mean he is cheating because the evidence is so limited and not clearly. Not like Hans, there is a lot of strong evidence against him.
Hans was being accused of cheating OTB. This is different. I don't think Hikaru was cheating either. But let hims suffer the accusation because he was such an a**h*** towards a guy that was innocent. OVER and OVER he really drove it in. Magnus too, slime ball. Terrible people. @@caudravan
Ridiculous comment by Kramnik. Why doesn't Kramnik look at the accuracy column and have an opinion on that? As far as Hikaru going to the bathroom and then making a move instantly when he returns, that what we would EXPECT. The position is in Hikaru's mind while he is away and he can analyze positions with lightning fast speed and come up with a move in that short break. The methods that online chess platforms use to determine cheating involve lots of details and information, not just a single notice of a winning streak. Why doesn't Kramnik accuse Capablanca of cheating during his 8 year winning streak without a loss? Because he speaks hastily without knowing all the facts. He should know better as a chess professional.
The problem is that Big Chess, Magnus, and all the top players want to make pro football style money for what is essentially a board game. Legitimization of online money chess tournaments is crucial to them. Not love of the game. This is the unspoken element here. Online chess is NOT and NEVER will be legitimate. You cannot stop cheating online. Can't happen, won't happen. So expect more of this stuff. They're all falling on their own swords over greed.
I think Kramnik is NOT accusing of Hikaru of "cheating at the game of chess"....but implying that by beatiing up on much weaker opponents to "solely increase his rating to get the record" is "form of cheating"/"gaming the system.
So him playing on blizz, with matchmaking he can’t possibly control, because it’s match making, some how he’s controlling the matchmaking? No shot you actually think that.
@@TheWhiteHorse6969 Have you never seen Hikaru “adopt” an IM/FM? He played same 3 guys, over 46 games….again, I love Hikaru…I think it’s BS….just saying what I think Kramnik was talking about
Hikaru streamed a few hours after this video was uploaded. I'm sure it'll be posted to his channel but he stated that Kramnik has suggested to other Grandmasters behind the scenes that Hikaru has been cheating in Titled Tuesday (and possibly OTB events) for months. He also stated that his relationship with Nepo was terrible, and that Nepo had very little respect for him - hence the assumption Nepo's tweet was backing Kramnik.
@@chrisbrown9036I dont think that's what Kramnik think tho.. The way he word it sound like he truly think Hikaru cheating with engine.. (Other chess player like Eric, Magnus and etc. Also does the same thing.. But no issue bcoz they rating lower than Hikaru in blitz? Why only Hikaru? I think Kramnik eye on him is bcoz of what Hikaru just achieve look impossible to him?. Except Magnus bcoz he's the goat, No one gonna accuse him at all lol) Btw, He already Accuse Hans (After scandal) cheating even tho Hans play terrible😅. I think Kramnik really think what i think it is. Just my opinion.
its imposible to beat such strong players around 2950 rating, with 45.5 out of 46 games..Kramnik is 2970 online...Do u think that Hikaru can beat him 45-0?of course not...and if they play blitz over the board it would be even closer result
Kramnik doesn't understand statistics. The chance of the streak is a little under 1 percent with that ELO difference, but that is if Hikaru had one attempt. These aren't the first 46 games Hikaru played online.
I think in general none of us really understand statistics when it can be so complex. Chess.com's analysis' in situations like these are usually pretty spot on and I wouldn't be surprised if they release some sort of statement backing what is virtually their #1 streamer. Its quite likely that Hikaru achieved his streak legitimately, but its also not impossible he had engine assistance - just improbable.
@@jacksarkisianimprobable? Impossible...
It's more that given the ratings of the players, the correct statistical analysis is the confidence interval of beating a similar group of players (ie not a single variable statistic analysis such as a win %). Even a 1% confidence interval would not be unusual if a 50-game session occurs every 3rd day of the year. But this is all speculation, as Kramnik didn't try to learn anything about statistics or epistemology. What was the accuracy rate across all games? That's an actual metric. @@jacksarkisian
@@jacksarkisianlol of course some people understand statistics. You just need to take a statistics course. For complex statistics needed for machine learning or data science you might need an advanced degree, but you do not need an advanced degree to realize statistical distributions have outliers. Moreover, people need to understand where the burden of proof lies in accusations, a fallacious argument does not constitute a serious accusation especially given all we know of Hikaru’s capabilities at blitz since he was a teenager and considering his opponents were FMs and IMs. The burden of proof always lies with the accuser since otherwise anyone could accuse anyone of anything on a whim and it is not serious to take accusations at face value even if a famous person makes them. It doesn’t matter who Kramnik is, his chess expertise as former champion is completely irrelevant to his argument here, invoking his authority in chess is just a red herring as the argument is not based in specific chess details but on the supposed atypicality of Hikaru’s run. Be serious. In the past people like Fischer accused Karpov and Kasparov of arranging their matches. That’s just how chess players are, generally paranoid.
@@ormerodp Hikaru covered the average accuracy across all 46 games in his most recent video: ruclips.net/video/Zm2R--_fBj0/видео.html
I do think its highly unlikely there was any kind of foul play - I think the odds of Hikaru cheating are closer to 0.001%. But if you told me in 2021 that Ding Liren would be the next World Champion after Magnus, I'd also have told you it was impossible. Stranger things have happened.
When people start accusing you of cheating then you know you doing something right in your life - the ULTIMATE compliment
edit: when 3 time World Champion's start accusing you of cheating then you know you're doing something right in your life
Not only getting called a cheater is a compliment but getting accused by some world champions? That's what we're talking about.
Nans Hieman
@@longofire236yea but he was actually cheating he is not tho
Hikaru can beat cramnik ass 69-0 for sure in offline games
If it goes anything like Kramnik's games against Niemann he'll just blunder Fool's Mate again 😭
A signifficantly large streak is the kind of thing that warrants a closer look with real statistics, but it is not evidence of anything by itself. Especially because winning beggets winning at least psychologically.
Most players are going to second-guess themselves more against a player that has a reputation for being strong which gives the player with the reputation the opportunity to throw bad looking moves to make their opponents have to waste precious mental energy and time.
Few players can consistently get away with that, especially at the highest level, but Hikaru (magnus being the other) can and does do that very often, from what I have seen. I would attribute the ability to do that to the size of his streaks more than playing suspiciously good at key times and positions. Though I would be interested in seeing actual statistics about that.
And how would you create a tool that helps you with that? I mean, I suppose you could, but I don't think that Hikaru would be able to afford the cost of developing such a tool considering current technology and costs.
And that doesn't even take into consideration that, if I understand things correctly, the accusation comes for a time where he was live-streaming. So you mean to tell me that he was reacting to chat, and looking at the tool, AND playing at a high level at the same time? Because a tool can only give you hints or confirm a suspicion without actually having to think it through, but you can't use a tool for every single move or you are going to get caught immediately.
Some people underestimate the mental tax of livestreaming. Being able to live stream and do a single thing well is impressive enough. Being able to do two other things well is even more unbelievable to me than the size of his streak.
Chess.com has these tools at their disposal to investigate such matters. But a major issue Chess.com face is the fact they could never share how they come to such conclusions.
Explicitly explaining the calculations and methodology used to draw conclusions on whether a player is cheating or not indirectly makes it a whole lot easier for people to cheat in a future scenario. If you know exactly how the fairplay process works on Chess.com, cheaters will actively avoid the invisible tripwires to avoid detection.
Also if you're curious, Hikaru uploaded the stream in which he achieved this incredible streak on RUclips. You can find it here: ruclips.net/video/B4Bs7dPUbyI/видео.html
"You have not been clickbaited"... meanwhile thumbnail shows Hikaru's profile showing "Closed: fair play" and showing 4150 rating for Hikaru...
If this isn't clickbait, what is???
I'll be honest, I hadn't made my thumbnail until after the video - by then I forgot what I said in the intro 😭😭😭
I think there is no shot of him cheating. I just do see it. 400 fide is a massive difference and also as have seen false accusations by gms before - I’m referencing the accusation on witty
why u look at their fide rating?look at thair online blitz rating and its 2950 strenth
@@drerskisame thing, hikaru is 3300 elo in blitz so about 400 difference still
@@nekorobin7028 he used FIDE rating because you a more realistic evaluation of real-life playing...so it is a 2400 player vs a 2780 player
@@seinundzeiten thanks man that’s why I did that :)
@@nekorobin7028 also correct cheers :)
Just wait until Kramnik notices how often someone wins the lottery and what those odds are…
7:24
If we exclude the game against Njal as Hikaru said his OTB elo performance is 3074.
We can use an average -350 Elo, but we have to create a table of trials to get a feeling as to whether the outcome is likely, but likely in pure statistical terms. Using an average is OK since it would be easy to beat a lower-ranked player in the group, but easier to tie with the closer-ranked players. There is a very high probability of a tie with close Elos.
NOTE 1: I could be wrong about this, but from what I have gathered, the win rate for a +350 Elo is not 88%, but 78.7% because we are talking about wins and not wins + ½ ties. There appears to be some confusion over what "win rate" means, and in this case a 350 Elo gap means:
WIn: 78.7%
Loss: 0.8%
Tie: 20.5%
NOTE 2: The Elo ratings, thus p values, at this range are very sensitive. If you want to assume Hikaru's wins are 'sus' (as Hikaru called Hans' games) then use the known values. But if you want to claim the results are wrong, then assume the Elo ratings are wrong, thus the win% should be higher. If we assume an average Elo difference of 450 then Hikaru's streak is not very 'sus' as p goes from .78 to .88
NOTE 3: This is blitz chess, and it is quite possible that ties are far less common because of time constraints.
The typical chess definition of win adds ½ of the tie value to the win column. I am using stats on the odds of a streak of 45 wins within a range of 46 games.
Elo difference: 350
p result: 0.787 win based on Elo math
Desired Outcome: 45.5/46, but we look at 45/45 as well
Trials: Selections 1 to 10,000
45.5/46
One trial (start now, end after 46 games): Probability ≈ 0.00024 or 1 in 4,000 (not 1%)
2,600 trials (play 10x a day, every weekday, for a year): 9.5%
10,000 trials: 33%
45/45
One trial: 0.00002
2,600 trials: 1.1%
10,000 trials: 4.3%
I follow hikaru he uploaded the games he played against said opponents and you can clearly see he makes errors and he takes you through his thought process on every move so unless kramnik can come up with concrete evidence that hikaru is cheating im just gna assun5e his a bitter old man that is having a hard time accepting his past his prime
I don't think Hikaru cheated, but note that when Hans played a game (even OTB) he clearly makes errors. But people chimed in with the absurd "when you cheat, you cheat at the right time and fool people that way). This was a very common and absurd logical fallacy.
Let's see. Hikaru beats 45,5-0,5 players that are on average ~400 points lower than him. A bit unusual, but not completely unplausible. Knowing Hikaru and how much he plays, such insane runs bound to happen eventually. If he scored 45,5-0,5 against 3300 rated opponents, then I would be very suspicious
you dont gotta bait brah. reality is stupid enough that it keeps our interest 🤣
reality is the new clickbait 🤣🤣
If Hikaru plays better than Stockfish, doesnt that show he's not cheating. ;)
Actually a fair point I thought of while making this video, but the simple explanation is that Elo on Chess.com (especially Blitz) can be massively inflated. Kramnik's statistical analysis drawing the "3600+" figure is also questionable at best.
Hikaru unfairly accused Hans OVER and OVER for weeks with unfair comments like, "A 2700 does not analyze like that". Constant. Barely indirect. More direct than what Kramnik did. What goes around comes around. Hikaru is a terrible human being and deserves what he gets.
Hey genius. Hans was actually caught several times cheating. And admitted it. Try to use those few brain cells in tandem
Thanks for noticing, although I'm not quite a genius, but close. You on the other hand are extremely limited in cognitive ability. Hans was not embroiled in the controversy of online play. He was being accused of cheating OTB. Perhaps you should learn to read and discern before being confrontational. And btw, I should note that I am not even a Hans fan. Just a fan of truth and justice. As far as brain cells, anytime any venue you want to match wits with me on any subject at all I'd be happy to hand you your a**. @@theworldofwoo8320
Even he is a terrible person that doesn't mean he is cheating because the evidence is so limited and not clearly. Not like Hans, there is a lot of strong evidence against him.
Hans was being accused of cheating OTB. This is different. I don't think Hikaru was cheating either. But let hims suffer the accusation because he was such an a**h*** towards a guy that was innocent. OVER and OVER he really drove it in. Magnus too, slime ball. Terrible people. @@caudravan
@@supernoobsmith5718i believe hans was genuinely caught cheating several times
This youtuber is a different breed, smh always posting silly ig reels have some class in the chess world
it is Kramnik's new hobby
Ridiculous comment by Kramnik. Why doesn't Kramnik look at the accuracy column and have an opinion on that? As far as Hikaru going to the bathroom and then making a move instantly when he returns, that what we would EXPECT. The position is in Hikaru's mind while he is away and he can analyze positions with lightning fast speed and come up with a move in that short break. The methods that online chess platforms use to determine cheating involve lots of details and information, not just a single notice of a winning streak. Why doesn't Kramnik accuse Capablanca of cheating during his 8 year winning streak without a loss? Because he speaks hastily without knowing all the facts. He should know better as a chess professional.
Basically Hikaru is saying the pool of players he won against are significantly lower rated than him. So it's no surprise he won 45/49 of them
The problem is that Big Chess, Magnus, and all the top players want to make pro football style money for what is essentially a board game. Legitimization of online money chess tournaments is crucial to them. Not love of the game. This is the unspoken element here. Online chess is NOT and NEVER will be legitimate. You cannot stop cheating online. Can't happen, won't happen. So expect more of this stuff. They're all falling on their own swords over greed.
as a korean I cant understand what's going on...
hikaru cheating???
I think Kramnik is NOT accusing of Hikaru of "cheating at the game of chess"....but implying that by beatiing up on much weaker opponents to "solely increase his rating to get the record" is "form of cheating"/"gaming the system.
So him playing on blizz, with matchmaking he can’t possibly control, because it’s match making, some how he’s controlling the matchmaking? No shot you actually think that.
@@TheWhiteHorse6969 Have you never seen Hikaru “adopt” an IM/FM? He played same 3 guys, over 46 games….again, I love Hikaru…I think it’s BS….just saying what I think Kramnik was talking about
Hikaru streamed a few hours after this video was uploaded. I'm sure it'll be posted to his channel but he stated that Kramnik has suggested to other Grandmasters behind the scenes that Hikaru has been cheating in Titled Tuesday (and possibly OTB events) for months. He also stated that his relationship with Nepo was terrible, and that Nepo had very little respect for him - hence the assumption Nepo's tweet was backing Kramnik.
@@chrisbrown9036I dont think that's what Kramnik think tho.. The way he word it sound like he truly think Hikaru cheating with engine.. (Other chess player like Eric, Magnus and etc. Also does the same thing.. But no issue bcoz they rating lower than Hikaru in blitz? Why only Hikaru? I think Kramnik eye on him is bcoz of what Hikaru just achieve look impossible to him?. Except Magnus bcoz he's the goat, No one gonna accuse him at all lol)
Btw, He already Accuse Hans (After scandal) cheating even tho Hans play terrible😅. I think Kramnik really think what i think it is. Just my opinion.
@@Sweet_Jelly39 you could be completely on point here, I may have misinterpreted his "veiled" accusation. you make a concise point...well done
He is world #3 not #4
This video made me fail NNN 😭😭😭
love to see a fellow arsenal fan💪
#COYG
what do you use to make thumbnails?
cool video, love the Scottish Accent!
its imposible to beat such strong players around 2950 rating, with 45.5 out of 46 games..Kramnik is 2970 online...Do u think that Hikaru can beat him 45-0?of course not...and if they play blitz over the board it would be even closer result
anything is possible