How do you build immunity from choking? (a meditation on Carlsen and Kramnik)

Yesterday Magnus Carlsen and Vladimir Kramnik each played decisive chess games in the Candidates’ tournament, for the right to challenge Vishy Anand for the world championship title.  Carlsen held the tiebreaker, so he had only to match Kramnik’s result — draw or win — to proceed to the match with Anand.

Both lost.  Uncharacteristically, both fell into time trouble.  Both made bad mistakes even after time trouble was over.  The chess world was shocked.

Arguably both choked.

Yet Kramnik has won several world championship matches, including against Kasparov, and Carlsen rose to world #1 at a very young age of course.

How does one become immune to choking?

If you have mastered stages 1 through n, presumably you still can choke at stage n + 1.  Carlsen had never played in a world championship or candidates’ match before.  In 1997 Kasparov choked when he had to play an improved Deep Blue, a machine.

Is there mean-reversion in choking and immunity to choking?  If you play at a supremely confident level at the very top, nine times in a row, do you forget how to handle pressure and eventually revert to choking?  Immunity against choking can wear off, or holding a title and having to defend it can raise the fear of choking through a kind of endowment effect (Bobby Fischer).

Does a string of confident winning raise the stakes more rapidly than you can master a rising choke, thus bringing you to n + 27 too quickly?  (The Miami Heat just lost a 27-game winning streak.)  Joe DiMaggio’s 56-game hitting streak has proven so hard to break, as there is always a new and tougher choking margin.

Or can one ascend to n + 3 with sufficiently strong margins of error that perhaps the fear of choking is never overcome and remains in the background for when tougher situations come along?  Or one can ascend to n + 3 if everyone chokes along the way; someone must be choking least but still you are always a choker.

Under one theory, you become immune to choking at stage n + 2 only by at least once choking at stage n + 2 and then, on another occasion, overcoming your choke.  I call this the LeBron James theory.  Can it be that such loss/win experiences are required periodically and not just once up front?

The lessons are that it can be difficult to overcome choking, and that a complex mix of losing and winning may help you more with choking than simply lots of winning.


Is choking sufficiently well defined here? Sometimes I agree that you know if when you see it as perhaps is the case in the chess. But the Heat losing? And making mistakes under pressure, even bad ones doesn't mean a choke because people make bad mistakes and when it's not important we don't notice.

Is Roger Federer a choker? He's had a large number of high profile 5 set losses including matches where he had match points, but does losing those matches say more about choking than losing more surprise blowouts in earlier rounds?

I've looked a fair amount for evidence of choking. It's hard to distinguish from simple statistical failure.

I think you can see choking more clearly in sports where the competitor has to initiate the action. For example, in golf the major championships and the Ryder Cup bring out behavior that you wouldn't see at a weekly tournament (e.g., Mark Calcavechia hitting his tee shot on the 71st hole in the 1991 Ryder Cup barely halfway across the lake, or Ian Baker-Finch having to quit golf because he couldn't hit the ball anymore.)

Similarly, I can recall pitchers falling apart mentally in post-season games (Burt Hooton in the 1977 NLCS or Rick Ankiel more recently), but it's hard to tell with hitters since they just react. Ankiel, for example, astonished fans by returning to the big leagues as an outfielder.

+1 to your first paragraph

For your second paragraph, it's an interesting theory. Most people think "choking" happens more often when instinct is overcome and you're "thinking too much". Would make sense that this happens more often at times when there is no instinct and pace is self-initiated.

The consensus among most statistically-minded sports observers is that what we call "choking" is largely an attempt to apply a narrative to normal statistical variation. There is a lot of random variation to athletes' performances to start with. When playing on the biggest stages, they are usually facing the best talent, so it is no surprise if that is the occasion when they perform poorly. Still, we want to give everything a story.

If choking is a real effect, it should be easy to measure statistically. However, there is no evidence that it exists.

Are people like Ankiel exceptions? Conventional wisdom is that he mentally lost the ability to pitch due to his disastrous postseason. However, when looking at his stats I discovered something. In the season following his collapse, his stats were not very different from before: except that batters were swinging at many fewer pitches. It may be that he was the same as always, but that batters had learned his weakness (lack of control) and were taking advantage of it. In other words, a completely ordinary reason for a pitcher's decline.

On Ankiel, he had a disastrous postseason because he lost the ability to pitch, not the other way around. It's called Steve Blass disease because the same thing happened to him (and Chuck Knoblauch). Rarely, players just lose the ability to throw the ball straight in a game. It's obviously mental and not specific to pitching (Knoblauch was a 2B). I don't know that it's related to choking since it persists beyond pressure situations. It's more like a complete mental block on the ability to throw the ball in a specific context. Oddly for Ankiel, it did not persist once he moved to the OF, where he was known (unsurprisingly) to have a pretty good arm.

Both Chuck Knoblauch and Steve Sax were star MLB second basemen who suddenly lost the ability to throw the ball accurately 60 feet to the first baseman:

Since these psychological impairments of performance went on so long we can be pretty certain they weren't just a string of bad luck, I'd call them pretty good evidence that choking exists.


And as always, ably symbolized by xkcd

Hilarious, as usual.

The opposite of the choke is the killer instinct. The choker changes their mental state right as victory nears, the champion does not waver.

Overconfidence can look like a choke.

The first bet I ever made in my life was on the Cardinals in the 7th game of the 1968 World Series. Nobody had the killer instinct like Bob Gibson. He'd won 7 straight World Series games going back to 1964. And the Tigers had to use up their ace Denny McClain in Game 6. It was a lock for Gibson and me.

We lost.

I think choking in chess is easier to see than in other sports. Merely losing is not choking. The performance of Kramnik and Carlsen was below par for both of them and it was clear to lots of outside observers. In contrast, Korchnoi at his peak was good in high pressure situations and when he lost (as in the final game against Karpov in 1978) there was no evidence that this was a "choke" but more genuinely natural outcome/statistical variation, etc. Choking in some things like chess are missing things that would normally be unheard of by someone at or near their level in critical positions and getting into time trouble so that you are likely to blunder (unless your name is Reshevsky).

In my experience with sports you can choke based on the moment feeling "too big" for optimal performance, as well as "too small".

Sports psychologists say that visualization is the key to avoiding choking - focus on positive execution against the objective. Perhaps this is in line with your theory: it is hard to visualize what you haven't experienced before? This would suggest if you haven't experienced a certain stress level, it's hard to see through it.

In terms of sports, the data is out there. Most studies suggest that there is no such thing as a hot or cold streak. Perhaps the question could be reworded for a definition of choking.

I suspect that many streaks are misinterpreted. Dimaggio's could hold up because it is a true 4 or 5 sigma event.

People misquote the 4 minute mile as an example of confidence overcoming the obstacle. The reality is that many people were closing in on it, and if Bannister hadn't done it first, someone else would have within a year.

"In terms of sports, the data is out there. Most studies suggest that there is no such thing as a hot or cold streak. "

As both a sports fan and a statistician, this oversimplifies greatly. What we have is lack of affirmative evidence in most cases. There's no evidence that these things don't occur. And that's an important distinction.

For a long time, folks maintained there was no "hot hand" in basketball. (Think "Marv Albert saying 'He's on fire!" in NBA Jam back in the day.) But more sophisticated techniques, looking at the data in a less naive fashion resulted in a publication showing some serial correlation with free throws. The conventional wisdom in the stats crowd was to sneer at the play-by-play announcers claiming someone was "hot". In fact they just hadn't been clever enough in their analysis.

That doesn't take away from the fact that we as humans like to construct narratives where none exist and assign streaks resulting from random variation to be a result of someone just "being hot" and perhaps requiring a "heat check". The truth is no doubt in between. "Hot streaks" exist, but the change in P("Success") is probably a lot smaller than we give it credit for conventionally. And those changes probably vary substantially from player to player, depend on many factors, and are very hard to quantify. It's a vague, unsatisfying conclusion, but it's probably the correct one. It boils down to human psychology, and we probably understand less about that than we do even economics.

I strongly believe the "hot streaks don't exist" dogma is overstated. It's obvious that cold streaks exist (causes include: the player is injured, sick, mad at his coach, getting divorced, hung over, has fallen into bad mechanics, too young, too old, etc etc). At minimum, hot streaks happen when the causes of cold streaks aren't present.

Would too young or too old really be causes of cold "streaks"? That would seem so general as to be able to include "inferior talent" as a reason for a cold "streak," which makes the entire idea meaningless.

Another bet I made was in 1974 I bet a naive Pittsburgh Steelers fan even money that the Steelers wouldn't win their next five games. You see, the Steelers weren't bad, but if they won their last two games of the regular season, then they'd be in the playoffs, and if they won three more games, they'd win the Super Bowl and that was extremely unlikely.

Of course, I lost because the Steelers hit a hot streak. But, actually, that wasn't a fluke streak, it was a great dynasty (who went on to win 4 Super Bowls in 5 years) suddenly maturing (i.e., getting their pathbreaking use of steroids just right) just as I bet against them.

That's not uncommon -- a little-known team goes on a long win streak. It sounds like a fluke at the time, but in hindsight it's actually what had been a too-young team not being too young anymore and living up to their potential.

Also: why do the "narrative" and "normal statistical variation" have to be mutually exclusive? What if "choking" and "having the killer instinct" are real but relatively uncommon phenomena? Might not the statistical reflection of that be: some (few) top-level sports players lose a disproportionately high number of important matches while other (few) top-level players win a disproportionately high number of such matches?

Just like there is no hot hand in sports (although there's a cognitive bias to see them) there's probably no choking in the highest levels of sport, in the true sense of the phrase. There will always be improbably bad performances, but what doesn't show up in the data is athletes performing any worse or better in the clutch *in a persistent way from one year to the next*. In other words there is no clutch or (not choking 'skill'). It turns out athletes mostly perform exactly in line with their career averages when in the "clutch," but in any given high pressure moment, there is random variation that we later interpret as clutch or choke performances.

Also, you were wrong about the Heat after the Decision.

Most people who can't perform well at high levels of pressure never make it to the big time in the first place, but occasionally you see athletes excel, then crater when the mental pressure starts to get to them:

Yes that's partly true, which is why I said the highest levels of sport (just like elite chess). They've been through the fire, so 99% of the "meltdowns" we see are actually just rotten luck, but memorable for that reason.

And one of the examples you cite isn't really helpful. Ankiel re-invented his career and has since produced 'clutch' moments as a hitter (youtube seach 'Ankiel playoffs home run'). Is clutch/anti-choke something that applies for pitchers but differently for hitters? Is losing all ability whatsoever to perform a physical task really 'choking'? If Ankiel can't even throw strikes in a Single A game with no pressure, is he really choking, or has he just, however improbably, lost a skill he once had.

"Is clutch/anti-choke something that applies for pitchers but differently for hitters?"

Yes, I think so. Pitching is like golf, you have to initiate the action, so there's more chance for your brain to get in the way.

"...there’s probably no choking in the highest levels of sport"

You've clearly not followed Andy Murray's tennis career.

I actually disagree with this. There are not a lot of examples of him choking in big matches, more just being beaten by superior talents. Andy Murray is having a great career but I think it's safe to say that he's not quite up to the level of the other 3 (well, he's better than Federer now).

I followed the match. IMO what happened was both players--Carlsen and Kramnik, who were tied for first but Carlsen would advance if they remained tied (and they did remain tied, as both lost their last games)-- bought the hype that they "must win"--and tried too hard. Both overreached. In the case of Carlsen, he did choke, but perhaps because he was looking at Kramnik. Note this, from, Kramnik speaking: “I had to play for a win, to burn bridges in a way, because of course I didn't think that Magnus was going to lose. I thought I got what I wanted at some point. It was an interesting position but terribly complicated. Somewhere around 20…Nhf4 I liked my position and then somehow I lost a bit of concentration because I didn’t know what to do,” said Kramnik, who also kept an eye on the other game. .. In time trouble the Russian missed an important tactic, and then his position was lost. Ivanchuk agreed that the position was at some point drawish. “But I noticed that my opponent started to play a bit risky and he gave me chances.”

In chess sometimes it is hard to win--much easier to draw. Kramik tried a new opening--the Pirc--which he rarely if ever plays. It's a flexible opening but if you've never played it before it is not the time to play it, in a last round game that decides the title. Ironically if Kramnik had drawn rather than tried for a win, he'd have won outright, but by losing along with Carlsen, he lost on tiebreak (and he knew this going in, so it makes his effort to win all the more ironic).

In short, don't play to win when there's no win on the board--just play good chess. Speaking as a chess fan, though I would again have loved to see Kramnik (approaching 40 yrs old) play the defending chess champion Anand (approaching 50), it is better for chess that the young talent Carlsen won, since the last championship match between Anand and Gelfand was between two 40-somethings and for most chess fans that's too boring (Gelfand btw choked in that match--like the above, he won a game and overreached on the following game, blundering badly trying to force a win and losing the 12 game match. Had Gelfand simply gone for a safe draw he would have won the 12 game match)

Good synopsis, although Vishy is 43, which one doesn't normally think of as "approaching 50".

Kramnik thought he had to go for a win and for a long time during the game that held true. I'm not sure that qualifies as a choke.

I'm not sure that Kramnik didn't play the right strategy. It's not unreasonable to think that Carlsen had a 45% chance to win 35% chance to draw and only a 20% chance to lose. Throw a little Game Theory on that and assume he has two strategies:

1) Go for Win (which gives him 30% chance to win, 60% chance to lose, and 10% chance to draw)
2) Go for Draw (which gives him 10% chance to win, 30% chance to lose, and 60% chance to draw)

If those probabilities (and my math) were correct, you'd be right to "play good chess" and go for the draw if you were trying to maximize your "points" for the game. But if you were playing to beat Carlsen, you improve your chances by going for the win. Which makes me wonder if the players actually go through the math in situations like this? If not, they probably should. It would be relatively easy to get fairly accurate probabilities in Chess (or at least get probabilities that would be more trustworthy than their gut) and they could probably improve score and result slightly by doing so.

Good observation, and let me add this is classic Prisoner's Dilemma game theory (or Chicken, or any variant thereof having two payoffs). You don't even need to know the exact probabilities, except to know that if both Carlsen and Kramnik 'go for the win' there is no advantage to either player in doing so--assuming, as is reasonable, both players have the same probability profile (that is, assuming that if Carlsen plays for a win he gets the same or roughly the same breakdown of probabilities as Kramnik, which seems fair). However, if Carlsen 'goes for the win' and Kramnik 'goes for the draw', clearly Kramnik wins. So you can have a Nash Equilibrium and I'm pretty sure the optimal strategy is "go for the draw"--and hope your foolish, rash opponent tries to 'force a win' in a drawn position which any chess player knows is fatal (as a matter of fact, if you want to weaken your chess engine--if you play like I do against chess software--check the box 'don't draw' and notice how often it will lose, from being in zugzwang from trying to avoid a draw).

So, using your probabilities, and regardless of what they actually are, both Kramnik and Carlsen blundered in their game strategies--which is ironic, given they are the #3 and #1 players in the world in chess.

Remember also that there was pressure on Carlsen to outright win, since Ivanchuk had flagged 5 of his previous games, so Kramnik winning on time was a significant possibility. So Carlsen had to press for a win in a drawish game.

Beyond the choking issue, Carlsen (according to his interview) and probably Kramnik were suffering from fatigue at that point in the tournament. Arguably, Carlsen-Ivanchuk two rounds earlier was a choke, though.

Under one theory, you become immune to choking at stage n + 2 only by at least once choking at stage n + 2 and then, on another occasion, overcoming your choke. I call this the LeBron James theory. Can it be that such loss/win experiences are required periodically and not just once up front?

This has my vote. More generally, you must learn to "play through" situations in which you are conscious going into the situation (and indeed during the situation as well) that the risk of choking is real.

The definition of choking is subjective, but I interpret it as competing at a substantially lower quality level than usual (still subjective, I admit). We could think of the ability to focus or concentrate as a stock, that gets depleted at a rate that is proportional to the stakes at each point in time. That's why we see choking towards the end of seasons or tournaments.

I don't think it was choking. Carlsen knew that he had to get at least as good of a result as Kramnik. He was in a fairly simple, equal position, then he saw that Kramnik's game was getting complex and was worried that Kramnik would win, so he got more aggressive and went for unnecessary complications. He then spent lots of time thinking trying to find a win in an equal position and ended up making himself worse when he got in time trouble. Kramnik, on the other hand, saw that Carlsen was in for an easy draw and so he thought he had to push extra hard for a win (because if Carlsen drew then a draw did him no good). So he strained extra hard for a win as well, thus creating the complications mentioned above. He missed a move or two and subsequently lost. This all came out in the press conferences. Between the psychology and the fact that it was the 14th round of the tournament, I don't think it had anything to do with choking.

Yeah, what Ray Lopez said.

I think you said it more succienctly. Choosing a high variance (high risk) strategy to get the win you need, but failing, is not a choke. If you take the risk and it doesn't pay off, you may look especially foolish in hindsight which might be seen as a 'choke.'. Like shooting a 3 when a 2-point shot would tie it. Or going for 2 in football down 1 for the win rather than the tie.

One way to overcome choking is to put two other MVP candidates on your team. LeBron James' demeanor and verbiage hints to me that he needs someone in his life to give him some stress inoculation. One way to do this would be to not run away from your situation. Oh well. For that he signed away any possibility of attaining GOAT status.

Disagree, this season (and last) he's putting up GOAT numbers. And he's only 28. When LeBron's career is done and he has 5-6 rings he'll be considered on the level of the alltimers, like Jordan, Magic, Bird, Chamberlain, etc.

If you enjoy basketball, enjoy LeBron. We haven't had a player this dominant since MJ.

You clearly have not followed his career, seen him play much, or seen much of his demeanor and verbiage. He's as skilled, clutch, and classy as they come.

You might want to read Brandon Sanderson's short story Firstborn in that regard. Arguably (to be as spoiler free as possible), the solution is about learning to deal with choking and with the possibility of failure.

Everyone is ignoring the opponents and their incentives. This isn't like trying to break the record for free throws with no one but Guinness and a photographer on the court.

The omitted variable here is cognition. Choking means over-thinking and looking too far ahead. Chess is a pursuit which stresses thinking and looking ahead.

This post is why I read this blog. The comments section is also superb.

Is it "immunity to" or "immunity from" or both (see original post)?

a complex mix of losing and winning may help you more with choking than simply lots of winning.

Also good advice for raising children so they are resilient.

Usually when we see someone choke we focus on the loser rather than the winner, but it takes both a winner and loser to produce a choke.

I have a friend who was a relief pitcher in college. He hated anyone who ever used the word "choke" or "choker." He blew a big game for his team once, but he decided that the other team had just adjusted to his pitches and he couldn't readjust in time. That's how it works.

Depends. This is mainly correct in team sports. But individual sports like golf, tennis, chess, etc can have some instances of 'choking'. It's not scientific, you just kind of know it when you see it.

Choking is about not being able to handle pressure. In tennis more older players choke because they are aware of the stakes more, and they know more about how bad it is too lose. One way many athletes deal with choking is to convince themselves they are the underdog and that they have nothing to lose. That is why you hear so many succesful teams spout nonsense about how nobody believes in them, even though millions of people bet on them to win.

I don't see why this would be named the Lebron James theory at all. Starting with his first trip to the playoffs he dominated against an average Washington team and then took the defending EC champs to 7 games- a 60 win team that took SA to 7 games the season before in the finals and a very similar team to the one that dominated the Lakers the year before that in the finals- without having a bad game in the series (especially considering DET was a top 5 defense that season).

His second year in the playoffs he again dominated against a couple of average teams and then again against DET he had 4 dominant games out of 6 to run them out of the playoffs. In the finals he had two very strong games and two below average games (for him) against a top 2 defense, not exactly unexpected given his competition and surrounding talent.

In year 3 after he again ran over a mediocre team in round 1 and then played 5 terrific games out of 7 against BOS including a game 7 in which he scored 45 with 6 reb and 5 assists against 2 turnovers against the #1 defense in the league that season.

Year 4 he crushed two mediocre teams and put up 3 40+ pt games (and 5 of 6 games of 35+ pts) that was ranked 1st defensively (including a 49 pt game in a loss in game 1) in losing in 6 games.

His first 4 seasons in the postseason put him in a variety of different situations- as an underdog putting up an unexpected fight year one, an underdog winning in year two and getting farther than anyone thought, on evenly matched teams year three against BOS and as a favorite on a 66 win team year 4 without exhibiting any signs of choking by any definition of the word.

"I don’t see why this would be named the Lebron James theory at all."

It's because LeBron got outplayed by a white guy in the 2011 NBA finals:

In golf, by the way, older, battle-tested players tend to have their putting break down as they age. Fresh young players choke less on the greens than veterans.

Ben Hogan was the most famous example of this. In his late 40s, after he'd won four U.S. Opens, he was still the best ballstriker in the world, but he putted like Wilt Chamberlain shot free throws. Was this psychological or physiological? Hogan was enormously respected for his character (he was almost crippled for life in head-on highway car crash in 1949, but came back to win the U.S Open in 1950, 1951, and 1953), but after 1953 his bad putting kept him from ever quite getting to the top again.

This may be an example of the choking-initiation of the action connection. Hogan's big swings continued to be superb, but he had increasing difficulty initiating delicate motions on the green.

Young golfers, in contrast, tend to have psychological problems managing their emotions after bad shots. Bobby Jones was a teen phenom from age 14 onward, but didn't win a major championship until age 23 when he finally overcame his temper. An embarrassing public meltdown at St. Andrews, the capital of golf, was the catalyst for him to rebuild his character on more self-disciplined lines. Jones's maturation was so appreciated by St. Andrews fans that he was the first America to be made an honorary citizen of St. Andrews since Benjamin Franklin.

There's no question that "choking" exists as a real phenomenon separate and distinct from a run of bad luck. Clearly some players are better than others; these are games of skill. Give a player strep throat, a stretched ligament, a late night out on the town, or a fifth of JD and see how well they play.

There is a distinct asymmetry to game play - it is very hard to sustain good play, but it is very easy to slip into poor play. And even temporary lapses of poor play can put you into holes too big to dig out of or force you to take bigger risks. This makes the choke even more obvious.

Let's also remember that a team that is leading into the final minutes likely exerted itself to a greater degree earlier in the game, thus the team coming from behind has an energy reserve to tap. Chess players can become mentally fatigued. Certainly poker players do, and in that case there really are black swan events that are decisive.

WOW just what I was searching for. Came here by searching for vitiligo skin disease pictures

Hi there to every single one, it's truly a nice for me to pay a quick visit this site, it includes important Information.

As some commenters above have alluded, I think it was "symbiotic choking".

Comments for this post are closed