The world chess blitz championships have just ended. As you would probably expect, the age profile of the top players is noticeably younger than in classical (i.e.. slow) chess. But the difference is actually only about 4 years. More surprisingly, I think (if you are a go player), there are only two teenagers in the top 48 in blitz, and just one in classical - and they are way down near the bottom. But there are three over-forties in the blitz list, and the classical list has not just more but several in the top dozen.

This strikes me as a sharp and surprising contrast with the professional go scene, where the teenyboppers are numerous and successful - my sense is that at the amateur level, the proportions of young players in chess and go, and their degrees of success, are somewhat similar.

A curious state of affairs. I've been racking my brains to find an explanation, but I've only come up with partial reasons. Please suggest better reasons if you can, but a couple of those I've mulled over:

1. Chess respects classical time limits rather more than go, and has fewer blitz events. I infer from another thread that my long-standing contention that Mickey Mouse time limits are not good for go in Korea is being conceded, but China has stepped back from the brink and is re-focusing on longer events - yet teenyboppers are doing well there, so the parallel with chess still fails.

2. The style of play has changed in go. In chess there is really only one way to win a game - checkmate. In go you can win either large-scale, by killing something big or amassing a huge territory, or small-scale - eking out a small territorial win. It seems from the record number of B+R/W+R results that large-scale now dominates. Maybe this is something that favours young people. If so, it may not be something inherent. It may just be that older players trained themselves to rely on endgame techniques rather than tactical sharpness (longer time limits would explain both those factors); young players just go straight for the jugular.

I hesitate over both those explanations, though. A third possibility is that intrinsically go and chess differ in some way that favours young brains. Beats me what that might be, but speculating wildly: do young people just absorb patterns and play intuitively (i.e. rather like AlphaGo) whereas older players start to think too much about the game and so lose that intuitive edge?

There is some evidence in bridge that intuition does not fade much with age. Primarily intuitive players tend to maintain their strength better with age than players who primarily rely upon calculation, at least at the top levels. We certainly know in general that conscious calculation fades with age. People slow down.

_________________
The Adkins Principle:
At some point, doesn't thinking have to go on?
— Winona Adkins

Visualize whirled peas.

Everything with love. Stay safe.

Fast time limits are one of many different interpretations of reasons for the state of go in Korea. There are several other variables: AlphaGo, rise in number of strong young Chinese players, cultural and political factors, etc.

You can blame everything on "Mickey Mouse time limits", but I'd say there's at least a chance that confirmation bias is coming into play.

_________________
be immersed

For fun, I looked at the top 40 live chess ratings for classical, rapid and blitz. Average age: classical 31.0, rapid 31.7, blitz 31.2. So that seems within noise.

I think chess itself has recently been moving away from classical time limits and towards rapid (which I think is more the equivalent of "blitz go" than is blitz chess, where flurries of fraction-of-a-second moves frequently take place).

I agree with Bill that an intuition-driven game favors the older player, in general. Of course, due to computers, a 25-year-old can pick up more experience to build her intuition than someone in the 20th century could have picked up in a lifetime. It will be interesting to see if the same thing happens with go. It probably already has started to, with people playing tens of thousands of games on Tygem or wherever.

I also think that chess has analogies to the two kinds of approaches that you mention in go (large-scale and small-scale). Of course you eventually win (if resignation has not come first) by checkmate, but you can do this by attacking the king early in a tactical melee, or by playing a slow positional game and trying to grind out the win in the endgame. I get the impression you are a strong enough chess player to be familiar with this, so perhaps you are getting at something else.

I do think that on a whole-board scale, chess lends itself to blitz-style intuition more than go. The board is much smaller, and the positions have more "texture" (different sorts of pieces, pawn structure, constellations of pieces and pawns that suggest strategy), making it easier to assess the nature of the entire position in a second or less than is the case in go. Of course, maybe I just think this because I'm a stronger chess player than go player.

I have to admit that the rapid and blitz tournaments made for pretty gripping viewing. I was glued to my seat every morning all week.

By the way, is there a more precise definition of "Mickey Mouse time limits" than "faster than I think appropriate"? I can't easily find a use of the term by anyone else.

With some trepidation I offer the idea that the term "Mickey Mouse" as used in these boards derives from sports, such as baseball, basketball, or American football, where the difficulty or playing situations are diluted somehow so the game is not as serious or demanding. For example, in basketball the height of the hoop is lower, in baseball there is no pitcher's mound, or in football there is no tackling. In Go the fast games might be considered less serious because there is less time to think and, therefore, the standard of play is lower.

When interviewing a golden oldie, it might be a sensible question to ask about this. Chances are, somebody was asked already. Somebody know an interview?

One difference between chess and go is the way to become professional. The harder it gets for kids to become pro, the stronger they will be.

I don't think there is the equivalent of insei classes for chess (at least in most countries).

_________________
If something sank it might be a treasure. And 2kyu advice is not necessarily Dan repertoire..

I think this is backwards. The better the players are, the stronger the competition is, the harder it is for kids to become pro.

_________________
Occupy Babel!

As far as I know, John coined the phrase when proposing a reason for successful performance by Korean players in international tournaments. I've explicitly stated that I dislike the term, but I think it simply resulted in repeated use. Nowadays, you can check any thread here related to time controls, and you'll probably find the term.

"Faster than appropriate" seems aligned with the usage on L19, without what I perceive to be a mocking tone.

The problem with both expressions are that they are generalized terms that gloss over the details of the actual time settings, players, and so on.

If John argued against the use of a particular time setting without adding a comical character (which I find to be mocking), I imagine that I would have much less contention toward these types of, in my opinion, disrespectful threads.

But alas, this discussion has gone on for years, and I expect it to continue for years to come (maybe until I die from high blood pressure).

_________________
be immersed

If you have high blood pressure I suggest you refrain from using Mickey Mouse time limits.

_________________
Still officially AGA 5d but I play so irregularly these days that I am probably only 3d or 4d over the board (but hopefully still 5d in terms of knowledge, theory and the ability to contribute).

According to the pros themselves, younger players are much better at reading, but older players have better intuition.

While not trying to compare myself to a pro I think that makes a lot of sense. Being definitely an older player (64 year old, playing for 43 years) I rely almost totally on intuition and find extensive reading to be much less enjoyable these days. Of course, this leads to gross blunders sometimes.

_________________
Still officially AGA 5d but I play so irregularly these days that I am probably only 3d or 4d over the board (but hopefully still 5d in terms of knowledge, theory and the ability to contribute).

Doesn't it simply suggest that chess is simply more complicated for humans, it takes them longer to learn all of its intricacies.

_________________
North Lecale

I don't think so. I think it shows that Go is in the middle of a period of rapid change. Much of the intuition that older players have developed no longer applies at the cutting edge level of top professionals. This seems to match what we have seen in recent developments in the game. A period of rapid change this late into the history of the game is a testament to the remarkable depth of go. It may be that the game will stabilize as computers become more proficient and we will see greater parity among younger and older players. (Or perhaps there are alternate explanations, and Go will continue to be a game dominated by the young.)

One other explanation is that go simply is a much longer game and takes more endurance. Go games normally take a lot longer than Chess games, and many more moves are played. If we assume that the same amount of energy is spent per move by top professionals, whether its go or chess, then a game of go will require a lot more energy.

Perhaps for chess, older players still have enough endurance for a game. But for go, the difference in energy between younger and older players becomes more obvious.

I remember when I was in high school, a family friend who's in his 50s talk about how his strength (like phsyical strength, not go playing) isn't that much weaker than when he was young in absolute terms. It's just that he can't keep it up for nearly as long. Might apply to go as well.

I believe the record number of B+R/W+R results is due a behavioral shift in players choosing to resign in the endgame when behind by a few points rather than play it out to the end. You rarely see B/W+4.5 or higher anymore since they usually resign before it gets to that point.

I don't think a change in the style of go is responsible for either the increase in resignations or the shift in age. If you look at the distribution of ages in top 10 go players before the recent explosion of the depth of Chinese talent, it resembles that of chess. In the 70's and 80's in the Japanese era, you rarely had teenagers in the top 10. Even prodigies such as Cho Chikun and Kobayashi Koichi needed some time until their late teenage years to the early twenties to climb into the top 10. There was usually a 40 year old or two like Fujisawa Hideyuki in the 70's or Rin Kaiho or Otake Hideo later in the 80's. The age 20's and 30's usually dominated with most players generally peaking in the late 20's and typically maintaining most of their strength into their 30s.

Same pattern roughly held during the Korean era. Even when Lee Changho burst on the scene and reached #1 as a 15 year old, the rest of the top 10 was mostly guys in their 30's - Kobayashi, the two Chos, Nie Weiping, Masaki. Yoo and Ma were up and comers but already in their mid 20's by the time they established themselves in the top 10.

As recently as 2000, Cho Hunhyun was ranked #2 in the world at the age of 46. But things have changed quickly where 40 year old players just cannot compete at the highest level anymore. Cho Hunhyun was still one of the strongest in his 40's back then but in today's game Lee Changho can't even crack the top 50. Lee Sedol is the only player in his 30's left in the top 10 and even he is starting to fade. So why the sudden shift in age profile among the top go players?

I think there are two inter-connected reasons. The first is the depth of competition. When the talent pool was shallower, an exceptional talent such as the Chos or Kobayashi can still remain in the top 10 even if their reading ability declines slightly with age. In today's environment, the competition is so fierce that even the slightest slip in form will knock you out of the top 10. Even guys in their 20's who have won international titles like Kim Jiseok or Jiang Weijie have fallen out of the top 10 with just a little slump. Top 10 in the old days were fairly static because the number of people with that level of talent was far smaller. Now, the separation between 10 and 20 is almost non-existent and the pool of hungry talented players looking to leapfrog is a hundred deep.

The second reason is internet go. Back in the day, you had much fewer opportunities to play against quality opponents and hone your skills. Even a player with the best possible environment such as Lee Changho, he had the benefit that he could play his teacher Cho, but how many games would he get with Cho? Since a teacher would only give a limited number of games to a student, the quality of the rest of the games were limited by the quality of other students at the go school that was physically there and looking for a game at the same time as you. These days, a skilled young pro can basically play against top competition 24/7. A kid like Shin Jinseo has probably played more high level practice games at 16 years of age as many other Go pros did by their 20's in the pre internet days. Look at the all the tygem games between Ke Jie and Park Junghwan. That kind of play improves both their play and lifts the standard of play as a whole. Those kind of practice games between the top pros would be so much harder if they had to be physically present in the same place. I would compare it to how talented Asian footballers become so much better once they move to quality European teams and train/play with the best talent.

So these two reasons are inter-connected because it creates a feedback loop. China's exploding middle class expands the talent pool for professional go players exponentially, and the internet allows all that talent to constantly play each other and improve at a quicker pace.

Yes, it's a bit sad that we no longer get to see the wise old veterans compete on the top stage as in past eras. But that is in my opinion, a blessing for us to be able to be spectators to a historic game that is continuing to evolve and reach higher and higher levels.

Great post - and agree with your analysis of these trends.

But I do wonder about this:

China's exploding middle class expands the talent pool for professional go players exponentially...

Will the rise of AI in the Go world disrupt this? Perhaps being a go professional will soon be seen as a much less stable career to aim at, and wont be supported by families in the way it has been in recent years.

Brooklyn: Your ideas are interesting, thank you, and possibly correct, but I do have some reservations:

1. To posit a change in behaviour I think you need to offer a reason for it. Maybe the rise of tv go is enough of an explanation, maybe the increasing number of games under Chinese rules (tedious counting) is a factor. But I would have thought that if you use reasons like that you'd have to ask why games under slow time limits, where counting can be accurate, are more often played out.

2. Not sure about calling the likes of Kobayashi and Cho Chikun in evidence works. Young players were artificially kept down then through giving seeded places in higher levels of preliminaries on the basis of dan grades.

3. Your arguments on the internet make sense, as far as I can see, and they have also been made to explain the quick rise of Magnus Carlsen, but none of that seems to explain the discrepancy in age profiles between chess and go (which was my main question).

Although javaness's suggestion got shot down quickly, I'm attracted by it enough to reserve judgement - there has been a bit of an argument recently on a chess.com site about pattern recognition in chess (hard or easy, possible or delusional?). But it could be that pattern recognition is both more important in chess than go, and at the same time is easier to shape into a useful tool because of the different textures of the game positions.

On the question of what effect AlphaGo will have on the popularity of go, I'm sanguine except in the case of Korea. I'm old enough to remember the time when computers first became prominent and the consequent death of mathematics as a profession was predicted (admittedly among the ranks of not too well informed people like careers advisors). In chess, computers seem to have boosted interest in the game. In Japan, despite almost fanatical resistance in some shogi pro quarters, shogi computers are now being taken in their stride. I expect to see that in the case of go.

The reason I think Korea may be different is that the recent rise of go was to a large degree driven by a worthy kind of nationalism - both a "let's beat the Japanese" urge and also a small-nation pride in becoming a world leader with its telecoms superhighway and being part of being a world tiger economy (translating into early use of the internet and sponsorship opportunities). On past experience, though, I would have to question whether go will retain its popularity there if the nationalist impulse wears off. AlphaGo may contribute to that. It seems strange that, for all its telecoms and computer skills and resources, Korea has not really kept up with AI go.

If you watch current games regularly, this is very obvious. It is almost standard for these days for a game with a difference of only 4.5 to get resigned. Kang Dongyun is well known for almost always resigning if he's losing rather than finishing it out. I remember being very surprised to see a game record where he lost by 3.5 as it was the first time I saw him finish out a game where he lost by more than 0.5 or 1.5. This is not a case of me stating an opinion, this is very well accepted fact.



I am using ELO ratings so yes this evidence is applicable.



Again, I totally disagree with your opinion. Korean nationalism is fierce in many competitive arenas but to attribute the rise of Korean go "to a large degree" by a motivation to "beat the Japanese" is misguided. Go was hugely popular as a recreational pastime in Korea before Cho Hunhyun's international titles and Lee Changho's rise. What sparked Korea's dominance in international go from the early 1990's was a developing middle class that expanded the talent pool for kids that have the talent, interest, and most importantly the resources to devote full time study to try and turn pro. While Koreans may watch sporting events against Japan with passion, the motivation to beat the Japanese has not been even a tiny factor in Koreans becoming professional go players. They became professional go players due to a love of the game and the desire to become successful, period.

In fact, Korea is going on the same trajectory as Japan in that the youth of today as a whole have many other forms of entertainment and few are picking up go. Smart young kids that have the potential to become professionals at go are more likely playing starcraft or other e-sports and not even bothering to give go a shot. Go is considered an old man's game in Korea and the thinning out of young talent that happened in Japanese Go is well underway in Korea as well. This trend started years ago way before AlphaGo.

Whose ELO ratings? Are you sure they're accurate for that time period?

_________________
Occupy Babel!

As I pointed out in my previous post, the discrepancy in age profiles between go and chess is a fairly recent trend. So logically, the reason for the recent discrepancy will likely be due to a recent change in the go scene. My strong suspicion is due to the increase in competition in the go scene.

For arguments sake, let's break down skill in chess/go to two factors - experience and reading ability. As experience goes up, reading ability generally goes down. Internet go has made it easier for kids to get a whole lot of quality experience in a short period of time so that has shifted the age younger. If you think about the once in a generation type young go prodigies, Cho Chikun, Lee Changho. They were both identified as potential stars at the age of 5 or 6, and apprenticed to top pros where they were able to devote full time study. Compare to a player such as Yoo Changhyuk. He had no teacher or formal training. In the pre internet days, the quality of his games were limited to strong amateurs at the local go club. He would be hard pressed to even get a few games against professional quality players until his teenage years. Yoo Changhyuk turned pro at 18 and started climbing the ranks in his early twenties. If he had the same environment as Lee Changho, he would've had a strong chance to turn pro in his early teens.

Let us also think about how many kids would be in the position to devote full time study with a professional from the age of 6. How many parents would even want that for their child? Yoo Changhyuk's path is much more common than Lee Changho's and Yoo in this day and age would have improved a lot faster with the internet and probably broken the top 10 in rankings as a teenager.

So then if chess has the same access to internet play as go, why has go ages skewed younger while chess has remained the same? Again, let's think about the factors that it takes to even turn pro. You generally need to have been raised in a developed country, or be the child of an elite in a poor country. To be able to devote the time needed, one generally needs the encouragement and support of parents from an early age. In the western world today, balanced lifestyles are encouraged. Very few parents in the US would encourage a 6 year old child to obsessively study chess at the cost of other interests and pursuits. The pro path of a chess player usually starts later and at a slower pace than in go.

With the sheer number of Chinese that are dedicating tremendous time and resources in go training at an early age, the pool has gotten very deep and I would posit that it is far deeper than serious young chess players in the west.