Is Improvement possible?

[ben.graff]

I buy a lot of Chess books. I even read some of them. I play many games and try and take the "learning," from the defeats and the further pointers from Fritz as to how the wins could have been more convincing. Yet the amount of time I spend thinking about and playing Chess does not seem to equate to actual improvement. So my question to the forum is, "Is improvement (for an adult player) actually possible?" If so what's the secret? If not, does it matter?

[Roger de Coverly]

So my question to the forum is, "Is improvement (for an adult player) actually possible?" If so what's the secret? If not, does it matter?

Better results are certainly possible, although does that count as improvement?

So look at the games you should have at least drawn and the winning of games that turned to draws. You may even wish to emulate titled players and specifically prepare for your opponents. Making sure you never come out of the opening with a lost position is another one. Improving your clock handling is an issue. Avoiding spending a long time in the early stages of the game on decisions that doesn't make much difference to your position can help in games with faster time rates. So provided you know that the normal moves are x,y and z, play one of them quickly.

[Eamonn O'Molloy]

Yes, of course it is possible. Here's one extreme example: http://www.ecfgrading.org.uk/?ref=174270L

[Joey Stewart]

Here is a less extreme example then Jon Hawkins - I didnt start playing until I was 20 and this is my rating progress (I am 30 now)

http://www.ecfgrading.org.uk/?ref=240800E

[ben.graff]

Wow Joey - that is amazing improvement.

What was your "secret?" Did you have a training plan/ strategy to improve?

[Joey Stewart]

I found that i got better after starting a full time job - chess became a luxury I could not play anytime I wanted, so it made me take each game more seriously. Also I did (and still do) tactical puzzles online each break and lunchtime, which seemed to help improve my gameplay, even though I tend not to be overly aggressive myself it allows me to see the attcking possibilities of my opponents and better prepare my defences to outlast them (usually after nicking a dodgy pawn or something...)

[Joey Stewart]

Actually, I have just noticed an interesting thing - all those big jumps of improvement of mine came after I had just finished with a long term girlfriend (2004 and 2008 were where I found I was playing at a noticably higher level then before) so maybe there is something in it....

[Andrew Collins]

I like to think that improvement is possible at any age,but that perhaps the older you get, the comparatively more effort you have to put in to improve. I also think it depends on how you want to quantify your improvement, will you see massive jumps in your rating over time, its possible, but of course not necessarily so. I also think it is easier to improve certain aspects of playing chess than others.

I personally find that when I study the endgame, I can take it in quite well, I can learn certain winning techniques and plans and understand quite a lot of concrete analysis. Just as an example I can win queen vs rook, but what does this show, technique yes, but how often is this going to actually benifit my game. It is improvment of course, but additional knowledge that will help in certain specific situations.

The two things in my opinion that should have a overall benifit to every game you play, is learning positional understanding, and better tactical calculation. I find learning general position understanding a lot harder and although I have books I look through and attempt to study, regarding the middle game, I stuggle much more to take a lot from it and don't feel like i'm improving much in this area. But I agree solving daily tactical puzzles, can only improve that area of your game.

Actually I struggle to stomach certain opening books as well, but I guess the point is overall improvement is a balance of all aspects of chess. Something I myself am attempting to work on.

[David Gilbert]

The evidence says .......


Age (Dordr). 2012 Aug;34(4):1001-9. Epub 2011 Jun 22.

Exponential growth combined with exponential decline explains lifetime performance evolution in individual and human species.
Berthelot G, Len S, Hellard P, Tafflet M, Guillaume M, Vollmer JC, Gager B, Quinquis L, Marc A, Toussaint JF.

Source
IRMES, INSEP, 11 avenue du Tremblay, 75012, Paris, France, [email protected].

Abstract
The physiological parameters characterizing human capacities (the ability to move, reproduce or perform tasks) evolve with ageing: performance is limited at birth, increases to a maximum and then decreases back to zero at the day of death. Physical and intellectual skills follow such a pattern. Here, we investigate the development of sport and chess performances during the lifetime at two different scales: the individual athletes' careers and the world record by age class in 25 Olympic sports events and in grandmaster chess players. For all data sets, a biphasic development of growth and decline is described by a simple model that accounts for 91.7% of the variance at the individual level and 98.5% of the variance at the species one. The age of performance peak is computed at 26.1 years old for the events studied (26.0 years old for track and field, 21.0 years old for swimming and 31.4 years old for chess). The two processes (growth and decline) are exponential and start at age zero. Both were previously demonstrated to happen in other human and non-human biological functions that evolve with age. They occur at the individual and species levels with a similar pattern, suggesting a scale invariance property.

[ben.graff]

All really interesting. So we've established that improvement is possible. Why is it then that most players don't really improve? Where are they going wrong in their approach to development? (I include myself in this category.)

Andy - whatever else we establish - try not to improve too much before our game on Tuesday!

[AustinElliott]

The evidence says .......

Age (Dordr). 2012 Aug;34(4):1001-9. Epub 2011 Jun 22.

Exponential growth combined with exponential decline explains lifetime performance evolution in individual and human species.
Berthelot G, Len S, Hellard P, Tafflet M, Guillaume M, Vollmer JC, Gager B, Quinquis L, Marc A, Toussaint JF.
Etc.

Wow. Must get hold of that one. Shall be interested to see if it predicts my playing strength on starting again at 50 after a 32 yr 'break'...

Actually, come to think of it, I guess a more interesting prediction is whether my rating should now be capable (hypothetically!) of improving, or whether it will trend inexorably downward as my ageing brain cells pop off.

More seriously, the problem with applying the analysis described in the abstract of the paper to "John Q Chessplayer" is that the population they derive the model from appears to be GMs. Who might not be all that representative of the average wood-chucker...

[David Robertson]

That is indeed the main methodological problem. The data addresses peak performance and decline at the species level, not at the sub-population level. So the cohort selected is the best the species can offer. But, by the nature of 'best', once you've reached that level, you only have one direction of travel available.

Some (possibly much) of this will apply to cohorts at lower levels of performance. But again, by the nature of 'lower level', there's more room 'above' to improve into. So the data may not apply with the same force at the sub-population level.

An equally (more?) fascinating dataset - I've posted it before in here - concerns age-related decline. The rate of decline over time for standard chess players is relatively slow. This might be compared with, say, pole-vaulting or 100m sprint where decline after a certain (youngish) age is quite steep. Chess players might lose 20% of their peak performance by age 70; and 50% by age 90. For pole-vaulting, the age 'steps' are something like 33 and 45.