Stylistic bias in computer chess programs

[Aser Huerga]

But I have heard people from other fields say that hard work is more important than natural ability ...

Genius is 1% talent and 99% percent hard work...
Albert Einstein

And the humorous version:
Genius is one per cent inspiration and ninety-nine per cent perspiration
Thomas Alva Edison

Inspiration exists, but it has to find us working
Pablo Picasso

And my favourite, this time related with those special abilities from the best chess players:

Logic will get you from A to B. Imagination will take you everywhere.
Albert Einstein

[Uri Blass]

I think correlation between time and success proves nothing and it is expected that people with more talent are encouraged to work harder.

If you want to investigate the problem you need to give motivation also for less talented players to work hard and practically it does not happen.

I read for example that Magnus Carlsen's sister left chess for 4 years after Carlsen beat her(so Carlsen worked harder than her but the reason is obvious and Carlsen simply had success when she failed by losing against a younger player or at least this was probably her feeling)

I guess that my opinion is somewhere in the middle between Don and Sam.

I believe that people can do better than what they do but I also think that Don is too optimistic if he thinks that any reasonably intelligent person could (in theory) become at least a low level master.

Based on my experience in teaching chess
I know that there are humans with visualization problems that simply do not see what is obvious for people who play in tournaments and the main question is if they are able even to get a level of 1400 with hard work.

I talk about players who can play chess by the rules but I beat without queen and 2 rooks.

They may calculate wrong and think that the opponent pawn can capture backward in their calculation or confuse between white and black in their thinking and avoid a move because they think that a friendly piece is going to capture them.

I do not know about one of these players who has a big motivation to try to improve in chess and it may be interesting if one of these people can get to a level of 1400 assuming that (s)he works hard in order to improve and get the right teacher.

[Don]

I think correlation between time and success proves nothing and it is expected that people with more talent are encouraged to work harder.

Yes, I thought of that too as a kind of chicken and egg thing. Do they work because they want to improve or do they discover they are good and are encouraged to work at it?

However, this correlation does prove your point either. I never claimed it was a proof of anything.

The principle of Occams razor makes me like my own theory a lot more than yours as it seems like a more direct and straightforward explanation. (Which of course does not make it automatically correct, but it does makes it the primary one to explore first.)

If you want to investigate the problem you need to give motivation also for less talented players to work hard and practically it does not happen.

You have no way of really knowing that. Once someone gets to the top you assume they are extremely talented and thus believe what you just said. But how do you know they were super talented? It's just an assertion, a circular argument.

Even in sports, where I think talent really does make a lot more difference, it's widely believed that some players have much less talent but worked much harder. Ivan Lendl in tennis is a classic example, they say he was ordinary and had NO backhand for a number of years but had one of the strongest work ethics of all.

I read for example that Magnus Carlsen's sister left chess for 4 years after Carlsen beat her(so Carlsen worked harder than her but the reason is obvious and Carlsen simply had success when she failed by losing against a younger player or at least this was probably her feeling)

I guess that my opinion is somewhere in the middle between Don and Sam.

I believe that people can do better than what they do but I also think that Don is too optimistic if he thinks that any reasonably intelligent person could (in theory) become at least a low level master.

Based on my experience in teaching chess
I know that there are humans with visualization problems that simply do not see what is obvious for people who play in tournaments and the main question is if they are able even to get a level of 1400 with hard work.

I talk about players who can play chess by the rules but I beat without queen and 2 rooks.

They may calculate wrong and think that the opponent pawn can capture backward in their calculation or confuse between white and black in their thinking and avoid a move because they think that a friendly piece is going to capture them.

I do not know about one of these players who has a big motivation to try to improve in chess and it may be interesting if one of these people can get to a level of 1400 assuming that (s)he works hard in order to improve and get the right teacher.

[JuLieN]

I think that you might all be right and under the false assumption that you disagree. Why? Because chess strength is probably distributed like a Bell curve (just like IQ, and probably like many other human fields). If we imagine that everyone would train in chess up to the point that they reach their maximal ability, we probably would get such a curve:

The question is: what would be Homer's level? I wouldn't be surprised if such a curve would be centered around 2200-2300 Elo. But to reach over 2600 you would have to get a superior amount of natural talent.
This explains why people can train themselves to a very high level, if they have the correct training method, and why only a few of us can really achieve very high success anyway. There will always exist some genetical limitations, just like not any kid can run as fast as Usain Bolt, however hard a work he puts in.
So Don's example of this player that reached 2300 after some work is probably a good example of an average chess player that trained himself up to his natural potential, that he will hardly be able to exceed. I doubt that he would be able to continue and become a 2800+ for instance, whatever the amount of time and dedication he puts into.

[Uri Blass]

I believe 2200-2300 elo is too optimistic for Homer.

I do not think that Don's example is good and I suspect that the player in his example has an high natural talent.

I do not think that you can find the natural talent of chess players based on their level without training and I think that there are players who have more talent to get advantage of training.

If you want to prove that 2200-2300 is realistic for the average then then we need some sponsor that give people money in order to work hard on chess(when the money that they get is not dependent on their result but only on the fact that they work hard) and see what is the percentage of them who get into the level of 2200-2300)

My guess is that Homer's level is 1800 and I think that it is more optimistic than what most people believe.

[Don]

I believe 2200-2300 elo is too optimistic for Homer.

I do not think that Don's example is good and I suspect that the player in his example has an high natural talent.

I do not think that you can find the natural talent of chess players based on their level without training and I think that there are players who have more talent to get advantage of training.

If you want to prove that 2200-2300 is realistic for the average then then we need some sponsor that give people money in order to work hard on chess(when the money that they get is not dependent on their result but only on the fact that they work hard) and see what is the percentage of them who get into the level of 2200-2300)

My guess is that Homer's level is 1800 and I think that it is more optimistic than what most people believe.

There are too many unknowns to really resolve this and as Marcel implied we don't even know for sure what we are agreeing or disagreeing on since the concepts we are talking about are so poorly defined. "Natural talent" is a pretty abstract and ill-defined concept and probably has many aspects.

After thinking some more on this, I have to say that I believe "motivation" is a huge part of the equation. Everyone wants to be a better player, but some people are really motivated or driven to succeed and others put varying degrees of effort into it. Only the smallest fraction of people are seriously dedicated to improving but there are various degrees of dedication too.

[PK]

I think that there is the third factor beside natural ability and training, namely self-assesment ot "aha-factor". I'm in my thirties now, and last week I secured the second required norm for the so-called 1st cathegory (an Eastern European grade below the candidate master). It took approximately 4 years since my first tournament after a long hiatus. This looks reasonable (no great talent, some work, recently mainly tactical puzzles at chesstempo.com, no external training).

The triggering factor might have occured during two holiday tournaments, finished with serious Elo losses. The first one abounded in opening disasters, which caused me to switch to much more solid defences as Black. During the second one I cracked physically after three tense encounters with the candidate masters (scoring 50% against them, but then losing 3 games in the row).

The point is that this set of 18 games thaught me a couple of important things: taking less risks (why fight a difficult defensive battle in the Pirc against a junior player, if I can play old fashioned Ruy Lopez instead), insistence (getting a defensible endgame against a stronger player is actually easier than holding it), allowing the opponent to make his plans and his mistakes ( it's often easier for the opponent to outwit himself/herself than for me to cause it), trying to be creative only in the right frame of mind (i.e. going for an unclear exchange sacrifice only when one feels like going all out afterwards), doing a lot of easy tactical puzzles (complex combinations are built out of easy ones).

Now, most of it is about attitude, not about talent or work. If somebody told me to play like that earlier, I would consider it boring. The fact is that I started to enjoy playing like that, and better luck/risk assesment added a whole new dimension to the game. And I still can pull stunts like seeing an interesting Dragon Sicilian game in the round robin, played by my main contenders, and then... play Dragon, which is not my opening, against both of them, one as black and one as white.

This is not universal improvement program - but probably every player that had not had a good coach has some reserves which can be put to use due to some insight ("aha"). My crude estimate is that normal amount of talent + amount of work one can put into chess as a hobby can bring You to roughly 1900-2000 Elo, but then there are another 100, perhaps even 200 Elo in some cases, to be gained due to "aha factor". And this amount depends on the number of bad habits one trained himself into.