how does "guess the move" / reviewing GM games compare to actually playing

[rockyrook]

was just curious to see what others' opinion of "guess the move" and reviewing GM games compares with actually playing against other people.

is this an apples vs oranges discussion in that the two are not comparable?

or do you think going over GM games is better than than actual play?

or does actually playing other people trump going over GM games?

and when i say "better", i'm talking in the context of trying to improve - actually get better at chess.

... discuss ...

[dfan]

Playing guess the move (aka solitaire chess) is a good way to learn actively (only if you're really trying and not just guessing), and as such is generally more productive than just reading through a book. Playing actual serious games is the most active form of learning of all, though, provided that you try to learn from your games afterwards.

[andreacoda]

I think using the GTM feature is one of the best way to improve, if done PROPERLY, for the following reasons:
- Rather than ending up playing your "usual" moves, you are forcing yourself to understand why a Master played a specific move, and this way you learn new plans
- The feedback you get on your actual move is excellent, because whenever you don't play the correct move, you will know if you played an acceptable alt (which is still very good), or a slightly inferior move (which is still ok, we are not GMs!), or a blunder (which is not good - then you should assess why you played that move, and try to understand why you didn't see why it's wrong)
- Going through a master game gives you the chance to see, in just one game, all the phases (opening, middle game and endgame), so it is a complete training
- With the statistics the site is providing, you can see if over time you are improving, or not, and reviewing your games, you can try and understand where your weaknesses are
- Finally, this kind of stretching is really stressing the concept of "active learning" - you only get what you put in. If you just skim through a master game from a book, it's unlikely you will get any real value out of that. With GTM, done seriously at every move, you don't have this problem.

[rockyrook]

thanks dfan and andreacoda for your thoughts.

so to summarize, what i heard is that solitaire chess is just as good (assuming the player really is trying to guess the best move and is seeking to understand the GM moves) as actual play.

in my mind, the feedback from solitaire chess occurs much more quickly, since in actual play, you have to finish the game, perform the post-mortem, analyze, annotate, etc.  whereas in solitaire chess, you are getting an immediate education.

[richard]

Hi rockyrook,

I think another way of looking at the two may be that reviewing master games is showing you what good moves look like in the context of a coherent master level game plan. While you can analyse your own game afterwards to see where you've gone wrong, you are not really emedded within a master level game from start to finish like you are when reviewing a master game. Of course reviewing your own games has the advantage that it is very specific to the current weaknesses in your game, and probably easier to relate to - some of the concepts in master games can be hard to fathom sometimes (at least for me :-) ).

Regards,
Richard.

[rockyrook]

thanks richard - that makes very good sense!

[asterion]

Playing is the very best form of practice, especially if it's long OTB games and you review them with strong players.

GTM is a good training as it somewhat emulates the conditions of OTB play. But given the choice, I would pick OTB long games at any time.

[hanelsker]

The essence of the guess-the-move experience is to grasp the totality of the chess game from start to end. The point is to be able to pick up patterns of thought that relate to whole scenarios of how games develop. Each game is like a coherent story with an introduction, a building-up of the tension along certain lines and a culmination. This phenomenon is spanned across the whole game and we need to be able to appreciate this aspect of playing chess when doing this exercise.

[fulmine]

I believe GTM training works if:
1. The player doing it already studied some middlegame, and endgame books. So he/she knows how to evaluate a position, and how to find the right candidate moves. (I don't believe a player at 1200-1500 level can really benefit, when he/she cannot sense most combinations, or doesn't know many endgames)
However, I believe at a certain point it will become clear, that if the player is not at GM level, and the games are at GM level, then that player cannot have the same grasp and understanding of the middlegame, that a GM has, consequently I believe there will be a 50% maximum success no matter how much time one think upon a move.
I have trained some young players, and I've noticed that time is not the "factor" in finding the right move in a certain position, but the right way of thinking is.

2. If the game collection is not pointless. For example, let's say that in order to have an understanding of a certain historical period, we need to examine a minimum of 50 games (maybe the number is higher), then we would need likely 50 games by Steinitz, then 50 by Lasker, then 50 by Capablanca, and so on. And through such exercise become able to distinguish the various features synthesized by that player, who in turn synthesize the period in which he was living. We would also see how opening ideas changed from one period to another, and how some openings were re-evaluated.

3. Is GTM more Active learning? Again, this is a difference of learning styles. If when one watches a game from a book, he knows how to look for some pawn structures, or catalogs the tactics, or the positional themes, there is no such big difference. GMs, before computers, would use a file system for cataloging the position they saw in opponents' games. They would catalogue the openings, try to find surprises/novelties, see in which stage of the game the opponent would err more easily, and so on. I believe the Informator was the publication used for such work.

4. Another point which makes me curious is the actual benefit from such exercise. I estimate it could give a 0.2 elo points for game studied/guessed. Which means that if a player makes all the 200 games, the improvement will be 40 points in his/her playing strength.
Of course this will be demonstrated if we have a random sample of players interested in doing such exercise, and record their Elo rating before and after, without doing anything else.

5. I also believe OTB games are really important, because it is where everyone gives the best, especially when the entry fee at the tournament is high. But especially if the player consistently goes to play in tournaments, where the opponents are at least 1-200 points higher than him/her.