@tpr said in #4:
The better player wins more than the weaker player.
The opening that the better player happens to play has a higher win margin.
If the better player happens to play another opening, then that opening gets a higher win margin.
The win margin of the Scotch was low, until Kasparov took it up and other strong players followed.
The win margin of the Berlin was low, until Kramnik picked it up and other strong players followed.
The win margin of the London was low, until Kramnik, Jobava and other strong players picked it up.
The win margin of the Giuoco Piano or Italian was low, until Caruana picked it up and other strong players followed.
Win margin just tells what is fashionable among strong players.
@tpr said in #2:
Win margin tells nothing about an opening, it tells something about the players playing it.
Your conclusion is very obviously mistaken. If Carlsen plays the Grob against other top players and loses it isn't just because he "happens to be a worse player" it absolutely is saying something about the opening.
The question is what should we do with that. We can very likely use information surrounding win rates to gauge to some degree the ease of playing the position and practical chances for each side. There are positions where one side can be objectively worse but practically much better and this can occur in openings though usually toned down.
To challenge your conclusion, look at the masters database in the Portuguese Gambit (1. e4 d5 2. exd5 Nf6 3. d4 Bb5!? 4. f3! Bf5. Even at the masters level, even despite Black of course starting with a disadvantage, despite Black playing an (as far as the computer is concerned) inferior opening, and despite Black playing a potentially dubious gambit within all of that, Black maintains winrates nearby Whites in the gambit. And to your conclusion that simply Black must be the better players, White is on average about 50 rating points higher than Black (average of ~2400 vs ~2350) and in White's response with highest rating average 5. c4 Black has a higher winrate than White despite the Black players being lower rated and the computer giving White an advantage.
All of this to say, winrates can help give us some information as far as what works. It doesn't tell us necessarily what we should do (playing an opening that isn't really a gambit and is rather based on a trap/traps for example will stunt player development.)
@tpr said in #4:
> The better player wins more than the weaker player.
> The opening that the better player happens to play has a higher win margin.
> If the better player happens to play another opening, then that opening gets a higher win margin.
> The win margin of the Scotch was low, until Kasparov took it up and other strong players followed.
> The win margin of the Berlin was low, until Kramnik picked it up and other strong players followed.
> The win margin of the London was low, until Kramnik, Jobava and other strong players picked it up.
> The win margin of the Giuoco Piano or Italian was low, until Caruana picked it up and other strong players followed.
> Win margin just tells what is fashionable among strong players.
@tpr said in #2:
> Win margin tells nothing about an opening, it tells something about the players playing it.
Your conclusion is very obviously mistaken. If Carlsen plays the Grob against other top players and loses it isn't just because he "happens to be a worse player" it absolutely is saying something about the opening.
The question is what should we do with that. We can very likely use information surrounding win rates to gauge to some degree the ease of playing the position and practical chances for each side. There are positions where one side can be objectively worse but practically much better and this can occur in openings though usually toned down.
To challenge your conclusion, look at the masters database in the Portuguese Gambit (1. e4 d5 2. exd5 Nf6 3. d4 Bb5!? 4. f3! Bf5. Even at the masters level, even despite Black of course starting with a disadvantage, despite Black playing an (as far as the computer is concerned) inferior opening, and despite Black playing a potentially dubious gambit within all of that, Black maintains winrates nearby Whites in the gambit. And to your conclusion that simply Black must be the better players, White is on average about 50 rating points higher than Black (average of ~2400 vs ~2350) and in White's response with highest rating average 5. c4 Black has a higher winrate than White despite the Black players being lower rated and the computer giving White an advantage.
All of this to say, winrates can help give us some information as far as what works. It doesn't tell us necessarily what we should do (playing an opening that isn't really a gambit and is rather based on a trap/traps for example will stunt player development.)