I did not win a game since 3 years


Back to forum

Stanislas Gounant    (2022-11-21)
I did not win a game since 3 years

Cela fait 15 ans que je joue sur ce site, assisté par ordinateur. Il y a 15 ans pour gagner une partie il fallait utiliser plusieurs programmes d'analyse suivant les phases de jeu. Je crois me souvenir que Hiarcs était meilleur en finale que les autres programmes. Il y a quelques années encore, il était possible de trouver des forteresses qui résistaient aux assauts de l'adversaire, même quand mon programme me disait que j'avais perdu. Mais cela fait 3 ans que je n'ai gagné que contre des joueurs qui jouaient sans l'aide de l'ordinateur ou qui se sont trompés de case en jouant leur coup. Ce que j'aimerais c'est qu'une solution soit trouvée pour redonner de l'intérêt au jeu par correspondance.

Thibault de Vassal    (2022-11-22 02:31:19)
I did not win a game since 3 years

Voilà un débat qui pourrait être intéressant...

Stanislas' message in english:

"I've been playing on this computer-assisted site for 15 years. Fifteen years ago to win a game you had to use several analysis programs depending on the game phases. I seem to remember that Hiarcs was better in the final than the other programs. Until a few years ago, it was possible to find fortresses that resisted the onslaught of the adversary, even when my program told me that I had lost. But it's been 3 years since I only won against players who played without the help of the computer or who made a mistake when playing their move. What I would like is that a solution be found to restore interest in the game by correspondence."

Thibault de Vassal    (2022-11-22 02:33:19)
I did not win a game since 3 years

By the way, the title would be:

"I did not win a game for 3 years"

Anyway, the topic is interesting... and tough!

Scott Ligon    (2022-11-22 14:49:30)
I did not win a game since 3 years

Correspondence chess with modern engines is a draw. The game isn't solved in the game theoretic sense, but from a practical standpoint it plays like a solved game. Anyone who runs the latest version of Stockfish on a computer with decent specs is unbeatable. That's been my experience anyway.

Scott Ligon    (2022-11-22 14:53:59)
I did not win a game since 3 years

As for ways to address this, if engines are allowed you could have tournaments from a starting position where the game theoretic value isn't clear. Where maybe it's a win for one side or the other, or maybe it's a draw, but it's right on the border and it's not obvious which side we're on. Then you play two games from that position against each opponent, once as white and once as black.

Scott Ligon    (2022-11-22 15:10:32)
I did not win a game since 3 years

In that case you'd need some way to address the possibility that a player could always draw their two game match by mirroring the moves their opponent plays in one game to the other game.

Scott Ligon    (2022-11-22 15:22:57)
I did not win a game since 3 years

Here's another idea. We start with two pools of starting positions. One pool where white has an advantage that may or may not be winning. Another pool where black has the advantage. First player picks a position from either pool. Let's say the first player picks from the white advantage pool (on the border between winning advantage for white and draw). Second player gets to choose which side of that position they want to play. They can either play white but then they have to win, or black but then all they need to do is draw. So no matter the outcome, each match will be decisive. We'd just have to populate the pool of opening positions first.

Scott Ligon    (2022-11-22 16:00:33)
I did not win a game since 3 years

Even the idea of having preset pools of starting positions isn't necessary. Going off my previous post, the first player selects any starting position they like (has to be a position that can be reached in an actual game of chess, so let's say the first player gives a sequence of opening moves that results in the position). First player wants to pick a position right on the border between win and draw, but they don't even need to stipulate which side is playing for the win, because a chess engine can run a quick search and determine that automatically. So the first player submits an opening sequence of moves, the FICGS server runs a quick evaluation with Stockfish or whatever just to decide which side has advantage, and the second player chooses between playing the advantaged side for a win or the disadvantaged side for a draw.

Scott Ligon    (2022-11-22 16:25:45)
I did not win a game since 3 years

Yet another amendment. From the previous post, we can eliminate the FICGS server evaluation and let the first player stipulate which side is playing for the win. If they assign the advantage incorrectly, this only helps their opponent, so the first player has no reason to lie. Example: First player picks the King's Bishop Gambit as the starting position (1. e4 e5 2. f4 exf4 3. Bc4) but erroneously claims that white has the advantage, so the second player has to choose between playing white for the win or black for the draw. Second player happily chooses to play black for the draw and should have no trouble holding the draw.

Stanislas Gounant    (2022-11-23 20:17:39)
I did not win a game since 3 years

The problem with the starting positions in FICGS thematic tournaments, one of the player can open the position and it will be draw. Some years ago, someone show me a youtube video about a game played on TCEC between leela chess zero and Stockfish.
The engines start to play at move 7, white had more space and black can't open the position

Scott Ligon    (2022-11-23 21:22:50)
I did not win a game since 3 years

Surely there are some positions where it's difficult to determine, even with engine assistance, whether the position is a game theoretic win or a draw. And then playing that position out would be interesting. That's all my proposed variant would depend on. The first player looks for a position that would be interesting to play, and if they've done their job well, the second player has a difficult task in deciding whether they want to play the side with advantage for a win or the other side for a draw. I think this would work up until the point that chess is actually solved.

The starting position could be as simple as 1. g4 (a terrible first move of course). Maybe black has a forced win and maybe with careful play white can hold the draw. If I did a lot of Stockfish analysis the answer might become clear but with a quick analysis I'm not sure. But if I knew the answer either way for 1. g4, I could always look at other positions. At least this would be a game where the outcome isn't immediately obvious.

A. T. S. Broekhuizen    (2022-11-25 10:20:51)
I did not win a game since 3 years

I think you don't have to go as far as to play dubious variations, as long as the chosen variations (for a thematic tournament) still have some 'music' left in them. I also have two suggestions: the Rio gambit against the Berlin wall and the following transposition into the Rubinstein variation of the French defence: 1.e4 e6 2.d4 d5 3.Nc3 Nf6 4.Bg5 dxe4 5.Nxe4. Besides, lately I still have won a game against the King's gambit with black. But this is much harder than holding a draw with white, so I would not define success for white as a draw and for black as a win in this case.

A. T. S. Broekhuizen    (2022-11-25 10:22:11)
I did not win a game since 3 years

I think it would be best to let every player take an equal amount of games with every colour.

Scott Ligon    (2022-11-25 16:20:13)
I did not win a game since 3 years

The problem with the two opening you suggested is that they're too good. Neither side made a mistake, so the engines will hold the draw every time (barring human error). The King's Gambit is a step in the right direction for a thematic tournament (with engines), and as long as everyone has an equal number of whites and blacks that's fair enough. I still think it's too easy for white to draw in the KGA with Nf3. The KGA with Bc4 is a bit more treacherous though I've explored it enough to be confident that white holds the draw. Anyway, I do think the openings need to be somewhat dubious or it's too easy for engines to draw.

A. T. S. Broekhuizen    (2022-11-25 16:48:19)
I did not win a game since 3 years

In case of the openings I mentioned earlier, I think there is a possibility for white to hold on to an advantage. The problem is that with the superhuman playing strength of engines nowadays, one will have to study harder than before to keep up with the engine. But now, after I studied these variations in detail, it has become clear. In practice one nowadays will have to have the right approach before the game to the played opening, otherwise I think it is not doable anymore to find the way to any white advantage during the game. This can be frustrating, but also in OTB chess more and more engine designed defences are played, so it is prudent to study these "unbeatable" defences.

Scott Ligon    (2022-11-25 17:10:39)
I did not win a game since 3 years

What does it even mean to have an advantage? Engine analysis has changed my perspective on this issue. From a human perspective, we can say that white has an advantage at the start of the game, and the statistics support this. At the highest levels of human chess white wins more often than black, but it's more often a draw. So white has a slight advantage.

From the perspective of correspondence chess with modern engines, the advantage is shown to be an illusion. It's just a draw. The engine evaluation at the start might be +0.15 or whatever, but if both sides are using an engine and there's no severe time constraint, it doesn't mean anything. By move 20 or so of a competently played correspondence game the engine analysis will have converged to 0.00 and it will stay there for the rest of the game.

In the final analysis, there's no such thing as a slight advantage. Every position is either a forced checkmate for one side or the other, or it's a draw. Even modern engines haven't pushed things that far, but they're strong enough to obliterate our human concept of an advantage.

A. T. S. Broekhuizen    (2022-11-25 17:26:08)
I did not win a game since 3 years

I am aware of these issues, nevertheless I think I could give any black player a difficult correspondence game in these two variations.

A. T. S. Broekhuizen    (2022-11-26 09:36:07)
I did not win a game since 3 years

Btw, here is the link to the gsame where I won against a KG with Nf3: https://ficgs.com/user_page.php?page=viewer&game=137701&auto=1&flip=1

Scott Ligon    (2022-11-26 14:55:46)
I did not win a game since 3 years

I couldn't use the link but I found game number 137701. You won against the King's Gambit but your opponent played 3 d4. Which is actually the kind of position I'm talking about. Right on the edge of outright losing and I'm not sure if white is already over the edge. Anyway I haven't been brave enough to try that one. 3 Bc4 is as far as I go, at least for now.

Stanislas Gounant    (2022-11-27 21:46:50)
I did not win a game since 3 years

I think it's better to play a position with players can't open the center. But i'm ok to play a thematic tournament with black and white on bishop's gambit. Rated if it's possible Thibault

Stanislas Gounant    (2022-11-27 21:48:49)
I did not win a game since 3 years

But John said he had refuted the bishop's gambit

Scott Ligon    (2022-11-27 22:06:35)
I did not win a game since 3 years

If you mean John Shaw's book on the King's Gambit, I have that book and he did say that the Bishop's Gambit is refuted. But he qualifies what he means: "In this context I define the term 'refutation' as Black being better in all variations, not winning by force." Not much of a refutation. If you search my game history, I have played the Bishop's Gambit several times and I haven't lost. That's why I'm confident white can hold the draw.

I'd be willing to play in that thematic tournament, and I'd be very well prepared.

Stanislas Gounant    (2022-11-27 23:06:56)
I did not win a game since 3 years

Yes John Shaw, I forget to write is name

Ilmars Cirulis    (2022-11-29 19:50:02)
I did not win a game since 3 years

Heh, the discussion reminds my efforts to refute Traxler (with Bxf7). :D We had some thematic games where draw counted as loss for white.

Ilmars Cirulis    (2022-11-29 19:50:19)
I did not win a game since 3 years

That was gazillion years ago.

Thibault de Vassal    (2022-11-30 00:51:58)
I did not win a game since 3 years

I still believe that Traxler counter attack could bring really good matches, tournaments & championships :)

Thibault de Vassal    (2022-11-30 00:59:29)
I did not win a game since 3 years

Did anyone "calculate" what opening leads to the lowest rate of draws in engines tournaments? (or Stockfish vs. Stockfish)

Ilmars Cirulis    (2022-11-30 01:06:20)
I did not win a game since 3 years

> I still believe that Traxler counter attack could bring really good matches, tournaments & championships :)

I'm currently using my e-points for Big Chess matches, but when I get enough of them (Big Chess matches), I would like to play white against Traxler (for e-points, draw counts as loss for white). :D

Stanislas Gounant    (2022-11-30 19:55:08)
I did not win a game since 3 years

In my openning tree of 11 700 000 games with games of big database 2023 + correspondence database 2022 + games play on FICGS in 2022 :

Bishop's Gambit : 6123 Games

Traxler Counterattack : 4293 Games

Stanislas Gounant    (2022-11-30 20:00:22)
I did not win a game since 3 years

In my openning tree of 200 000 games with games played since 2019 and players rated 2400 and more in big datatbase 2023 + games played since 2019 and players rated 2300 and more in correspondence datatbase 2022 + games played in 2022 and players rated 2300 and more on FICGS :

Bishop's Gambit : 27 Games

Traxler Counterattack : 29 Games

Thibault de Vassal    (2022-11-30 22:12:10)
I did not win a game since 3 years

Of course, it is rarely played (and maybe you count thematic tournaments)... but the aim is to find complex lines that decrease the rate of draws, right?

Stanislas Gounant    (2022-11-30 23:08:17)
I did not win a game since 3 years

I think the meanning of this topic is now to find some positions where theory is unclear I think it's the case of the bishop's gambit, and for the traxler, i think there is not a lot of games because it is refuted

Ilmars Cirulis    (2022-12-03 21:58:05)
I did not win a game since 3 years

Maybe someone wants to play Bishop gambit thematic games/matches? I'm interested.

But not the fast time control... instead at least +1 day/move, if possible. To ensure greater quality of the games. :)

Right now I have almost no e-points, though. If that matters.

Ilmars Cirulis    (2022-12-04 02:06:52)
I did not win a game since 3 years

Let's solve King Gambit finally. :P

Stanislas Gounant    (2022-12-04 23:28:09)
I did not win a game since 3 years

Thibault, is it possible to have thematics tournaments, double round robin and in a starting position unbalanced ?