Xiphos x Ethereal: Who has the best style?

[BrendanJNorman]

People attribute fantastic "style" to the engines they happen to like.

Wrong. It is in fact the reverse.

People LIKE the engines with a fantastic style.

They invent a "style" out of a subset of the games they see.

I guess I'll just use Ziggurat from now on then. No difference at all!

Time and time again it has been shown that people cant distinguish games played by different engines like SF and Leela.

This has nothing to do with engines, but the players observing, themselves.

The truth is that most players (statistically) are only around 1500 Elo.

At this level, they also couldn't tell the difference between Tal or Morphy or Petrosian or whomever.

They see a positional player do a knight fork and call the guy a tactical genius.

In a very well-played game, even a strong player might not recognize Tal's play.

This occurs also when observing engine games.

It doesn't mean the engine doesn't have a style.

Engines don't have a style. They have a raw calculation. Nothing more.

If this is the case, why don't we all just use a Stockfish dev/Leela combo and give up on this silly hobby?

[BrendanJNorman]

Long live the 'cannon fodder' engines! That's exactly what I do when I need a quick assessment of engine style.

It is the simplest, most effective way.

Mike Tyson looked most brilliant when he was knocking out bums, not when he was grinding against Lennox Lewis.

That's where we saw his style in its true glory.

Same with Kasparov/Tal in chess. They looked greatest when sacking pieces against a guy 200 points weaker.

Finding out who is stronger is one thing, where you want engines to be as closely matched as possible, but the identifying style is a totally different goal from this.

[matejst]

Brendan,

Tal looked great also against Smislov, Panno, Spassky, Fisher, Botvinnik...

But back to the topic... I do not know if I would call it "style", but side-by-side analysis allows you to spot differences among engines. Then, watching matches between some weaker engines I like and the beast rated 3200+, games are decided most often at the end of the middlegame when an advantage in search depth of 20+ plies makes all the difference. In the opening, despite Ed (Shroeder)'s results, I find the new engines, using Texel tuning, in general weaker. After the moves 1. d4 Nf6 2. c4 e6 3. Nf3 Bb4+ 4. Bd2 Qe7 5. g3 Nc6 engines like Xiphos, Ethereal, will chose 6.e3 even at depths 20+, Wasp 4.0 too (when Wasp 1.02 would chose Sc3).

Texel tuning-like methods brought something new, but they have their limits, and, without improvement and evolution, I think they are detrimental for further improving engines. I would add -- most modern engines play a similar ugly chess, and it is especially noticeable in well analyzed quiet opening positions. Engine moves often makes no sense. One could say that engine moves make no sense to me -- but when I switch to SugaR NN, Winter, RubiChess (although Rubi was also trained with the TexTM; I would be very interested to hear from the author if he tuned his engine by hand too), suddenly they chose the moves from opening books, or moves with plans I understand from similar positions.

For me, it is clear now that the future is a mixture of nets and alpha/beta, with learning/training abilities, a right balance between search and evaluation. So far, the evaluation is lagging behind.

[AndrewGrant]

Wrong. It is in fact the reverse.

There is a reason why GMs and super GMs typically don't care for engines in the slightest.
Its not because they are ego tripping. Its because engines don't play in an interesting way.

[BrendanJNorman]

Brendan,

Tal looked great also against Smislov, Panno, Spassky, Fisher, Botvinnik...

Hey Boban, long time no talk, my friend...

Of course I know Tal also looked great against those guys...but I am talking about extremes.

Observe this game for example...

[pgn][Event "Simul"]
[Site "Chicago, IL USA"]
[Date "1988.??.??"]
[EventDate "?"]
[Round "?"]
[Result "1-0"]
[White "Mikhail Tal"]
[Black "Jack Miller"]
[ECO "C55"]
[WhiteElo "?"]
[BlackElo "?"]
[PlyCount "65"]

1.e4 e5 2.Nf3 Nc6 3.Bc4 Nf6 4.d4 d6 5.dxe5 Nxe4 6.Bxf7+ Kxf7
7.Qd5+ Be6 8.Qxe4 Be7 9.O-O d5 10.Qd3 Qd7 11.Re1 Raf8 12.Nc3
Ke8 13.Ng5 Bc5 14.Nxe6 Bxf2+ 15.Kh1 Bxe1 16.Nxf8 Rxf8 17.Bg5
Nb4 18.Qe2 Nxc2 19.e6 Qd6 20.Nb5 Qe5 21.h4 Qg3 22.Rd1 Rf2
23.Qxf2 Bxf2 24.Rxd5 Qxh4+ 25.Bxh4 Bxh4 26.Nxc7+ Kf8 27.Rf5+
Bf6 28.Rd5 a5 29.Rd7 Nb4 30.Rf7+ Kg8 31.Rxf6 Nc6 32.Rf7 g6
33.e7 1-0[/pgn]

Tal did not play such sacrifice-filled, magician-like games against elite opponents.

Nice games, but not amazing like this...their defensive level is too high.

Same goes for engines.

But back to the topic... I do not know if I would call it "style", but side-by-side analysis allows you to spot differences among engines. Then, watching matches between some weaker engines I like and the beast rated 3200+, games are decided most often at the end of the middlegame when an advantage in search depth of 20+ plies makes all the difference. In the opening, despite Ed (Shroeder)'s results, I find the new engines, using Texel tuning, in general weaker. After the moves 1. d4 Nf6 2. c4 e6 3. Nf3 Bb4+ 4. Bd2 Qe7 5. g3 Nc6 engines like Xiphos, Ethereal, will chose 6.e3 even at depths 20+, Wasp 4.0 too (when Wasp 1.02 would chose Sc3).

Texel tuning-like methods brought something new, but they have their limits, and, without improvement and evolution, I think they are detrimental for further improving engines. I would add -- most modern engines play a similar ugly chess, and it is especially noticeable in well analyzed quiet opening positions. Engine moves often makes no sense. One could say that engine moves make no sense to me -- but when I switch to SugaR NN, Winter, RubiChess (although Rubi was also trained with the TexTM; I would be very interested to hear from the author if he tuned his engine by hand too), suddenly they chose the moves from opening books, or moves with plans I understand from similar positions.

For me, it is clear now that the future is a mixture of nets and alpha/beta, with learning/training abilities, a right balance between search and evaluation. So far, the evaluation is lagging behind.

I agree for the most part here too. Modern engines play super strong, but boring and weird chess.

Although I do like messing around with a few of these Leela nets.

I also still prefer Wasp 1.02 as well.

[BrendanJNorman]

Wrong. It is in fact the reverse.

There is a reason why GMs and super GMs typically don't care for engines in the slightest.
Its not because they are ego-tripping. Its because engines don't play in an interesting way.

Actually...your thinking is flawed here again too.

1. Most GMs are strong enough not to need the engines eval and simply use it for "blunder checking"...this is in fact, a humble form of ego-tripping. Sometimes they poo poo the engine suggestion, even when the engine is right, because their ego is getting in the way of truth-finding.

Their "classical chess education" refuses to believe that a certain type of position might be playable. Lots of ego-tripping going on with GMs re: computers. Nothing against the GMS, but this is a fact that even Vishy Anand has admitted to.

2. Also...If engines do not have style...

Why does Lev Aronian prefer an engine from years ago because as he says "it moves match his style" ...

and why did 2650 Elo GM Lenderman when asked to create a course for iChess, fill it with beautiful, stylistic examples of engine play?

https://www.ichess.net/shop/chess-gambi ... lenderman/

I know why...and it has nothing to do with engines "not playing interesting chess".

[Ovyron]

Engines don't have a style. They have a raw calculation. Nothing more.

Andrew, I urge you to check the linked thread for comparison. I stopped at 10 games, but if I continued for another 10, Fury and Lunatic personalities would show some amazing chess in their games! Even Xiphos or Ethereal would show amazing chess because only one engine requires style, so "neutral style" engines like Ethereal would still produce firework games against others (other "anti-style" engines like Naum 3 could probably dull the games against them).

Also, saying engines don't have a style is like saying no chess player has a style, since the Turing test has been passed by chess engines, and, unless you have memorized all the games of Karpov, you wouldn't be able to tell the difference between the style of a Rodent Karpov chess game and a real Karpov chess game, you'd either have to say "Karpov didn't have a playing style" or "engines have a playing style", since Rodent is playing with his style.

Finally, I think I could clearly tell who was who on a blind test where a group of games were played between Xiphos and Ethereal, but their identities were concealed. Ethereal's obsession with doubling pawns was apparent, as was Xiphos's disregard of it. In the elo world this could bring more elo to Ethereal. In the style world I make Ethereal weaker so the levels are matched, and this doubling of pawns difference does nothing but showcase their style differences.

[AndrewGrant]

Why does Lev Aronian prefer an engine from years ago because as he says "it moves match his style" ...

Well that is an easy one. Because the engine from years ago is far weaker.

[Daniel]

That Tal game always seemed funny to me. Tal had a sense of humor that was like what we call trolling. He plays e7 with the sole intention of sacrificing his rook; it’s totally unnecessary because he can play Ne8 Ø Nf6+ Kh8 Rxh7++, but he still did it because he wanted to keep sac'ing.

As far as style is concerned, it exists. One can call it what he wants — style, personality, algorithm — but it’s there. I’m going to study it more in depth.

[BrendanJNorman]

Why does Lev Aronian prefer an engine from years ago because as he says "it moves match his style" ...

Well that is an easy one. Because the engine from years ago is far weaker.

If you say so Andrew, I'm not gonna waste my morning arguing.

But isn't it interesting...

Komodo, the engine designed by a human GM is known to have a more "human-like" playing style.

Shashchess, designed to emulate 3 GMs, notably Tal, has a very dynamic playing style.

And Ethereal - the engine made by a programmer who says engines don't have a style - plays boring (albeit super strong) chess.

Once you pass 3000 Elo or so, strength isn't everything bro...unless of course, that's all you're aiming for...

Would be a shame not to work also on style, in addition to strength...

...but you'd have to admit that style exists first.

You do you though, you're obviously doing a great job at the strength side of things and I did see a few more nice games from Ethereal in version 12.0.