Emulating player's styles

Discuss anything you like about chess related matters in this forum.
stevencarr

Emulating player's styles

Post by stevencarr » Sun Apr 13, 2014 8:52 am

When I was young, I naturally wanted to play like Tal. So I had Modern Benoni's and sacs against the Sicilian in my repertoire.

There was only one snag about trying to emulate Tal.

I wasn't Tal.

I would also like to play like Karpov or Adams. Adams has such a smooth style, so admirable.

But I can't play like them.

I've found in the past year that the player I should be emulating is Simon Williams.

With all due respect to Simon Williams, who is a far better player than I, his style is more suitable to players like me than the games of Tal, Karpov or Adams. I don't have to be a genius, which is one of the requirements of trying to play like Tal. I just have to be a bit of a nutcase to play like Williams - again, no disrespect intended.

So do other players have role models to try to model their chess on?

MartinCarpenter
Posts: 3048
Joined: Tue May 24, 2011 10:58 am

Re: Emulating player's styles

Post by MartinCarpenter » Sun Apr 13, 2014 9:58 am

Mostly content with managing to emulate a coherent version of myself :) Hard enough a lot of the time!

Barry Sandercock
Posts: 1356
Joined: Tue Sep 11, 2012 10:52 am

Re: Emulating player's styles

Post by Barry Sandercock » Sun Apr 13, 2014 10:34 am

I try to emulate any players who make a habit of winning. However, at 83 it is not so easy !

David Robertson

Re: Emulating player's styles

Post by David Robertson » Sun Apr 13, 2014 10:36 am

I'm modelling my style on Houdini.

It's work in progress

Craig Pritchett
Posts: 91
Joined: Mon Nov 29, 2010 7:54 pm

Re: Emulating player's styles

Post by Craig Pritchett » Sun Apr 13, 2014 12:17 pm

Keen to hear about Houdini's take on "style". That style has some real meaning in chess is clear, at least in my view, but it's hard to define, pin down to any actual chess-player and is chimeric. In the introduction to my recent Everyman Chess book "Chess Secrets: Great Chess Romantics", I quote the late British champion (and Scottish lawyer), R.F. Combe, "who once wrote interestingly on this theme in “The British Chess Magazine”, in 1948":

“In chess I was soon faced with a crucial problem, should I play like Capablanca, or … in the style of Alekhine? Or again, was the teaching of Tarrasch the true faith, or should I follow the eclectic Lasker? … It was not so bad when I got through a whole game as a disciple of Lasker, but often I would experience a change of faith in the course of a single game. Unknown to my opponent, Lasker would get up from my chair … leaving Capablanca to take his place ... all very disconcerting.”

Following on from this, I go on to suggest that "Combe’s words neatly emphasise that we can learn a lot from studying others’ styles. They also imply that ultimately we should seek to develop our own unique style of play to be truly comfortable. Moreover due to the game’s essential sporting character - so that, say an “attacking” player may frequently choose a “positional” solution, if he or she feels that it’s clearly the best way to win a game - we will often display a combination of sometimes quite different styles even “in the course of a single game" ... So, in reading this book, please don’t unrealistically expect to find “romance” over-simplistically written into every move of my five selected players’ games. Never forget, that in all of them, they were, like you in your own games, primarily out to win points ... Try more generally to appreciate, learn from and simply enjoy their “romantic” essence."

On second thoughts, perhaps more over to "you" than Houdini. Tough game chess! But (still) endlessly fascinating!

David Robertson

Re: Emulating player's styles

Post by David Robertson » Sun Apr 13, 2014 8:27 pm

Craig Pritchett wrote:Keen to hear about Houdini's take on "style".
Aye, I bet :-) My comment above was somewhat facetious. But not entirely so, as you probably gathered.

For some time now, I've been brooding over the question of what exactly is Engine doing to the game. Will we look back on Engine in years hence, and say: "ah, yes that's its contribution(s) from which we've learned; that's what it's added" in the way we've absorbed the principles demonstrated by Morphy, or Steinitz, or Tarrasch, or Nimzovich and so forth.

The contribution, if any, may not be simply technical either. It may be one of style. Is there an aesthetic that defines Engine's treatment of positions, and perhaps distinguishes itself from a Kasparovian or a Karpovian aesthetic? And will a future high proficiency at chess have to satisfy a reversed Turing test - in the standard model, a computer's response is 'intelligent' when indistinguishable from a human's; in chess, when ours is indistinguishable from Engine's?

This, for starters anyway :-)
Craig Pritchett wrote:Tough game chess! But (still) endlessly fascinating!
Damn right!

Roger de Coverly
Posts: 21315
Joined: Tue Apr 15, 2008 2:51 pm

Re: Emulating player's styles

Post by Roger de Coverly » Sun Apr 13, 2014 8:53 pm

David Robertson wrote: The contribution, if any, may not be simply technical either. It may be one of style. Is there an aesthetic that defines Engine's treatment of positions, and perhaps distinguishes itself from a Kasparovian or a Karpovian aesthetic?
If there's a direct tactical line, engines will naturally find it. But in quieter positions, what it comes up with is influenced by which little positional rules the programmers thought important and how they implemented them. It was the chessdom site during the Candidates who displayed the top moves suggested by three different engines. Whilst more often than not they would come up with the same move, they could differ quite a bit in the numerical assessment particularly if there was an imbalance in the position.

Mick Norris
Posts: 10362
Joined: Tue Apr 17, 2007 10:12 am
Location: Bolton, Greater Manchester

Re: Emulating player's styles

Post by Mick Norris » Sun Apr 13, 2014 9:41 pm

David Robertson wrote:
Craig Pritchett wrote:Keen to hear about Houdini's take on "style".
Aye, I bet :-) My comment above was somewhat facetious. But not entirely so, as you probably gathered.
It would be interesting to see if players who use Houdini develop a different style than those using Rybka or Fritz
Any postings on here represent my personal views

David Robertson

Re: Emulating player's styles

Post by David Robertson » Sun Apr 13, 2014 10:11 pm

Mick Norris wrote:It would be interesting to see if players who use Houdini develop a different style than those using Rybka or Fritz
For me at least, this is uninteresting because it is simply beyond us. We already lack adequate terms to describe the style of players we've observed for decades. 'Attacking'/'tactical' (Alexhine, Tal, Keres, Kasparov); 'positional'/'defensive' (Capablanca, Petrosian, Karpov, Carlsen) - even here, the categories only do rudimentary justice to the players' style. The prospect of meaningfully distinguishing between Houdini, Rybka, Stockfish, Komodo and so on is too daunting a task.

For this reason, I use the umbrella term 'Engine' to describe all non-human 'players'; and from that concept, to try and tease out what, if anything, can be said about Engine style.

I play a lot of CC-style games online (3500 over the past five years). And a colossal number of those have involved opponents using Engine. With that kind of exposure, you begin to recognise Engine style if you're attuned to look for it. That said, I'm a long way from being able to describe what it is I believe I recognise as distinct from human style. As/if the thread develops, I'll try to find some words that edge me towards what I'm getting at.

Arshad Ali
Posts: 704
Joined: Mon Feb 22, 2010 12:27 pm

Re: Emulating player's styles

Post by Arshad Ali » Sun Apr 13, 2014 11:35 pm

David Robertson wrote:For this reason, I use the umbrella term 'Engine' to describe all non-human 'players'; and from that concept, to try and tease out what, if anything, can be said about Engine style.
This won't illuminate anything, but engine "style" is some emergent property of the set of algorithms used by a chess engine programmer, and one he probably couldn't anticipate himself. Words such as tactical, positional, strategic, etc. simply have no meaning here (if indeed they even have any in the context of high-level human chess). I use the word "strategic" for moves made on general grounds, where the tactical consequences either cannot be worked out (too complex) or it's redundant to do so (e.g., K+P v. K ending, where you know you have to get hold of a key square without any calculation and the game will win itself).

MartinCarpenter
Posts: 3048
Joined: Tue May 24, 2011 10:58 am

Re: Emulating player's styles

Post by MartinCarpenter » Mon Apr 14, 2014 10:08 am

I've certainly had the impression from various annotations etc in NiC and the like that the new brand of very strong young players, who have of course been brought up using engines, do tend to play slightly different chess to the older ones. Not sure how easy it would be anyone who wasn't very strong to fully clarify though.

Roger de Coverly
Posts: 21315
Joined: Tue Apr 15, 2008 2:51 pm

Re: Emulating player's styles

Post by Roger de Coverly » Mon Apr 14, 2014 10:35 am

MartinCarpenter wrote:I the like that the new brand of very strong young players, who have of course been brought up using engines, do tend to play slightly different chess to the older ones.
One characteristic is an appreciation that what earlier generations sometimes termed "laws of chess" are no more than general guidelines and in some cases mitigate against finding the best move. This would be comments such as one I saw in a recent issue of Chess "Don't lose time with your queen by getting her out early". It's something frequently broken even at amateur level. The other weekend, my game started 1. Nf3 d5 2. g3 Nc6 3. d4 Bg4 4. Bg2 Qd7 5. Ne5 Nxe5 6. dxe5 0-0-0 7. c3 Bh3 8. Bxh3 Qxh3 and now my opponent correctly countered my lone excursion with the Queen against his Kingside with one of his own, namely 9. Qd4 with a menace towards a7 and the Black King.

A ridiculous assertion of supposed "laws" was one I saw, possibly on a forum for turn based chess, that a standard sequence in the Scotch running 1. e4 e5 2. Nf3 Nc6 3. d4 cxd4 4. Nxd4 Nf6 5. Nxc6 had to be incorrect because White moves the Knight three times in five moves thereby breaking the "law" that says you should only move a piece once in the opening.
Last edited by Roger de Coverly on Mon Apr 14, 2014 5:26 pm, edited 1 time in total.

Arshad Ali
Posts: 704
Joined: Mon Feb 22, 2010 12:27 pm

Re: Emulating player's styles

Post by Arshad Ali » Mon Apr 14, 2014 3:39 pm

I like to kid myself into thinking my timid, craven style of play emulates Petrosian and likewise my confused and dithering maneuvers camouflage profound prophylaxis.

MartinCarpenter
Posts: 3048
Joined: Tue May 24, 2011 10:58 am

Re: Emulating player's styles

Post by MartinCarpenter » Mon Apr 14, 2014 5:20 pm

The most likely result I can think of of trying to emulate Petrosian's style would be a sucession of quite horrible losses :)

David Robertson

Re: Emulating player's styles

Post by David Robertson » Mon Apr 14, 2014 5:46 pm

Arshad Ali wrote:
David Robertson wrote:For this reason, I use the umbrella term 'Engine' to describe all non-human 'players'; and from that concept, to try and tease out what, if anything, can be said about Engine style.
engine "style" is some emergent property of the set of algorithms used by a chess engine programmer
Nah, I'm not allowing that! In my concept, all pre-conditions for Engine are dismissed. And so are all 'techie' arguments. We don't explain Kasparov's style by reference to his Mum & Dad; same applies to Engine.

Moreover, Engine is the Universal Other, a supplier of moves of non-human origin. There may be different subsets of engines. But we're mainly interested for the time being in Engine, the undifferentiated Other. To underline my point, imagine we stumble across a race of Martians. Being who we are, our first question after meeting them will naturally be: do you play chess? We will not be concerned to differentiate between those with orange eyes at the end of purple stalks, and those with green fingers at the end of red tentacles. They will all be Martians to us.

And so to Engine.

My first thought on Engine-style is in the vein of Roger's example(s) above. Engine will accept and play positions which we have learned, wisely, to treat as inferior or suspect, demonstrating that a greater variety of positions have more resilience than we have hitherto assumed. Accepting doubled pawns and compromised pawn structures might be an example. Unless a position is completely shot, Engine will find something in it. Take a look at Robertson - Martin Carpenter (2014; 4NCL Rd 6) where I consciously chose to create an Engine-style position (as I judge it) simply in order to test my 'resilience theory'. As you'll see, I accept one tripled pawn; another doubled pawn; and four pawn islands. Not recommended of course. And I'm not the best-placed person to test the theory anyway. But I was playfully attracted by the 'structure' after looking at Dominguez Perez - Caruana (Tata 2014) which had some similar bizarre characteristics.

Sorry to mention it in the same breath as my game with Martin, but to give a further illustration of what I think is one aspect of Engine-style, consider 44...gxh5 (Anand - Carlsen, 2013, WC game 6). For me, that capture is pure Engine. It would not have been played in decades past, by champions past. Yet Carlsen snapped it off, doubled his h-pawns, and strolled home.