Are Humans Starting To Play More Like Computers?

     Computers "think" in Base 2, assigning each "bit" of information a 0 or a 1.
Humans do not think strictly in Base 2 and play chess much differently than
computers.
     Garry Kasparov has suggested that the future of chess will be the Computer-
Human hybrid.  Is this already happening?
     Since most chess players are using computers to study and analyze games,
doesn't it make sense that humans are beginning to "absorb" some of the
characteristics of a computer, or are the two still mutually exclusive?

Yes I think players go into more risky defense situations today. Players can defend better now then they could before and maybe some of that comes from learning the ice cold defenses of computers in tough situations.

I don't know if studying with computers was the cause of developing a tendancy to willingly go for material at the cost of having to defend against the worst of initiatives.

Might also be because I really like Korchnoi.

Humans don't play chess like computers. We can't calculate as fast or as accurately, therefore we make our move decisions using a whole different criteria.

I wonder if a sample of amateur OTB games could be analyzed to see how many moves are "computer" moves and then compare that to a sample of games played
30 years ago.

Chess learning computers are quite interesting - provided that they have started out from not even knowing how the pieces move - in the same way that people do.  www.chess.com/forum/view/general/when-chess-gets-solved

A typical pro spends most of the time analyzing with chess engines, so it's quite natural that more and more people are starting to play computer-like chess.

Speaking only for myself, I am as bad today as I ever was.  I aspire to be able to play more like a computer.

In the only way that I can see that humans are playing more "computer-like" is in our more accurate defense of positions but that doesen't nesesarily mean we play like computers. On that note I remember an interview with Anand a while back and he did say something about this subject when a reporter asked him the same thing. He responded that "Grandmasters are learning to be less fearful of "ugly moves" the computers have taught us that if something works tactically then it simply works no matter how odd it looks" (not an exact quote).

Err Schachgeek you exclude moves from all db games played prior to those being analysed.

As to book lines & engine analysis, well this is why you must have a large sample size of non-database moves from multiple games. If for instance your games were analysed & high matchup found in say 500 non-db moves from 20 games, you could argue 'ah yes but I have extensive book analysis for all these openings that goes right into the endgame' but you'd be laughed out of court!

Unassisted humans can indeed achieve very high matchup in a few individual games, but simply cannot maintain this in many games over time.

Of course humans are playing more like computers. hundreds of years ago, grandmasters tended to play with a "swashbuckling" style, setting traps that would bring about a spectacular and quick win against an opponent who didn't see the trap, but could also be refuted by maybe one very subtle and non-obvious move. However, a computer sees every single move that can be played. So obviously this wouldn't work against computers... at the same time, through interaction with computers chess masters now have more practice at seeing these subtle moves. In the sense that the top masters analyze more moves at a time rather than playing purely positionally, it can be said that they are becoming more like the computers, which "brute-force" many possible moves.

At the same time, computers are playing more like humans. 

The specific details of precisely how a chess engine functions is quite complex so I won't bother explaining... but let's say this:

Rybka, one of the strongest chess engines, differs from most other engines in that it has a more sophisticated function to evaluate a position. This slows it down, so it can look at less moves, less lines, but it can play more accurately since it would know which positions are better than others. Of course, it is easy to see how it is now more like a human (i.e. more positional and less tactical). Also, modern chess engines can selectively "deepen" analysis of certain very interesting lines, though it has only enough power to do this for a few lines (this is a bit different from alpha-beta pruning, wherein the computer rejects moves that immediately lose, so that it only needs to search a tree that is about 1,000,000,000,000,000,000 times smaller). This is like a human, who only looks only at a few lines (albeit carefully) and not all approx. 35 possible moves for any position.

Top 3 matchup criteria:

All analysed games must have 20 or more non-database moves.  A game is classed as "out of book" once the position cannot be found on the chesslive.de database in any game played prior to that being analysed.

Fritz 11 @ 30 seconds per ply
Pentium 4 2.93GHz 1GB RAM
Hash Table 192MB
Database used  www.chesslive.de

Capablanca-Alekhine 1927

Alekhine
Top 1 Match: 467/852 (54.8% )
Top 2 Match: 622/852 (73.0% )
Top 3 Match: 685/852 (80.4% )

Capablanca
Top 1 Match: 470/853 (55.1% )
Top 2 Match: 632/853 (74.1% )
Top 3 Match: 703/853 (82.4% )

Fritz 11 @ 30 seconds per ply

4x AMD Phenom 2.30Ghz

4GB DDR2 RAM

Hash Table 512MB

Database used www.chesslive.de

Kramnik-Topalov 2006

Kramnik
Top 1 Match: 306/552 (55.4% )
Top 2 Match: 417/552 (75.5% )
Top 3 Match: 461/552 (83.5% )

Topalov
Top 1 Match: 309/555 (55.7% )
Top 2 Match: 417/555 (75.1% )
Top 3 Match: 447/555 (80.5% )

Although admittedly performed on different systems, the analysis results for the matches which are separated by 80 years of research & technology fit perfectly into a large body of evidence which shows remarkable consistency in engine matchup levels for top-flight unassisted human play.

Similar top end results for many other OTB Super GM & pre-computer era CC World Championship matches using the same methodology have been generated independantly on different systems using modern versions of Fritz, Rybka & Deep Shredder. 

No matchup results between legitimate human players using this methodology has ever yielded total average results greater than:

top 1 match = 60%

top 2 match = 75%

top 3 match = 85% (all 3 figures)

Pretty simple really, ain't it?   

From what I have read, computers then to be more material hungry...seeing this as a stronger position based on points. Humans at the top level will sac a piece for a positional advantage...even though below on points.

When Kasparov played Deep Blue, this is how he defeated that super-computer...and why he challenged IBM after he was beaten by Deep Blue (because he saw the super-computer had changed it's "thinking" and was not as material hungry as games before.

I would think that GMs by training a lot with software, are seeing more of these computer lines etc. But when it comes to OTB play, most want to move away from known lines...and send their opponents into unknown territory, or prepared stuff...like Anand-Topalov 2010 WCC. I followed Susan Polgar's coverage of this...and was amazed when these guys were actually selecting Rybka's#1 choice in certain key plays. 

I am always reading also, that computers are not as strong in certain middle and end-game scenarios...though this has not been clearly revealed why?

Not sure, but computers are beginning to walk a lot more like animals.

http://www.youtube.com/watch?v=nUQsRPJ1dYw

The last game on Topalov Anand the blunder Topa made was a typical computer move... No other master in history would ever even consider to take that pawn and leave is king wide open to attack. As for the openings they already are at least 90% computer (in master games).

At really master chess 2600++ yes you need to think like a computer to get there at patzer level 2500-- we can still see one or two guys making some wild sacrifices and playing like real humans

 Not from what I have read...in this sense, computers do not go for piece sacrifices for positional advantage; this is the only part of chess computers lag behind humans. The strongest programs evaluate positional piece sacrifices unfavorably...this means that the "art" form of the game is lost to a computer program and at the moment will most likely never be adopted by a computer...at least for the time being; who knows if that will change in the future when computers become more human in nature. Computers do not play artistically beautiful (chess) games, they play in a scientific way.

Are humans playing a little more like computers? In some ways...

Are computers playing more like humans? No...not yet.

I have a friend who got banned from this website for "cheating", because somehow whoever runs the website tracked his games and thought that he played too much like a computer for him not to be using one. His playing style along with his rating (he's rated about 2120 ACF) was enough for him to get banned. I know for a fact that he wasn't using one, but he does train with Rybka 4, so it would be no supprise to me that he was just thinking all along "What would Rybka do?" I also train with Rybka 4, and after a while you get to know pretty well what it's thinking and what it will do next.