Survey: Ruy Retreaters - Why Repeatedly Eschew Exchange?

The chess.com game explorer indicates of the over 100,000 Ruy Lopez, master games White captured the knight 11.5% of the time.

Retreating leads to 38.5% wins and 24.9% losses

Capturing the knight leads to 30.6% wins and 24.4% losses.

Draws do not really come into the equation for players on this site.

On the surface the exchange is the poorer choice but it can be successful.

Your opponents will be at your level not Fischer's so why do you retreat?

1 - Black's Bishop pair

2 - I never exchange a bishop for a knight

3 - ?

(Advertising: There are stringent restrictions but in this unrated tournament you can experiment with the exchange variation where Black captures with the wrong pawn. Is that a big enough advantage for White?) 

3 - Because theorotically retreating offers more chance of white to claim a slight advantage, and it offers white more active chances.

I would also choose number  1, by the way.

3. I like to retreat the bishop, game is more fun.

In the exchange variation you want to quickly go to the pawn endgame and it is not so interesting for me.

It doesn't really give white a huge advantage and mainly because of the Berlin Wall people feel more inclined to retreat.

according to FM Eric Schiller in:

http://www.chess.com/article/view/how-to-develop-an-opening-repertoire

"...ignore statistical information about the winning percentage. In most chess games the stronger player will win. It has nothing to do with the opening move. "

IMO Bxc6 has much narrower scope than Bh4 - please see this game as it demonstrates one of the possibilities:

http://www.chess.com/forum/view/game-showcase/do-not-do-this-otb

"...ignore statistical information about the winning percentage. In most chess games the stronger player will win. It has nothing to do with the opening move. "

This statement is a clear nonsense simply because statistically there is an equal number of stronger players playing on both sides of the board. Thus if in a given position white wins 80% (for example) of the time we can suspect that the 20% of the other cases is a weak white player playing against a strong black player. But when skills are about equal (and no gross errors are made, which assumes that both players play high level chess) the position should be considered as strongly advantageous for white and leading to a very likely win for the white side. As an extreme example we could take a position in which white can win by a forced checkmate in 3 moves but if the correct combination is missed the position would lead to a quick defeat of white. A small percentage of players (depending on the difficulty of finding the solution) would fail to find the correct moves and would lose the game so we would have perhaps 90% of wins for white. In this instance to claim that the stronger player will always win is just daydreaming. The opening variations are not so clear cut but the ideas are similar. If the trend is strongly favourable for white in a given position it will take a strong black player to beat a weak white player but when the skills are about equal black will be in trouble most of the time. In correspondence chess where everyone routinely uses computers there is no such thing as weak players. If you go into a position that is statistically strongly disadvantegous you are asking for trouble. I would do this only if I thought I have found a novelty (which accidentally if quite easy with the help of computers - it is amazing how many good moves in the openings are still waiting to be played).

On the other hand I am ready to acknowledge (in view of the above remark) that the statistical evidence may be biased because the majority of players may be following well known move sequences (as recommended by books or approved by "theory") which in themself may be faulty. If another move were used at some point the results could have been different. At any rate a conscientious approach is required to the openings choice. When in doubt however, choose variations which are statistically in your favour.

1200 strength players should think thrice before trying outright refutations of statements made by FM Eric Schiller. Statistics only works with large numbers - when we are dealing with not so large numbers, statistics invariably give wrong conclusions. Statistics work only if other factors are made equal. Also here Eric criticizes only the win percentage - he does not say to ignore other percentages like percentage of players who play that move.

Also anybody with a decent playing strength would realize how narrow the scope of Bxc6 is compared with the rich lines of Ba4 - attested by numerous players and authors of no mean repute - works only as a novelty against unprepared players, not against prepared opponents.

Both can be interesting, both are reasonable.  Fischer played both, annotating the exchange variation with an exclam on at least one occasion.  The reason you rarely see white take the knight at the master level is because of the drawishness of the line.  If black tries to draw the game, he very likely will succeed at master level. At the same time, nonmasters tend to have a very skewed idea of what is involved in holding a draw. 

Personally, I play both.  Recently I've done well with the exchange.

I actually played an exchange variation as black a couple of days ago on the LSS correspondence server. I know nothing about the variation. I was just using the moves from the ChessOK opening databases, always using the most statistically advantageous moves and most popular ones. When were eventually out of the book, we played some 5-6 moves more and agreed on the draw as the position was totally pawn-blocked, most of the pieces exchanged. But I am not going to play this anymore relying on the white to exhange his bishop. In my opinion a6 favours white too much and I am going to fix my mind on Berlin Wall as being a more certain way of drawing with black. And besides we should develop our pieces in the opening rather than move pawns  Steinitz said that a6 is wrong because it sends white bishop where it wants to go anyway. He might well have been right about it.

In the exchange variation, black usually gets the initiative (with proper play, of course). e.g. see this game:

http://www.chess.com/forum/view/game-showcase/do-not-do-this-otb

In the other lines excepting Marshall Attack, white usually retains the initiative.

The only reason the Exchange Variation can be played is to surprise an unprepared opponent.

I play both 4 Bxc6  and 4 Ba4 and do ok with both but score 60% with the exchange and 68% with 4. Ba4.  Whats most interesting, to me, is that with the exchange ( 10 games here ) none are draws..... all 10 games were decisive.

You make it sound like the exchange is a fundamentally poor opening -- only useful against an uprepared opponent.  Of course this is not true. 

4.Ba4 is simply more popular because it leads to more dynamic play.  At the highest level, the precentages effect popularity for sure.  To someone like me, I choose the exchange because I don't have to remember as many opening moves black may play and I like endgames.

To have all these sub masters talk about percentages as if they're relevent is being fairly oblivious to the fact that they will loose to tactics or endgames 100% of the time, and never because of falling behind in opening theory.

I did not imply that - I meant Bxc6 has much narrower scope of play - Ba4 has wider scope of play - as you say more "dynamic". With Bxc6, if black is prepared, he faces much fewer problems. What happens is, Bxc6 can catch some people by surprise (they are not much prepared for it) - in those cases Bxc6 really shines, and gives high chances of winning.

Donating the initiative to my opponent and gifting him the bishop-pair just for the sake of variety is not my cup of tea. Yes, Bxc6 gives a better endgame - but more chances of a draw against a strong opponent (Bxc6 gives higher chances of winning against weaker opponents, who have no clue about the nuances of the position).

Uh... I think that is because he's Fischer, not because of the opening...

In other words against competition hundreds of points below him?

It's not an unsound opening, I think dsarkar is just saying both statistically and practically you have better winning chances with 4.Ba4 (unless your opponent is unprepared for Bxc6, in which case, your opponent will of course do poorly).