04 August, 2014

T61 Openings Summary

Another TL season over, so an appropriate time to look again at what Amateurs play in the opening phase in TL chess.

Hardly changed over the seasons, as the Sicilian Defence is firmly rooted at the top, and the remaining openings played seem to remain more-or-less static, just varying in popularity, although General Queen's Pawn Openings ( ie not specific Queen's Gambits or English or often not with an automatic c4 ) seem to be on the increase at the moment.

Counted in this ( General Queen's Pawn) are 'systems' such as the London or Colle, which I would think are often used by amateurs to cut down on amount of theory required: certainly that applies to me, as I have adopted the London recently.

Other openings remaining most played are the Caro-Kann and French, from a Black point of view, and the English and Spanish from White's perspective.

However, if you consider the most played move, then, technically, 1.e4 is the most used opening move with the reply of e5 just beating c5 followed by e6, c6, d6, d5.

So, if you played e4, manage to avoid the Sicilian, and you want to play the Spanish Game, after 2. Nf3 what will happen after the 90% chance that Black replies with 2...Nc6 and you respond with 3. Bb5, ending up in your favourite opening ?

The Morphy defence with 3...a6 is the most likely reply ( 75% chance ), with the Berlin ( Nf6 ) or Steinitz ( d6 ) the most played alternative, but both with significantly fewer games than 3...a6.

Only one player in T61 tried the Classical Defence of Bc5 and he won. ( Note, however, that Black was rated 250 points higher than White, and White made the last, and major blunder, although they stuck to theory up to move 9 ) !

Likewise a single player rolled out the Schielmann ( 3...f5 ), and this time White won. Again, I doubt due to the opening ( although theory up to move 9 ), rather a mistake in the endgame.   

The Cozio Defence of 3...Nge7 also saw the light of day, but after 4. 0-0 a6 5. Ba4, Black played the frankly bizarre 5...Na7

5...Na7 ???
... and lost within 20 moves. 

Another branch at move 4 gives a 10% chance of Black seeing the Exchange variation ( Bxc6), but with poor results for White, only 33% ( a loss and 2 draws ) with the loss running to 'ratings-form', but also being a long King and pawns endgame.

The mainline is 4. Ba4 with the biggest reply ( 73% ) being Nf6, but also seen is an early b5, plus a single instance of d6, the Steinitz Deferred, which turns out to be played by me !  Theory for 8 moves, followed by a fairly level and typical ' see-saw' amateur game, decided by White making the last mistake.

Black could enter the Open Spanish here with Nxe4, but both games played ended in White's favour, and this time both games were with evenly-matched players.

However, the most likely reply to 4. Ba4 is 4...Nf6, and things run to form from there with 5. 0-0 Be7 6.Re1 b5 7. Bb3 and then Black either castles or plays d6 ( equal chances of both replies happening). 

After 7. Bb3, do you choose 0-0 or d6 ?

Intriguingly, 7. Bb3 d6 was always met in T61 by 8. c3 0-0 9.h3 with good results for Black ( 4 wins, 1 loss = 80%).

However, playing 7. Bb3 0-0 8. c3 ( the Marshall Attack ) also gives Black a nice 75% ( 3 wins, 1 loss ), and a 100% if the gambit move of 8...d5 is played, although you have to say that the rating difference in both games was (significantly)  in favour of Black.

If you compare this to all TL games ( not just T61 ), Black generally does better choosing d6, with 8. c3 0-0 9. h3  leading to lots of choices for Black, with Nb8 ( Breyer Variation ) scoring 67% for him and the most popular Na5 ( Chigorin ) down at 43%.

Does the above show anything for the TL-player? 

In reality, I don't think so, its just statistics, and on a small sample of 900 games, but it is intriguing.

TL-players obviously still enjoy playing e4 and Classical openings such as the Spanish, and some of them know their theory as well ( one Spanish game went to 15 moves of theory, for example, and those moves were played quickly, as both players had 50 minutes on the clock at that time, from an initial 45 ! ).

In T61, White did better by not playing 3. Bb5, ie by choosing openings such at the Italian, the Three/Four Knights, or the Scotch Game, but that is on a small sample, and not looking into ratings differences between players.

In fact, White did OK with most of the 9 opening moves played in T61.

Only 1. Nf3, 1. b4 and 1.g3 scored less than 50%, all the others (in order,  e4, d4, c4, b3, a3 and f4 ) scored ( in order ) 51%, 52%, 51% and the rest 50%.

It probably points to the relative truth of Kasparov's, " I think all openings are 100% sound - all normal openings that is ! Its just a question of your mood and preparation".

Its hard to disagree with that, although the discussion on what "normal" openings are will doubtless eat up many column inches.

Plainly, looking at the percentages, apart from 1. Nf3 , which scored only 45% for white, all the normal e4/d4/c4 openings give at least a statistically even chance for both players. 

What you play depends on what you like, how you feel, what you think the other player will do ( and if you prepared by looking at his games ! ), and maybe even if you just want to play like you chosen chess ' hero' !

If you want to guide your study/preparation, then statistics are useful, and can give you a fair idea of where to concentrate for the best returns, but if you don't end up with the statistically expected moves, well, that's just "unlucky" isn't it : just don't waste the study you did !

A long ramble, but below you'll find the top 10 T61 Openings ( with special mention to the Réti Opening in 11th, with the same number of games as the Grunfeld ).

The Wing Gambit in the Sicilian seems to have disappeared from TL ( not seen since T53 !) , but there are still King's Gambits out there, with the Falkbeer Counter-Gambit keeping its head, but not doing very well on the two occasions it was played in T61 !

For my own interest, I also compared T61 to the first two rounds of the current Olympiad. Different, but similar. 

For example, although the Sicilian is top in both, it appears in a remarkable 23% of games in the Olympiad , while TL manages a "poor" 14%. 

In case you wondered, so far in the Olympiad, no Wing Gambits, no Falkbeer Counter Gambits (in fact no King's Gambits at all ), no Réti Gambits, only 2 Bird's Openings ( both lost ) and only two severely ' unorthodox'  opening moves ( possibly meaning with a lack of respect ) of 1.h3 and 1.Na3, although in his defence, he does seem to play such openings regularly.

         Comparing TL61 with Olympiad Rounds 1 and 2

Rank Olympiad Round 1 and 2 T61
1 Sicilian Sicilian
2 Queen's Gambit Declined Queen's Pawn
3 English Opening French Defence
4 French Defence Caro-Kann
5 Queen's Pawn ( General) Queen's Gambit Declined
6 Spanish Game English Opening
7 Torre Attack and Old Indian Torre Attack and Old Indian
8 Caro-Kann Spanish Game
9 Réti King's Indian
10 Pirc Defence Grunfeld Defence

                           T61 Top 10 Played Openings

B20-B99Sicilian 117 (14%)
C00-C19Queen's Pawn (General) 86 ( 9.8%)
B10-B19French Defence 66 ( 7,5% )  
D00-D05Caro-Kann  61 (6,9 % )
D30-D69 Queen's Gambit Declined61 ( 6,9 % )
 A45-A50English 53 ( 6 %)
E60-E99Torre Attack/Old Indian46  (5,2%) 
A10-A39Spanish Game (Ruy Lopez)  42 ( 4,8 %)
C60-C99King's Indian 29 (3,3%)
A04-A09Grunfeld   26 ( 2,9% )

For those interested, all games were re-indexed fr ECO code using Chess Assistant 11,. 

The Opening classification ranges ( ie grouping ECO codes), I borrowed from Chessville (many thanks, although it seems to have disappeared again :(

See also T56 Summary,  T55 SummaryT51 Summary,  T50 Summary and T47 Summary and T57-59 Summary.