Increasing your chess knowledge with database programs
I was just watching a game online the other day and I thought it would be interesting for improving chess players to get some ideas of how you can use database programs to increase your chess knowledge.
Anyway….I was watching this game online. A FIDE Master was playing against a computer and he reached a rook ending like this:
Here, the Master played …h5 ….well actually the white rook was elsewhere and I am too lazy to put up another picture 🙂
The point is he played …h5 and white couldn’t capture it. The general question here is how to set up the pawns! While my general feeling is that this should be easily drawn either way, while watching the game I wondered if my intuition that …h6 was “more” correct was in fact right.
My initial feeling was that with the pawn on h5 black might have more trouble holding the defence since the pawn could more easily become weak and if you have to bring your king to …g6 and play …f6, now you are introducing the option of checks from behind. This was something i felt that black shouldn’t have to worry about if he simply set up with 1…h6!
How can we figure this out? Well one option is to place the position in a computer and see what it says. I was more interested in seeing what other strong players have done here! So I fired up my ChessBase program. And opened up my Database of 5 million+ games!
I clicked on ‘Filter List’ this allows me to enter criteria for a Database search. I select first to set up part of a position.
This is what I set up. I put the pawns on f7 and h7 and the King on g7. Note that if I am interested in the pawn structure, placing the King there will yield worse results because I still wanna see examples if the King is on g6, g8, h8 etc…….anyway in this example I had it on …g7.
Next I click the material tab. I have underlined in red what I changed there. I set the rooks for both sides to be exactly ONE, and the pawns for each side, 3 for White and 2 for Black. I also too the selection of ignore colors and selected ‘one wing’ because I want all the pawns on the Kingside.
This gave me a list of games which met my criteria and below I have clicked on ‘Elo B’ which sorts the results by the rating of the Black player.
I then proceeded to check some of the games at the top, you can see that a big majority of them are drawn and after opening up some of them I saw that the majority of strong GM’s such as Cheparinov, Epishin and other indeed did play …h6 in these structures!
Ok, great! I already learned something, well that or I confirmed what I thought I knew! But how can we also use this search to increase our knowledge even further? How about trying to see how people actually lose such endings so you better know what to avoid…..AND…..you also learn some tricks about how to win!
One idea that caught my eye was executed by Julian Hodgson:
Here he just played 39.Rg4+ here Black can probably safely keep everything tidy with 39…Kh7 since the white King can never cross the 3rd rank without the help of the rook and then we just go back to g7. Instead Black played 39…Kf6 this is probably still drawn but does allow white to set up an interesting winning attempt. 40.Rg8 Rb3 41.g4 Rb5 42.h4 Ra5 43.Kg3 Ra3+ 44.f3 Ra5 45.h5 Ra3?? 46.g5!
And here it is! Black absolutely fell for it. If he takes then the h-pawn runs down the board. Note that 45…Ra3? was the losing blunder. The rook also loses on squares such as a2 but others such as a1 and a4 would have still saved the game because in both cases the rook can swing to the h-file (hxg5 covers h4 square).
Another winning attempt is trying to set the king up on h5.
This is from Cheparinov-Papin, Plovdiv 2012. This looks drawn still and Papin should have just kept his rook on the 6th, similar to a Philidor defene and it’s hard to see a way through for white. Once white goes g5 we just take and then go …Kg8 since moving the pawn to g6 allows easy draw with passive defence on the 8th rank.
Papin however blundered 47…Ra5?? and after 48.Rb6 resigned since the pawn on h6 is lost.
Also very similar and very instructive was how Oleg Korneev (a great player and winner of countless open tournaments!) beat 2507 rated Cabrera in 2009.
Similar to the Hodgson game, Korneev played 39.Rg4+ again, though it feels very passive, I think 39…Kh7 might be the safest defence. Cabrera went for 39…Kf6 this allowed Korneev to systematically advance his forces:
40.Kg3 Rb5 41.Kh4 Ra5 42.g3 Rb5 43.Rg8 similar to Hodgson, cutting the king off, but not with the same idea. Hodgson brought his pawn to h5 to trick with g5 but Korneev wants the king there… 43…Ra5 44.g4 Ra3 45.f4 Rf3 46.f5
Cabrera has been more or less helpless against the expansion of the White pawns, but now with the pressure on h3 he is preventing the white king from reaching h5. Korneev fixes that problem by swinigng the rook in a nice circle (or square if you will) and blocks the 3rd rank on g3. From there h3 is covered and the king can go to h5.
47.Rb8 Rc3 48.Rb1 Ra3 49.Rg1 Rf3 50.Rg3 Rf1 51.Ra3 Kg7 52.Kh5 Rf2?? 53.Ra6
Just like in the Papin game, the 6th rank is under control and the …h6 pawn falls. So where did Cabrera go wrong? To me for instance 48…Rc6 49.Kh5 Kg7 should be good enough. Black hold on to the 6th rank similar to the Papin game. Another ugly defence was possible. And that was to immediately pressure h3 with …Rh1 and once white plays h4 to switch over to the g-pawn. That is assuming black can hold this:
But ok, to me holding the 6th like in the Papin example seems more safe to me!
I hope this little blog was helpful to you. If you liked it, share it with your friends and please hit the little heart one the bottom right for a like so that my best blog posts appear on the right sidebar 🙂