Monday, November 17, 2014

Theory vs Practice - Queen vs 2 rooks

In my last blogspot , I posed a question on who wins in that battle of Q vs 2P; I was more interested in 'how' though. AOX replied with Qh2, there was no Qh2 as it was black to move (and not white).  AOX sealed the discussion though, by mentioning that the King should move towards f4. Correct. That game was one I played as White. Either way, White had a very strong chance of win but definitely had to know 'how'. There was a fair chance of a misstep, taking into consideration that we are dealing with players rated around 1500, and one wrong step may cause difficulty for white and end up as a draw. I don't think there was a win for Black either way as the best chance was a draw.

I was reviewing game 5 of World Chess Championship with my students. At the above position, it is obvious that Carlsen had a knight vs Anand's bishop. Carlsen had double pawns as well. When I review games with my students, we do it OTB 1st. After we finish the game review OTB, we switch to  the computer review. When we were reviewing the above position OTB, obviously the students mentioned the 2 'weaknesses' about Carlsen's position. When we switched on the computer review, Houdini didn't show any significant disadvantage for Black.

Again in game 6 (position below), it is obvious that Carlsen is battling with double pawns on his side. The machines did not show any weakness though.


I was playing a tournament game with a ~ USCF 1600 rated player (higher than my rating).  I won this game comfortably. When we were reviewing this game after it was over, we paused at this stage where I was left with a Queen & a Rook whereas my opponent was left with 2 rooks and 2 bishops (position below, White to move). We calculated the piece values and my opponent said he should won if not for the mistakes he made. The value of the pieces was higher for him, but I said I will win 9 times out of 10 if I have the queen. We argued back and forth, he refused to believe that having a Queen was superior to having 2 rooks. Because theoretically, 2 rooks have more value than a Queen. Does it work like that though?



I will show you one more example. This time I am illustrating a higher rated game, not at the 1500 - 1600 level. A tournament game between a US National Master vs another similar rated player (not NM, but slightly below). Black had 2 rooks, a bishop, a knight and 4 pawns. White (National Master) had a queen, a rook and 5 pawns. Picture below, Black to move. They did not have any advantage with the positioning of the pieces or the pawns. Black made a mistake very soon, was down a piece and never recovered and White won comfortably from that stage.



In all the above games, the machines are showing that neither side has significant advantage. It is very easy to look at the computer analysis and say that the positions are even. But, is it that easy and straightforward? What do you think?

At the level (1600/ 2200/ Super GM) these games were played, does the inadequacies (having knight vs bishop/ double pawns/ Queen vs 2 rooks) matter? Or not? My opinion is that the inadequacies do matter in practice though in theory the positions are even.

May not matter so much at the Super GM (Carlsen/ Anand) level or the Koumoudo/ Stockfish 3500 rated level. But, the inadequacies do matter below the 2500 rating level, they make a difference.

Even though the positions are even, it is hard for the player with double pawns to chalk out a win. It is easy for the player with a bishop to win vs the opponent's knight in the end game. It is easy for the player with the queen to win vs the opponent with 2 rooks. In my honest opinion, it is an excuse for my 1600 rated opponent to think that "I made a mistake, otherwise 2 rooks can match the queen". My opponent had to coordinate 4 pieces while I was at ease to slide my queen all over the board to take out the pawns and pieces with forking the king and the pieces. I think for a 1600 rated player, it is very hard to coordinate 4 pieces when playing with an opponent with a queen and a rook. FYI, me and my opponent replayed the game from that position and I won again. I even offered to play with 2 of my pawns off the board. Not just 1600 rated level, I have shown you above a game involving a NM as well to support my point of view.

Discuss......






2 comments:

  1. In my opinion it is much more complicated. At least it comes from my experience and analyses. For the long time I was AFRAID to exchange Q for R+B (or R+N) because I could lost 95 out of 100 games. Nowadays I simply KNOW when I can make such an exchange and do not lose the game. However it requires a lot of practice and analysis. For example: when I extracted (created) the theory of fortress - I understood that the stronger side is this side that can FORCE winning from the position after the exchange (Q for R+B or R+N). However - unless the weaker side knows how to play and can hold the position - playing with "2 pieces vs 1" - is much easier than it seems.

    From my perspective and experience: most often it does NOT work - it is better NOT to exchange Q for R+B. And it depends on your understanding of such concept as comensation and exploiting the holes in the position. I have played about 20 positions of this kind - and only in 2-3 cases I was wrong (that I could safely exchange Q for R+B). To be honest - I like playing with material deficit and defend difficult positions.

    You probably may be interested in explanation how to teach students when this kind of exchange is/may be correct. There are a few factors that may help to make such evaluation (assesement) correct:
    1. Pieces coordination - if weaker side can coordinate forces and play these with a plan - it is hard to lose.
    2. Passed pawns - if the stronger side has a passed pawn (or pawns) - you have you think how you may block these pawns in the future.
    3. Weaker side pieces activity and "ways to break the position". If you control all the lines and diagonals and obtaines a grip on the vertical and/or horizontal lines - it is very hard to lose.
    4. Having a plan and checking if the stronger side may win material in a few moves (or with simple plan that may not be stopped). If yes - the exchange is/was a bad idea.
    5. Obtaining 7th/2nd row with the rooks. Most often when the opponent cannot exchange these rooks (or at least one of these) - it is the "stronger side" (with the Q against R+B) that is obliged to defend!

    I am not sure if I convinced you my friend, but at least I tried. And of course it is a good idea to test our skills having stronger side (no matter how much) and try to outplay the engine (set at 2600-3500 level - depending on the size of the advantage) being the side "which should win most times". I have to admit you may be very nice surprised of the resistance our silicon friends :).

    ReplyDelete
    Replies
    1. Hi Tomasz, I doubt if a player wants to relinquish his queen for a couple of rooks. However, it is forced. I was merely mentioning the player sweating bullets to force a win when you play against the Q. Thanks for your comment.

      I agree that with piece coordination, passed pawns and having a plan to obtain the 7th/ 2nd row with the rooks can force the win.

      Delete