Opposite color Bishop endgames... & two games I've played
After tracking a game of Henrique Mecking against Taimanov, for a previous blog, I got really interested in opposite color Bishop endgame theory. And started studying it as thoroughly as I could [I had already realized that my analysis of that game was somehow 'low' on this part]. However, I think that before I tried or applied it in a game, there couldn't be some consciousness of the gained info. So here're two of my dailies, where I chose to try such an endgame; they are played about 1,5 years ago, and the material was gathered since then. For various reasons I didn't post it and left behind. But tracked it again, refiltered it and here it is...
But before the games, some theory on opposite color Bishop endgames along with some positions [endgame-studies and from recent games] that might be useful [if you are familiar with and want to skip it, press this to go to my games].
Opposite color Bishop endgames; some theory
This kind of endgames has been analyzed by theory; rules have been written, though maybe without being possible to exhaust the topic. I'm trying to give the most basic of them accumulated and organized somehow, at least according to my point of view. The basis I use is the B + 2ps vs B endgame type as a limit. Its principles can be found and in endgames with more pawns. However it should be underlined that everything has to do with the position of the pieces:
1. Generally
In order to win such an endgame, it should be at least of the kind B + 2ps vs B. If the strong side has just one pawn, it's a draw; the other B takes pawn. If the strong side has just +2 pawns, meaning that the weak side has at least 1 [eg. B + 3ps vs B + p], it's still uncertain; it depends on the position.
As a rule in a +2p adv: more pawns are in favor of the strong side. They can make the defensive moves of the weak side really more difficult, and the possible B sacrifices pointless. So more times are winning for the strong side; the more the threats, the better usually. Sometimes will be a draw. And fewer times the position can ensure a win even if the strong side is just 1 pawn up.
Thus, being on the strong side, a +2 pawn adv should be always a start of thinking the win.
2. The basic R-pawn exception
In B + 2ps vs B endgames, either pawn of the strong side should not be a rook pawn of the file of the wrong color B; i.e. of the file where its final promotion square is of the opposite color of the B of the strong side. In this case it's a known draw position; where the weak side sacrifices its B for the other pawn, and sends its K to that final corner of the rook pawn. As B of the strong side can't control this final promotion square, it's either draw by repetition or stalemate.
3. Distance between the 2 pawns of the strong side
As a rule but with exceptions in B + 2ps vs B endgames: The 2 pawns of the strong side should be separated at least by two files. If this distance is shorter, the weak side could manage to control a square by K + B, while the weak K will be close enough to catch up the other pawn, too; with ultimate goal to take a pawn for free or to sacrifice B with a drawish result. Thus the strong side can't make progress. Unless of course these two pawns had advanced far enough [eg 6th rank]; in this last case the weak side is missing 1 rank behind so to spare a tempo with B.
Anyway, back to the draw position... the Key: the weak K to wait on a safe square. Here're two diagrams depicting this main strategy of the weak side:
| d01: = | d02: = |
 |
 |
| separated by none file | separated by one file |
And it should not be forgotten that a Bishop can deal alone successfully against two connected pawns.
Nevertheless, it should also be noted that, though the above positions are a draw, this kind of pawn setup can be winning if there're more pawns in the game; even equally for both sides [eg B + 4ps vs B + 2ps].
In these cases the possible sacrifice of weak B could be just pointless...
... or the weak B could be busy with other tasks too, making defense impossible
In opposite color Bishop endgames with more pawns, the strategies of the two games above, can be applied even when pawns have a distance greater than 2 files.
4. The basics in B + 2ps vs B endgames
So let's say that, in B + 2ps vs B endgames, the minimum distance of these 2 pawns of the strong side should 2 files between them. The greater is usually in favor of the strong side. Yuri Averbakh was writing on this [in his Comprehensive chess endings, vol.1, 1983, Bishop endings & Knight endings, p. 119]:
When there are two or more files between the pawns, the stronger side wins if it is possible to break through to one of the pawns with his king, and ensure its advance to the queening square.
I believe that he's totally right; and I think that obviously it's a rule applied and in cases where there're more pawns for both sides. Generally King's activation is an important factor in this kind of endgames. However when firstly read it, I understood it but couldn't exactly embrace it. I was missing at least some visualized patterns [?], positions [?], that came with the solving process of endgame studies-positions. Here're some I've picked or made [+ with the help of tablebases], trying to give the most main characteristics.
4.1. Basic strategy: Break in with King etc
Firstly let's see how the break in with K, supporting an advancing pawn could work or doesn't, in a really basic form; just to get a clue. Two diagrams:
| d03: + | d04 |
 |
 |
| white to play and win. Here Kf5-Kg6 makes pawn advancement with f5-f6-f7 certain. Then, creating threats with the other pawn, white wins | here white's 1.Kf5 is met with black's 1... Bd2 pinning the pawn. So 2. Kg6 is impossible, and 2.Kg5 pointless. It's still a winning position for white but there should be another strategy. |
It should be remembered that when a pawn, is a Bishop one or a center one, the break in of K can be achieved and by the outer side. However the break in from the middle is usually stronger.
4.2. Exceptions in the King breaks: Knight pawns & Rook pawns
4.2.1. In general
I can't actually know if it's an exception or the rule [based on what appears more often], but these pawns could be a real problem for the strong side. And this cause if they are stopped advancing with the weak B in front of them, strong K can't easily kick B, cause one side of its route [the side route] is closed; and so it's a draw. The following two diagrams, with just a file difference, depict the situation via comparison:
| d05: + | d06: = |
 |
 |
| Here the winning strategy for white is Kc5-b5-a6-b7, kicking B and supporting pawn. | But now this is not possible, as strong K can't go round Qside. Here the only winning strategy could be to go around the other Kside, with eg K@f5 controlling e6. But with a free weak B and opposition of the weak K, this seems really really difficult; generally. |
In the following example it can be seen how hard could be these endgames, even if the strong side succeeds to break in, but from the other side...
And here is an endgame study-position found in Averbakh's book. It hasn't always unique solutions so to be set in puzzle form; however it depicts the defense of the weak side, even if weak B can stop the advancing pawn in the last promoting square. Something also noticeable is that this blocking of the pawn can work also as a threat. And it should also be noted that if it was black to play, he could win; showing how crucial is tempo in these endgames.
4.2.2. The break ins with R-pawn or N-pawn
Thus as a rule: the strong side should avoid advancing firstly this knight or rook pawn; it should advance the other first. Or better is to break in with K from the middle, between this pawn and B. There're following two positions, a file next to, so to see the difference between knight and rook pawn; examined for both sides to play first. In some cases tempo is crucial.
| d07: +/= | d08: +/+ |
 |
 |
| Rook pawn: white to play and win [Averbakh 1950]: strategy 1. Kb6 Kd7 2. Ka7 kicking Bishop 2... Bd5 3. Kb8 securing the advancement of rook-pawn. black to play and draw [Lisitsin 1956]: 1... Kd7 2. Kb6 Kc8 3. Ka7 Bd5 but now white can't play Kb8 to secure advancement. It's stuck on the side-rook file | Knight pawn: whoever plays first white wins with Kc6-b7-a8 |
When there're more pawns on the board, with the balance of the +2 on, the strong side can actually create this break in more easily and successfully. Two played examples...
4.2.3. More files between the pawns
And back to the B + 2ps vs B setups, even if B has stopped a knight pawn, in the case where the difference between the two pawns are 3 files or greater, things can be winning for strong side; but not always.
| d09: + | d10: = |
 |
 |
| Whoever plays white wins. Strategy is to try to bring battle on f pawn. If weak K closes with opposition the side, strong K breaks in from the middle winning; if strong K manages to go from the side, he sacrifices the pawn as decoy and then promotes the other pawn; neither weak B or K can catch up cause of the longer distance. | However if the non-knight pawn is far behind from the knight-one [at least two ranks], there could be situations where the strong side can make no progress. In the above position [Averbakh 1950], white could break with K only for Kside, via h4, but this is not possible as weak K has opposition. |
Anyway, a played position depicting some approach in a longer way...
5. A special exceptional case: Draw by controlling the pawns with the same long diagonal
Sometimes the Bishop of the weak side can control with a longer single diagonal, squares in front of both pawns of the opponent. In this case, a draw is possible, and this is one reason for why the advancement of a single pawn is usually avoided. The weak K will try to move in parallel with this diagonal, following strong K according to the rules of opposition; with ultimate goal, always to be there in time, and control the squares in front of the pawns, so to prevent advancement. Key: the good placement of B.
Check this Cheron's study, found in Averbakh's book.
However, this doesn't happen when the B of the weak side is controlling with this diagonal just the final promotion square of one pawn. The comparison of the following two diagrams, with just one rank difference, can show why. On d12 [right], even if it's black's turn, the played advancement of c7 can't be stopped...
| d11: = | d12: + |
 |
 |
6. Trying to create break ins from the beginning wih tempo, zugzwang, etc: A game-analysis against Stockfish 10
I had already tried playing such positions against the engine in the past, but my pgns are a mess; not being able to recognize the played lines or the analysis etc. So I created anew a setup with two Bishop pawns, making things easier for me; while tried to place the rest of the pieces closer to their initial positions, so no obvious advantage to occur. And then I played it against Stockfish 10. Analysis was written afterwards. With this setup I was successful in my 2nd try; having of course already studied and refreshed theory and positions recently, along with the help of tablebases. In the first I hadn't lost hope, but I had just repeated many times some moves making the pgn useless.
Regarding the game: the intended break in with K came in the end, when weak K had less options in moving; however the threat of a break gave me the chance of advancing pawns. Controlling squares with B was critical. Sparing tempo and zugzwang were, also, sometimes crucial for progress. As a strategy I tried to push equally my pawns supporting them, while a left side-right side movement of K would be proved winning; meaning bringing the threat to one side so to advance the other. Take a look. It contains I think, many basic patterns-strategies.
Through these examples, I tried to create a place to refresh the basics accumulated in this kind of endgames. I know I left out patterns like double pawns, B + 3ps vs B, and others maybe more exceptional; but the topic can't be exhausted of course [not only by me but generally]. I just hope that I gave a sufficient idea of. So now to my games...
My games
Both they were endgames of opposite color Bishops, but with more pawns from each side. Maybe it isn't the 100% accurate strategy, however, I started thinking on this kind of endgames since I seemed having secured at least a +2 pawns advantage. Nevertheless this didn't seem enough; I thought I needed either +3 pawns or specific exchanges.
1st game:
It was the game where I deepened my study more. With the opponent of this first one, I had another game where a draw was agreed, cause of opposite color Bs, again; but this time each with 6 pawns on the same files. I don't know if this kind of endgame occurred again due to the way of our gameplay.
2nd game:
In this game I decided to cause an ending like this, maybe a little influenced of my previous one. It started a little after my first game. My study here was somehow less on endgame theory; however the key solution was of the really basics. And maybe it took me a while to see it, as I was probably trying to think more advanced.
....thanx for reading
