What sort of results can you expect when playing the Blackmar-Diemer Gambit? And more importantly, which defense are you the most likely to face? Let’s have a look.
In all cases, the number in brackets are the score for White, followed by the number of games in the database.
1. d4 d5 2. e4 (55% / 12.3k) dxe4 3. Nc3 Nf6
3. … e5 (48% / 0.9k) 4. Nxe4! (58% / 0.3k). White has crushing results with the less frequent 4. Nge2 (65% over 61 games), but I believe this variation to be somewhat weaker than 4. Nxe4. More on this another day.
4. f3 exf3 (53%, 8.6k)
The most frequently played deviation is 4. … Bf5 (58%, 1.3k). Fortunately, the much stronger 4. … c6! is rarely played (only 0.5k games, but only 47% for White). Of the rare alternatives, 4. … c5 scores best (51%), just before 4. … e3 (53%).
5. Nxf3 (7.7k, 54%)
You might, like me, have forgotten that 5. Qxf3 is relatively often played alternative (almost 900 games with a decent practical score of 52%). I doubt that this romantic double pawn gambit is any good though. Another one to explore one of these days…
5. … Bg4 (54%, 2.7k)
While 5. … Bg4 is the most popular move, it is probably not the best. 5. … c6 has proven very challenging for White (44% only) but is rarely played (only 0.7k games). It is much less popular than 5. … g6 (50%, 1.8k) and 5. … Bf5 (55%, 1.2k). I was surprised to see that Black scores very poorly after the reasonable 5. … e6 (59%, 1.3k) games.
We will not go in depth in the statistics of the sub-lines at this stage. As far as 5. … Bg4 goes, there is a further subdivision after
6. h3 Bxf3 (56%, 1.1k)
This is marginally more popular than 6. … Bh5 (55%, 1k games).
As you can see, while the BDG provides acceptable pratical results, it branches out fairly quickly, meaning the gambiteer out there will need to prepare for a large variety of lines. If you must start by one, the line with 5. … Bg4 is the one you can expect to meet the most.