FREE - In Google Play
FREE - in Win Phone Store
Here is Kasparov's Evans Gambit win over Anand:
From most to least sound:
1) Evans Gambit - played at all levels. White has good compensation.
2) Smith-Morra Gambit - White gets some compensation, so it's playable. But it's difficult to justify. Not great, but overpowered at club level because of tricksies.
3) Blackmar-Diemer Gambit - Similar to Smith Morra. White struggles to justify the pawn in some lines but isn't outright -/+ in any line. Again, overpowered at club level because of tricksies.
4) Blackburne Shilling Gambit - Very bad for Black. Not a true gambit in the sense that if White tries to take the pawn on e5, he will lose. But if white simply castles or takes the knight he will have a significant developmental advantage.
Here's Spassky playing a delayed Blackmar-Diemer Gambit in the Candidates.
So Spassky played a good version of the BDG in a 1956 game and could only draw with it. Short, as white, drew with it. The Blackmar Deimar probably isn't losing, but it throws away white's advantage entirely.
Having said that, I like Fear-Itself's quote, "The BDG is the quickest way to 2000, in both directions."
I do hate to face it against a well-prepared BDG specialist in blitz games. In slower time controls, I like my chances, even against the specialists.
Avrukh in his new book effectively dismantles the BDG completely by using the usual recipe (The Ziegler/ O'Kelly system).