I wouldn't go for 8...e5
in this position. When black plays e5
in Scheveningen/Najdorf and even in Smith-Morra gambit, it is usually connected with a win of tempo (attacking the knight on d4) or black is forced to do so because white is threatening e5
. This simply isn't the case in the position above, as 9. e5
won't do much harm.
Finally, the 8...e5
move can cause you a lot of trouble because of the weak d5 square and the a2-g8
diagonal. Therefore, I would suggest you leaving e5
for future.
8...Nf6
is a much more logical developing move and I would play it. Not surprisingly, it stands as the move #1 in my database with a great statistic (72.1% for black). For example:
[fen "r1bqk1nr/pp2bppp/2npp3/8/2B1P3/2N1BN2/PP3PPP/R2Q1RK1 b kq - 3 8"]
8...Nf6 9.Qe2 O-O 10.Rfd1 {and only then (when white threatens e5) play 10...e5 yourself} 10...e5
Another try for black is 8...a6
(covering the b5 square), but I think there will be enough time for that after finishing the development (if needed at all).
EDIT: As @Evargalo noted in the comments, in the 8...Nf6
line I gave above, you can also try 9...Ng4
with the idea of blocking the e5 break by putting a knight on e5:
[fen "r1bqk2r/pp2bppp/2nppn2/8/2B1P3/2N1BN2/PP2QPPP/R4RK1 b kq - 4 9"]
9...Ng4 10.Bf4 Nge5
The key idea of the manoeuvre is that black doesn't mind doubled pawns on the e-file, because he is pawn up and thanks to those pawns he can control the central squares d4
and d5
.
Other possibilities shown by @Evargalo are to deviate from standard set-ups in Smith-Morra gambit by playing 10...Qa5
or 10...Bd7
, I guess they are at least as good as 10...e5
so thanks to Evargalo for pointing them out.