<Jimmy Cagney> stars in the original, and it's an over the top performance, just as <Pacino> is over the top.

I concur,

<Fischer vs Larsen, 1958> can be very instructive for us I think in a prophylactic sense. They both played solid book up to move 9 (Sicilian Dragon) and then Larsen chose the obscure 9...Nxd4 which does not have very good stats in the CG data base. I think the equally obscure 9... Na5 looks interesting, although with few games in the data-base, it might be risky, the stats are actually better than they appear Opening Explorer , look at the next page and although there are fewer games(20), the stats compare favorably with the main line 9... Bd7.

Then again at move 10... Be6? which has been thoroughly refuted since 1958 and not just by Fischer, but we don't have to play that line, 10... Qa5 shows 14 games with stats dead even!

Once again, move 11, Larsen chooses the move which is listed as the main line (Qa5), but which has been refuted by the stats and once again there seems to be a viable option (Qd7) although less than a handful of games to choose from.

I think we have a lot of potentially robust but unproven variations to choose from in the Dragon. I have had some success with it myself in correspondence chess (albeit VERY limted, just a handful of games).

Then I decided to take a long look at the position after 15.Bxd5, the last "book" move of the game. Larsen here plays 15...Rad1, which looks sound on general principles, but he seems to have overlooked what to me is a superior move, 15... Nxd5. The tactical complications of this move are considerable (to me at least), and perhaps that is why larsen avoided it, but it seems to give a slight advantage to Black after the skirmishing is over (minor pieces are all gone and black has a superior pawn structure), then again it is highly likely that I missed some critical move ....

I have no idea, I was wondering that myself. I hope he didn't OD from chess :-))

3.Bc4 is the other main one.

I have been having a few thoughts on these, since our immediate reply to 3.d4 doesn't seem to be taxing people's brains too hard ;)

(i) 3.Bb5+ (Moscow) gives us a few dilemmas. ...Nd7 seems the most logical, but it is tricky, and I for one don't know any theory. I have tried it once but got into a very closed and uncomfortable position. If we are going to play it, homework is necessary. ...Bd7 seems the most popular reply, and the only one I have theory on (from Rogozenko). Any thoughts?

(ii) 3.c3 Nf6, then either Bd3 (Kopek) or Be2 (Basman). (4.h3 is also playable.) I have been doing a bit of reading up on these, which I will put in a separate post.

(iii) 3.Nc3 intending transposition to an Open Sicilian but trying to confuse us. We need to be clear about what Open Sicilian line we want to play. If white seriously means it as an anti-sicilian it is a relatively harmless one. Our main immediate choice is whether to allow the transposition to the Open Sicilian or to block it with ...e5.

(iv) 3.Bc4 - an idea of Adams, planning d3 and a closed struggle in the centre. Rogozenko recommends 3 ...Nf6, 4.d3 Nc6, 5.c3 g6. Surprisingly (to my amateur mind) ...e6 does not figure in this, leaving the white bishop free to range on its diagonal.

These seem to be the only important ones, but there are also:

(v) 3.b3 Rogozenko doesn't think highly of this, but gives 3 ...Nf6, 4.Bb5+ Nbd7!?.

(vi) 3.b4 transposes into mainline Wing Gambit, where white virtually has to follow up with Nf3 anyway. I guess we just take it?

<(iv) 3.Bc4 - an idea of Adams, planning d3 and a closed struggle in the centre. Rogozenko recommends 3 ...Nf6, 4.d3 Nc6, 5.c3 g6. Surprisingly (to my amateur mind) ...e6 does not figure in this, leaving the white bishop free to range on its diagonal.>