I can't really comment on 5.g4 as nobody has ever played that against me- a quick check on Fritz shows that it is certainly interesting.  It wouldn't surprise me if Black does have improvements over the aforementioned analysis though.

How do you meet 3.Nf3?  Good question.  I personally meet it with the combative but dubious 3...Bg4?!, as opponents generally don't know White's ways to emerge with a strong += against it (many just reply with 4.d4 or 4.Be2).  If I reach a standard where opponents start giving me some turbulence, then perhaps just 3...Qxd5 transposing to 1.e4 d5 2.exd5 Qxd5 3.Nf3 Nf6.  This should not be a problem for the Portuguese gambiteer, as after 3.d4 Bg4 4.Nf3, Black usually plays 4...Qxd5 transposing to the same line with 4.d4 Bg4 inserted.  I think White has an edge in this line, but a combative game with castling on opposite sides usually results.

Okay, I'll send you the issue. Is your address still the same? With mobile Americans, you never know. E-mail or message, please.

According to an analysis by Michiel Wind in Kaissiber #32, pp. 26-30, 4. f3 Bf5 5. g4! followed by c4 is critical; the resulting ending is clearly better for White (+/-). So it seems that the line 3...Bg4, introduced in 1877 by Dr. Carl Göring (of Göring Gambit fame) and later reinvented by Pascal Vandevoort and Michel Jadoul from Belgium, is dubious.

Before I am going to delve into this - how do the Portuguese meet 3.Nf3 ?

I've looked back over the Portuguese and come up with some ideas- maybe not enough for equality, but perhaps enough to give Black active play instead of passivity and a riskless edge for White.

A) 1.e4 d5 2.exd5 Nf6 3.Bb5+ Bd7 4.Be2 Nxd5 5.d4

A1) 5...Nb6 (a suggestion of Tim McGrew) 6.Nf3 Bg4 (here the bishop may be better placed on g4 than on f5) 7.0-0 (or 7.a4 a5 8.Nc3 e6 9.0-0 Nc6 followed by ...Bd6 and ...0-0) 8.Re1 Nc6.  Here Black can again follow with 9...Bd6 and 10...0-0 with what looks to me like a decent game, or with the a-pawns not advanced, the riskier approach with ...Qd7 and ...0-0-0 may be playable.

If instead 6.Nc3 or 6.c4, then 6...e5 (McGrew) may playable, e.g. 6.Nc3 e5!? 7.dxe5 Nc6 8.Nf3 Qe7 9.0-0 0-0-0 with a combative game.

A2) Simply 5...Bf5 6.Nf3 Nc6 (instead of 6...e6) followed by 7...Qd6 and 8...0-0-0 and I think Black is okay in the resulting positions with castling on opposite sides.  If White tries to thwart this idea with 7.c4, 7...Ndb4 is strong, e.g. 8.Na3 e6 9.0-0 and Black can go ...Qd7 and ...0-0-0 after all.


B) 3.d4 Bg4 4.Bb5+ Nbd7 5.Be2 Bxe2 (the crunch line, as discussed in a thread a while ago)

B1) 6.Qxe2 Nxd5 7.Nf3 e6 8.0-0 N5f6 and now for example 9.Bg5 c6 10.Re1 Qc7, breaking the pin on the f6-knight.  The bishop goes to d6 and Black can castle on either side, with a fairly active position, some pressure down the c7-h2 diagonal.  Perhaps the most critical is 9.c4:

B1a)  9...Bd6 10.Nc3 c6 (10...0-0 11.Bg5 c6 12.Ne4 forces the bishop to retreat to e7) 11.d5 Qe7 12.dxe6 fxe6 (not 12...Qxe6 13.Qxe6+ fxe6 14.Re1 and Black's e-pawn is hard to defend) 13.Re1 e5, and Black can safely castle on the queenside.  If 12.dxc6 bxc6 Black should go kingside in view of the shattered Q-side, e.g. 13.Rd1 0-0 14.Bg5 h6 15.Bh4.  Black does have an isolated c-pawn but will gain control of the half-open b-file.

B1b) 9...c6 might be playable as well, e.g. 10.Bf4 Qa5, with the idea of transferring the queen to the kingside in some lines; bishop goes to b4 or e7.

B2) 6.Nxe2 (less critical, in my view) Nxd5 7.0-0 e6 8.c4 N5f6 9.Qb3 b6 10.Bf4 Bd6 11.Bxd6 cxd6 12.Qg3 Qc7 followed by ...0-0, and Black is slightly passive but has a strong centre.  Or 9.Nbc3 Bd6 (riskier, but maybe playable, is 9...c6 10.Bf4 Qa5 and 11...0-0-0) 10.Qb3 b6 with similar play.  Perhaps White can bag the bishop-pair with 11.Nb5 0-0 12.Nxd6 cxd6 but Black gets a strong centre and chances of an ...e5-break.

I might be missing something here, but Black seems to me to get fairly active play just by plonking the bishop on d6 (instead of e7), the queen goes to c7, e7 or a5 and the king can often go either side.

C) 3.d4 Bg4 4.f3 Bf5 5.Bb5+ Nbd7 6.c4 and I prefer the approaches with ...e6, for example 6...e6 7.dxe6 fxe6 (7...Bxe6 8.d5!) 8.Ne2 c6 9.Ba4 Qa5+ 10.Bd2 Qa6 (eyeing c4).  Is it enough?  I don't know.  What I do know is it looks like Black should have most of the fun.