Depth 15 is nothing...I use Houdini 3 and on my hardware it reaches depth 25 pretty quickly. Here is the most prominent suggested Houdini improvements for White:
17.Bxd4 (Irontiger's idea) cxd4 18.Nd2 gives White +0.68 at depth 25. Even 17.b4 is better than 17.Nd2.
How much did you pay Critter to play that move? :)
Depth 15 took forever...sounds like I ought to use better hardware. 
I doubt that many humans are calculating every variation 25 moves deep.
GreenCastleBlock wrote:
There is also the so-called "Anti-Portugese" 3.Nf3, with the idea that if Black is taking on d5, White will play d4+c4 as normal, while if Black gambits the pawn White can profit from his ability to play with a pawn on d3 and not d4.
Specifically 3...Bg4 4.Bb5+ Nbd7 5.h3! has been offered by Silman as being good for White.
Please post Portuguese lined in the Portuguese thread.
I think you have that Silman line wrong by the way, as White's queen is en prise!
Depth 15 took forever...sounds like I ought to use better hardware.
I doubt that many humans are calculating every variation 25 moves deep.
No, but their static evaluation is much, much better than computer's.
Example : if I reach the following position in my calculation, I know the line is refuted, even if the material is equal and will possibly be for a few moves. This just looks too awful.
GreenCastleBlock wrote:
There is also the so-called "Anti-Portugese" 3.Nf3, with the idea that if Black is taking on d5, White will play d4+c4 as normal, while if Black gambits the pawn White can profit from his ability to play with a pawn on d3 and not d4.
Specifically 3...Bg4 4.Bb5+ Nbd7 5.h3! has been offered by Silman as being good for White.
Please post Portuguese lined in the Portuguese thread.
I think you have that Silman line wrong by the way, as White's queen is en prise!
I was discussing the position after 3.Nf3 and not 3.d4. Read whole post please.
Has anyone mentioned this line? To me, it looks better for White…
1.e4 d5 2.exd5 Nf6 3.Nf3 Nxd5 4.d4 Bg4 5.Be2 e6 6.O-O Nd7 7.c4
Ah OK. Sorry.
You'd need to look at the lines where Black pulls the Bishop back to d7 of course, as well as not playing Bg4 in the first place.
You are likely to have the options
1) to take the pawn and get the usual cramped position in the mainline Scandy Modern (with white's N already committed to f3 if that makes a difference)
2) go the Bd7 route and end up saccing the c6 pawn to attack the hole on d4 - which would probably be my preference in blitz
3) follow Silman's Bg4, h3.. line.
I'll have a look at those lines tonight - but will put them in the Portuguese Gambit specific thread that is already running not this one.
FYI, the variation noted in post #48 is the Modern Variation, Gipslis Variation...
Back to the Icelandic Gambit:
I got Rybka 2.2 and ran analysis after ...Bxe6. Here is Rybka's line. Surprisingly, it showed c5! as a response to Nf3. I had never considered this line before, but Rybka seems to think that black has some compensation:
I don't understand all the subtleties of this position, especially why 7. Be3 was played followed immediately by 8. Bf4. Does anyone understand it?
And Expertise87, what's your refutation to 5. ...c5?
Silman's line.
This looks quite fun:
[Event ""]
[Site "Shredder Android"]
[Date "2013.06.06"]
[Round ""]
[White "Player"]
[Black "Shredder"]
[WhiteElo "1909"]
[BlackElo "1899"]
[Comment "8:08"]
[Result "*"]
1.e4 d5 2.exd5 Nf6 3.Nf3 Bg4 4.Bb5+ Nbd7 5.h3 a6 6.hxg4 axb5
7.Qe2 Nxd5 8.Qxb5 e6 9.Qxb7 Nb4 10.Kd1 *
Silman's line.
This looks quite fun:
[Event ""]
[Site "Shredder Android"]
[Date "2013.06.06"]
[Round ""]
[White "Player"]
[Black "Shredder"]
[WhiteElo "1909"]
[BlackElo "1899"]
[Comment "8:08"]
[Result "*"]
1.e4 d5 2.exd5 Nf6 3.Nf3 Bg4 4.Bb5+ Nbd7 5.h3 a6 6.hxg4 axb5
7.Qe2 Nxd5 8.Qxb5 e6 9.Qxb7 Nb4 10.Kd1 *
As a diagram:
Wow...I wish I could make an entire thematic tournament from that position.
Back to the Icelandic Gambit:
I got Rybka 2.2 and ran analysis after ...Bxe6. Here is Rybka's line. Surprisingly, it showed c5! as a response to Nf3. I had never considered this line before, but Rybka seems to think that black has some compensation:
I don't understand all the subtleties of this position, especially why 7. Be3 was played followed immediately by 8. Bf4. Does anyone understand it?
And Expertise87, what's your refutation to 5. ...c5?
...c5 is not surprising to me. After Nf3 (unlike after d4) Black has the possibility to blockade on d4, which is what this move achieves.
This being said, 7....Nd4 ?? is completely inconsistant with the plan to pressure on d3/d4 because it closes the file. White is completely winning in the end position, because the d4 pawn will be a pain to defend (even if the computer does not see it, I can se it).
...c5 is not surprising to me. After Nf3 (unlike after d4) Black has the possibility to blockade on d4, which is what this move achieves.
This being said, 7....Nd4 ?? is completely inconsistant with the plan to pressure on d3/d4 because it closes the file. White is completely winning in the end position, because the d4 pawn will be a pain to defend (even if the computer does not see it, I can se it).
I see your point and agree, but I can't find a better move. Do you see any better alternatives?
...c5 is not surprising to me. After Nf3 (unlike after d4) Black has the possibility to blockade on d4, which is what this move achieves.
This being said, 7....Nd4 ?? is completely inconsistant with the plan to pressure on d3/d4 because it closes the file. White is completely winning in the end position, because the d4 pawn will be a pain to defend (even if the computer does not see it, I can se it).
I see your point and agree, but I can't find a better move. Do you see any better alternatives?
Well, anything that does not block the d file, for instance ...Qc7 followed by ...Rd8.
Don't ask me to find a good plan for Black, I'm not sure there is one. But certainly he could try to get something for the pawn at least.
Perhaps a kingside fianchetto would benefit Black?
I still like my original line. I've added a response to 8. a3.
Join the 3...Qd6 or the 2...Nf6 Scandinavian Club at Chess.com. These are probably the most active clubs for this Defense.
I play both Icelandic and panov. I love them both. I think panov is slightly stronger, but it keeps it fun and interesting to switch it up. My main hope is always to get into the Marshall Kiel lines, that is as fun as it gets if you ask me. 3 brilliants in the first dozen moves and so many different ways to win.
There is also the so-called "Anti-Portugese" 3.Nf3, with the idea that if Black is taking on d5, White will play d4+c4 as normal, while if Black gambits the pawn White can profit from his ability to play with a pawn on d3 and not d4.
Specifically 3...Bg4 4.Bb5+ Nbd7 5.h3! has been offered by Silman as being good for White.