Semi Slav and Computers

So....Im reading the chapter on the Semi-Slav Botvinnik and Anti-Moscow Variations-2 of the sharpest openings in chess in my book Starting Out:1.d4. I play out the moves on CB and turn on Fritz.  Many times, I notice the computer giving evaluations of -/+ or even -+ (decisive black advantage) when black is up material. OFten Cox (the author), states something like "...and white gets a strong attack...etc" but I see the computer evaluating the position as -/+ (~-1.1) which really troubles/scares me.

So my question is, is this normal and when should I trust the computers evaluation and moves b/c this evaluation is kinda just scary as sometimes white is down 3 pawns (with compensation based on the author) and the position is -/+...

Im not an avid pawn sac'er but I don't mind taking these lines up if white is at least slightly better (or is = but has better chances, etc etc)

Computers are pretty good for evaluating middle game. However, opening and endgame moves have consequences 20 moves down the line, so unless you set your computer to test every line that far out (which gets to be very long time), it is not giving you an accurate judgment.

Neither human nor computer evaluations play out the position for you. In practical terms this means if you don't understand how to play the position and don't feel the compensation: it isn't there!  But they sound like really interesting lines -- I'd work with the engine and do some analysis of your own and see if you can figure out how you feel about them. 

Computer preferences are notoriously unreliable in the opening -- but at the same time if the computer sees a +1.0 or more for the opponent... it's hard to imagine that's an easy position for you.

Can you put up some lines -- I'm curious about them and I'm probably not alone.

Computers generally are not great at evaluating sacrifices, but an important point is you will not be playing these positions against computers, so the defence will never be as accurate. You will find that in reality the attacker will win a far higher proportion of the games, despite what the computer says.

Here is a sideline that Black can use on move 9 where white is down so much material (3 full pawns...) and the author regards black's sideline as very dangerous for black but im not sure what white's compensation, other than development-but is this really worth those pawns??!

Its such a crazy position:

 

I think I agree with Cox: black's h-, g- and c6-pawns are weak, and the centre is about to open to white's advantage with black's king caught in the middle.  Black would be very lucky to come out alive.

you can't really trust the computers evaluations in positions that arise after the anti-moscow varation or the botvinnick. They are really irrational positions, I think black is fine in both variations. I stopped playing those lines, just because it's to difficult to play them accurately.

---

I think you mean white :P

Im not scared of being down material per say, im just scared of aimlessly wondering if I have sufficient compensation positionally with a good plan to follow. Ex)That above game that Rigamagician posted seemed pretty good for this variation-It seems that Black must return some material in order to attempt to hold his position.

But I think playing these types of positions will help me conquer my material deficit fear- Ive never ever played gambits (exclude QG) before so this "Anti-Moscow Gambit" will be the first somewhat 'real' one I might take up.

How do people who play the Botvinnnik or Anti-Moscow as white and/or black feel about them?

I prefer the white side of that position even down 3 pawns. Black has four very weak pawns (c6, f7, g4 h6), and he's behind enough on development he can't hope to hold onto more than 1-2 of them. Black has also fallen behind in development, has a huge dark-squared weakness, and found himself with knights instead of bishops in a rapidly opening position. The real deciding factor though (and the bit that leads me to disagree with Adam's statement that a Super G would prefer black) is that black honestly doesn't have any play. There's no attackable targets for him on the kingside, so he can't play there. He doesn't have any good squares for his knights in the centre. Playing there will be possible but difficult. His biggest advantage is queenside, but his pawns are really hindering his bishop there, and if he plays Qside he'll have to either castle short or keep his king in the centre, and neither of those options are attractive. So black's game plan is going to be to try and force weak squares in the centre, get some milage out of his knights, and try to head for a pawn up endgame with a q-side majority. The odds of this actually happening when he's got that many weaknesses in his own position and no initative are slim to none, and I think a super G would VASTLY prefer white here. *addendum* I was curious, so I fed the position into Stockfish (a stronger engine than fritz) and let it analyze while I posted. At present it's giving white a +0.44 and suggesting 16... Qc5+ as black's strongest continuation.

I played the black side for awhile, I think theoretically the openings lead to an equal position, but practically they favor white. blacks position is so awkward. Plus there are a lot of different lines for white to play that can be really dangerous if you are prepared. So you will have to know a ton of theory.

I play a line of the Modern that Fritz gives -1.67 to after 9 moves, good line interesting games fun and challenging to play.

So my question is are you a man or a machine ?