Are gambits solid at all?

WHY TAKE A PC

SERIOUSLY-JOIN THE

FANZY LAND WITHOUT

BIAZISM & Y START

KICKING PROGRAMZ MADE

FOR THE SAMME

CEASEFIRE......CW

PS IN THESE DAYS OF

FANZY SECRETIVES & LOTS

OF EMK FOR PROGRAMMES

GIVE THEM A CHANCES

TO...DIALOUGE Y.

I think Fritz is far more over-optimistic on pawn sacrifices than Rybka, I trust Rybka much more in its evaluation.

Interestingly I read somewhere that Rybka doesn't value pawns as +-1 in the opening. For example, play 1.Nf3 e5? and after 2.Nxe5 Rybka only evaluates it as +0.5/0.6 for White, despite the fact Black has no compensation whatsoever.

Even if they don't play well known opening gambits often, at the top level of play plenty of gambits are employed (pawn or piece sacs for other positional advantages)

Not all gambits are created equal...

Haha Reb :) Or, some gambits are more equal than others...

One reason for the relative lack of gambit play at the highest levels is that most gambits are very committal. GMs more often want games with the possibility of shifting from one plan to another depending on the response of their opponents. Many gambit-laden openings and early middle games simply blow away all but a single plan which, if incorrect or slightly misplayed, leaves the author out on a very long and thin limb with nowhere to retreat.

Still, there are those who don't mind taking that risk. (I'm one.) And at lower levels of play, the risks are diminished, especially if the gambit has given you activity and initiative because defensive skills are relatively weaker amongst amateurs.

So, I say, "Gambit on!" and let the GMs decide their own fate. We - the amateurs of the chess world - are still allowed to play ALL the openings, to try ALL the gambits, and to have a truck load of fun in the process.

Interesting insight. I really wanted to start a debate here and listen to peoples thought on gambits and so far it's going well . I personally haven't experimented too much with them, but I agree on the commitment issue you point out. And of course the fact computers don't play them too often may just mean that Gambits don't quite fit with their type of "thinking"; which is so different from ours.

Stick the first few moves of the Morra, Evans or KG into Fritz & leave for 10 minutes.  You'll see that engines do actually rather like 'em.

Tim McGrew of chesscafe's interesting gambit cartel articles theorizes a rather cute concept regarding gambits. He calls this the Caltrop-coefficient and it is  a measure of how "sharp" an opening's lines can get and how much precision they require one side to play or face total anihillation.

To explain what he meant, let me present an example that every engine/beast user can relate to.

Say you have a position 5-8 moves into the opening A that just happens to be good for white in certain lines.  Let's say uncle Fritzy was evaluating it across the top 5 lines in a position and the eval. scores read something like this....with say White to move.

1. +0.50 (Best move)
2. +0.32
3. +0.31
4. +0.25
5. +0.18 (5th best move)

Notice how the drop-off by playing the sub-par or sub-best moves is not completely punishing => 0.50-0.18 = 0.32 => 3/10ths of a pawn.

You might agree, that Opening "A" might have such an eval-spread for mostly ALL of the moves in the tabiya, i.e. main-line or variation. Nothing fancy.

Now let's take Opening "B". Evidently "B" is a shoddy unsound opening of ill-repute (perhaps a shady gambit...) that is initiated by White.

Now B is a tactical landmine for Black, who has to step carefully. On a large PERCENTAGE of positions in the opening tabiyas, the eval-spreads look like this (again 5-10 moves into the opening, with Black to move).

1. -0.60 ---------> Clearly indicating that the gambit in theory is unsound!!
2. -0.20    (better but still Black's favored)
3. +0.60   (White is happy)
4. +1.50   (white is really happy!)
5. +3.00   (murder! death! Kill!)

Okay...the spread above may be exaggerated and not common across all of opening B's tabiya, but the point to get is that => If the best move (-0.60 in favor of black) is EASY to find, then the opening is pretty much a lame duck.

but what if.... WHAT IF...it was TOUGH to find moves #1 and #2 and Black could play natural-looking moves and walk into the 3rd/4th/5th best moves most ofhe time.

Imagine if your opponent had to deal with this (find the best move or lose! ) 5-10 times out of say 15 moves in an opening he does not know too well?

A Strong Master or GM will probably either know the theory or come close to #1 or #2, but a sub-2000 player is very likely to step on these landmines, or caltrops as McGrew calls them.

So with that hypothesis, consider a choice between playing A and B. If you knew B better than most people and you didn't always face master-level players, you could really post a lot of wins with lines such as B's, however bad of a reputation the opening might have.

Gambits are clearly openings with such "high" caltrop-coefficients compared to stodgier openings like the Caro or Slav.

Just wanted to chime in ... I played gambits (and ONLY gambits) only until recently. Had a lot of success (and rating point scalpfests) because of them.  :)

On gambits:

There are three general methods in which a gambit can help a player's position. For a gambit to be sound it will typically have some degree of at least two of the following:

• Time gain: the player accepting the gambit must take time to procure the sacrificed material and possibly must use more time to reorganize his pieces after the material is taken.

• Generation of differential activity: Often a player accepting a gambit will decentralize his pieces or pawns and his poorly placed pieces will allow the gambiteer to place his own pieces and pawns on squares that may otherwise have been inaccessible. In addition, bishops and rooks can become more active simply because the loss of pawns often gives rise to open files and diagonals. Former world champion Mikhail Tal, one of the most extraordinary attacking players of the 20th century, once said that he had sacrificed a pawn just because "it was in his way."

• Generation of positional weaknesses: Finally, accepting a gambit may lead to a compromised pawn structure, holes or other positional deficiencies.

Does that proffessor has an IQ of 3217?

If you only play what computers play you arent a chess player imo. What would you have done before there were chess playing computers ? 

     I never implied I only play what computers do.  I guess you missunderstood the point.... I don't even use computers to analize my games.

   If you can take your time to read again my original post I'd love to know your opinion on gambits; since THAT's where I wanted the focus on. Thank you.

The main problem with a gambit, is that you "show your hand to early"

The opponent can make a long term strategi on a eraly stage.

 I think gambits have their place in chess. They are certainly fun/exciting to play and they teach us to play with the initiative/attack because usually you cannot enter a safe endgame since you will be down a pawn, or more. I have read about Spassky's life/career in chess and his start was a little unusual compared to most of the greats. Spassky himself said he played " as an old man" even when he was young. He didnt take risks and relied on his strong endgame ability and often ground his opponents down. I think it was his trainer Tolush who changed all this by teaching Spassky to attack and play with the initiative. In order to assure he wouldnt head into endings Tolush once insisted at one event Spassky played that he MUST gambit a pawn in all his games ! Spassky learned to sharpen his game and attack..... mainly throug the use of gambits it seems. Some gambits are far more risky than others. The Marshall counter attack in the Spanish for example is probably the most "correct/solid" gambit out there. Most GMs seem to agree that its good enough for a draw for black and when even Kasparov has been known to avoid it with a4 I think it must be true. I think Spassky's love for the king's gambit was born during his work under Tolush.

I think the primary issue here is that moves which are good for computers are not always good for humans. Just because Rybka can usually get compensation for the pawn sac does not mean you usually will be able to. Computers in general attempt to reach tactically rich positions, because they obviously thrive in these positions. On the other hand, computers are usually not nearly as effective in closed positions where their horizon severely hampers their ability to effectively orchestrate their pieces.

Of course, there are some tactics humans can find that computers cannot, but these are very uncommon.

I believe this is also the source of computers becoming "materialistic." This is a virtue, not a vice. In a large number of positions, humans will avoid taking material because the resulting position is "complicated," "uncertain," "trappy," or otherwise difficult for a human to evaluate. There is also the implicit assumption that your opponent knows what she is doing. Computers do not worry about this. Because they can see many moves ahead, they can usually determine if they can, for example, spare the tempo to take that pawn and still defend, or capture the knight with the h-pawn and not get crushed under the weight of h-file mating threats. If a dangerous line has only one refutation, the computer will find it, brilliantly saving an apparently lost position. But to the computer this is not a comeback, this is a line it has planned all along in a position that was winning from the start.

This will occasionally backfire, because computers do not actually search EVERY node, and they will sometimes miss mating lines, but usually these only show up in chess problems and other novelties, not actual games.

So I guess my point, at least originally, was that you should not naively apply human rules to computers (especially ones that always beat you!) or complain about them valuing pieces differently than you would. They are not valuing incorrectly, just differently, and hopefully in a way more suited to their style of play.

Queen's Gambit - Although many say it isn't a gambit, if you go in to variations such as the Catalan or the Botvinnik it is - and they are played by Kramnik and others.

I believe the Budapest also shows up at ~2600 ratings at times.

There are also poisoned-pawns variations, where sometimes a side could hold on to it.

I am not a prticularly strong player yet so I still fall into them and dont realize until afterward that it was bad move. I just see "oh look i'll be up a pawn" or even better "oh look, I'll be up two pawns", etc. I tend ot be overly agressive anyway - its a big problem of mine. Dumb...

If you are playing against a weak-to-intermediate player you can always get away with more gambit-plays ;-)

Interesting, I didn't know that. Maybe I should give gambits a shot too.... without expecting to become anything close to Spassky, of course.