What % of the time is taking 3 pawns sufficient compensation for losing B or Kn?

I understand the joy of sacrifice.  Some of the most exciting games come from taking risks and sacking pieces and winning.  But I'm just wondering if taking out three pawns would be a good exchange for a 3 point piece as a typical strategy would tend to work out well in the hands of a reasonably competent player.   I guess that would be a good use of unrated blitz games ... to just try it a bunch of times and see how it works out.

It depends on the mobility of the pawns. Three mobile pawns are stronger than a piece, three blocked/restricted pawns are worth almost nothing.

And it opens up files. Is that what you need at time in the position? has your opponent rooks waiting to eye down them?

Call me crazy, but I'd say there is no compensation for losing a K.

if pawns arent being blocked by other pawns, thens its probably worth it. 

You're crazy.  Of course there is.  You could drink, smother your woes in wanton sex, gorge yourself on chocolate, any number of things.  But, :-} I corrected the subject line.  Thanks.

3.927%.  Roughly.

I don't understand questions like this, unless they're strictly out of pure curiosity. Every single position and situation in a case like this needs to be evaluated on its own merits, so what difference would any  percentage(s) make? 

I suppose one of the brand-name DBs like ChessBase or Chess Assistant may have a way to filter games with this particular criterion out of their DB and I know CB will work out the percentages for you from any list of games. As to determining the soundness of the sac, I'd just assume the games in which the sac won made it "worthwhile" and the losing games make it "unsound" but even that really requires an inspection of each game for errors and mistakes.

This is actually a rather interesting question. 

I would guess that to answer it effectively would require a fairly complex database query.

I'd want to exclude all known opening lines where such sacrifices are a matter of opening theory. Because we already know there is sufficient compensation in those lines. 

I'd also want to exclude all matches where the players are grossly out-matched merely by chess skill, so I'd want to only include games where the players are within 200 points or so of each other's ratings.

We'd want to toss out those games where we have reason to suspect the quality of play, so making sure at least one participant is 2500 would limit us to GM v GM and GM v IM games.

Then we'd want to get rid of games where the game ended in a loss within a few moves of the material imbalance being established -- because we could rightly assume that in those cases the material imbalance wasn't the point, but the tactical win was.

Then we could look at how many games that leaves, and see who wins most of the games, the side with the piece or the side with the pawns. We could also then look for structural similarities in the positions to draw out what features make for "equality" and what features favor one side or the other.

If anyone knows a (nearly) complete list of ECO codes that have such imbalances as a matter of theory, I can build the rest of the query in CQL and come back here with the results. 

Barring odd positions ,3 pawns can force draw end-game.Of course where game stands, players, all have something to do w/ it.

Oh, I know what you're saying, but you know what I mean, right?

I know each game is different, but I'm just wondering if there's any kind of 'rule of thumb', which I fully understand have plenty of exceptions even if they exist.  There are all kinds of chess wisdoms that are only guidelines.

I'm just wondering how often it is effectively used to sac a piece for a few pawns.  I win a lot of blitz games after losing a 3 point piece early on, but I'm wondering what kind of traction I could get if I considered destroying pawn structures more in sacs.  I know with stronger players that's a good way to get taken out back and smacked around, but anyhow.

That is way cool!  

I've never messed with chess DB's much besides the chess.com opening DB, and wouldn't know where to begin besides what you posted, although I would love to see the result and how it was achieved and start conducting my own investigations (and sharing), based on what I learned.

If you can't get a few more decimal positions into it, it isn't even worth posting.  Get real.

I'm supposing "N or Kn" means just Knight.

As a mean, in a non-specified position, If you give a Knight + 3 Pawns the cumulative value of 100%, then the Knight alone accounts for 52% and the 3 Pawns together for the remaining 48%.

(N=3.25; 3*P=3*1=3 ; 3.25:13=3:12 ; 13+12=25 ; 25:100=13:52  => N=52%)

In actual play it depends on the position (Is any of those Pawns passed? How advanced are those Pawns? Do they shield an otherwise underprotected King? How far is the endgame? Does the Knight dominate, or is dominated by, other pieces? Are there central or deep outposts for the Knight? etc.etc.etc.).

Oh damn. I just see what I want to see sometimes. I really wanted to post B and Kn, but I didn't notice my error until you said something. I'm definitely not GM material.  

Thanks for the feedback. I will think about all replies of quality like that.  Interesting to see how other people think about the problem.

Phase of the game is important.  Early on I'd rather have the piece.  Toward the end, I'd probably rather have the pawns, especially if they're connected/passed/far advanced or otherwise strong.

No problem, since the result doesn't change, at least if the exchanged Bishop is the last one: I added my mental calculations, which are based on the statistically derived mean values of N=B=3.25*P (the mean value of the Knight or Bishop is equal to 3¼ times the mean value of a Pawn).

But when you still have the Bishop pair, the mean value of the two bishops is 2Bs=7*P, so when you lose one of the Bishops of the pair you lose 7-3.25=3,75*P.

Hence, still speaking about some undetermined "mean" position, if by exchanging the Bishop for the three pawns you lose the Bishop pair (hence one Bishop remains on the board after the exchange), the proportion becomes 55.555% for the Bishop and the remaining 44.444% for the 3 Pawns.

-------

p.s.

I failed to find the way to edit my thread(s) title: how did you do it?

I like to look for these, sometimes they're good.  Much better chance of working in an endgame situation though.