Bishops vs knights. Stockfish prefer bishops!

There has been a couple of discussions regarding this in the forums and I wanted to settle this for the players that question this. Although stockfish isn't flawless in it's suggestions it is obviously better than any human player at this moment. So what does stockfish prefer? BISHOPS.

I went to the board editor and removed the white knights and the black bishops and the evaluation was +0.7 in whites favor. Then I repeated the experiment but this time I removed the white bishops and black knights, this time the evaluation was -0.3 in blacks favor. (Please watch the images below).

I repeat, obviously stockfish isn't flawless, but if the priority of bishops is good enough for stockfish it should obviously be good enough for human players.

That might be because Stockfish evaluates how cluttered the board is, and bishops have an advantage when it is less cluttered. Add some clutter and knights become better

And as usual, this simply isn't relevant. It is trivial to show Knights are better than Bishops in some positions. Bishops are better than Knights in others.

Open Positions tend to favor Bishops. Computers tend to play for open positions where Bishops would be better.

BTW, one of the better ways to beat computers is to play for closed positions.

Also, because of the way you tested, it is quite possible there is some inherent weakness about the setup position that the computer is exploiting. An obvious one that comes to mind is that without the White Bishop the b2 and g2 squares are very weak (b7 and g7 for Black). There is no such corresponding weakness with the Knights gone in this position.

I get what you two are saying, but depending on the position knights are even better than rooks. I am not talking about unique position, I am talking about the fact that every piece is given a value and that the value of the bishops should be higher than knights because they are statistically a better piece. Fisher for an example wanted to increase the value of bishops to 3.5.

Someone recently did a test and showed the results here using a large number of actual game positions. Their conclusion was that basically Knights and Bishops should be valued the same. What did have a distinct value was having the two bishops. That was worth about .5 pawns. Incidentally, in the position you have setup, the favored side has the two bishops! So much of the better positional evaluation could be attributed to that.

It might be interesting to repeat your test but using just one bishop instead of the two and see what happens. You would still have the weakness I pointed out though.

I like when people value bishops at 3.5, so they're clingy to their bishops and I get an amazing position.

Stockfish prefers bishops because its easier to open a position than it is to close it. You cant force a position closed unless your opponent is willing, however you can open it even if they are unwilling. In the endgame its also more common for pawns to be on both sides of the board, which makes bishops a lot stronger than knights.

The study I was thinking of is actually fairly old now and was based on something like 8,000,000 positions. It was done by GM Larry Kaufman of chess computer fame. I have seen an article with more specific breakdowns, but here is what I could find more immediately.

https://www.chess.com/article/view/the-evaluation-of-material-imbalances-by-im-larry-kaufman

This is stockfish's evaluation!  Click on "Material column>>> then piece square value mg(middle game) and piece square value eg(endgame)

Here is full evaluation page.

https://hxim.github.io/Stockfish-Evaluation-Guide/

In right chess board, in every position you move, you can see changes in static evaluation per move. ( Actual stockfish evaluation is dynamic evaluation at the end of principal variation which may varies every second and every change in depth)

In middle game, kt value 764  vs B value 826

In endgame ,     Kt value  848  vs B value  891.

Bishops are mostly better as they mostly control a greater amount of squares than knights.

Based on the position that I had setup stockfish seem to agree with your statement. Meaning that one bishop versus one knight seems to be equal. 

Interesting, will check it out later!

Thanks for sharing! Very interesting!

Thanks for sharing!

Having delivered numerous Knight forks, and also having been the victim of countless Knight forks, I can tell you that Knights are tricky beasts!

Yes they are tricky. But based on the info shared by jonathanpiano13 and drmrboss my original title is correct. Stockfish does in fact favor bishops.

Oh yes, StockFish and Bobby Fischer both favored bishops.  In fact, Bobby Fischer prejudiced me against Knights, lol.  I would pin knights with my bishop, and when my bishop got kicked by the rook pawn, I never, NEVER, captured the knight.  

But now, sometimes I do.  I've become more appreciative of knights, and their abilities.   I wonder if there are players who know how to play against players who have too much fondness for bishops over knights, and what the method or strategy is to exploit Bishop-loving fanatics.