I generated a tree diagram of a minimax search versus an alpha beta search. Interesting to see how much gets pruned.

[JimmyRustles]

[zenpawn]

Very cool.

[amanjpro]

This assumes a very good move ordering... I hope my engine is somewhere in between the two

[j.t.]

This assumes a very good move ordering... I hope my engine is somewhere in between the two

On reddit he mentioned this uses no move ordering at all, which I find impressive. How impressive will depend on the concrete position this was generated from.

[amanjpro]

This assumes a very good move ordering... I hope my engine is somewhere in between the two

On reddit he mentioned this uses no move ordering at all, which I find impressive. How impressive will depend on the concrete position this was generated from.

It is indeed impressive

[Vinvin]

That remember me this video :
Comparison : Minimax vs A-B + move ordering

[pedrojdm2021]

Holy ... that is amazing! i always wanted to do something like that, great work!

[hgm]

This assumes a very good move ordering... I hope my engine is somewhere in between the two

On reddit he mentioned this uses no move ordering at all, which I find impressive. How impressive will depend on the concrete position this was generated from.

That something is mentioned doesn't mean it is true.

The diagram shows a tree where the best move is searched first, so that all other moves could be refuted by a single move, and that this refutation was found on the first try. The odds that this would happen with random move ordering are less than astronomically small.

Unless, of course, all moves are really equivalent. Which is what you would get when you evaluate every position as 0. Then every order will automatically have the/a best move fist in every possible permutation of the moves, as every move is the best one. But then they are ordered by score.

[Rein Halbersma]

This assumes a very good move ordering... I hope my engine is somewhere in between the two

On reddit he mentioned this uses no move ordering at all, which I find impressive. How impressive will depend on the concrete position this was generated from.

That something is mentioned doesn't mean it is true.

The diagram shows a tree where the best move is searched first, so that all other moves could be refuted by a single move, and that this refutation was found on the first try. The odds that this would happen with random move ordering are less than astronomically small.

Unless, of course, all moves are really equivalent. Which is what you would get when you evaluate every position as 0. Then every order will automatically have the best move permutation of the moves will automatically have the best one first, as every move is the best one. But then they are ordered by score.

The Knuth Moore paper that proved that perfectly ordered ab-searches have a branching factor b^(1/2) for a minimax tree of branching factor b, also showed that randomly ordered ab-searches have branching factor b^(3/4). So instead of searching 2x deeper you still search 4/3 deeper.

[chrisw]

This assumes a very good move ordering... I hope my engine is somewhere in between the two

On reddit he mentioned this uses no move ordering at all, which I find impressive. How impressive will depend on the concrete position this was generated from.

That something is mentioned doesn't mean it is true.

The diagram shows a tree where the best move is searched first, so that all other moves could be refuted by a single move, and that this refutation was found on the first try.

Look a little more carefully and you’ll observe this is not in fact true.

The odds that this would happen with random move ordering are less than astronomically small.

Unless, of course, all moves are really equivalent. Which is what you would get when you evaluate every position as 0. Then every order will automatically have the/a best move fist in every possible permutation of the moves, as every move is the best one. But then they are ordered by score.