Evaluation of coding skills

Discussion of chess software programming and technical issues.

Moderators: hgm, Rebel, chrisw

Fguy64
Posts: 814
Joined: Sat May 09, 2009 4:51 pm
Location: Toronto

Re: Evaluation of coding skills

Post by Fguy64 »

MattieShoes wrote:Branching factor is *huge*.

Lets say you have an A/B search with a branch factor of about 6. Lets say you search somewhere around a million nodes per second. That means the 10th ply of a search would take about 60 seconds to complete.

Would you rather double the speed of the engine or drop the branching factor from 6 to 5.5?

Well, doubling the speed gives you that 10th ply in 30 seconds. Reducing the branching factor to 5.5 gives you that 10th ply in 24 seconds. With each successive ply, the branching factor becomes more and more important relative to raw speed.

Of course, we want both, but generally with new engines, it's easier to shrink the branching factor a little bit than to speed the engine up a lot.
Right this makes a lot of sense. I guess this thread related to the fact that for me, sometimes the coding itself is a prime motivating factor. Coding can be an art form, having nice efficient code can be a goal in and of itself, in the same sense that musicianship and technical skill, as opposed to rhythm and melody, can be prime factors for appreciating a piece of music.