Page 3 of 3
Re: Piece values
Posted: Wed May 26, 2021 4:50 pm
by soheil_hooshdaran
Nick Burrows wrote: ↑Thu May 20, 2021 8:59 pm
John Upham wrote: ↑Wed May 19, 2021 11:56 am
Does EK discuss the changing value of a pawn as it progresses from the 2nd to the 7th rank?
He has a paragraph saying in certain positions a pawn on the seventh rank is worth at least 3 pawns, but trying to work out when its worth 2 pawns etc is not worthwhile.
A fast/dangerous/well supported passed pawn is likely more valuable than an extra pawn for the other side
who is EK?
Re: Piece values
Posted: Fri May 28, 2021 11:37 pm
by John Clarke
Erik Kislik. Read the earlier posts.
Re: Piece values
Posted: Sat May 29, 2021 12:08 pm
by MJMcCready
It does beg the question that if the numerical valuation of pieces is so easily drawn into question, why we persist with it. I did overhear Howell mention in the commentary the other day that Alphazero does not adhere to that system.
Re: Piece values
Posted: Sat May 29, 2021 5:58 pm
by IM Jack Rudd
We persist with it because it's a decent first-level approximation that can help us get to the level where we start realizing its shortcomings.
Re: Piece values
Posted: Sat May 29, 2021 11:05 pm
by MJMcCready
Yes you have a point there, a sort of ''rough guide'' if you like.
Re: Piece values
Posted: Sun May 30, 2021 2:05 pm
by Nick Ivell
No one pretends it's anything other than a rough guide.
If it stops beginners going a4 e5; Ra3, it serves its purpose.
Re: Piece values
Posted: Mon May 31, 2021 4:32 pm
by MJMcCready
Well yes and no, computers use it and have a numerical value system in place, well except Alphazero according to James Howell.
Re: Piece values
Posted: Mon May 31, 2021 6:10 pm
by Wadih Khoury
MJMcCready wrote: ↑Mon May 31, 2021 4:32 pm
Well yes and no, computers use it and have a numerical value system in place, well except Alphazero according to James Howell.
Computers take also into account their activity, so they do nuance it quite a bit.