calculating seldepth

Discussion of chess software programming and technical issues.

Moderators: hgm, Harvey Williamson, bob

Forum rules
This textbox is used to restore diagrams posted with the [d] tag before the upgrade.
Post Reply
AxolotlFever
Posts: 50
Joined: Sun Nov 11, 2018 8:26 pm
Location: Germany
Full name: Louis Mackenzie-Smith
Contact:

calculating seldepth

Post by AxolotlFever » Sat Apr 27, 2019 12:38 pm

Hi all,

One aspect of the uci protocol I am not 100% sure about is seldepth. How do you calculate this? Is it the maximum depth qsearch has reached? How does it combine with extensions?

Side question: should the depth we send in info string should be the iterative deepening depth, or should we also account for extensions?

Thank you very much,
Louis

abulmo2
Posts: 186
Joined: Fri Dec 16, 2016 10:04 am
Contact:

Re: calculating seldepth

Post by abulmo2 » Sat Apr 27, 2019 3:24 pm

AxolotlFever wrote:
Sat Apr 27, 2019 12:38 pm
One aspect of the uci protocol I am not 100% sure about is seldepth. How do you calculate this? Is it the maximum depth qsearch has reached?
In Amoeba, I compute it as the maximum depth reached in qsearch.
Side question: should the depth we send in info string should be the iterative deepening depth?
I use the iterated depth.
I also agree that both numbers, depth & seldepth, are barely related to the real depth reached in average during the search, due to extensions & reductions.
Richard Delorme

elcabesa
Posts: 806
Joined: Sun May 23, 2010 11:32 am
Contact:

Re: calculating seldepth

Post by elcabesa » Sat Apr 27, 2019 9:03 pm

some engine calculate seldepth as the max depth reached in search/qsearch, some other (i.e. stockfish) calculate the max depth reached by a PV node

Post Reply