# Computer Chess Club Archives

## Messages

### Subject: Re: Fail high reductions

Author: Omid David Tabibi

Date: 08:09:21 07/02/03

Go up one level in this thread

```On July 02, 2003 at 10:49:41, Gerd Isenberg wrote:

>On July 02, 2003 at 07:22:00, Georg v. Zimmermann wrote:
>
>>On July 01, 2003 at 17:34:47, Russell Reagan wrote:
>>
>>>From "Fail High Reductions by Rainer Feldmann"
>>>
>>>"...a fail high node is a node 'v' with a search window of [alpha,beta] at which
>>>a static evaluation function 'c' produces a cutoff. The FHR-algorithm reduces
>>>the search depths at these fail high nodes thus searching their subtrees with
>>>less effort."
>>>
>>>Their subtrees? I thought fail high nodes didn't have subtrees, and that you
>>>return beta at a fail high node. I must be misunderstanding something. Could
>>>someone give a simple explaination of how fail high reductions work?
>>
>>IMHO Rainer Feldmann uses bad terminology. A fail high node is - at least by my
>>definition - indeed a node where one subtree returns a value above beta, you
>>therefor "fail high" and return (value or beta, depending on if you use fail
>>soft).
>>What he intends to say is probably : " a fail high REDUCTION node is a node 'v'
>>with a search window of [alpha,beta] at which
>>>a static evaluation function 'c' produces a cutoff. "
>>
>>The technic he describes sounds a lot more error prone than null move to me, at
>>least in tactical situations.
>>
>>Georg
>
>Hi George,
>
>If i remember well, Rainer Feldmann's FHR is based on the NullMove observation.
>Instead of foreward pruning, FHR reduce depth if a NullMove fail high occurs.

FHR reduces the depth if static eval >= beta. I think you are confusing FHR with
verified null-move pruning; the latter reduces the depth when null_move_score >=
beta (and cuts off immediately if null_move_score >= beta in the subtree of a
fail-high reported node).

>
>Regards,
>Gerd

```