The only engine I know that can "dumb itself down" successfully is HIARCS. It does have a rep as having very human like play. You might want to check it out.
Chess engine minicing below 1000
Sort:
Feb 22, 2020
0
#4
Yeah, the commercial HIARCS probably does it best. The now defunct commercial Delfi engine was almost as good as HIARCS. There is a free version of Delfi still available that runs only at full strength and 1000 elo. The freeware engines Rodent and Ufim weren't terrible.
The free Delfi engine and various low-rated versions of Rodent and Ufim are available in the Lucas Chess GUI:
https://lucaschess.pythonanywhere.com/
Feb 22, 2020
0
#5
suunnistus wrote:
I am rated at around 800 and what irritates me is that chess engines at lower level play so bizarre. I played the computer at like 1350 difficulty and fooled around and I could leave pieces undefended and under attack and the computer would not take it. I placed my queen in line with computers bishop and it didnt take it.
But this got me thinking. It should be possible to create a computer that actually plays like players at lower levels. Most of the computer seems to have been created in order to be as good as possible and calculate many many steps ahead. And it us really hard and complex.
It should, however, be easy to program a computer to play like a 600 or 800 ranked player. We see like 2-3 moves ahead, not more. Manuallt teach computer to sometimes detect pins, skewers, forks. And just mimic played gaimes. And teach it beginner opening strategys and behaviour. Like do too many pawn moves in the beginning and run around with the queen at 600.
An engine playing at 1350 elo should just about never miss a hanging queen. <cringe>
Feb 23, 2020
0
#6
EscherehcsE skrev:
suunnistus wrote:
I am rated at around 800 and what irritates me is that chess engines at lower level play so bizarre. I played the computer at like 1350 difficulty and fooled around and I could leave pieces undefended and under attack and the computer would not take it. I placed my queen in line with computers bishop and it didnt take it.
But this got me thinking. It should be possible to create a computer that actually plays like players at lower levels. Most of the computer seems to have been created in order to be as good as possible and calculate many many steps ahead. And it us really hard and complex.
It should, however, be easy to program a computer to play like a 600 or 800 ranked player. We see like 2-3 moves ahead, not more. Manuallt teach computer to sometimes detect pins, skewers, forks. And just mimic played gaimes. And teach it beginner opening strategys and behaviour. Like do too many pawn moves in the beginning and run around with the queen at 600.
An engine playing at 1350 elo should just about never miss a hanging queen. <cringe>
Yes its so wierd! Im at 800 and if I meet someone ranked 900+ they sometimes crush me, if I play against 1000+ I have like no chance. Yet I beat the 1350 engine.
I am rated at around 800 and what irritates me is that chess engines at lower level play so bizarre. I played the computer at like 1350 difficulty and fooled around and I could leave pieces undefended and under attack and the computer would not take it. I placed my queen in line with computers bishop and it didnt take it.
But this got me thinking. It should be possible to create a computer that actually plays like players at lower levels. Most of the computer seems to have been created in order to be as good as possible and calculate many many steps ahead. And it us really hard and complex.
It should, however, be easy to program a computer to play like a 600 or 800 ranked player. We see like 2-3 moves ahead, not more. Manuallt teach computer to sometimes detect pins, skewers, forks. And just mimic played gaimes. And teach it beginner opening strategys and behaviour. Like do too many pawn moves in the beginning and run around with the queen at 600.