Training for OTB play

[glav]

I would love to use a chess engine for training my over the board (OTB) play. However, they are often well above my level of comprehension (1900 Elo) and a I would need to seriously weaken them in to level the field. I am afraid that making some simply adjustment, say lower the number of searched nodes, would not make an ideal sparring partner. Do you have any recommendation for an engine I could use? Also, what type of adjustment would you make to make it a reasonable opponent?

[matejst]

The best software is probably Lucas Chess, where there are already a plethora of chess engines already prepared for OTB training. I can advise you to use the Maia engine, from Elo 1600 up. There are additional settings possible for Maia, but choosing an Elo is enough. For my part, I usually play against Maia 1900 with depth 4 added and some time bonus for the human player. I find it very realistic, and should be like a human player rated 2000 on serious time controls.

[CornfedForever]

A good bet would be the current version of Dragon - you can't set it to whatever level you want and Larry K put a good bit of time into getting the rating levels to about as good as one can.

[alfilete]

play to depth 5 and at the end (with only one piece, which is not a queen) go to depth 8

[BluRayVertex]

i have an elo adjustable engine with a myriad of GM personalities, which are adjustable, and as human as it is, it also uses a neural net trained on human games

[JohnW]

I like using old chess computers like a Mephisto or Fidelity product. Just seems more satisfying knowing I beat something that was playing as strong as it could versus playing against an engine I had to dumb down.

[BluRayVertex]

I like using old chess computers like a Mephisto or Fidelity product. Just seems more satisfying knowing I beat something that was playing as strong as it could versus playing against an engine I had to dumb down.

the one i use has customizable parameters via txt file so technically you could play it at its max for a certain personality

[PeterO]

Hello Blueray,

What ist the name of your engine? Where can I download it?

Thanks

Peter

[Chessqueen]

i have an elo adjustable engine with a myriad of GM personalities, which are adjustable, and as human as it is, it also uses a neural net trained on human games

Can you make it available for Android ?

[Chessqueen]

I would love to use a chess engine for training my over the board (OTB) play. However, they are often well above my level of comprehension (1900 Elo) and a I would need to seriously weaken them in to level the field. I am afraid that making some simply adjustment, say lower the number of searched nodes, would not make an ideal sparring partner. Do you have any recommendation for an engine I could use? Also, what type of adjustment would you make to make it a reasonable opponent?

You could start with Ajedrez Tactico 2.73b and if you can beat it, select the next up, and so on, and on until you find the engine that can challenge you the best and by selecting different engines you will play different style, for instance some engine are very tactics and other play solid strategical games Or better yet download Komodo 14.1 and set it to play at a skill level close to 1900, Mr. Larry Kaufman can give you the setting, by selecting Komodo it will play at your level , but more like a human chess player
=====> https://komodochess.com/
489 Deepov 0.4.1 64-bit 1901 +24 −24 44.1% +46.0 17.3% 677
58.6%
490 Surprise 4.3b13 1897 +22 −23 39.5% +82.1 19.4% 768
55.5%
491 Sapeli 1.02 64-bit 1895 +21 −22 47.9% +14.9 20.4% 796
69.7%
492 MOBMAT 0.903d 64-bit 1886 +25 −25 37.5% +110.3 16.0% 668
56.5%
493 GChess IV 0.2.0 1884 +21 −21 41.2% +64.4 28.8% 855
59.3%
494 CDrill 1800 build 4 1880 +23 −24 41.4% +73.0 15.5% 741
76.8%
495 Predateur 2.2 64-bit 1862 +41 −42 39.6% +79.3 18.1% 216
77.1%
496 Skiull 0.3 64-bit 1843 +23 −24 45.2% +35.7 24.8% 660
52.8%
497 BeRoChess 1.00 32-bit 1842 +37 −38 37.4% +98.2 22.9% 266
52.4%
498 Sayuri 2018.03.06 64-bit 1840 +26 −26 42.8% +57.0 16.0% 600
52.2%
499 Roce 0.0390 1839 +27 −27 43.4% +49.9 20.4% 539
51.6%
500 Pigeon 1.5.1 64-bit 1838 +27 −27 40.5% +70.1 26.1% 502
56.7%
501 Ajedrez Tactico 2.73b 1835 +26 −26 42.4% +61.9 18.6% 598