personal experience regarding "self learning"

[deeds]

ah ok so it's no longer to know if you're on the right path or if you're far from the mark, it's just a story of curiosity.

In this case, you should test your experience file with the experience files of Khalid or Kevin (infinitychess), Alexchess or Igor (banksia), Sylvy (talkchess), etc.

[rulleeeee]

Well, the test was supposed to be out of curiosity how far I am from you. I'm on the wrong path

But first your training, and if you find time, I would like you to test my experience file

P.S. Your website is down. The one with your shop. It won't load the page

[deeds]

Shop hours : daily from 7:15 a.m. to 9:30 p.m. (=hours of my own nas in France).

[sarona]

How can the use of experience files/engines/books be considered cheating insofar as OTB play is concerned?

I suppose it's the same behavior between players on the board when the highest elo didn't won.

Disagree here when it comes to OTB play. I do not care about the servers or correspondence, so I have no comment on this regarding them.

Self learning is a tool that is accessible to all who seek it. Players use engines, books (and experience files) for opening preparation. I know a few FIDE rated players who have included them in the pre-tournament prep.

IMHO, they are no different than videos and books. As long as they are not used directly during a game, I see no ethical issues. Since there are no longer adjournments, I see no problems.

[deeds]

Disagree here when it comes to OTB play. I do not care about the servers or correspondence, so I have no comment on this regarding them.

Hey Sarona, maybe you didn't see that I made a joke when I said :

All the technologies which can beat sf will be firstly considered as cheating devices, unfair conditions, sss, etc until they will be added to sf (e.g. lazysmp, syzygy, nnue, etc.).

I suppose it's the same behavior between players on the board when the highest elo didn't won.

All engine's trainers don't care about whether Stockfish is beaten or not, a few months/years later, the new technology will be integrated into SF loool

[sarona]

Yes, sorry Chris. . .

My newborn seems to live a nocturnal life and I have not slept a full night since her arrival home. She sleeps peacefully in the daytime, though. . .

So my concentration level/focus may be compromised.

[deeds]

Hi Sarona,

You should reread the manufacturer's documentation, maybe it's just a small adjustment of its time zone. It still happens on newer models with default settings.

At OCF too, they no longer sleep normally since the birth of the "personalities trading bank" loool

[rulleeeee]

This was a little test done: my exp vs exp deeds

if there is interest, it will provide games.pgn

[deeds]

Hi Chris on your website there are screenshots that show the engines changing moves long before they've learned 500 games
How did you speed up their learning process

Hi, i assume you are referring to (e.g. ls3 - a00 c2c4) :


where the engines changed their main moves after only about 100-140 games...

In fact, I often use one of the tricks well known by all engine trainers :

...before starting the training, (sometimes) i quickly pre-analyze* the opening with Eman to get a kind of "main line" in its experience file...This tells Eman immediately there are 1 or 2 main moves in each "key" position of this opening rather than having 7 or 8 moves where Eman will have to play several dozen times each before having an idea of their effectiveness...

*pre-analyze :
- some make a small tournament between engines on an opening from scratch (=i mean their experience files contain no experience data on this opening)
- some run "selfplay" analyzes at fixed depth (e.g. D28 for me when i train Eman/BrainLearn) or with a very short time/move (e.g. 10sec/move for most of the AM's engines trainers i know)
- i use Commenter PGN (10min/move) to analyze the lost games
- i also use Experience Viewer (1 : all legal moves)
- etc

[deeds]

In line with this :
- concerning those who still haven't understood anything, anyway now they have missed the train of learning so might as well put an end to this sharing of experience in public
- regarding those who understood the threat that engines with learning feature posed to their engines or their opening books, no interest in explaining it again and again, the next big thing in chess is already here

Good luck to the new engine trainers !