2500! Not such an exaggeration. And how computer chess helps

Discussion of anything and everything relating to chess playing software and machines.

Moderators: hgm, Rebel, chrisw

BrendanJNorman
Posts: 2526
Joined: Mon Feb 08, 2016 12:43 am
Full name: Brendan J Norman

2500! Not such an exaggeration. And how computer chess helps

Post by BrendanJNorman »

A few weeks ago, I mentioned that my current chess goal was to reach a 2300 rating on chess24.com, but when this was achieved, I remarked that 2500 was a realistic target.

Somebody, as is usual on this forum, attacked me, twisted my words (mentioned my FIDE rating, which is irrelevant when talking about a goal for an online rating) and labelled me a liar, saying that I was exaggerating my chess ability.

Instead of saying "hmm that seems a bit high - good luck!", as a normal person would - he called me a liar, twisted my words (changing topic of discussion from online rating to FIDE rating) and mocked me. Standard TalkChess practice.

Nevertheless, and regardless of what cynics say, within a few more games, I hit my target.

Image

How does this relate to computer chess?

I have improved my chess by a few hundred points in the last few years, despite not playing a tournament since 2010, all as a direct result of only training with engines.

This is all as a result of three things.

- Playing against them (this results in a strong law of contrast coming into effect when you return to playing humans - in effect, the human "feels" weaker.)

- Analyzing their games played in my openings (from deliberately created theme tournaments)

- Creating personalities of all playing styles (mostly with Rodent) and seeing how strong a given personality can become, no matter how extreme the settings are.

A direct result of all of this has been a wide broadening of my knowledge of the many varied ways of handling positions, that previously, I'd thought could only be played in a specific way.

Especially ways that a position, even a quiet one, can quickly be turned dynamic in a way that human players quickly drown in.

Of course, the next step is to learn some reliable openings, and enter some LTC tournaments again.

When I have the time. :?

P.S Here is a game from this morning, where, I blundered my queen in the opening by playing quickly "without looking" - and when the splash of cold water that was the queen loss took effect, I focussed.

The tactical onslaught that came from my opponent's only real blunder, could well have been played by one of my engine "personalities". :twisted:

[pgn][Event "chess24 online game | blitz"]
[Site "chess24.com"]
[Date "2018.02.24"]
[Round "?"]
[White "ChessnCognac"]
[Black "ArturoMarnCorresa"]
[Result "1-0"]
[BlackCountry "es"]
[BlackElo "2308"]
[ECO "A80"]
[Time "01:03:35"]
[TimeControl "300"]
[WhiteCountry "au"]
[WhiteElo "2486"]

1.d4 f5 2.g4 { [%clk 0:04:57] } 2...fxg4 { [%clk 0:04:56] } 3.h3 { [%clk 0:04:55] } 3...d6 { [%clk 0:04:54] } 4.e4 { [%clk 0:04:50] } 4...c6 { [%clk 0:04:51] } 5.hxg4 { [%clk 0:04:47] } 5...Nf6 { [%clk 0:04:47] } 6.f3 { [%clk 0:04:43] } 6...e5 { [%clk 0:04:43] } 7.g5 { [%clk 0:04:38] } 7...Nfd7 { [%clk 0:04:29] } 8.f4 { [%clk 0:04:09] } 8...exd4 { [%clk 0:04:08] } 9.Qxd4 { [%clk 0:04:01] } 9...Qb6 { [%clk 0:04:06] } 10.Qc3 { [%clk 0:03:50] } 10...d5 { [%clk 0:04:01] } 11.Be2 { [%clk 0:03:40] } 11...Bb4 { [%clk 0:03:58] } 12.Nd2 { [%clk 0:03:24] } 12...Bxc3 { [%clk 0:03:57] } 13.bxc3 { [%clk 0:03:22] } 13...Nc5 { [%clk 0:03:52] } 14.Ngf3 { [%clk 0:02:57] } 14...dxe4 { [%clk 0:03:47] } 15.Ne5 { [%clk 0:02:54] } 15...Bf5 { [%clk 0:03:45] } 16.Ndc4 { [%clk 0:02:49] } 16...Qd8 { [%clk 0:03:36] } 17.Ba3 { [%clk 0:02:42] } 17...Nbd7 { [%clk 0:03:33] } 18.Nd6+ { [%clk 0:02:39] } 18...Kf8 { [%clk 0:03:17] } 19.Nxf5 { [%clk 0:02:35] } 19...b6 { [%clk 0:03:12] } 20.Ng6+ { [%clk 0:02:28] } 20...Kf7 { [%clk 0:02:56] } 21.Bc4+ { [%clk 0:01:55] } 21...Kxg6 { [%clk 0:02:52] } 22.Nh4+ { [%clk 0:01:52] } 22...Kh5 { [%clk 0:02:51] } 23.Be2# { [%clk 0:01:49] } 1-0[/pgn]

The opponent is over 2300 by the way.
carldaman
Posts: 2283
Joined: Sat Jun 02, 2012 2:13 am

Re: 2500! Not such an exaggeration. And how computer chess h

Post by carldaman »

BrendanJNorman wrote:A few weeks ago, I mentioned that my current chess goal was to reach a 2300 rating on chess24.com, but when this was achieved, I remarked that 2500 was a realistic target.

Somebody, as is usual on this forum, attacked me, twisted my words (mentioned my FIDE rating, which is irrelevant when talking about a goal for an online rating) and labelled me a liar, saying that I was exaggerating my chess ability.

Instead of saying "hmm that seems a bit high - good luck!", as a normal person would - he called me a liar, twisted my words (changing topic of discussion from online rating to FIDE rating) and mocked me. Standard TalkChess practice.

Nevertheless, and regardless of what cynics say, within a few more games, I hit my target.

Image

How does this relate to computer chess?

I have improved my chess by a few hundred points in the last few years, despite not playing a tournament since 2010, all as a direct result of only training with engines.

This is all as a result of three things.

- Playing against them (this results in a strong law of contrast coming into effect when you return to playing humans - in effect, the human "feels" weaker.)

- Analyzing their games played in my openings (from deliberately created theme tournaments)

- Creating personalities of all playing styles (mostly with Rodent) and seeing how strong a given personality can become, no matter how extreme the settings are.

A direct result of all of this has been a wide broadening of my knowledge of the many varied ways of handling positions, that previously, I'd thought could only be played in a specific way.

Especially ways that a position, even a quiet one, can quickly be turned dynamic in a way that human players quickly drown in.

Of course, the next step is to learn some reliable openings, and enter some LTC tournaments again.

When I have the time. :?

P.S Here is a game from this morning, where, I blundered my queen in the opening by playing quickly "without looking" - and when the splash of cold water that was the queen loss took effect, I focussed.

The tactical onslaught that came from my opponent's only real blunder, could well have been played by one of my engine "personalities". :twisted:

[pgn][Event "chess24 online game | blitz"]
[Site "chess24.com"]
[Date "2018.02.24"]
[Round "?"]
[White "ChessnCognac"]
[Black "ArturoMarnCorresa"]
[Result "1-0"]
[BlackCountry "es"]
[BlackElo "2308"]
[ECO "A80"]
[Time "01:03:35"]
[TimeControl "300"]
[WhiteCountry "au"]
[WhiteElo "2486"]

1.d4 f5 2.g4 { [%clk 0:04:57] } 2...fxg4 { [%clk 0:04:56] } 3.h3 { [%clk 0:04:55] } 3...d6 { [%clk 0:04:54] } 4.e4 { [%clk 0:04:50] } 4...c6 { [%clk 0:04:51] } 5.hxg4 { [%clk 0:04:47] } 5...Nf6 { [%clk 0:04:47] } 6.f3 { [%clk 0:04:43] } 6...e5 { [%clk 0:04:43] } 7.g5 { [%clk 0:04:38] } 7...Nfd7 { [%clk 0:04:29] } 8.f4 { [%clk 0:04:09] } 8...exd4 { [%clk 0:04:08] } 9.Qxd4 { [%clk 0:04:01] } 9...Qb6 { [%clk 0:04:06] } 10.Qc3 { [%clk 0:03:50] } 10...d5 { [%clk 0:04:01] } 11.Be2 { [%clk 0:03:40] } 11...Bb4 { [%clk 0:03:58] } 12.Nd2 { [%clk 0:03:24] } 12...Bxc3 { [%clk 0:03:57] } 13.bxc3 { [%clk 0:03:22] } 13...Nc5 { [%clk 0:03:52] } 14.Ngf3 { [%clk 0:02:57] } 14...dxe4 { [%clk 0:03:47] } 15.Ne5 { [%clk 0:02:54] } 15...Bf5 { [%clk 0:03:45] } 16.Ndc4 { [%clk 0:02:49] } 16...Qd8 { [%clk 0:03:36] } 17.Ba3 { [%clk 0:02:42] } 17...Nbd7 { [%clk 0:03:33] } 18.Nd6+ { [%clk 0:02:39] } 18...Kf8 { [%clk 0:03:17] } 19.Nxf5 { [%clk 0:02:35] } 19...b6 { [%clk 0:03:12] } 20.Ng6+ { [%clk 0:02:28] } 20...Kf7 { [%clk 0:02:56] } 21.Bc4+ { [%clk 0:01:55] } 21...Kxg6 { [%clk 0:02:52] } 22.Nh4+ { [%clk 0:01:52] } 22...Kh5 { [%clk 0:02:51] } 23.Be2# { [%clk 0:01:49] } 1-0[/pgn]

The opponent is over 2300 by the way.
Congratz to you Brendan! The best way to beat back those pesky detractors. ;)



Pls check out the latest Schooner and Winter engines when you get a chance - some fresh 'blood' on the horizon. :) I've only recently found out about how well Schooner plays here: http://www.talkchess.com/forum/viewtopi ... 84&t=66599

I might post some sample games over the weekend or whenever I have more time.

Cheers,
CL
User avatar
velmarin
Posts: 1600
Joined: Mon Feb 21, 2011 9:48 am

Re: 2500! Not such an exaggeration. And how computer chess h

Post by velmarin »

BrendanJNorman wrote:
No one called you a liar.
There was a question as to how the chess24 rating could be compared to the FIDE rating.
Perhaps a normal joke among Spaniards that you didn't understand, apologize if you were so offended to say Grand Master. I apologize.

My congratulations, I think it's a good achievement.
On the other hand, if it seems that the chess24 classification is a little inflated,
don't you think?



Image
BrendanJNorman
Posts: 2526
Joined: Mon Feb 08, 2016 12:43 am
Full name: Brendan J Norman

Re: 2500! Not such an exaggeration. And how computer chess h

Post by BrendanJNorman »

carldaman wrote: Pls check out the latest Schooner and Winter engines when you get a chance - some fresh 'blood' on the horizon. :) I've only recently found out about how well Schooner plays here: http://www.talkchess.com/forum/viewtopi ... 84&t=66599

I might post some sample games over the weekend or whenever I have more time.

Cheers,
CL
Thanks Carl.

Yes, I seen your post and downloaded Schooner.

It really is an excellent engine, and I'd already experimented with Winter as well and like it also.

When I get time, I'll review both for my site.

Very interesting styles, both of them!
Henk
Posts: 7218
Joined: Mon May 27, 2013 10:31 am

Re: 2500! Not such an exaggeration. And how computer chess h

Post by Henk »

I didn't know you read Lyudmil's book.
BrendanJNorman
Posts: 2526
Joined: Mon Feb 08, 2016 12:43 am
Full name: Brendan J Norman

Re: 2500! Not such an exaggeration. And how computer chess h

Post by BrendanJNorman »

velmarin wrote:
No one called you a liar.

On the other hand, if it seems that the chess24 classification is a little inflated,
don't you think?
Why can't you be a man and just say "Sorry, I was wrong" instead of constantly altering what your original argument was.

What truly happened was this (direct quotes from the thread):

1. I told Uly the following:
I'm now at 2450 Elo on chess24...making progress.
2. He asked:
Do you have any idea about how Chess24 rating compares to FIDE
3. I responded:
It seems to me, and I might be wrong, that chess24 is about 200 points higher than chess.com, so ideally, I should be able to crack 2500+ on chess24 assuming a good internet connection and sober play
4. You entered and starting making jokes, and when confronted, you feigned innocence and some sort of moral outrage, saying:
I don't understand your attacks, I already gave you some private information.

... if Brendam (sic) claims to be a 2500 ELO, he could be a Grand Master.....
This is not insulting... Oh, my God.
As we can see...

1. Yes, I DID mention that chess24 ratings are inflated, you are not winning any contests by pointing this out.

2. No I NEVER claimed to be 2500 Elo, nor a GM, this was your twisting of my words to discredit me.

Now be a man and admit to the truth of the matter, only a child would keep making excuses.

I like your Bouquet engine, but not your disingenuous way of handing interpersonal communication. One is excellent, one needs improving.

Everything I ever (truly) claimed with regards to ratings and playing strength, has been proven as sincere - everything the TalkChess trolls have tried to throw at me has missed by a mile.

What type of person would you like to be?

A mean-spirited troll, or a nice guy who programs an even nicer engine?

:lol: :wink:

BTW... My estimation of the chess24 inflation is almost completely spot on. chess.com and ICC are about the same, while chess24 is about 200 points higher.
Last edited by BrendanJNorman on Sat Feb 24, 2018 10:24 am, edited 1 time in total.
BrendanJNorman
Posts: 2526
Joined: Mon Feb 08, 2016 12:43 am
Full name: Brendan J Norman

Re: 2500! Not such an exaggeration. And how computer chess h

Post by BrendanJNorman »

Henk wrote:I didn't know you read Lyudmil's book.
Please enlighten us, what are you truly trying to say here?

Don't be vague. :roll:
whereagles
Posts: 565
Joined: Thu Nov 13, 2014 12:03 pm

Re: 2500! Not such an exaggeration. And how computer chess h

Post by whereagles »

I think blitz ratings are always inflated to over 3000 because of low percentage of draws. It's the same with Go, where top players have ~3500. It's very rare to have a draw in Go (same nr. of stones for each side at the end).

Probably Kai can confirm this :wink:

And of course, congratz on an achievement. Always good to see efforts pay off.
User avatar
velmarin
Posts: 1600
Joined: Mon Feb 21, 2011 9:48 am

Re: 2500! Not such an exaggeration. And how computer chess h

Post by velmarin »

BrendanJNorman wrote:
I think what I've written is written.
Four words deserve your replica of an entire letter.
Certainly my intelligence is not enough to dialogue with you.
Don't worry, this kid's giving up the thread.
BrendanJNorman
Posts: 2526
Joined: Mon Feb 08, 2016 12:43 am
Full name: Brendan J Norman

Re: 2500! Not such an exaggeration. And how computer chess h

Post by BrendanJNorman »

whereagles wrote:I think blitz ratings are always inflated to over 3000 because of low percentage of draws. It's the same with Go, where top players have ~3500. It's very rare to have a draw in Go (same nr. of stones for each side at the end).
Yep, indeed, blitz ratings always stack on at least a couple hundred, but I had no idea it was related to number of draws. Makes sense though.
whereagles wrote:Probably Kai can confirm this :wink:
Most likely, the guy is a genius. :lol:
whereagles wrote:And of course, congratz on an achievement. Always good to see efforts pay off.
Thanks, I'll see if I can keep pushing it above 2500>>> :twisted: