lichess.org
Donate

Follow up on recent rating changes and possible wide bot use

@mcgoves said in #20:
> In chess we can quantify things with a number. In most areas of life we cannot, so small changes can be unnoticable.
I thought I have quantified chess numbers already. Do you want to hear me quantifying how well I design software components or write code? What is it that you are asking here about?
<Comment deleted by user>
@MrAWD said in #21:
> Do you want to hear me quantifying how well I design software components or write code?

This actually looks like a good sanity check idea. Download Lc0 and some human-strength networks
like Maia
github.com/CSSLab/maia-chess
or chess sparring networks from Dietrich Kappe
like Mean Girl
github.com/dkappe/leela-chess-weights/wiki/Mean-Girl:--the-most-fun-leela-style-net
or Bad Gyal
github.com/dkappe/leela-chess-weights/wiki/Bad-Gyal

Then once a month run a quick gauntlet tournament against the fixed set of bot opponents of well known strengths.

Then post your en.wikipedia.org/wiki/Performance_rating_(chess) so we can actually gauge the variance of your own performance against unchanging opponents. I presume you will not be cheating or making rating changes in your own tournaments run under your total control.
@kalafiorczyk said in #23:
> This actually looks like good sanity check idea. Download Lc0 and some human-strength networks
> like Maia
> github.com/CSSLab/maia-chess
> or chess sparring networks from Dietrich Kappe like Mean Girl
> github.com/dkappe/leela-chess-weights/wiki/Mean-Girl:--the-most-fun-leela-style-net
> or Bad Gyal
> github.com/dkappe/leela-chess-weights/wiki/Bad-Gyal
>
> Then once a month run a quick gauntlet tournament again the fixed set of bot opponents of well known strengths.
>
> Then post your en.wikipedia.org/wiki/Performance_rating_(chess) so we can actually gauge the variance of your own performance against unchanging opponents. I presume you will not be cheating or making rating changes in your own tournaments run under your total control.

I like that!! Thanks for suggestion!
@newlinkwave , I read English well. Indeed, I have been an editor, professionally. And I am ridiculously over-educated.

Here is a cut-and-paste quote of what YOU recently typed: "Please read my comment again. If you don't understand English well, then use a translator (I recommend DeepL). What I said is that before "agreeing" or not with something, do some research."

But what did you ACTUALLY type even BEFORE that? Here's another quote, in pertinent part, from one of your other posts: "... but don't say you don't agree with something without a bit of research."

In other words, as I previously noted, correctly, you didn't merely tell me what I could "agree" with -- you outright tried to tell me what I could "say." My reading is accurate, and I can cut-and-paste just fine.

In any event, I must, apparently, repeat myself: I HAVE done some research, as I described expressly. So telling me, condescendingly, to "do some research" simply ignores what I said.

I'm sure you can find a "youtube" that suggests how cheating is "possible." No doubt cheating is "possible." So what? It doesn't appear to be widespread here, at least at the faster time control that I almost exclusively play.

In actual practice, HERE, based on my own numerous, repeated checks of my ACTUAL games over a period of YEARS, I have apparently NOT encountered cheating. Which means, regardless of how many "youtubes" you find that show HOW to cheat, or that cheating is "possible," the practice doesn't seem to be of a great deal of practical moment HERE, at least not at the faster time controls that I have played.

Can you find SOME cheating here? I suppose. They do, after all, deal with offenders from time to time. But is cheating a terribly widespread, common practice here? That has not been my ACTUAL experience, over YEARS. Which I am entitled to say without being insulted.

Let me quote what I said at the beginning of all this:

"I fear the possibility of some form of "cheating" at slow time controls. But at blitz? Or even more so, at bullet? Naaaahhhh."

I didn't say cheating was not possible. I said I DIDN'T FEAR THE POSSIBILITY of cheating at faster time controls here. Because I don't. Since I have YET to encounter cheating, the possibility that I "might" encounter it sometime, because it's "possible," does not cause me concern. I don't fear it.

At this site and at another site, I have sometimes (although of course rarely) had people accuse ME of "being a bot" just because I played a particularly good game and their egos apparently couldn''t handle the loss -- so they accused me of being a "bot." It was laughable -- but they were just CERTAIN. I, of course, know rather intimately that I am not a "bot." Bots don't own dogs and ride bicycles or make a good burrito. At least not yet.

You can be disagree with me, and hope to be informative, without being rude. But telling me what I may "say," or snidely suggesting that I might not understand English, is not only laughable, it is unnecessarily rude.

However, I forgive you. We all sometimes would like to be able to reword something, in retrospect.
@Noflaps said in #14:
> I've found (anecdotally, of course) that even reducing sleep by 2 hours a day can have a surprisingly suppressive effect on my chess.

I remember it, and you're right, it is a chessy enough subject to live here.

Not only that, but common-life worries, poor focusing, back to school/work... can noticeably (and consistently) affect your performance. Not to mention purely chessy factors like studying some opening or endgames (usually after a row of related defeats). I think all of these are far more likely to happen than cheaters/bots.
How can you comment on something you don't have the slightest experience at??
You play unrated games only. Of course cheaters play rated games...
@mrbasso, if you wish to believe cheating is a significant problem here, what evidence do you have? If you don't have any you can articulate, then aren't you merely telling me, in a more subtle way, to shut up?

I have played a vast number of different people here. Nearly all have current ratings. Do the cheaters among them (if there are some) just decide not to cheat when they play me? If so, that's very nice of them! I wish I knew whom to thank!

Isn't it better to state your own position, and provide evidence for it, whatever it might be, than to try to suppress my own observations?

I have played (for a long, long time) at another well-known site, too. Almost entirely rated. Literally thousands of games. I saw no indication of rampant cheating there, either. Is this site somehow less guarded against that? I frankly doubt it.

Again: I don't think cheating is impossible. But I think some overstate how common it actually is, at least at quicker time controls. You're welcome to disagree, and give me contrary evidence. I eagerly await that moment.
Man, this lag switching video is hard to watch. Great comeback by Tang.
So, just like that , after having to have hard time winning against players in 1800 and 1900, few days later I am in 2100s.

I am just curios if my game is fluctuating this much on its own, or something else is changed underneath (something on the site, my wifi connection, placement of the Moon, anything else...)

This topic has been archived and can no longer be replied to.