Originally posted by wormwoodI think certain graphs ARE enough. Back in 2005, when you had a James Woodley or a Blobby go from a 900-1000 rated player to all of a sudden winning all their games and jumping 500 rating points or so in less than a month, I'd say that's "overwhelming evidence" in and of itself. And there's been other players of a more recent vintage who have done similar things, yet escaped banning because their match ups didn't meet the criteria of the Game Mods even though said match ups had jumped sharply in a very short period of time.
of course it is, but that doesn't mean it can't happen. atticus is claiming looking at graphs instead of statistical evidence is enough. it's not enough, it's pure guesswork.
ALL evidence should be considered.
Originally posted by no1marauderso you think someone like quirine should be banned without even checking his games? the guy goes periodically all the way down to 1700s without mass resignations nor timeouts, but can beat any of the (human) top dogs when needed.
I think certain graphs ARE enough. Back in 2005, when you had a James Woodley or a Blobby go from a 900-1000 rated player to all of a sudden winning all their games and jumping 500 rating points or so in less than a month, I'd say that's "overwhelming evidence" in and of itself. And there's been other players of a more recent vintage who have done simila ...[text shortened]... arply in a very short period of time.
ALL evidence should be considered.
Originally posted by wormwoodWhat part of "ALL evidence should be considered" did you miss?
so you think someone like quirine should be banned without even checking his games? the guy goes periodically all the way down to 1700s without mass resignations nor timeouts, but can beat any of the (human) top dogs when needed.
1 edit
Originally posted by wormwoodNo I'm not. By all means, let the mods do the stats stuff. But it's a long, slow and pretty inefficient process though - a bit like trial by jury. And tough on the mods too.
atticus is claiming looking at graphs instead of statistical evidence is enough. it's not enough, it's pure guesswork.
So I guess I'm pitching for a bit of summary justice - some knee-capping, the occasional lynching maybe. The graphs provide most of the evidence one needs for this.
Frankly, I've already started in private. I'm having a 'quiet word' with some of my current opponents, asking them politely if they can explain astonishing variations in their performance ranging from patzer-level incompetence through to GM-standard precision.
The responses so far are instructive. One guy (2150) has flounced out and hasn't been seen for five days; another (2300) has admitted engine use but is shameless about it; and a third (2200) is considering his position. So far, I haven't accused anyone of anything. I've merely asked an innocent question 😛
Originally posted by atticus2okay then. but back in the days when we still knew who the mods were, they said there was no backlog with crunching the analysis. the problem was always convincing the admins to drop the axe...
No I'm not. By all means, let the mods do the stats stuff. But it's a long, slow and pretty inefficient process though - a bit like trial by jury. And tough on the mods too.
So I guess I'm pitching for a bit of summary justice - some knee-capping, the occasional lynching maybe. The graphs provide most of the evidence one needs for this.
Frankly, I've ...[text shortened]... o far, I haven't accused anyone of anything. I've merely asked an innocent question 😛
all the stronger humans on the site spot the engines quite well, and we all know who they are. but I wouldn't mention in public anyone just because I 'felt' I was sure just from looking at the games. we can be sure about most of them through statistical analysis, so there's really no reason to not use that option.
getting them to talk chess is a great way to blow their cover. it doesn't work on strong players who cheat like cludi though, but those are really a marginal group anyway. the most fun is getting them to do it here. but many of them also avoid talking chess in public like plague, probably after being burned before with a different account, so it's not that easy.
Originally posted by wormwoodthe irony of it is, there are Patzers like me that like to talk chess, and like to look at the games posted and are still rubbish!
okay then. but back in the days when we still knew who the mods were, they said there was no backlog with crunching the analysis. the problem was always convincing the admins to drop the axe...
all the stronger humans on the site spot the engines quite well, and we all know who they are. but I wouldn't mention in public anyone just because I 'felt' I was ...[text shortened]... lague, probably after being burned before with a different account, so it's not that easy.
- Joined
- 16 Feb 07
- Moves
- 27653
Originally posted by atticus2That's really funny and probably a sensible approach. I don't like turning in people myself unless I've played a few games with them and I'm pretty sure.
No I'm not. By all means, let the mods do the stats stuff. But it's a long, slow and pretty inefficient process though - a bit like trial by jury. And tough on the mods too.
So I guess I'm pitching for a bit of summary justice - some knee-capping, the occasional lynching maybe. The graphs provide most of the evidence one needs for this.
Frankly, I've ...[text shortened]... o far, I haven't accused anyone of anything. I've merely asked an innocent question 😛
Originally posted by robbie carrobiewe're all rubbish here, and like to look at games. 🙂 (anybody else following uscl games right now? at least icc is relaying them...)
the irony of it is, there are Patzers like me that like to talk chess, and like to look at the games posted and are still rubbish!
Originally posted by no1marauderI'm wary of the idea that a sudden improvement in a players ability should in itself be seen as evidence of engine use. Although I do agree that there needs to be a more robust response to the problem from the site admins as I get the distinct impression that I'm running into more engine use on the site than I did a year or so ago.
I think certain graphs ARE enough. Back in 2005, when you had a James Woodley or a Blobby go from a 900-1000 rated player to all of a sudden winning all their games and jumping 500 rating points or so in less than a month, I'd say that's "overwhelming evidence" in and of itself. And there's been other players of a more recent vintage who have done simila ...[text shortened]... arply in a very short period of time.
ALL evidence should be considered.
I think that graphs are grounds for initiating an investigation, but not for a banning if no other evidence is found. I think there's a pretty clear problem that engine users are going to get more sophisticated and manage their match-up rate to avoid banning. It's not clear to me how to combat this, except that there are some mistakes an engine will make that a human just won't.
The other problem is that the the list of banned players was removed, so it's difficult to see that players are being banned for engine use. I feel that this should be reinstated for its deterrent effect.
1 edit
Originally posted by wormwoodWhat a joke. The idea that I "ignore statistical analysis" (I've probably run more on cheaters on this site than anybody) is so ridiculous it doesn't warrant a detailed response.
the part where you ignore statistical analysis and are ready to condemn people based on graph alone.
1 edit
Originally posted by no1marauderyou claimed in this thread that looking at graph is enough. I blew that right out of the water with a simple counterexample. you didn't even comment that counterexample but started trying to change the subject. nuff said.
What a joke. The idea that I "ignore statistical analysis" (I've probably run more on cheaters on this site than anybody) is so ridiculous it doesn't warrant a detailed response.
what makes your stance even more silly, is that you of all people should know better than implicating that a drastic rating difference between different time controls means cheating.