4NCL Online

[Roger de Coverly]

It would be slightly akin to an arbiter taking a player aside at an OTB tournament and asking them to solve some puzzles of a certain standard as part of the process (which I think happens?).

There was a case like that in the USA in the 1990s, but I don't think it happens now.

I don't thing any UK over the board chess organiser or organisation would have the courage to ban or remove a player from an over the board event without hard evidence. When they can physically observe the player, establishing that the player is consulting external sources can be relatively easy. You might suspect they are using the Regan analysis as a tip off.

The problem organisations like the 4NCL and ECF have with partnering the likes of lichess and chess.com is that they lose control of the process of enforcement of rules.

[Roger Lancaster]

You've understood everything fine from what I can tell

Mildly surprised at that if I'm honest! But it's good to hear!

I suppose the background for my query is that we did have a player banned during the season when their account was marked as receiving computer assistance. It leaves me in a tricky position of trying to understand objectively what has happened - and this experience may be of benefit to the thread in general.

Our player denies cheating (but of course, no-one ever admits it). Both 4NCL and Lichess have said that the player's 4NCL games were not actionable. (Those games scored high, but from reading this thread, it seems this could be explained by a double lag in FIDE rating in an improving player - a natural lag anyway as not all games are FIDE-rated, and then a covid lag as there is now a disconnect between rating and strength)

Aside from the 4NCL 45+15 games, the player only played rapid or blitz games on the site - so the implication is that the computer assistance was at a time control of 10 minutes or less. That strikes me as unlikely, although I accept it's possible.

The experience of having players banned from Online 4NCL and/or Online J4NCL due to "chess engine assistance" which, once one enquires, is alleged to have taken place in non-4NCL events [but just where is never disclosed} was an unfortunate feature of the 4NCL events just ended. Not only were team captains obliged to understand the theoretical basis behind these bans, they were also required to spend hours trawling through back games in an effort to uncover where the alleged offences might have occurred. I can say this from personal experience as a team captain. The recent 4NCL changes are in no way retrospective. The loss of goodwill, particularly towards Lichess, must be considerable.

[JustinHorton]

Our player denies cheating (but of course, no-one ever admits it).

Chess.com demand you admit it as the price of having your account re-opened. Obviously this is particularly unpleasant if it's something you haven't done, but I am guessing that some guilty people must over time have availed themselves of that service.

Our player is in the position of being asked to appeal against something they've been given no information on. That's kind of hard.

From experience, it's impossibly hard. You have absolutely nothing to go on

[Joseph Conlon]

Aside from the 4NCL 45+15 games, the player only played rapid or blitz games on the site - so the implication is that the computer assistance was at a time control of 10 minutes or less. That strikes me as unlikely, although I accept it's possible.

Computer assistance is possible at 1+0 bullet with some technical sophistication, and at rapid or blitz time controls it doesn't require any sophistication. As alluded to on these pages, cf the recent U12 mega final where a couple of the top finishers were quickly flagged for engine assistance (and to head off one line of debate, I don't think the reason <100 ECF players suddenly start playing like Carlsen is because they have memorised lots of analysis).

[Caoimhín de Búrca]

Yep. I should clarify that I think it's unlikely in this case, knowing the player. I know that's involves a leap of faith of course.

(Actually, as an aside, I did see a player banned from lichess for cheating in ultrabullet (15 seconds) recently. He had evidently not only hooked an engine up to his account, but the engine was able to make the moves in real time on the board as well. He would checkmate 2100+ players with the full 15 seconds still on his clock. Amazing to watch - if a little bit too obvious!)

[MartinCarpenter]

This seemed a matter of some concern so I enquired of someone, non-participant in this discussion, more familiar with Ken Regan's methods than I am. I was told, and I have absolutely no reason to doubt the integrity of my informant, that - in such cases - Ken Regan is informed of the higher figure and uses this as a safeguard against producing false results. I'm not totally reassured, as human errors {for example, in overlooking someone's national rating} can still occur but in principle it seems to answer the "tiny bit tricky" point raised by Martin.

That's expected but definitely good to know!

I do worry a little bit about the theoretical validity of the statistics if you're using a converted ECF grade. The ECF <-> FIDE conversion formula is a ball park thing, and ECF grades are definitely more seasonally volatile than FIDE so you probably need to use a larger standard deviation.

This could all be calibrated for. Maybe it has been You'd get a really accurate, firmly based, ECF <-> FIDE conversion formula out of it as well.

[John McKenna]

Aside from the 4NCL 45+15 games, the player only played rapid or blitz games on the site - so the implication is that the computer assistance was at a time control of 10 minutes or less. That strikes me as unlikely, although I accept it's possible.

Computer assistance is possible at 1+0 bullet with some technical sophistication, and at rapid or blitz time controls it doesn't require any sophistication... SNIP

Doing that in such a way as not to arouse suspicion does require a significant amount of it and therefore more time to implement in a game no matter how smooth or practised the operative gets.

Stopping the do-or-die charge of the samurai is a lot easier than avoiding assassination by ninjas. (Japanese proverb)

[Nick Burrows]

http://www.4ncl.co.uk/data/Article_on_Fair_Play.pdf

http://www.4ncl.co.uk/data/Questionnair ... ason_1.pdf

[Kevin Thurlow]

http://www.4ncl.co.uk/data/Article_on_Fair_Play.pdf

http://www.4ncl.co.uk/data/Questionnair ... ason_1.pdf

The first seems to be anonymous.

The second - in questions 2, 3 etc. has people voting 1,2,3,4 or 5, but doesn't seem to state what that means! If someone would rectify that, I might read the rest of it.

[MartinCarpenter]

Those are interesting documents, thanks. Very positive feedback overall Surely 1 is 'I hated this' and 5 (or 10) is happy, happy? Very normal.

This graph from the fair play article, comparing Z scores from 4NCL Online vs Real world is interesting.

4NCL_Chess.png (100.62 KiB) Viewed 1305 times

Yes, there's an obviously worrying cluster at the right hand side - some people are playing alarmingly well online.

There's a huge excess on the left hand side as well though. There's a mass of people playing rather worse than expected (especially at ~2 SD.). The online grades <-> performance is just a different distribution.

At a guess, the mean is moved a bit to the left - overall lower standard of play from online 45/15 vs long OTB - and the standard deviation is a bit higher for the online results. That probably leaves the basic conclusions solid enough, you'd really want to do the calibration work to make sure.

You can very definitely see why you need Ken's move quality data rather than 'raw' performance data to conclude anything. A lot of people badly unperformed vs their OTB strength.

[Roger de Coverly]

You can very definitely see why you need Ken's move quality data rather than 'raw' performance data to conclude anything. A lot of people badly unperformed vs their OTB strength.

I think I saw a disclosure that the case in the 2018 Telford Minor showed off the scale values. It's a different sort of cheating, but what sort of move quality score would a 2000 player without external assistance who sneaked into a Minor be able to achieve?

I believe it's also disclosed that there's no great sophistication in excluding "the opening" from analysis.

But treating the regular 4NCL as a control, there are anomalies in the results.

Having to set up video and audio surveillance is off putting to potential participants but is there any other answer to playing "serious" chess online?

[Ian Thompson]

http://www.4ncl.co.uk/data/Article_on_Fair_Play.pdf

The first seems to be anonymous.

It's linked to from the 4NCL website which says the author was Alex Holowzak and that it will be published in the August 2020 Chess magazine.

[Ian Thompson]

There's a huge excess on the left hand side as well though. There's a mass of people playing rather worse than expected (especially at ~2 SD.).

I wonder what effect mouse slips have? They would explain some really bad moves and upset results. How many though?

[Mike Gunn]

Or playing someone who is cheating?

[MartinCarpenter]

Or playing someone who is cheating?

No. This is measuring the quality of the moves people are playing vs that expected from their grades.

It entirely removes the opposition.

I presume that its simply much harder for a lot of people to fully concentrate when playing on a computer than over a board. A very well known thing with, say, analysis vs real games.