4NCL Online

[Brian Valentine]

I agree with Nick

[Nick Grey]

off pic entry = Some of us MANY awful losses in 2020 no time to get bak.

[Roger Lancaster]

"Sorry for an off topic entry, but I thought I would set out my preliminary findings. I have compared the grades/ratings in the first investigation with the current situation. A lot depends on which players one samples; to try and get rid of some of the obvious distortions I have selected seniors in both lists. These lists are of those with both a category A-E grade and a no inactive flag on the rating record. This applies separately for both lists.

"There are 772 players on both lists. In the old list the means are 183/2066 updating to 184/2022 now.
There are 934 players only on the old list with means of 179/2031
These are replaced by 1371 only on the new list: means 175/1936

"There appears to be a slow tectonic shift going on."

First reaction to this, if the earlier list were based on 2014 data, is to express surprise that the mean ECF grade for those 772 seniors has not decreased - as one might have expected with advancing age - over the five or six years in question. [Same applies, to a lesser extent, with only slight decline in ELO's].

Second reaction, if I understand the figures correctly, is that - while in 2014 the {ELO = 7.5xECF + 700} relationship applied, plus or minus a few points - currently the conversion formula yields figures some 50 points higher, on average, than actual ELO figures. In other words, a more appropriate formula might be {7.5xECF + 650} although further investigation might instead indicate that the problem lies with the multiplier rather than the constant.

[Roger de Coverly]

In other words, a more appropriate formula might be {7.5xECF + 650} although further investigation might instead indicate that the problem lies with the multiplier rather than the constant.[/size]

There was a debate some years ago before the formula was introduced. The key point of evidence being that if you compared the calculations on games between players with a 0-200 Elo difference and 0-25 ECF difference, 7.5 could be inferred from the tables.

For most players resident in England, their FIDE rated games are also ECF graded, so no assumptions of independence can be made between FIDE ratings and ECF grades.

[Brian Valentine]

I expect to have a summary of my further investigations on the ecf website by next week. I would point out that seniors in 2013 (the data are from that year) includes all players over 18, many still improving. Also many seniors will no longer be active on one of the list so missing in 2020. It is plausible that these effects cancel out the expected drift.

[Matt Bridgeman]

Alex's update from yesterday shows that 12 players from the online 4NCL and J4NCL have been added to the barred list for cheating. https://www.4ncl.co.uk/

[Adam Raoof]

Alex's update from yesterday shows that 12 players from the online 4NCL and J4NCL have been added to the barred list for cheating. https://www.4ncl.co.uk/

What does "- 2 were attempting to cheat using other methods" mean?!

[Roger de Coverly]

What does "- 2 were attempting to cheat using other methods" mean?!

It could be attempting to hack the software so as to adjust clock times.

Consulting published theory whilst playing or getting advice from a third party is also possible.

[Ian Thompson]

Consulting published theory whilst playing or getting advice from a third party is also possible.

But unlikely when the statement was that the cheating involved getting assistance from a computer, but not the use of an engine.

[Jonathan Rogers]

Alex notes

"It is interesting that the pool of people who think their opponents’ moves are worthy of investigation, and the pool of people who have been caught, are two distinct sets with no intersection."

Indeed it is. But not a total surprise. I recently started blitz on LiChess, and one day was refunded five points with the explanation that I had lost to a cheater (someone subsequently discovered, presumably). No names or details given, and I must say that I have no idea which game it was!!

[Kevin Thurlow]

"It is obviously disappointing that we have experienced cheating in these competitions, but we hope that by publishing these numbers, people can see:

- this issue is being taken seriously by the 4NCL
- players should not use computer assistance in these competitions, because they will be caught"

But as we have seen elsewhere, you can get caught when you haven't done it (although with a different provider).

[Ian Thompson]

- players should not use computer assistance in these competitions, because they will be caught"

I'd be interested to know how it is ensured that someone consulting an openings book, or following strong players' games, for as long as they could (i.e. until their opponent varied and there were no longer any books or previous games to follow) would be caught.

In a similar vein, how could tablebase use in an endgame be detected?

[Richard Bates]

From my reading it’s actually 7 people not 12? (accused of cheating in relation to the 4ncl, some of which are under appeal)

5 players are prevented from playing because they are already barred from lichess. So for example, had it been held on chess.com then presumably Justin would have been prevented from participating (until his recent reprieve/exoneration).

It is I suppose possible that some people might cheat in casual online blitz, but set a more higher standard for themselves when playing in the 4ncl competition. And the motivations for themselves may be different. There’s no real rating element in the 4ncl, you’re more likely to stand out if you have unexpected results. And (despite the anonymity guarantee) the consequences are likely to be greater in the longer term.

[Roger Lancaster]

I expect to have a summary of my further investigations on the ecf website by next week. I would point out that seniors in 2013 (the data are from that year) includes all players over 18, many still improving. Also many seniors will no longer be active on one of the list so missing in 2020. It is plausible that these effects cancel out the expected drift.

Have to confess that, when Brian originally used the term "seniors", I misunderstood his meaning and my comment partially reflected that.

[JustinHorton]

So for example, had it been held on chess.com then presumably Justin would have been prevented from participating (until his recent reprieve/exoneration).

Well yes, and I did wonder whether one motivating factor here might be that now they're running competitions on behalf of other bodies rather than just their own private show, they realise that their previous way of going about things will cause them a world of problems very quickly.