English Chess Forum

I've just listened to a report on BBC Radio 4 news about students using specially designed watches to provide assistance in exams. Ring any bells?

NickFaulks wrote:I've just listened to a report on BBC Radio 4 news about students using specially designed watches to provide assistance in exams. Ring any bells?

BBC News article here: http://www.bbc.co.uk/news/education-35716523

Well, yes, watches are seriously dangerous things to allow

Does also rather illustrate just how objectively dumb doing tests which principally evaluate people's raw memory skills is nowadays.

The trouble is that if you test their ability to think you know that most of them will fail.

NickFaulks wrote:The trouble is that if you test their ability to think you know that most of them will fail.

I think the real problem is inventing new questions each year to test thinking: you can't just have the same questions with slightly different numbers (thinking in terms of Maths exams)

Michael Farthing wrote:

NickFaulks wrote:The trouble is that if you test their ability to think you know that most of them will fail.

I think the real problem is inventing new questions each year to test thinking: you can't just have the same questions with slightly different numbers (thinking in terms of Maths exams)

I think the exam board we use for Maths A level (OCR) has been very good in recent years at inventing new questions, involving the required stretch and challenge.

Nick Faulks >I've just listened to a report on BBC Radio 4 news about students using specially designed watches to provide assistance in exams. Ring any bells?<
If it does, then that results in a loss in chess.

Bulgarians are back in the news.

http://en.chessbase.com/post/new-case-o ... g-in-chess

It remains my belief that you need to detect the method of inputting the moves to the engine and receiving suggested replies back, before throwing accusations around. In many cases, the circumstantial evidence is obvious enough, that a player had a device in their possession or hidden somewhere to which they had access and periodically appeared to retire to consult it. They may only have been reading their emails or checking the football scores, but such behaviour during a game is now outlawed.

The article raises the question as to whether computer matching should also test use engines from twenty or twenty five years ago. Something I've wondered is whether a test using a modern engine can detect whether a game played by an old engine was computer or human.

Seizing on individual moves as engine inspired may be correct, but doesn't prove cheating as pre game preparation can reach similar positions even if the position itself hasn't appeared in published analysis or an actual game.

The latest story on this.

http://en.chessbase.com/post/cheating-i ... under-1600

As a method of cheating, it was easy enough to detect, provided physical searches were permitted. The report doesn't say whether the accomplice was found.

"On every stroll I took to the restroom during my games, I could see scores of people discussing and analyzing games on smartphones, right outside the tournament hall."

"But one has to feel bad for those players who lost a point against Shrang. Nothing can be done for them. Some of them asked the arbiter for some sort of compensation in the end, but to no avail. All he could do was to report the incident to FIDE."

I don't understand why this should be an insurmountable problem. Surely where a player has lost to a proven cheater, then a notional game could automatically be added by the tournament organiser or FIDE to all the players who played the cheater during the tournament which would correct the rating points lost? ie it would effectively put the offended player in the position as if the game had never happened. Of course one would have to make the assumption that the cheater cheated in every game in the tournament but I would be comfortable with that assumption even if one couldn't prove that it was true.

That seems roughly right to me.

Chris Rice wrote:"But one has to feel bad for those players who lost a point against Shrang. Nothing can be done for them. Some of them asked the arbiter for some sort of compensation in the end, but to no avail. All he could do was to report the incident to FIDE."

I don't understand why this should be an insurmountable problem. Surely where a player has lost to a proven cheater, then a notional game could automatically be added by the tournament organiser or FIDE to all the players who played the cheater during the tournament which would correct the rating points lost? ie it would effectively put the offended player in the position as if the game had never happened. Of course one would have to make the assumption that the cheater cheated in every game in the tournament but I would be comfortable with that assumption even if one couldn't prove that it was true.

I'd assume if a cheater has been found during a tournament, all previous game in that tournament of the cheater should be given as a win to the opponent (similar to exams, if you get caught cheating in 1 exam, you fail the lot)

Mick Norris wrote:I'd assume if a cheater has been found during a tournament, all previous game in that tournament of the cheater should be given as a win to the opponent (similar to exams, if you get caught cheating in 1 exam, you fail the lot)

For rating purposes that seems to make sense, but in terms of results counting in a Swiss Tournament, adjusting results from earlier round may well distort the overall result and in extreme circumstances could encourage cheating

David Shepherd wrote: For rating purposes that seems to make sense

I couldn't agree less. A game between two players, one of whom is cheating, adds no information regarding how good either of them is at chess. Lest we forget, that is what the rating system is supposed to be about.