The future of online go tournaments

[Javaness2]

I think online tournaments will just become prize less - or at least small prizes only. Maybe in Europe they will also become unrated.

[Tryss]

It's just not convenient and too much of an imposition on the vast majority of honest players. Applying such conditions to only suspicious players (e.g. CM) would seem more reasonable.

But if a non monitored player is accused of cheating, what happens?

[Uberdude]

It's just not convenient and too much of an imposition on the vast majority of honest players. Applying such conditions to only suspicious players (e.g. CM) would seem more reasonable.

But if a non monitored player is accused of cheating, what happens?

Evaluate the accusation to see if it has any merit. If so analyse that game (and others if those suspected too) for similarity to bots and if level of play is very unusual, both with reference to decent control groups (we can't do this well yet). If that provides sufficient evidence of cheating then forfeit and punish, else require future games of the suspicious player to be monitored. If their good results continue then that is evidence they weren't cheating before, if they stop it's evidence they were.

I would rather the PGETC continues with me being able to play at home, and lets say 2% of players will cheat and get caught and 2% will cheat and not get caught, than either: cancel PGETC or continue but I and everyone else have to go to some secure monitored location to play every game. I would be willing to buy and run a webcam if that was thought useful, so long as it wasn't too much of a faff (it wasn't fun being IGS technical support when I was captain). I don't know how widespread my view is, maybe others think cheating is already or will be much more prevalent, or are less willing to accept some may happen but continue with the event nevertheless. Perhaps a survey of participants would be a good idea.

[John Fairbairn]

I start with the handicap of not really understanding why anybody would want to waste three or so hours of their precious lives by creating a fake game with no money involved, but if I was forced to find reasons, I expect two would come out at the top of the pile.

One would be to ensure winning a match (e.g. not letting mates down).

The other would be a desire to play with their shiny new toy. I suspect this may be the dominant reason, but in any event both these reasons would seem to cover most cases.

Now at some point, probably not too far away, everybody will have this new toy. That means it will no longer be possible to ensure winning as it will just be AI versus AI, and, more significantly, the dull old toy will have lost its glamour. Just as no-one now tells us "Hi, honey, I'm on the train" in a loud voice (meaning "Hi fellow passengers, I've got a shiny new iPhone"), no go player will specially want the world to brag that he's got LeelaZero. Having AI thus becomes a bit pointless, except for entirely private study.

There doesn't seem to be much danger of an arms race in this field, so online cheating could more or less wither on the vine, while private study takes off.

How does that possible scenario fit what has happened in the chess world?

[Uberdude]

Look what event just popped up in my inbox:

https://www.pandanet.co.jp/event/iwag2018/index_e.htm

[Bill Spight]

Thanks, Uberdude. If you enter, good luck!

From the announcement:

When the organizers finds out that you have violated the rules, you may be disqualified, even if the game result has been confirmed.
With players who keep winning and who keep rising in strength, there may in some cases be players who are disqualified on the grounds that their exact rank cannot be judged.
Also, players may be disqualified when there is an obvious disparity between their registered rank and their real strength.
If it is confirmed that you are using Go software, you will be disqualified from this tournmanent and forbidden to participate in all other tournaments.

Emphasis mine.

They don't say how they might confirm the use of Go software, especially since you normally play from home.

The disqualification based upon disparity between registered rank and real strength might be used in cases such as Metta's where he was apparently playing above himself, but the use of go software was not proved.

[sybob]

Also, players may be disqualified when there is an obvious disparity between their registered rank and their real strength.

They don't say how they might confirm the use of Go software, especially since you normally play from home.

Wouldn't this scare off people?
What is an "obvious disparity"? How would Pandanet determine that? Who to trust?
Today, I played 3 games with 2 players who are definitely stronger by 2 or 3 stones than me (in real life/over the board). I won all three games. Could happen online too?
This all sounds a bit vague to me, and not very inviting.
(just kyu speaking here, but same may apply to better players?)

[jlt]

Perhaps the paragraph about "obvious disparity" primarily concerns sandbaggers. Sandbagging is not easy to detect when people play 2 or 3 stones below their level to deflate their rank, but becomes "obvious" when the strength disparity is larger.