Some tweaks to the CCT rules

[Ferdy]

The only trouble is, having the same book author for different engines opens the tournament up to abuse, so why not remove any possibility of that?

Can you elaborate further, who were the abusers?

[garybelton]

If I sit on *both* sides of the board, I can do lots of fun things, haven't you watched any Freestyle games lately?

The rule has now been passed for CCT so I will stop talking about it, I don't want to give anyone any ideas.

[Ferdy]

If I sit on *both* sides of the board, I can do lots of fun things, haven't you watched any Freestyle games lately?

The rule has now been passed for CCT so I will stop talking about it, I don't want to give anyone any ideas.

I did not get it, and I don't watch freestyle.

[CRoberson]

I was wondering myself why people don't do that.

IIRC Winboard sends "ics Hostname" on startup when in ics mode....

Then the engine can simply use if(ics) [tellopponent],[tellall],[ tellothers] as it needs.

Regards

Richard

That is simple. Sometimes I don't want the engine to kibitz. If I have it online and it is not in a computer chess event, then I don't want it announcing its plans to the humans playing it.

[Suj]

how about also increasing number of disconnects to 3

[Dann Corbit]

Hello everyone,

Before, during, and after CCT12, there were many rules "tweaks" that were suggested, and I think they are a terrific idea.

In reference to rule:

Code: Select all

2. Only the original author, Operator or a team member of the original program may enter and operate. The definition of a team member is as follows:

An individual who has internal knowledge of the program and is recognized by the author as a member of the development team.. i.e. Developer, book maker. This person must have advanced knowledge of the program, such as how the program is designed, hashing schemes, book making procedures..

2a. Operators will only be allowed under the following conditions:

2a1. The author must be online during the tournament to participate in discussions in channel 64, and in the event there is an issue with their program they can provide real time support to their operator.

2a2. The operator must use the account that the author has signed up. No exceptions will be made.

2a3. No operator can operate more than one program. No exceptions will be made.

Over the course of the event, we had several authors online, but their operators were AFK. Unusual, but true.

I propose that Author's pick an "operator" or two NOW, train them on engine parameters and which interface/books you would prefer to use. This way, if an event comes up where you need hardware, you have someone readily available, and can perform. They essentially would become "Team Members", thus nullifying the need for an "operator".

You have a full year now to get one or two people ready for the next CCT event, and you have the ACCA events coming up. So I would HIGHLY recommend this practice.

In reference to rule:

Code: Select all

2. No manually operated programs, and all programs must kibitz their evaluation, and book moves/TB hits if possible. Providing as much information as possible for the viewers and participants is key. It should also be noted that 1-3 lines of text is sufficient. No need to scroll out an entire page.

Whispering is NOT what the rule states. It states kibitzing it mandatory. We had several participants who whispered instead of kibitzing, or kibitzed nothing.

Next year, it will be very simple to avoid any complaints. If you are not kibitzing at the beginning of your game, you will forfeit. If you decide that you do not want to abide by this rule, do not enter. It will be strictly enforced.

As to the book makers:

Two teams will not be able to use a book created from the same book maker.

i.e. Ted Langreck makes Crafty's books. He would be prevented from making a book for another participant. Just as a participant would be prevented from using a book created by him.

If we only allow one engine per author, then we only allow one book per book maker. That is fair right?

This sounds utterly absurd to me.

Why? Should it be OK for me to work on several programs? We have five members working on Crafty. We could split it into 5 different programs that are almost identical, and each of us enter a different version and greatly increase our chances of winning. This is a team vs team competition using programs developed by each _team_. The WCCC adopted this rule several years ago and its a good one. One person per program, whether he codes or writes the book does not matter. Can't be on two different teams at the same time. Huge conflict of interest.

I agree about the engines (to some degree) but not about the books.
Professional books often have books written by professional book writers.
Should that prevent an engine from entering the competition or forcing them to remove their book?

Even with the engines, I think it requires case by case analysis. For instance, John Merlino worked on ChessMaster and has now written Myrddin. If ChessMaster enters a contest can Myrddin?

I do not think that the rules should be applied to the point that the tail is wagging the dog.

Ok one to this one:

Code: Select all

5. In the event of a disconnection, the party will be given 5 minutes to return to complete the game and no more than 2 disconnections  per game will be allowed. On the third time, the game will be a forfeit.

This one is being changed to 10 minutes, instead of 5. Nothing too significant, but should correct some issues.

More are coming, as I decide how to word them.

Peter

[CRoberson]

I don't think live kibitzing is a good idea as that allows the other program to see what the engine is thinking and play appropriately ... unfortunately we have reached a situation where there are some who would do that. Maybe whispering is better, but really that only adds one extra hurdle for someone unscrupulous to simply have another account filter out the whisper information.

Correct, whispering doesn't solve the problem. Kibitzing is so that each competitor can keep an eye on his opponent.

I guess something can be configured to have output only to a TD with a total log after the game. Also I do not think that an operator should have to decide if his opponent is following the rules or to react a certain way if his opponent has broken certain rules ... the TD should make an immediate decision and keep both players out of it. This would avoid uncomfortable situations where players would have to make decisions that can cause ill feeling among participants.

That solution doesn't scale. One person can not watch all the games in real time, especially in a tournament with 40+ competitors. Logs are ok, but decisions on most rules need to be when it happens not after the game. One thing I thought of is a group of unbiased referees. Look at a football, soccer or a basketball game, there are multiple refs not just one for each game.

At physical site events, it seems easier and it seems people are less likely to cheat because all are watching them. I've been to a WMCCC event and 3 ACCA events where all or some of us were in a room together. It really makes things easier.

I am not saying that CCT should become a nononline event, I am saying that online events are very difficult to handle and the type of ideas that have been offered in the recent threads have not been well thought out nor are they sufficient.

The recent rule changes are steps in a good direction.

[Dirt]

As to the book makers:

Two teams will not be able to use a book created from the same book maker.

i.e. Ted Langreck makes Crafty's books. He would be prevented from making a book for another participant. Just as a participant would be prevented from using a book created by him.

If we only allow one engine per author, then we only allow one book per book maker. That is fair right?

This sounds utterly absurd to me.

Why? Should it be OK for me to work on several programs? We have five members working on Crafty. We could split it into 5 different programs that are almost identical, and each of us enter a different version and greatly increase our chances of winning. This is a team vs team competition using programs developed by each _team_. The WCCC adopted this rule several years ago and its a good one. One person per program, whether he codes or writes the book does not matter. Can't be on two different teams at the same time. Huge conflict of interest.

I doubt all the competitors are interested in developing a custom book. Should Mark's performance.bin be banned just because he's no longer around to give someone explicit permission? I think an exception should be made for free publicly available books. At the least I see no conflict of interest there.

[bob]

Hello everyone,

Before, during, and after CCT12, there were many rules "tweaks" that were suggested, and I think they are a terrific idea.

In reference to rule:

Code: Select all

2. Only the original author, Operator or a team member of the original program may enter and operate. The definition of a team member is as follows:

An individual who has internal knowledge of the program and is recognized by the author as a member of the development team.. i.e. Developer, book maker. This person must have advanced knowledge of the program, such as how the program is designed, hashing schemes, book making procedures..

2a. Operators will only be allowed under the following conditions:

2a1. The author must be online during the tournament to participate in discussions in channel 64, and in the event there is an issue with their program they can provide real time support to their operator.

2a2. The operator must use the account that the author has signed up. No exceptions will be made.

2a3. No operator can operate more than one program. No exceptions will be made.

Over the course of the event, we had several authors online, but their operators were AFK. Unusual, but true.

I propose that Author's pick an "operator" or two NOW, train them on engine parameters and which interface/books you would prefer to use. This way, if an event comes up where you need hardware, you have someone readily available, and can perform. They essentially would become "Team Members", thus nullifying the need for an "operator".

You have a full year now to get one or two people ready for the next CCT event, and you have the ACCA events coming up. So I would HIGHLY recommend this practice.

In reference to rule:

Code: Select all

2. No manually operated programs, and all programs must kibitz their evaluation, and book moves/TB hits if possible. Providing as much information as possible for the viewers and participants is key. It should also be noted that 1-3 lines of text is sufficient. No need to scroll out an entire page.

Whispering is NOT what the rule states. It states kibitzing it mandatory. We had several participants who whispered instead of kibitzing, or kibitzed nothing.

Next year, it will be very simple to avoid any complaints. If you are not kibitzing at the beginning of your game, you will forfeit. If you decide that you do not want to abide by this rule, do not enter. It will be strictly enforced.

As to the book makers:

Two teams will not be able to use a book created from the same book maker.

i.e. Ted Langreck makes Crafty's books. He would be prevented from making a book for another participant. Just as a participant would be prevented from using a book created by him.

If we only allow one engine per author, then we only allow one book per book maker. That is fair right?

This sounds utterly absurd to me.

Why? Should it be OK for me to work on several programs? We have five members working on Crafty. We could split it into 5 different programs that are almost identical, and each of us enter a different version and greatly increase our chances of winning. This is a team vs team competition using programs developed by each _team_. The WCCC adopted this rule several years ago and its a good one. One person per program, whether he codes or writes the book does not matter. Can't be on two different teams at the same time. Huge conflict of interest.

I agree about the engines (to some degree) but not about the books.
Professional books often have books written by professional book writers.
Should that prevent an engine from entering the competition or forcing them to remove their book?

A professional book author is perfectly legitimate to add as a team member. But he is a member of _one_ team, supporting _one_ program. There is no way you can allow a single author to provide books for more than one program.

(1) the book is an integral part of the engine. It provides as many moves as the engine provides in the typical 40-50 move game.

(2) conflict of interest? I could put some ugly back-doors into a book, and save them for a case where I need a win, or I want my "other" program to win. Not good.

Even with the engines, I think it requires case by case analysis. For instance, John Merlino worked on ChessMaster and has now written Myrddin. If ChessMaster enters a contest can Myrddin?

I don't think John was ever a programmer on the chessmaster engine. So I don't see a problem. I do see a problem with a programmer jumping around, because again, this poses a conflict of interest with potential misbehaviour.

I do not think that the rules should be applied to the point that the tail is wagging the dog.

There are actually good reasons for restricting a single person to a single team.

Ok one to this one:

Code: Select all

5. In the event of a disconnection, the party will be given 5 minutes to return to complete the game and no more than 2 disconnections  per game will be allowed. On the third time, the game will be a forfeit.

This one is being changed to 10 minutes, instead of 5. Nothing too significant, but should correct some issues.

More are coming, as I decide how to word them.

Peter

[bob]

Is there a compliant way of making xboard kibbitz book moves from a uci engine? I'm not aware of a way to make this work, currently.

Crafty started the "online kibitzing" stuff back in 1995. It has always done its own thing using the "tellics kibitz ..." approach. Any programmer can use that and not be lazy enough to depend on the various GUIs to do everything for them... Or else they can choose a GUI that works properly. Or they can contact the GUI author and ask for a fix so that the GUI can be used in online tournaments. Just saying "the GUI won't do this" is not good enough.

There's no need to be so combative. I'm not trying to avoid kibbitzing book moves, I'm only trying to find the best way of doing so. If I just add "tellics" printouts to daydreamer, I think polyglot will eat them. If polyglot needs to be modified to support this feature, no one is well-served by everybody running off and hacking up their own incompatible solution to avoid your charges of laziness. We (and by "we" I mean the community of UCI engine authors that wish to participate in tournaments through polyglot) should take the time to design a proper solution that all UCI engines can benefit from.

I'm not "being combative." I am simply tired of seeing this problem every time we have an event. Even though we have had the rules in place for several years now. Yet we keep coming back to "I can't kibitz book moves" or "I can't kibitz a move on a ponder hit." Or "I can't kibitz an instant reply move" or "I can't kibitz when escaping check and there is only one legal move."

"can't" is bogus. This _is_ computer software, and we can make it do whatever we choose to make it do. Some just don't _choose_ to make it comply with the rules because it takes a little effort.

If polyglot is the problem, don't use it. UCI is not mandated as a protocol, the xboard protocol handles this perfectly. If UCI can't deal with this, and the GUIs that use UCI can't deal with this, then there is an obvious solution. I leave that as an exercise for the reader to work out.