CCT: New rules for CCT Events #2

[Stephan Vermeire (Brutus)]

...it's the complete opposite to what you say, correctly said it is:

"Unlimited books add to the untrue differences between engines"

By this I mean the only way an iPhone can draw with a cluster is because of an unlimited book. Therefore an unlimited book makes an untrue difference (of zero) between the iPhone engine and the engine running on the 100 core cluster. In reality the true difference between the engines is enormous.

If your iPhone would really make a chance against a cluster, then why doesn't anyone use an iPhone? Because engines DO make a difference!

Don't worry, I see your point. THIS is what I am aming at: I consider a book as part of the 'known and precalculated data' that limits the task of the engine to a specific part of the game. CCT is the type of tournament that has very few limits and I like that a lot. It isn't cheating when you can obtain a cluster. It is just about getting the best circumstances for your engine. Getting the biggest EGTB or the best book is just another one of those circumstances. Using these, you WILL compare engines at the part of the game where they are making the real difference.

If your iPhone is able to draw against a top-engine at CCT, then it teaches us an important lesson: Books make a difference at top level.

From what I hear though people are not interested in the difference between Engine A and Engine B, they are interested in the difference between Engine A+Book A and Engine B+Book B.

PS. Engine A+Book A and Engine B+Book A is also being stamped out, quite correctly I might add.

You are right about that one. Personally I am more interested in:
- finding the best available book using any available engine
- finding the best engine using any available book

Of course you can restrict books to engines or engines to books if authors choose to do so. In many cases the book+engine has been tuned as a combination. But what if a book-author doesn't explicitly 'provide' his book to a single engine? There is no rule about that yet. You could choose to grant the book to the first engine to claim it. That however sounds pretty silly to me...

[garybelton]

By the way ,after r2 we used my book

Er, now I'm confused, your book but with help from Erdo right? He did say that the line played by Sjeng against us in the blitz was "mostly homework". I assumed it was his homework, as when I congratulated him he took it. Congratulations to you (as well??)

Not my problem.

[garybelton]

But what if a book-author doesn't explicitly 'provide' his book to a single engine? There is no rule about that yet. You could choose to grant the book to the first engine to claim it. That however sounds pretty silly to me...

I guess in this case it should be up to the book author to decide? If he is not contactable (or deceased) then that raises another issue if two teams want to use one of his books.

I guess when it comes down to it I don't want to see the contestant go from book to EGTB without the engine getting involved, which is what we are starting to see in some lines on Playchess, maybe this is what people want but I don't find it exciting. The problem will only get worse unless something is done about it, having engines play from the end of ECO lines sounds like it has some promise. Maybe pick a position randomly from a set of large and balanced EPDs (normal chess not 960) and play from there? It comes down to "what is this competition actually trying to test here?"

[garybelton]

The other thing I wanted to say about not having the same book with two participants is this. Any good bookmaker who knows his trade will make sure that any lines in his book have refutations in his book. So when I play Engine+Mybook vs. SameEngine+Mybook I want to see as close to 100% draws as I can get. This is not so good for an online event, and authors always have their favorite lines, even in different books that they create.

[hgm]

Why is a player that does not kibitz required to disconnect?

Wouldn't it be more logical to just have him switch on the kibitzing?

For some using Polyglot + Winboard, or Chess Partner, it is required to disconnect to switch to kibitzing.

With Crafty it is as easy as sending a zippyPassword2 tell.

Peter

That depends on who does the kibitzing, not? The engine, Polyglot, or WinBoard. The engine _might_ implement setting it at any time, Polyglot _does_ implement setting it at any time. And future versions of WinBoard might support switching it on at any time too.

So the question remains, and is actually diversified now:
1) Why do you propose to force Crafty to disconnect, if the problem would be much quicker solved by sending it a tell?
2) Why do you make rules for future tournaments based only on currently existing (or, more accurately, currently already obsolete) software?

Neither of that seems to make very much sense, IMHO.

[hgm]

Revision:
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 is recognized by the author as a member of the development team.. i.e. Developer, book maker. The operator of the team must have knowledge in running the program online.

The rules were formulated as they were for preventing authors to just designate anyone that offers his services just before the tourney as 'team member'. There must be some definition of what 'team member' means, or this would become a completely empty condition.

I think the rules are still wanting, as it is not completely clear what the relation is between the persen that can enter of rule 2 is to the author that must be on-line as per rule 2a1. Does the latter mean the actual programmer, or would it be any of the categories that 'can enter' according to rule 2?

[jdart]

Under rule 2a, I think the participation of Amyan would not have been allowed. So either this rule is going to be enforced or it should be modified.

I suggest instead of 2a:

- the author may designate an operator for the event. It is the author's responsibility to ensure that the operator is familiar enough with the program to operate it in the tourney. The TD must be informed of the operator's identity prior to the tourney and have a means to contact that person.

I suggest removing the requirement that the author be online. However desirable this might be, it should not be an absolute requirement IMO. Stuff happens and sometimes a person can't be there although the program could be - if this rule were in place.

I am ok with the other subparts of rule 2.

[hgm]

Well, this is a clear choice by the organizers. It is a tournament for authors, not a tournament for programs. If the author can't be there, he can't be there, and he is out. We don't want his program, we want the author. If he cannot be there, having the program there is pointless.

[Ferdy]

Revision:
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 is recognized by the author as a member of the development team.. i.e. Developer, book maker. The operator of the team must have knowledge in running the program online.

The rules were formulated as they were for preventing authors to just designate anyone that offers his services just before the tourney as 'team member'. There must be some definition of what 'team member' means, or this would become a completely empty condition.

I think the rules are still wanting, as it is not completely clear what the relation is between the persen that can enter of rule 2 is to the author that must be on-line as per rule 2a1. Does the latter mean the actual programmer, or would it be any of the categories that 'can enter' according to rule 2?

Revision 2:

2. Only the original author, Operator or a team member of the original program may enter and operate. The operator and team member must be recognized by the original author.

2a. Operators will only be allowed under the following conditions:
2a1. The author or team member or both 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.
2a4. The operator must have enough knowledge in running the program online, preferably with experience in online tournaments.

I think in 2, it is preferable that the author will register for the event.
In 2a1, an author is also preferable to be there to provide supports to his program and operator, however in his absence a team member can also do so. This can be considered as a scheme to make sure the tourney goes well. Of course authors and developers should also be in the channel for discussions. But I think it is not necessary that they will be there all the time. I may add that the operator shall be there all the time

[jdart]

Well, this is a clear choice by the organizers. It is a tournament for authors, not a tournament for programs. If the author can't be there, he can't be there, and he is out. We don't want his program, we want the author. If he cannot be there, having the program there is pointless.

Well, clearly we have different opinions on this. I will defer to what the majority wants. But I would point out that, as far as I know, none of the CCT events prior to CCT 12 had this rule in place.