2009 WCRCC: Bright/Spark issue

[Harvey Williamson]

Do whatever you guys want - this will ensure that engines like Hiarcs do not play any more. I am sure this will apply to Rybka also. Even though Rybka also always supply an operator that is 100% committed to the Rybka project.

Why can't we have a fully automated event without the need for operators? If such were possible, would you still object to participating?

Isn't that called SSDF?

SSDF only handles Winboard/Windows programs nowadays, doesn't it? Symbolic will never run on a Microsoft O/S.

Anyway, the SSDF doesn't allow for customized hardware, or for distributed processing.

So you guys go ahead with the primary programmer only. Does that mean if I can not operate Hiarcs that Shay can not operate Junior or Larry Rybka? Your rule, proposal, has a lot of merit but just needs a bit of common sense.

[CRoberson]

How about a new rule to be used for future events:

"Each program must be operated by its primary author, and no author may operate more than one program."

This works for me, and has worked well many times in the past.

Well what if the primary author is not available because he has a family? Or a family with special needs? I agree the operator should be someone who has a permanent relationship with the engine and knows about the engine and gives far more hours to the engine than they do the job that pays his salary. This is a rule I could accept.

Hi Havey,

You are a very smart guy, so I know you must have thought of this: many of us don't have teams - we are individuals on our projects.

Thus, your rule (as good intentioned as it is) would give the rest of us an unfair disadvantage.

If you could figure out how to restate it in a way which covers all the bases, I'd be very happy to consider it. We could all discuss it here.

A thought I had was that new entrants must not use an operator. That is too strict - last year Dieter had an operator and he was online himself.
So, changing it to Author must be online their first year is better. However, I am not completely sure that is a good rule either.

[Harvey Williamson]

How about a new rule to be used for future events:

"Each program must be operated by its primary author, and no author may operate more than one program."

This works for me, and has worked well many times in the past.

Well what if the primary author is not available because he has a family? Or a family with special needs? I agree the operator should be someone who has a permanent relationship with the engine and knows about the engine and gives far more hours to the engine than they do the job that pays his salary. This is a rule I could accept.

Hi Havey,

You are a very smart guy, so I know you must have thought of this: many of us don't have teams - we are individuals on our projects.

Thus, your rule (as good intentioned as it is) would give the rest of us an unfair disadvantage.

If you could figure out how to restate it in a way which covers all the bases, I'd be very happy to consider it. We could all discuss it here.

A thought I had was that new entrants must not use an operator. That is too strict - last year Dieter had an operator and he was online himself.
So, changing it to Author must be online their first year is better. However, I am not completely sure that is a good rule either.

maybe I should add AND ARE UNPAID. That probably rules out Shay and Larry. So yes a sensible compromise is needed.

[sje]

So you guys go ahead with the primary programmer only. Does that mean if I can not operate Hiarcs that Shay can not operate Junior or Larry Rybka? Your rule, proposal, has a lot of merit but just needs a bit of common sense.

Perhaps the rule could allow for a single designated operator by the primary author while still having the condition of one author/one entrant.

Having in the past hauled my computer and myself to any number of events, the effort of participating in an ICS event pales in comparison to the work required in the Old Days. I guess I don't understand why spending a few weekends per year is too much for some authors, at least considering how much time was actually spent developing the program.

[Harvey Williamson]

So you guys go ahead with the primary programmer only. Does that mean if I can not operate Hiarcs that Shay can not operate Junior or Larry Rybka? Your rule, proposal, has a lot of merit but just needs a bit of common sense.

Perhaps the rule could allow for a single designated operator by the primary author while still having the condition of one author/one entrant.

Having in the past hauled my computer and myself to any number of events, the effort of participating in an ICS event pales in comparison to the work required in the Old Days. I guess I don't understand why spending a few weekends per year is too much for some authors, at least considering how much time was actually spent developing the program.

That seems reasonable.

[CRoberson]

So, I think there are 2 or more rules that need reconsidering and both rules are allowed in the CCT and ACCA events.

What is the point of reconsidering the rules, or in fact have any rules at all for future events, now it turns out that the rules are treated as a joke? This seems just pure misleading of prospective participants.

If you want to be honest, for next year's addition you should announce:

For this event you will be at the total mercy of the TD. He can impose any demand on you, nuke you whenever he wants, or do otherwise as he pleases, without having to observe any rules whatsoever. You will have to trust that he knows what is best, as he is a good guy.

H.G. it is not that bad. Rules have to be adjusted almost yearly for events such as this. The issue is that every year somebody tries
to cheat the rules or maybe not cheat but take extreme advantage of them.

Case in point: Every year I get entry requests from people none of us have every heard of and they have a full featured chess program that they claim is around 2200.
I tell them to go on ICC and get rated. They do and guess what? They have a rating in excess of 2800.
At that point, I congratulate them on their excellent GM strength program and proceed to ask questions about their data structures and algorithms. Its quite amazing how many times I get nonsensical data structures answers to my algorithm questions and vice-versa. At this point, they get nervous and I don't hear from them again.

Last year I added a rule that the tournament doesn't allow unheard of authors to enter unheard of programs. They must become
recognized first and they can do that via ICC, FICS, CCRL, RWBC, WBEC, OpenWar, ChessWar or any of those. This new rule has
worked rather well in multiple tournaments and has been adopted by CCT.

Also, each year there is one person that changes his name and the name of his clone and tries to enter. There are some discrepancies on
his part that allow me to recognize him. I always give him the benefit of the doubt and he gets caught anyway (except in one event). But, he was caught as a cloner a few months later.
It is a little difficult to catch people cloning programs in the 2300 range.

I am not fond of a rule that restricts you and I from entering two programs, but it may be the best overall solution for future events.

I am not fond of a rule that restricts all operators/authors from operating a second program, but it may be the best overall solution for future events.

[bob]

All completely irrelevant.

As it pertains to what would be the best rules, which is not the issue here. The rules have been chosen month ago.

And as I said, this particular point was (a) not noticed by me (or by many others it would seem) and (b) is contrary to all other CC tournaments. So it is not "completely irrelevant". It is an important consideration.

I agree it is an important issue and I am considering it.

Also, another issue came up that I handled through email and nobody here knows about it. The issue has been allowed for sometime
in ACCA and CCT events. The issue of one operator for multiple programs. Often an author will enter and operate another program as well.
The reason this happens is that several authors only feel good about letting another author run their program. Experience is the factor.
We have had operators that ran two programs, because they have the equipment. This year we had a potential issue
where one person was going to operate two programs both with CCRL ratings > 3000. I decided to disallow this as it would create
the potential for manipulating the tournament outcome.

So, I think there are 2 or more rules that need reconsidering and both rules are allowed in the CCT and ACCA events.

I view this as less of an issue, but it does offer the chance for unseen interference, since a single operator may have a vested interest in one program, and could easily interfere if he is running two engines on his computers. Be easy enough to run an _extra_ engine on one when the two are paired against each other to penalize one because it won't get 100% of its machine.

So it can end up just like the two entries from one author problem. A truly independent operator (non-author) could, I suppose, be acceptable, but how do you determine "true independence" from any engine???

My philosophy has always been "If a rule is difficult or impossible to enforce, then it is not worth having the rule in the first place."

I prefer one person, one program, to avoid the issue completely.

[bob]

All completely irrelevant.

As it pertains to what would be the best rules, which is not the issue here. The rules have been chosen month ago.

And as I said, this particular point was (a) not noticed by me (or by many others it would seem) and (b) is contrary to all other CC tournaments. So it is not "completely irrelevant". It is an important consideration.

I agree it is an important issue and I am considering it.

Also, another issue came up that I handled through email and nobody here knows about it. The issue has been allowed for sometime
in ACCA and CCT events. The issue of one operator for multiple programs. Often an author will enter and operate another program as well.
The reason this happens is that several authors only feel good about letting another author run their program. Experience is the factor.
We have had operators that ran two programs, because they have the equipment. This year we had a potential issue
where one person was going to operate two programs both with CCRL ratings > 3000. I decided to disallow this as it would create
the potential for manipulating the tournament outcome.

So, I think there are 2 or more rules that need reconsidering and both rules are allowed in the CCT and ACCA events.

Usually entering an event conveys agreement with the rules of said event. Failure to read the rules BEFORE agreeing to them comes across to me as a bad idea, and poor justification for arguing about them later.

You buy 30 cars from the same dealer over a 40 year period of time. Do you read the _last_ contract just as carefully as you read the first 29, particularly when the first 29 were always the same???

[bob]

So, I think there are 2 or more rules that need reconsidering and both rules are allowed in the CCT and ACCA events.

What is the point of reconsidering the rules, or in fact have any rules at all for future events, now it turns out that the rules are treated as a joke? This seems just pure misleading of prospective participants.

If you want to be honest, for next year's addition you should announce:

For this event you will be at the total mercy of the TD. He can impose any demand on you, nuke you whenever he wants, or do otherwise as he pleases, without having to observe any rules whatsoever. You will have to trust that he knows what is best, as he is a good guy.

This is a crock. You _appear_ to be claiming that the rule allowing two programs from one author led you to expend significant effort to produce that second program. I don't believe that for a minute. Which means this is only about the ability to have two entrants from one author, something that has not been allowed previously in almost 40 years of computer chess tournaments.

That makes your argument a bit silly. A change in this rule costs you exactly nothing. And prevents some potential problems that could cost us all something tangible.

To claim a rule change somehow "damages" you requires that there actually be some sort of "damage". I don't see it myself.

[bob]

And as I said, this particular point was (a) not noticed by me (or by many others it would seem) and (b) is contrary to all other CC tournaments. So it is not "completely irrelevant". It is an important consideration.

Well, next time pay attention before starting to make silly comments, then.

Even what you write here is apparently totally un-informed nonsense. Several earlier instances of this tournament have allowed multiple entries by the same author. CCT has always allowed multiple entries by the same author. ChessWar allows multiple entries by the same author.

So "all other CC tournaments" is only the ICGA WCC? Since when do you want to slavishly copy the rules of that one? Does it also mean you now are in favor of an 8-core rule for on-line tournaments, just because "all other CC tournaments" require that.

You are just totally out of touch with reality...

One of us is, that is for sure. Please cite an example where in prior CCT or ACM or WCCC or WMCCC events, one author has been allowed to enter _two_ programs. Then we can discuss it. However, you are going to have a small problem coming up with the programs and author names...

I don't care about ChessWar and such. We are talking about an online event with programs written and operated by the author. There is a difference when the event is "live". The "offline" events can run however they want, and authors can choose whether or not to allow their engine to participate.

One of us is really out of touch, I just suspect it is not the person you think.