Future of competitive speedsolving – participation of delegates

[Lucas Garron]

That's why I liked the idea of scrambles being auto-generated by the WCA. When you submit your competition to the database before the competition, you would tell what the events and rounds are, and the scrambles would be generated on a WCA server automatically, then served on the day of the competition. Since the scrambles would be stored there as they're generated, someone could verify the scrambles used at the competition with those on the server, to make sure they weren't manipulated. The only danger of manipulating the scramble program then should theoretically be hacking the server, right?

+ communication channels (who's interacting with the server at the competition and how?) and internet access.

 

[pjk]

That's why I liked the idea of scrambles being auto-generated by the WCA. When you submit your competition to the database before the competition, you would tell what the events and rounds are, and the scrambles would be generated on a WCA server automatically, then served on the day of the competition. Since the scrambles would be stored there as they're generated, someone could verify the scrambles used at the competition with those on the server, to make sure they weren't manipulated. The only danger of manipulating the scramble program then should theoretically be hacking the server, right?

If it gets to the point where we worry about people hacking the server to get scrambles before a competition, we've got it wrong. At any competition, there has to be some trust (with scramblers scrambling, judges judging, delegates providing real scrambles w/o looking at them, etc.). The delegate issue is just one of many places throughout a competition where there needs to be trust. Having 2 delegates would certainly reduce the chance of someone cheating, and having 2 scramblers verifying a scramble would reduce the risk of a fixed scramble too. Using a scrambler on the spot would be a good idea, assuming 2 delegates can verify the scrambles are being generated right then and there. Otherwise, generating scrambles on the spot isn't doing much to prevent 1 delegate from cheating compared to if the delegated generated the night before. Another issue with having 2 delegates is that competitions can sometimes have a hard time getting 1 official delegate to show up.

I think more delegates would do more good than harm. There are loads of people out there probably more qualified than some of the current delegates that could and would be willing to help out in this situation, and I think they should be allowed. Why not gradually expand on the number of delegates across the world? With so many people involved in scrambling and judging, the more qualified people watching, the better. If someone goes to a competition knowing they have a role (such as being 1 of 2 delegates), they will be more keen to making sure things flow properly than if they showed up and get pulled in last minute as a volunteer.

 

[Vincents]

I think that in high cuber density areas (e.g. California), having more than one delegate is feasible. However, just a few months back, Jeremy had an insane travel plan of competition hopping through the midwest just to delegate competitions because nobody else could make it. Sometimes I think the energetic toll can be larger a burden than the financial toll. In the end, this is a hobby, and if it starts affecting peoples jobs and schoolwork with no real benefit in return, you're going to get a lot more cuber apathy.

I think, though, that at least for competitions I run, we will start generating at the competition real-time. Benefits are obvious. Anyone running a competition with an active internet connection and nice battery life (or a nearby power source) can do it.

 

[Erik]

Lets go through it in a few substeps:

Situation now:
Delegates produce all scrambles for a competition, but fact is too that some delegates will compete in almost all events

Is this unwanted?
Like stated by several people here, yes. My opinion though: if we cannot trust the delegates in looking at the scrambles, how can we trust them in anything else they do? All delegates are chosen only if they have proven to be reliable enough. So unwanted? Yes maybe. Is it a serious threath to competitions? Nah.

It is never possible to make something 100% cheat-proof. If we already stop trusting delegates, then we might as well put all cubers in seperate prison cells as a waiting area so they can't communicate...

Possible solutions:
Lets say we want to change it though, it seems a majority would like this. Several options were brainstormed:
- organisers help (partly) in providing the scrambles: I am against this, only already for the fact that an organiser is not a delegate (thus not proven trustworthy) and sometimes not even a cuber. On the other side it also happens that the delegate is the organiser.
- 2 delegates: I would favor this, but is it feasible? At big competitions: certainly. World champs, US nationals, Aachen Open etc. all big competitions with several delegates present. The 2nd delegate wouldn't even have to do much, besides bringing a few scrambles. At smaller competitions, 2 delegates are not always present.
- the delegate can just not compete: honestly, which delegate would want to go to a competition, travel, pay for gas, pay for a hotel and be ok with the fact he/she couldn't compete? Like Pedro said, he'd consider not going at all. So I'm quite confident we can agree this is not a solution.


Btw.
@ Stefan: 'if you want to cheat you simply manipulate the scramble program': do you think that's the easiest way of cheating if you wanted to? ;-)

 

[TMOY]

Another possible solution: the delegate chooses another competitor he considers trustworthy enough (not necessarily a WCA delegate), and that person provides the scrambles for some events. For small comps, this is definitely more feasible than requiring 2 delegates.

 

[Stefan]

WCA could - in addition to the list of delegates - create a list of delegate assistants. Those would assist the delegates with scrambles and maybe a bit more, but not have the responsibility or authority to run a competition on their own. This would provide benefits of having more delegates without actually making more people delegates. Though those assistants could be likely delegate candidates, their assistant work acting as training where they get to know delegating a bit and the delegates get to know them a bit.

+ communication channels (who's interacting with the server at the competition and how?) and internet access.

That was my first thought as well, but for "manipulating the scramble program" that generates scrambles on the server, I'd say Mike was correct. And while the scrambles could me manipulated through the local network, people can still take them home and check them there and see the manipulation.

@ Stefan: 'if you want to cheat you simply manipulate the scramble program': do you think that's the easiest way of cheating if you wanted to? ;-)

Well, my point was just that witnesses wouldn't see manipulations in the program, so while it prevents simply bringing already prepared special scramble sheets, it doesn't ensure they're not generated there.

And yes, unless I'm willing to trust an accomplice, that's the easiest way I can think of.

 

[Erik]

Haha, I think there are a lot easier ways to cheat if you really wanted to, but lets not discuss that here.

Delegate-assistants sounds like a good plan!

They could also relieve the pressure on the delegate by:
- starting new rounds
- determining if something is a +2 or not.
- find people to scramble or judge if necessary

 

[Bryan]

WCA could - in addition to the list of delegates - create a list of delegate assistants. Those would assist the delegates with scrambles and maybe a bit more, but not have the responsibility or authority to run a competition on their own. This would provide benefits of having more delegates without actually making more people delegates. Though those assistants could be likely delegate candidates, their assistant work acting as training where they get to know delegating a bit and the delegates get to know them a bit.

It seems like this would be a lot of work screening people, etc. Plus some organizers would then have to secure two people. Or people who organize and delegate on their own now have another hassle.

That was my first thought as well, but for "manipulating the scramble program" that generates scrambles on the server, I'd say Mike was correct. And while the scrambles could me manipulated through the local network, people can still take them home and check them there and see the manipulation.

I guess the WCA server could generate some MD5 hash of the scramble along with a salt. But honestly, I don't like the idea of being forced to get the scrambles right before a competition. I'm busy with other stuff.

Also, I thought about the idea of not being able to compete, and I'm wondering how regular competitors would react to that situation. Let's say when the competition starts, a random set of competitors are chosen to lead an event rather than compete in it. It would help keep things on track easily.

 

[Escher]

Also, I thought about the idea of not being able to compete, and I'm wondering how regular competitors would react to that situation. Let's say when the competition starts, a random set of competitors are chosen to lead an event rather than compete in it. It would help keep things on track easily.

I think the phrase is '**** that ****'

Imo it's not even a question for any competition that doesn't have at least 200+ competitors. It (getting WRs, ERs, personal records) doesn't matter at all unless money is involved, or the competition is sponsored by the Rubik's brand, or until current WR holders get bought tickets to Worlds or regional championships or something.

Cubing is far too young for the kind of things people are suggesting. Until it actually matters; do nothing.

 

[Stefan]

It seems like this would be a lot of work screening people, etc. Plus some organizers would then have to secure two people. Or people who organize and delegate on their own now have another hassle.

Yes, it's certainly more work than now (though also has benefits). But I'm mainly suggesting this as an alternative to the other suggestion of having two full-blown delegates. I think it would be easier than that, still get most of the benefits, and help produce more delegates. TMOY's suggestion also goes in this direction, though even less far. It's like:

A) 1 delegate (current situation)
B) 1 delegate, 1 assistant chosen by delegate
C) 1 delegate, 1 assistant accepted beforehand by WCA
D) 2 delegates

 

[Bryan]

I think the phrase is '**** that ****'

Imo it's not even a question for any competition that doesn't have at least 200+ competitors.

No, it's not even a question ever, regardless of size. I was merely pointing out stuff in a way that allows all people to realize that they really don't want to give up competing.

 

[Escher]

No, it's not even a question ever, regardless of size. I was merely pointing out stuff in a way that allows all people to realize that they really don't want to give up competing.

Sorry, I understood but I wasn't clear in my reply - the first instance of 'it' was meant to be about any further measures we've currently discussed to prevent cheating.

 

[Meep]

Yes, it's certainly more work than now (though also has benefits). But I'm mainly suggesting this as an alternative to the other suggestion of having two full-blown delegates. I think it would be easier than that, still get most of the benefits, and help produce more delegates. TMOY's suggestion also goes in this direction, though even less far. It's like:

A) 1 delegate (current situation)
B) 1 delegate, 1 assistant chosen by delegate
C) 1 delegate, 1 assistant accepted beforehand by WCA
D) 2 delegates

This is somewhat how the competitions in our area are run. Despite the delegate being listed under a single name, a little group of three split the duties (Like having three delegates, or one and two assistants) at the competition, with one of them only competing if the BLD event exists. When scrambles are generated, they are done in a group (or with someone a delegate from elsewhere recommends). It's entirely possible for two or more people to work to cheat together, but less likely.

 

[kinch2002]

Delegate-assistants sounds like a good plan!

They could also relieve the pressure on the delegate by:
- starting new rounds
- determining if something is a +2 or not.
- find people to scramble or judge if necessary

I was under the impression that jobs 1 and 3 there were for the organiser anyway, not the delegate? I realise they are often the same person, in which case they should probably be getting other people to help organise rather than assist in the 'more important' delegate duties.

 

[Bob]

I haven't seen it mentioned here, so I will bring it up. There is something called a "main judge" for each event during a WCA competition.

1d) Each event must have one main judge. The main judge for an event is responsible for:
1d1) Making sure that the regulations are followed.
1d2) Decisions on disqualification of a competitor for an event, as described in other articles.
1d3) Decisions on starting a round later or earlier than scheduled. In case of changes in the time schedule, there should be a clear announcement to all competitors.

These regulations used to insist that the main judge could not compete in the event. However, we decided that if the WCA delegate is the main judge, he could compete in that event (see Regulation 1i). This decision was reached because the WCA delegate is trusted enough that he will not cheat, especially in the events that he is competing in. If he would, he should not be a WCA delegate. Before this regulation, I used to assign somebody who was not competing in the event as a main judge, but it was usually somebody who knew so little about that event that I would make the decisions anyway. It was really hard to find people who are knowledgeable enough about an event that they could be in charge of it but didn't want to compete in it.

Today, I think the quality of delegates has improved dramatically over the years. Sure, we've had a couple that were poorly chosen, but I think most of the delegates do a great job.

This leads me to a question. Do you think that I would generate scrambles and cheat with them? If I did, you'd probably never know because I couldn't even set a WR with them if I tried. What about Tyson Mao? Or Tim Reynolds? Or Bryan Logan? Or Ron van Bruchem? Would they cheat? If they would, we need to start looking at a new set of delegates.

 

[Erik]

I was under the impression that jobs 1 and 3 there were for the organiser anyway, not the delegate? I realise they are often the same person, in which case they should probably be getting other people to help organise rather than assist in the 'more important' delegate duties.

Technically, it is the organisers task to do 1 and 3 indeed. But like you already said yourself they are often the same person. Also, I have seen it happen that the organiser wasn't even a cuber (!) so the delegate had to do all things even though he wasn't the organiser.
I realise that it doesn't really matter than much if someone is an official asisstent or not, to be able to do 1 and 3 anyway, but for number 3 it would give him/her a bit more authority.

@ Bob, totally agree with this.