[WCA Regulations 2014] Mean of N for FMC

[toruu]

In my opinion, a mean of x (2 or 3) would reduce most of the complaining about lucky scrambles or finishes. Also good people like Sébastien could achieve a better rank than they have now. I mean Sébastien is one of the world's best in FMC,
but doesn't even have the NR. Don't get me wrong, I don't say Moritz 21 was totally unfair/lucky. Looking at the German FMC-Thread , Moritz often posted sub-30 results,so I can't say he got noob luck. But a system like that and a time limit of 60 min or one attempt per day would imo show who has real skill. And for god's sace the shitstorm of what is a lucky result and what isn't would finally stop.

 

[qqwref]

I would only be interested in a fewest moves mean if we can get 2 or 3 attempts done in one hour.

 

[okayama]

This has been discussed before, I stand by my conclusion that median of three (so basically just the second best result) would be almost as indicative of skill as a mean would, but without being so punishing if you DNF for what ever reason. But yeah, (sadly) few comps have 3 (or more) FMC attempts.

FYI: previous discussion thread: Mean of x for FMC
I also still hope mean of x (2 or 3) or median of 3 format.
NB: not banning "best of x", but just adding an option to choose for organizers.
If organizers do not have enough time, of course they can choose best of 1.

More NB: increment of x in "Best of x" does not prevent winning with a lucky solution,
but just enhances the odds of such an win by someone.

Another suggestion I have:
Everyone gets two (or three?) scrambles and 60 minutes time to submit solutions for both of them. The mean of both solves determine the winner of the event.
This adds some interesting tactical challenges (how much time do you want to invest in cube x?) while reducing the luck factor considerably.

FYI: previous discussion thread: 1 Hour FMC is a Luck Based Event

 

[ThomasJE]

I think we can just add Mean of 2 or 3 to available formats and have 3 separate WR lists (1, Mo2 and Mo3). This way, if there is time for two or three, we can use the means and where only time for 1 is available we can use the Best of 1 system; so it is flexible.

 

[Sebastien]

To have 2 different average rankings for one event makes absolutely no sense...

An Mean of 2 ranking would work quite well though I think.

 

[Noahaha]

I might be biased, but I think 3BLD deserves a mo3 category before FMC does. I mean, we already do the means. They just aren't ranked for some reason.

 

[Sebastien]

This is not only biased but also very unreflected.

Two things are absolutely obvious:

1. Luck is a bigger factor in FMC than in BLD.
2. It is easier to get a valid solve in FMC than in BLD.

Apparently this qualifies FMC more for a Mean than BLD.

 

[qqwref]

Maybe it does on the basis of the qualities of the solve itself. But you forget that for a top solver BLD takes a minute (including scrambling, even) and FMC takes an hour. If you want to hold mean of 3* that will take 3 hours, so now you're telling organizers that they can't give people an FMC average unless they have time to hold a 3 hour event. That is a lot of time to have competitors sitting in a room unable to compete in other stuff (or judge). I seem to remember multiBLD having to be fundamentally changed because it was getting into that kind of length.

As far as BLD is concerned, I would even consider averages of 5. Organizers can always do combined finals, for top people it's faster than 5x5x5 (which has been avg5 for ages), and it would allow for a DNF. But I think we'd really have to think hard about whether we should rank people by single first or not.


*Come on, let's be honest, no other event has ever even considered mean of 2 - for a reason. The fact that anyone thinks it makes more sense than mean of 3 just shows how much of a problem super-long events are. And, with mean of 2, if you get super lucky on one solve, you don't need to be all that good to beat a truly skilled person. Suppose someone gets a 20 move solution on solve 1; now they only need to get under 35 or so to beat someone else who averages just under 30. Mean of 2 doesn't solve the problem and is a silly idea by itself.

 

[Noahaha]

This is not only biased but also very unreflected.

Two things are absolutely obvious:

1. Luck is a bigger factor in FMC than in BLD.
2. It is easier to get a valid solve in FMC than in BLD.

Apparently this qualifies FMC more for a Mean than BLD.

1. I agree, but luck is a HUGE factor in BLD as well. At this point, even Marcel needs a lucky scramble to break the WR.
2. That's why BLD mo3s deserve recognition.

NOTE: I am NOT saying competitions should be based on mo3s, just that there should be a WR and world rankings for it. Just like how for 3x3 there is a single WR but having the best 3x3 time won't win you a competition.

Also... the FMC thing would fundamentally change the way FMC is run at certain competitions, whereas BLD would change nothing except make it more fun for top BLDers. Imagine if 3x3 were based on single and there were no average rankings... it would suck because everyone would just go to each competition hoping for a PLL or OLL skip. That's how I feel about BLD, like the results I get are very much out of my control.

 

[tim]

1. I agree, but luck is a HUGE factor in BLD as well.

There are actually two types of results in every event:
- the result of the competition itself
- the world rankings

I'd argue that BLD is the fairest (read: basically no luck involved) WCA event when you consider competition results. (Everyone benefits from the same kind of scramble since everyone uses roughly the same method. Compare that to 3x3x3 where you could get all kinds of skips depending on how you approach your solve.)
On the other hand, an easy scramble/luck obviously helps to improve your world ranking in BLD. But everyone who went to more than three competitions did very likely get one of those easy scrambles. I also don't see anyone in the top-10 who doesn't deserve to be in there (everyone's second best time is very close to his PB — except for Aaron's, he dnfed a lot).
Compare that to the top-10 FMC rankings (best solve, second best solve):
(20, 26), (21, 33), (22, 34), (22, 32), (23, 31), (24, 26), (24, 24), (24, 25), (24, 26), (24, 28)

At this point, even Marcel needs a lucky scramble to break the WR.

Actually, I don't think that's true.

Anyway: I don't see how we could get rid of lucky competition winners/record holders in FMC without either changing the nature of said event significantly or requiring more time for it. Since saving time was one of the major arguments for removing old multi BLD I don't see how adding more time could be justified for FMC.
Therefore I'd like to suggest an alternative to FMC:
Every competitor gets one scrambled cube and 5-10(?) minutes time. Every move the competitor does counts towards his result and he can't undo any move (so he does a so called linear solve). In the end an average of five attempts determines the winner. Here's a video of how it might look like (just ignore the stupid chess clock and the second cuber ^^). I can totally see myself watching a final on stage while rooting for my favorite block building geek.

 

[Noahaha]

Ok... you're right. I was biased after all, but I still think that adding mo3 rankings to the official rankings section of the WCA would be a very easy way to make a lot of people happy.

 

[irontwig]

Mean of 3 for BLD just makes sense as an interesting statistic rather than a way to decide who wins a competition, I mean would you really want to see a 1:57.21 2:01.22 2:31:98 win over 31.75 29.72 DNF?

 

[Sebastien]

Anyway: I don't see how we could get rid of lucky competition winners/record holders in FMC without either changing the nature of said event significantly or requiring more time for it. Since saving time was one of the major arguments for removing old multi BLD I don't see how adding more time could be justified for FMC.

Who talks about adding time? A Mean would of course just be an option as for all the other events. You know that most WCA events could be held as "Best of", right?

So no need to add time, just rank by "Mean of 2" instead of "Best of 2" when 2 attempts are planned anyway and do "Best of 1" when only one attempt is planned. There would be enough competitions where you can do a Mean I guess. Of course the important major competitions should have a Mean. But the upcoming WC for example has 2 attempt already.

Therefore I'd like to suggest an alternative to FMC:
Every competitor gets one scrambled cube and 5-10(?) minutes time. Every move the competitor does counts towards his result and he can't undo any move (so he does a so called linear solve). In the end an average of five attempts determines the winner. Here's a video of how it might look like (just ignore the stupid chess clock and the second cuber ^^). I can totally see myself watching a final on stage while rooting for my favorite block building geek.

That's for sure a funny event but no accurate replacement for FMC. FMC is absolutely not linear and as long as you don't want to ban the use ofall the amazing theoretical knowledge good FMCers have, please don't propose a linear replacement. Also, this event would we painful to judge.

 

[Kirjava]

That's for sure a funny event but no accurate replacement for FMC. FMC is absolutely not linear and as long as you don't want to ban the use ofall the amazing theoretical knowledge good FMCers have, please don't propose a linear replacement. Also, this event would we painful to judge.

I would be for changing to a linear format if we were changing to some kind of mean or average as it would greatly reduce the required time taken. It changes the event a lot, but it's no different than changing the format of multibld - like we have done and like the WCA are considering doing yet again.

Who are you to say that FMC is not linear? That is totally a matter of opinion.

 

[Sebastien]

Who are you to say that FMC is not linear? That is totally a matter of opinion.

Taking "FMC" as the current official event, it is a fact and absolutely not a matter of opinion.

 

[Kirjava]

Taking "FMC" as the current official event, it is a fact and absolutely not a matter of opinion.

The current event uses this format, but that is subject to change.

The idea of solving in the fewest number of moves does not itself denote a lack of linearity

 

[Sebastien]

Sure. I just wanted to explain what I was talking about before and that it differs from what you understood.

anyway: Proposing linearity for the official event is just not practical because of the judging aspect.

 

[tim]

Who talks about adding time? A Mean would of course just be an option as for all the other events. You know that most WCA events could be held as "Best of", right?

I'm aware of that and I think it's bad that different organizers could determine the winner of an event differently.

as long as you don't want to ban the use ofall the amazing theoretical knowledge good FMCers have

e.g.? I think you overestimate the amount of theoretical knowledge required for being good at FMC. I just looked through the last five pages of our local FMC thread and the only two theoretical concepts people seem to use are commutators (for insertions) and switching to the inverse scramble. Both are concepts which could be learned in a day. Being efficient at/experienced in block building seems to be much more important. And these are things which would be encouraged in a linear format.
(Besides, I always feel cheap for using theoretical knowledge (commutators and knowledge about permutations) when solving a new puzzle since it doesn't actually require me to gather a deeper understanding of the puzzle itself.)

Also, this event would we painful to judge.

That's the only point I could agree with. But it's not as bad as it might seem. You need one judge for each competitor and one mechanical counter. 10 minutes (the upper bound of my proposed time limit) is shorter than your average Multi/Big Cubes BLD attempt (where you theoretically also have look at your competitor's cube during memo). By using cutoffs (Best of 2 / Average of 5 (< 60 moves)) like we already do you can also quickly reduce the amount of required judges.

 

[Noahaha]

[WCA Regulations 2012] Mean of N for FMC

Mean of 3 for BLD just makes sense as an interesting statistic rather than a way to decide who wins a competition, I mean would you really want to see a 1:57.21 2:01.22 2:31:98 win over 31.75 29.72 DNF?

Not sure if you think I was implying that, but as I've already said, competitions would still be based on singles in the same way. There would just be a new WR and a new set of world rankings. It is already an interesting statistic, but is there any reason not to make it an official interesting statistic?

 

[Kirjava]

anyway: Proposing linearity for the official event is just not practical because of the judging aspect.

Agree.

However, I don't believe mean of 3 is practical either due to time restrictions. Some comps may have it, but entire countries will go without it indefinitely.

Are we honestly still bothered by perceived "luck"?