[WCA Regulations 2014] Mean of N for FMC

[Swordsman Kirby]

I'm just saying that the frequency will go way up because of the regulation.

It's also pretty easy to run FMC alongside other events, given that it only needs one judge.

What about the people competing in it? They're now unavailable to judge for three hours.

 

[porkynator]

Preferred format is what it describes: the format that is adviced/preferred/the better one (if time and resources allow).

So organizers can choose which one to use. 'Best of x' or 'mean of 3'.
Yes you can use a best of x format, but x can not be 3 anymore.
You can only do a 'best of 1' or 'best of 2' if you want to do a 'best of' FMC round.

Ok, this isn't very cool.
I like the idea of "Mean of 3" format for FMC, but I think there are many reasons one may choose to hold FMC as "best of 3" (for example, giving a chance for anyone who wants it to be ranked for the average without forcing everyone to do 3 attempts).
Conclusion: I voted for "Yes, I would like competitions to have this option." but I would also like to have the old format. Was the 9b+ addition necessary?

 

[Lucas Garron]

I was personally pretty neutral about the change.

For one, 3 hours of FMC are fairly impractical in the US.

However, I agree with the argument that the new format should help reduce the role of luck, *at big competitions*. While getting lucky is not a crime, it has not been uncommon for someone to win/podium at FMC at a major championship by pure luck. It is reasonable to try to reduce that.

9b+ is an interesting point that did not occur to me. I think it makes sense to have the mean of 3 preferred, but it is the first example of a "preferred" format that is not expected to be used the majority of the time.

 

[Laura O]

The ranking for the top cubers would have been exactly the same with the old and the new rules
The ranking for the mid-level fmc-ers changed. They became bottom level fmc-ers because of DNF's
The ranking for the low-level fmc-ers change. They became mid-level fmc-ers because they got 3 solves.

Because I was one of the competitors you are talking about, I would like to add something to that and explain my results.
I would classify myself in your "mid-level fmc-er" category since my official PB is 36. I am regularly (more or less) attending the FMC competition in the Geman speedcubing forum, where I did some sub30s and I am averaging approximately around 30-32 moves.

As I already told you yesterday, my current aim is to do a good FMC result in competition officially. I generally do not care about good competition rankings. To be honest they are just another piece of paper I put on a large pile. Hopefully that doesn't sound arrogant, but I think that's a common point of view for many people, who are attending competitions frequently.
That's also the reason why I DNFed the second scramble: after a good beginning I couldn't find a good ending and was angry with myself. If I aimed to do a mean I probably could have done this in about 45 moves, but as I said above, I just don't care.
If you take a look at my results, the majority of my FMC results were DNFs and the reason for that is just what I described above. So mean of 3 did not change anything for me.

Apart from the fact, that one small competition does not prove anything, your description of the changes in the rankings seem to be prejudiced. If you take a closer look at the results, the majority of competitors who got a DNF in Duisburg (Adrian, Andreas and me) also had a lot of DNFs in prior competitions. Thus it is just wrong to compare the results of Duisburg with single (not DNF) results from past competitions.

Sebastien didn't propose mean of 3 in 2011. He just made a remark that he would love mean of 2 or mean of 3. If you read this thread from the beginning you can see that he actually was in favor of mean of 2 at first. Then this thread died out. Then he reanimated this thread just after not becoming the world champion in Las Vegas. Jimmy Coll, István Kocza, Steven Xu and Devin Corr-Robinett are the reasons we are having this discussion.

Linus already pointed out that this is not true.
Furthermore I don't know how you made up the connection to the results of the world championship. This is an unfair statement not based on any facts.

And, btw, to forestall any assumptions: this is my objective view I formed since the FMC mean topic was brought up and is not based on any personal relationships.

 

[tseitsei]

So the first competition with Mean of 3 was held this weekend. It proved just how bad this Mean-of-3 idea is:

No it didn't. It was 1 competition with 9 competitors... So really really small sample.
Plus it was the first competition under the new rules, so maybe all competitors haven't yet realized how to change their tactics in order to do well.

This is what Sebastien said: FMC is the event where you are least likely to have a DNF if you pay attention to time and writing. For the top that turned out to be true but not for the middle-class
The amount of DNF-results is much bigger than normal, as is expected

What Sebastien said is true. Just the fact that people did DNF doesn't necessarily mean that it's hard not to DNF.
Just think about it: If you haven't figured out a good solution in (let's say) 50 minutes, you still have time to write down "easier but longer" solution. That shouldn't take more than 5min tops. And you still have 5 min to scramble your all three cubes and apply the move sequence and fix it if it isn't correct. So plenty of time.
People DNF because they try to the last minute to find a good solution and fail. Or because they are too lazy to check if their written solution actually works.

He should compare FMC to bigblind and there a single is what counts and DNF's are acceptable.

Now this line made me extra sad/angry.
Of course FMC IS NOT comparable to big blind in the least. As I explained above you will not DNF in FMC if you just accept a little longer result and CHECK YOUR solution.
But in BLD if you execute 1 move wrongly you will DNF and there is nothing you can do to fix that.
DNFing is and always will be a part of BLD but it doesn't have to be a part of FMC if you just take some precautions.

So all in all I like the change. (As well as 3bld mean of 3 being recognized as a record)

Just my two cents on the topic

 

[Erik]

I still feel a bit mixed about mean of 3 to be honest. One the one hand it's nice that at least WC and EC (I assume FMC will be at least mean of 3 there) will be less luck-based.
On the other hand there are the following things con and pro a mean of 3 format:

con-ish: at mean of 3 rounds you will have to change your approach. In a few of the FMC-rich competitions I organized the following happens: on solve 1 or 2 you get a mediocre result (30 or so). Then on attempt 2 and/or 2 and 3 you can go all risk and try to force-find an ending to a promising start instead of having to have to find a safety solution. This does change the nature of the event.
con: as competitions can be long and time schedules sometimes too tight for competitors it is nice to be able to skip the morning attempt for example, when knowing there will be another one later.
con: As I am a club-chess player as well I know how long periods of hard concentration can influence your performance later on the day. So if during a solve you find the scramble hard, it's nice that you can just quit and save your energy for later.
con: A preferred format should in my opinion be the format which will be used for the vast majority of competitions. Especially in regions where there are not a lot of comps, people will not be able to do a mean of 3.

pro: I like FMC so it would be fun to do more of it
pro-ish: the chances of luck interfering in the results drop, though only by that much. Not because a n00b can still win, but because the good FMC-ers mostly already win against a PLL skip.

All in all: I am not sure if the extra 2 hours of competition is worth the effort. IF you want a mean, then a mean of 2 would be a better compromise I think. The fact that mean of 3 would be consistent with other mean of 3 events is irrelevant.

Also: retroactive records are really stupid. The fact that a horrible BLD-cuber like me suddenly has had an ER in blind is laughable. ER's and WR's should still have value, even if it are old records.


EDIT:

Now this line made me extra sad/angry.
Of course FMC IS NOT comparable to big blind in the least. As I explained above you will not DNF in FMC if you just accept a little longer result and CHECK YOUR solution.

If you accept "a little longer result" and check your memo again, or just use a second memo method to double memorize your solution you will also not DNF in BLD.

 

[porkynator]

If you accept "a little longer result" and check your memo again, or just use a second memo method to double memorize your solution you will also not DNF in BLD.

There are pops, slippery cubes/hands and many other possible errors not related to memorization/recall.

 

[Erik]

And all of those can all be avoided by being careful. A pop? Really? Use a better cube. That's enough off-topic for now...

 

[kinch2002]

I'm not happy that best of 3 is no longer an option. I didnt realise this. That is not the only thing that was not made obvious at all during the proposal stage e.g. retroactive records, mean of 3 rankings.

Given that there is a mean of 3 ranking now, I'd like to hold a 3 attempt competition in the UK sometime. But I'd much rather rank it by best of 3 because tbh idc whether I get beaten by a lucky solve or another decent solver in a local comp, and it will of course ensure that most people get results.

 

[Carrot]

I'm not organising or participating in fmc anymore. (that's my opinion about the mean of 3)

 

[tseitsei]

EDIT:

If you accept "a little longer result" and check your memo again, or just use a second memo method to double memorize your solution you will also not DNF in BLD.

No. Just wrong...

You can make your chances better by memoing longer, but a good part of all DNFs are because of execution mistakes.

See the difference here, as I explained earlier, is that in BLD if you do an execution mistake you cannot fix that simply by spending a little more time in your solve, BUT in FMC if you can easily a) notice if your solution is correctly written, because you should always test if it works and b) you can easily fix it as well.

Can you see the difference?

 

[tim]

It wasn't just based on this sample, it was also based on historic knowledge. Sebastien basically claims that if you care for FMC and are somewhat decent at it you will not DNF. I think this previous World Record holder has a different opinion: https://www.worldcubeassociation.org/results/p.php?i=1982RAZO01#333fm

You can't use historical data for the reason you stated yourself:

mean of 3 changes the way you have to approach fmc when you compare it to best of 3.

Sebastien also claimed that safety solves are not needed, but yet he needed a 33 moves solution.

That's a sample size of like 3(?). Also: He had a 32 moves solution in November last year!

BTW. I should have an even stronger opinion against mean of 3 for FMC as you (I voted against it), but I think you're just jumping to conclusions here.

 

[Carrot]

Interesting fact:

Places with Mean of 3 results:
Japan, China: they have dedicated FMC competitions.
Chile: The weird Sudamericano competition where only one guy competed.
Spain: They seem to like FMC
Sebastien: Competed in 6, delegated half of them.

 

[AvGalen]

If you take a look at my results, the majority of my FMC results were DNFs and the reason for that is just what I described above. So mean of 3 did not change anything for me.

Apart from the fact, that one small competition does not prove anything, your description of the changes in the rankings seem to be prejudiced. If you take a closer look at the results, the majority of competitors who got a DNF in Duisburg (Adrian, Andreas and me) also had a lot of DNFs in prior competitions. Thus it is just wrong to compare the results of Duisburg with single (not DNF) results from past competitions.



Linus already pointed out that this is not true.
Furthermore I don't know how you made up the connection to the results of the world championship. This is an unfair statement not based on any facts.

And, btw, to forestall any assumptions: this is my objective view I formed since the FMC mean topic was brought up and is not based on any personal relationships.

Thank you for responding Laura, I was hoping you would write.
I don't understand how you can say "So mean of 3 did not change anything for me". Your approach to FMC seems to be the same as mine and the result is the same as mine: Sometimes you get good results, sometimes you DNF because you can't find a good enough finish. But in the past that didn't matter because your result for the event would be your best of the 3 solves. Now it immediately turns your result into a DNF. How can you still say that it didn't change anything for you.
And that is also why I compare the amount of DNF's from Duisburgs mean-of-3 to previously much lower amounts of DNF's for the entire event. Of course it is not an equal comparison, that is my whole point. Because of the new rule more people get DNF results and that is a bad outcome of a new rule.

Also, I stick with my point that the thread went dead and then came alive again by Sebastien. How can so many people deny that when it is so obvious just by looking back:
29th of April 2013, after about 40 posts in 1 week the thread dies; http://www.speedsolving.com/forum/s...n-of-N-for-FMC&p=848173&viewfull=1#post848173
12th of August 2013, after Sebastien didn't become World Champion he revives the thread; http://www.speedsolving.com/forum/s...n-of-N-for-FMC&p=887963&viewfull=1#post887963

For everyone that thinks it is easy not to DNF: History has shown that lots of competitors make small mistakes in writing down their solution. I mentioned Guus because he is a previous (undisputed/non-lucky) World Record holder that has often missed a podium because of an error in his solution. All it takes is just 1 apostrophe, or being as stupid as writing down the solution to the inverse scramble. Why don't people realise that DNF-ing FMC is extremely common? If you spend 3 hours, spread over 2 days and you are not allowed the slightest mistake that really removes the fun of FMC for me.

 

[tim]

Interesting fact:

Places with Mean of 3 results:
Japan, China: they have dedicated FMC competitions.
Chile: The weird Sudamericano competition where only one guy competed.
Spain: They seem to like FMC
Sebastien: Competed in 6, delegated half of them.

You are at least missing Germany and the Netherlands. And since when is Sébastien a place?

 

[Lucas Garron]

Japan, China: they have dedicated FMC competitions.

Anyone know whether they like/dislike/care about "Mean of 3" vs. "Best of 3"
?

 

[Carrot]

You are at least missing Germany and the Netherlands. And since when is Sébastien a place?

I'm just saying that of all competitions in the world with a competitor that succeeded a mean of 3, only 4 countries managed to not have Sebastien involved to some extend. 1 of them is an odd one out, in other words, only 3 countries in the world actively does FMC 3, the rest just happened to have Sebastien going to the comp. (as it happens, the only reason Danish Special 2011 had 2 attempts for FMC was because Sebastien decided that it should have it, so yes, Sebastien can be persuading, even when he is just a competitor)

Either Sebastien just happens to aim for the 3 attempts competitions else he may have had some influence in the decision of hosting a 3 attempt competition

Sebastien is a place because I wanted to emphasize a point.

 

[AvGalen]

You are at least missing Germany and the Netherlands. And since when is Sébastien a place?

Wasn't it clear what he wanted to say? Let me phrase the results for mean-of-3 this way (no mention of who is the delegate in my way though):
In the past 10 years we only have 8 countries, 22 competitions,
34 people, 54 results for mean-of-3. No big tournament has had 3 attempts.
There simply is no historical support for making the least popular format
(3 attempts) into the preferred format.

Anyone know whether they like/dislike/care about "Mean of 3" vs. "Best of 3"
?

I really don't know. Of course Tomoaki has expressed that he would like "Median" but other than that I haven't heard much from Japan/China.
From the start of this topic it has been clear that there is no support for holding mean-of-3 in the USA except for Mike Hugheys "missing averages" competition. It was actually mentioned in one of the very first posts that it would be extremely hard to fit it into big US competitions. This "no-3-in-USA" has been repeated over and over but there hasn't been a response.

Either Sebastien just happens to aim for the 3 attempts competitions else he may have had some influence in the decision of hosting a 3 attempt competition

I have heard from others that Sebastien has requested extra attempts for FMC and sometimes asked to adjust the time schedule. I want to make clear that he just asks and accepts no without hesitation. Although I think Sebastien is pushing FMC too much we are most certainly not talking about Hayan-situations! I see no problem at all with requesting an extra round or attempt. More fun!

 

[Lucas Garron]

I really don't know. Of course Tomoaki has expressed that he would like "Median" but other than that I haven't heard much from Japan/China.

I actually like "Median of 3". It just seems annoying to argue for it, though.

From the start of this topic it has been clear that there is no support for holding mean-of-3 in the USA except for Mike Hugheys "missing averages" competition. It was actually mentioned in one of the very first posts that it would be extremely hard to fit it into big US competitions. This "no-3-in-USA" has been repeated over and over but there hasn't been a response.

I think that's because Americans understand that it doesn't really affect us. When we know we won't be doing three solves in our competitions (except maybe Nationals), so maybe there isn't much more to say.

 

[cubizh]

Also, I stick with my point that the thread went dead and then came alive again by Sebastien. How can so many people deny that when it is so obvious just by looking back:
29th of April 2013, after about 40 posts in 1 week the thread dies; http://www.speedsolving.com/forum/s...n-of-N-for-FMC&p=848173&viewfull=1#post848173
12th of August 2013, after Sebastien didn't become World Champion he revives the thread; http://www.speedsolving.com/forum/s...n-of-N-for-FMC&p=887963&viewfull=1#post887963

Just a small note to remind that this thread was posted on the private forum for the majority of its existence, so not everyone was actually allowed to reply to it.