Proposal: allow a move limit for FMC

[Erik]

It'd be less time consuming for a delegate/organiser to enter an FMC solution into a system and have it automatically checked.

I think you are wrong here. How long do you take to type over a solution into the computer? It'd be max the same speed I estimate.

Would the time/effort to create a system that analyses the solution, beyond simply checking whether it solves the puzzle, outweigh the combined time/effort of people manually checking solutions all over the world?

I don't think so and even if it would, how would you program it?

What is wrong with the current way? Is 10-15 minutes of checking really that much of a burden for you?

Preferably not. I don't even see why this is under discussion as it isn't the point of this thread. If I was a delegate in Europe with lots of good FMCers I would ask some of them to assist with the judging process. The reason we didn't get help at this competition is because we were flat out, and concerned about various factors which we hadn't dealt with before.

Why not? If you didn't have time at the competition, why not ask a few fast people to do it right after it? And regardless of this, why don't you try the approach mentioned before out in the future, but instead keep arguing in favor of your approach?

You might call it name calling and rudeness. I might call it honestly and bluntness. I guess it depends on the sort of person you are how you see it. If you feel this makes me an unworthy delegate then perhaps you should talk to the Board. If they feel I should step aside then I'll respect their decision. Regardless, this is not the appropriate place to have such a discussion so it ends here now. If you wish to continue this topic I suggest you choose a different medium.

Are we talking in different languages? Or is calling someone a moron not namecalling? Defending your behavior is maybe even worse than the behavior itself. If you do continue this behavior I will contact the board, yes. Let's stop discussing this here now.

 

[Sebastien]

I think you are wrong here. How long do you take to type over a solution into the computer? It'd be max the same speed I estimate.

The time needed to type in a solution is not even the main problem I see. The real problem is, that by typing in a solution in somewhere, i.e. create a copy of that solution, you are just reducing the problem statement from "is this solution correct?" to "did I transfer the solution correctly?". So after performing that copying process that already requires at least the same resources than directly checking the solution, you will see yourself confronted with another maybe even more time consuming check. With this view, I don't think that any approach that includes transfering an FMC solution in order to check it is very senseful.

 

[Erik]

The time needed to type in a solution is not even the main problem I see. The real problem is, that by typing in a solution in somewhere, i.e. create a copy of that solution, you are just reducing the problem statement from "is this solution correct?" to "did I transfer the solution correctly?". So after performing that copying process that already requires at least the same resources than directly checking the solution, you will see yourself confronted with another maybe even more time consuming check. With this view, I don't think that any approach that includes transfering an FMC solution in order to check it is very senseful.

Oh yes, that as well! Very good point. So the only real alternative would be direct digital input, which we have seen earlier is not a practical or safe option either.

 

[kinch2002]

I don't see any intrinsic problem with allowing 'noobs' to compete, so am I right in thinking that the only problem we have here is that the checking process takes a while?
Sebastien and I have already explained how it doesn't take long for us, but I will elaborate on how I check.

1. I let other experienced competitors help if we want to get it done quickly.
2. Solved I only check once...what's the likelihood that I manage to do a wrong move (1 in 1000 moves?) which happens to be the same move that they wrote wrong (1 in 70)? And that assumes they wrote one move wrong (1 in 10?). It just isn't worth double checking everything for that sort of miniscule probability.
3. DNFs can be checked twice. I check it once and then let the competitor execute it while reading off the paper. They either say "cba whatever DNF is fine" or they try it and realise their mistake because they're usually doing CFOP and it's really obvious when it goes wrong (most people who write moves wrong are 'noobs' and use CFOP). This process is good because then the competitor is happy to accept the DNF, and I only have to check once myself - in fact I can even check another one while I keep an eye on them in my peripheral vision.

All in all, a few extra solutions takes very little time, and doesn't even hold up the competition (in the same way that no cutoffs in other events can) because you can do the checking at any point.

I feel this extra effort is massively outweighed by the positives of allowing everyone to compete (with the chance of getting a result).

 

[Stefan]

My issue is with the fact that half of the people doing FMC (at ausnats) hadn't ever done it before. I got questions like "can you do rotations?" or, after the fact, comments like "I didn't realise you can do rotations!"

Maybe it would help if the regulations included a link to an example FMC sheet? I think that would be a good idea anyway.

Or if we changed the FMC sheet, like print the scramble on the back side and give competitors a few minutes to read the instructions on the front side and enter their name and id there and to "get ready" in general, before the actual hour and attempt starts? The back side could also have a field for a replacement solution, so one could write a "safety solution" on the front early on and replace it with a better one later. Not sure that's useful, just an idea.

Btw: Can you tell the lengths of the submissions at your 2014 Nats that were DNF? That would tell a more complete story than what's visible now.

 

[Stefan]

The time needed to type in a solution is not even the main problem I see. The real problem is, that by typing in a solution in somewhere, i.e. create a copy of that solution, you are just reducing the problem statement from "is this solution correct?" to "did I transfer the solution correctly?". So after performing that copying process that already requires at least the same resources than directly checking the solution, you will see yourself confronted with another maybe even more time consuming check. With this view, I don't think that any approach that includes transfering an FMC solution in order to check it is very senseful.

I strongly disagree. With the computer...

• You can check and fix moves one by one. Without worrying that you might repeat your error or replace it with another one. The program could even show the entered moves overlayed onto a photo of the sheet, making comparison very comfortable and fast.
• You don't need to scramble.
• If something is unclear, like the "L vs U" mentioned before, you don't need to scramble and solve twice. You need to enter just once (telling the computer that that move could be L or U). If there are let's say four moves that can be read two different ways, you don't need to scramble and solve sixteen times and keep track of what you tried, you still only need to enter the solution once.
• As I said before, the computer can find errors and point out where they are, so that the judge only has to check that move or those few moves!

 

[TimMc]

I think you are wrong here. How long do you take to type over a solution into the computer? It'd be max the same speed I estimate.

2-7 keystrokes per second.

Someone could type x2rflifir2fl2ul2rb2ru2biuiruirururyuiru2riufiuifu2 and it can be converted:
x2 R F L' F' R2 F L2 U L2
R B2 R U2 B' U' R U' R U
R U R y U' R U2 R' U F'
U' F U2

...like entering "12345" into Cubecomps for 1:23.45

What is wrong with the current way? Is 10-15 minutes of checking really that much of a burden for you?

The scramble could be loaded for the attempt, and the solution could be applied to see whether it solves the cube. If the solution works, then further checks could be run to enforce: E2) The competitor's solution must not be directly derived from any part of the scrambling algorithm. This would make checking more consistent and remove some interpretations of what is and what isn't considered "directly derived".

10-15 minutes is fine. I'm happy volunteering 30+ hours on a weekend for those who really want to compete (i.e. improve their PB).

15-30 minutes will be allocated in the schedule for checking FMC solutions next time.

With this view, I don't think that any approach that includes transfering an FMC solution in order to check it is very senseful.

It'd give us a complete record of the results. It seems like we're manually transferring the solution to a puzzle and not keeping a record of what was submitted beyond "35 moves".

Tim.

 

[TimMc]

Can you tell the lengths of the submissions at your 2014 Nats that were DNF?

43 moves + 2 rotations
47 moves
52 moves
52 moves + 1 rotation
54 moves + 5 rotations
57 moves + 8 rotations
65 moves + 3 rotations
65 moves + 2 rotations
66 moves

Maybe it would help if the regulations included a link to an example FMC sheet?

Great idea.

Tim.

 

[Stefan]

It'd give us a complete record of the results.

Right, that's another nice advantage. Also, the program could count the moves. And analyze the solution, like find out whether the "inverse-scramble" technique was used or whether it might have abused the scramble. And it could display the solution like alg.cubing.net does (allowing to go through it slowly without losing track, going forwards and backwards, and freely rotating the cube without losing track).

 

[Stefan]

2. Solved I only check once...what's the likelihood that I manage to do a wrong move (1 in 1000 moves?) which happens to be the same move that they wrote wrong (1 in 70)? And that assumes they wrote one move wrong (1 in 10?). It just isn't worth double checking everything for that sort of miniscule probability.

You're ignoring muscle-memory and commonality of algorithms. I remember checking my solution more than once and subconsciously fixing an error every time, because I "knew" the alg and did what I knew, not what was written. If I have the time, I also check my FMC solution several times before I hand it in, even if it appears correct every single time, because I don't trust myself. And I beliieve I'm not the only one.

 

[Renslay]

You're ignoring muscle-memory and commonality of algorithms. I remember checking my solution more than once and subconsciously fixing an error every time, because I "knew" the alg and did what I knew, not what was written. If I have the time, I also check my FMC solution several times before I hand it in, even if it appears correct every single time, because I don't trust myself. And I beliieve I'm not the only one.

Checking without fingertricks (doing only "wrist turns") can help forgetting and avoiding muscle-memorized moves.

 

[kinch2002]

You're ignoring muscle-memory and commonality of algorithms. I remember checking my solution more than once and subconsciously fixing an error every time, because I "knew" the alg and did what I knew, not what was written. If I have the time, I also check my FMC solution several times before I hand it in, even if it appears correct every single time, because I don't trust myself. And I beliieve I'm not the only one.

True. Let's say a solve ends with a T perm that has a move written wrong. I might automatically do the T perm moves. There's no point me checking it again as I'll probably do the same thing. Anyone else might also do the same thing. So even handing every 'correct' solution to someone else to check won't be foolproof. Handy that we're discussing automated checkers already

I think I rarely notice algorithms that I actually know when I'm checking solutions. Most UK cubers have been told enough times that no rotations is better for them, so even CFOP-esque solutions will be upside down etc. If I do notice an alg I think I already read more carefully to try and avoid the mistake you mention. Not foolproof, but this problem is getting off-topic from the original post anyway.

At risk of straying more off topic and riling a few people, we could even move onto whether writing down a move wrong should be considered a mistake
After all, the essence of FM is to find a short solution, not to write one down

 

[Laura O]

• If something is unclear, like the "L vs U" mentioned before, you don't need to scramble and solve twice. You need to enter just once (telling the computer that that move could be L or U). If there are let's say four moves that can be read two different ways, you don't need to scramble and solve sixteen times and keep track of what you tried, you still only need to enter the solution once.

I don't think that's something we need or that we should aim for. It certainly saves time but it's also some kind of trial-and-error-checking.
E2c says that you have to "[... ]give the judge a legibly written solution[...]". So if moves can be read two different ways that's a DNF.

In case of doubt you can still ask the competitor about specific moves. If he isn't able to read his moves, that's clearly a DNF.

 

[Mollerz]

I've been tracking this thread since its inception and here's my input, most if not all of which has been said before but I believe its not really getting through.

Firstly, the process myself and Daniel do at competitions when it comes to checking. Once all of the results are in we will move onto the next event. Usually the organiser will call out the event and groups and such so myself and Daniel are not needed. Usually, if not all of the time, myself and Daniel are in different groups for events, and as much as possible, we will avoid scrambling/judging so we can actually delegate the competition. Obviously in some cases we will be needed, but most of the time we are not. Let's say Daniel is in the first group, and I am in the second group. If its 3x3 then it will be 20-30 minutes for each group. This gives me enough time to check all the solutions through. If the solution is correct, you don't really need to double check it, but I always do, to see what kind of method is used, what blocks they build, insertions, etc. If I find one that does not work, I will check it a second time, if it again does not work I will put it to another side. When Daniel finishes competing I tell him which scrambles are correct and which I could not get to work. Whilst I am competing, he will check to see if they are in fact incorrect or if I have made a mistake, and usually also check the correct solutions, first to make sure they are correct and also to see how the solution was made out of interest like myself.

When I check for solutions, first I just do the solution. I don't look for anything, no triggers, blocks, I barely look at the cube. The chances that if it is solved and the solution is written down incorrectly is so minute that usually it is good enough to see whether the solution is correct. After this I will write down the amount of moves. I will then check it again but this time looking at the solution and the method used, to see if there is anything interesting to do with the solve or something like that. If I then spot a mistake I may have made I'll leave it for Daniel to check, but usually I have no problem with this. If the solution does not produce a solved cube, I will do it again "blind" and see if it produces the same state as the first time I tried the solution. If it is solved then I obviously made a mistake the first time around, if it is the same state then the solution is very likely a DNF, if it is a different state I'm doing something idiotic and leave it for Daniel. I will then go through the solution step by step looking for everything. Usually there will be a U' written down as U or something similar, and I know this will be the cause of the DNF, and let Daniel know so he can try the DNF solutions as well.

If there are a lot of rotations in the solution, when counting moves I will just circle all of the rotations, and count them up separately. The ones that use rotations usually do CFOP so I will write at the end something like "63 - 8 = 55", showing that there were 63 moves including rotations, 8 rotations, meaning a solution of 55, which I will write in the box at the top.

Within the time of the next event, all solutions will have been checked thoroughly and to no dissaray of the competition proceedings. If you were slow to check the solutions or there were a lot of competitors for FMC, this could go into the next 2 or 3 events. If you are unable to check all solutions as 2 delegates before the end of the competition day then I recommend you seek help from trusted and experienced competitors, of which I know there are plenty in Australia.

Now, returning to the original reason. We've had a competitor in the UK try really hard at getting a good solution for FMC, and has struggled to get under 50, and at the last competition said person got a sub-50 and were really happy with the result. Just because the result has a lot of moves compared to world class FMC results, does not mean they are not trying their hardest.

What I gather from this is that if you think that someone who is getting 40+ solutions is a waste of time (obviously I am just using your example, as you said many times), then you should make qualification for 3x3 20 seconds, since anyone slower that should not be competing, as they are wasting your time too.

 

[Dene]

What I gather from this is that if you think that someone who is getting 40+ solutions is a waste of time (obviously I am just using your example, as you said many times), then you should make qualification for 3x3 20 seconds, since anyone slower that should not be competing, as they are wasting your time too.

First part of the sentence is not true, and baseless. Second part of the sentence is a strawman. Perhaps at the next competition I should have no time limits. We'll do 3, 4, 5, 6, and 7. Let's see how much we get through in 8 hours....


But thanks for detailing how you check solutions. At the very least, finally someone has acknowledged that we do make mistakes and solutions need to be checked over properly. I was starting to think everyone was judging FMC by checking once or twice quickly and going with it.

Regardless this isn't my point, and never has been. I don't care about the time taken. If you read near the end of my last post you'll see what my reasoning is.

 

[Mollerz]

First part of the sentence is not true, and baseless. Second part of the sentence is a strawman. Perhaps at the next competition I should have no time limits. We'll do 3, 4, 5, 6, and 7. Let's see how much we get through in 8 hours....


But thanks for detailing how you check solutions. At the very least, finally someone has acknowledged that we do make mistakes and solutions need to be checked over properly. I was starting to think everyone was judging FMC by checking once or twice quickly and going with it.

Regardless this isn't my point, and never has been. I don't care about the time taken. If you read near the end of my last post you'll see what my reasoning is.

But isn't the whole point and spirit of the WCA to encourage and support people, no matter how bad they may be, to compete and try their hardest? At UK competitions, someone is like, "Oh I don't know whether I should compete in this, I barely learnt to solve it this morning" we would still encourage them to compete. I know at least at my very first competition, I had been practicing 4BLD at home, and never got a success, but I still went into competition and attempted it. Before I competed in it I had never had a success, does that mean I should not be allowed to compete in it because I am incapable of doing it? Certainly not.

It is the same with these people competing in FMC for the first time. Obviously a lot of them want a result, especially if you only hold it ONCE A YEAR at AusNats. If you held it at every competition there would be one or maybe a couple of want their result. I guarantee those who are "Result getters" will come back next year and either A) Not compete because they have their result, or B) Will actively try and beat their previous best. People want results, that's why they compete. Would you deny a first time 3x3 solver their chance to compete even though they take 5 minutes to solve it? No, you would not (I would hope). Loads of people, the first time they compete in an event, compete to get a result.

This is something that I noticed recently and I think it has some value in this discussion. Have you ever noticed that in US Competitions, they usually have like, 2, 3, maybe 4x4, and a couple of side events, the competitions last for half a day and then its over? And have you also noticed how a LOT of the competitors, only compete in say 3x3 and another event, but damn they are fast. It's not the norm to be good at a lot of events in the US, obviously there are outliers, but for the most part this holds true. Think about how often you hear about young US kids who got really fast at 3x3 or 2x2 or something, right? It's quite common. Now if you go over to European competitions, usually we have long 2-day competitions and most, if not all, events are held. If you look at the competitors from these countries, lots of the competitors compete in lots of the events. Now again, obviously there are outliers, those that only compete in 3x3 or a couple of events, but for the most part, people compete in lots of events. A more specific example, we usually hold all BLD events in the UK at every competition, we have WR podiums for things like 5BLD, because we hold the events all the time, so people have incentive to practice. I am fully aware of the argument that, it's only held once a year so you should practice a lot for it so you get a good result. However I don't think this is very true, you could practice a lot just to mess up at competition, now you do not have a chance for another year, it's quite demotivating. But if you know you'll get a chance at every competition you attend, it's much more relaxing since you know you will have more opportunities.

I'm sure there are Magic: The Gathering players who buy a deck, have no idea how to play and decide to go to a Magic tournament, they get put in the first round against some person who's played once or twice before, get destroyed and then leave. Are they making a mockery of the magic tournament? Are they wasting peoples time? No, they just want to have some fun and get a result, even if that result is 0W 1L. Someone would have played that match anyway, might as well be them.

I really do think it just boils down to you being lazy. What are you doing all day at competition if you do not have 15-45 minutes spread over the course of a few hours to find the solutions? I'll admit, myself and Daniel are fast at scrambling and solving so we would obviously be able to check results faster, but we also compete in every event and always get it done reasonably fast. As I stated before, if you are really struggling with checking solutions, pick a couple of trusted people to help you. You chose to hold FMC, you should be fully aware of what comes with it.

 

[TimMc]

What are you doing all day at competition if you do not have 15-45 minutes spread over the course of a few hours to find the solutions?

Competing, judging, scrambling, checking results to generate scorecards for the next rounds and so that competitors know that they've made it through for the next day, resolving incidents, announcing rounds, tracking down missing competitors etc.

There's plenty to do with a tight schedule. It's easier to focus on other things that need to get done so that the schedule works.

If there were two rounds of FMC then it'd become more of a priority to allocate time to check solutions straight away.

It's be awesome if people spent 10min on CFOP solutions, and 5min checking what they wrote down before submitting it. Then all the CFOP solutions could be checked before the end of the round.

Tim.

 

[Stefan]

In case of doubt you can still ask the competitor about specific moves.

That takes extra time for finding them, getting them there, and doing it. Hitting two or three keys is faster. And what if they're not even there (went to toilet, smoke, restaurant, home, whatever)?

 

[Laura O]

That takes extra time for finding them, getting them there, and doing it. Hitting two or three keys is faster. And what if they're not even there (went to toilet, smoke, restaurant, home, whatever)?

And faster is better in this case?
As said before, hitting those two or three keys and making a program decide about the correct move is a trial-and-error method for me. If people can't write their solution down legibly that's a DNF.
Asking people about single moves is just the last option for me if the usual procedure (quite the same as James described) doesn't work out. I remember doing this only once when someone wrote down cryptic B and R moves and he immediately agreed that this is a DNF without any discussion.

Nevertheless illegible moves or solutions are more or less unusual in my experience. For example I helped to check the solutions at the European Championships and I think there was only one sheet (of about 30 or even 40) where I couldn't identify one move. I asked someone who was checking with me and he was sure that this was a L and not a U -problem solved without double-checking.

 

[Dene]

Before I competed in it I had never had a success, does that mean I should not be allowed to compete in it because I am incapable of doing it?

Your sentence is ambiguous. If you are incapable of solving 4bld, then obviously you shouldn't attempt it (you should be disqualified as per A1c). If you are "incapable", as in haven't done it before, but know enough to get it right eventually, then it depends on circumstances whether you should be allowed to attempt it or not.

I really do think it just boils down to you being lazy.

I think this is quite possibly the most unfair thing anyone has said to me in a very long time. I would expect such comments from all these American kids that feel like it's their right, and not a privilege, to be able to go to competitions. But from you? That's disappointing...



btw I didn't respond to the rest because I didn't feel there was anything especially worth mentioning. If you feel strongly about me responding to a particular point, please ask.




Anyway, I think this thread is mostly exhausted. I already submitted this proposal to the delegates and WRC so it's up to them to consider the merits and flaws now.