2009: FMC

TimS (2008-11-12 02:52:23 +0000)
[quote:9mrgldge]E2e) The solution of the competitor must not be in any way related to the scrambling algorithm. Penalty: disqualification of the solve.[/quote:9mrgldge] I feel that this regulation is not precise enough. I heard from Ville that at the Tampere Open, there was a very easy 3-move 2x2x2 start that could only be achieved by going one move back into the scramble (i.e. "...F L D" D' R B); the competitors were not allowed to do this, while I feel that it was perfectly acceptable to do it. Furthermore, the scrambling procedure varies over competitions. At US Nationals 2008, we received the long scrambles that include stuff like (U R D L')*4, etc. At Tampere, I heard they used the two-phase scrambles. I think the scrambles should be the same style across all competitions. Preferably, we use the longer ones such that the conundrum above wouldn't be a problem. -Tim
Shelley (2008-11-12 17:09:26 +0000)
We've been using groups of repeated moves to make scrambles longer and keep competitors from using the scramble in their solution. I agree, something needs to be standardized about the way FMC scrambles are done.
Pedro_S (2008-11-12 20:57:34 +0000)
but it is standardized (at least in my point of view) this is what Article E says about FM: E1) Regulations are as described in Article A (Speed Solving). this is what said in article A about scrambling: A2b) A scrambler scrambles the puzzle according to the regulations in Article 4. and this is what article 4 says: 4f) The number of moves to scramble a puzzle must be: (...) Rubik's Cube - Random position - Cube Explorer (version 4.30 or higher), by Herbert Kociemba so, to me, it's pretty clear that 3x3 for FM should be scrambled the same way as 3x3 for speedsolving, which is done with cube explorer
Tyson (2008-11-18 17:46:36 +0000)
But Cube Explorer was a new thing. Before, when the scramble was just 25 random moves, this was certainly inadequate for FMC competition. That's why we were using the 40 move scramble with repeated sequences so that there would be less probability of having a case where a 2x2x2 block was easy to see, but happened to reverse the last moves of the scramble. This problem does seem to be mitigated with Cube Explorer, though... what do people think? Because many turns in cube explorers, and the system that it solves the cube (the double turns and two-phase solving) do seem to help.
Ron (2008-12-21 21:31:03 +0000)
A simple solution may be to execute a random scramble twice. That would give a 40 move random position. Anyone else? Ron
Lucas (2008-12-22 08:30:49 +0000)
[quote="Ron":2ydc0g87]A simple solution may be to execute a random scramble twice. That would give a 40 move random position. Anyone else? Ron[/quote:2ydc0g87] I like this, because it allows the competitor to get used to the scramble more easily. However, your suggestions doesn't even generate half the positions odd corner/edge parity). Note that U D, U' D, U D', and U' D' all go to U2 D2, which is really bad. I think (unsubstantiated intuition) that something like () B () B (), where B is a random scramble, should work fine. () would be a few more random moves each; we'd have to a bit of thinking to make sure what is the minimum necessary to make sure all cases stay equally likely. Stefan, if you find this, what do you think? Got any simple suggestion with a proof of randomness, that essentially keeps a double-same scramble? Otherwise, two concatenated random scrambles are slightly less convenient for the competitors, but should work fine.