Why are "contacts" more important than novel structure?
I finally got on the board with a mediocre score and improved it as much as I could by rebuilding and moving around the sections of the protein.
Frustrated because I hit a dead end early, I randomly banded together unrelated sheets, and wiggled, hard. The exploration score shot up before the other score dipped below 1960.
When it was all done, I had mangled one of the sheets: bent it so badly that it no longer bonded completely with its next-door neighbors, and was hanging part-way out into oblivion... but closer to the unrelated sheets to which it was temporarily banded. And I now have 3rd place on this puzzle, with a hideously ugly protein.
Scoring fail? Or have I discovered the cure for AIDS?
Something else that I noticed yesterday was that someone appears to have changed the LUA get_score() to return the exploration score rather than the normal energy score, thus ensuring that scripts don't work properly. Team AD released a couple of scripts that do something like summing the segment scores to patch them.
If contacts are that important, why not lock off the secondary structure, and score only on beta sheet contacts, otherwise the result is going to be the 'my morning hairstyle' type structure.
Perhaps the way to do it is to turn the peptide into a giant helix. That might be what they're looking for :)
In other puzzles, merely trying to work it earns you a point for participation. Not so, here. The result is skewed awarding of global points. One point for 23rd place?? Was this intentional?
That is a bitter pill. Wasting my weekend for a weird puzzle and earning only 3 points leaves me with the conclusion that I should leave these puzzles alone and leave a script running on the other ones. I have better things to do with my spare time.
I'm posting a news item about this issue
This is unfortunately correct since the denominator used in the score calculation is the total number of players on the scoreboard:
22nd place out of 46: (24/45)^7 * 100 = 1.23 ==> 2 points
23rd place out of 46: (23/45)^7 * 100 = 0.91 ==> 1 point
That rapidly decaying scoring algorithm is pretty harsh when there are only a few participants.
And that's another measure of how messed up this was, less than 50 people able to get on the board.
I quite liked the puzzle in general, such a nice change from the QTTN where I have to wonder what the point is.
But yeah the implimentation could use a little ironing out.