y—h—-all–w
o-this-for–t-o
r-has—-each-r
keep——–and
-d-p—–list–
given-u—n—-
-t-news-bits—
job-week–some-
-r—i—–f–g
—c-g–want–a
–math–o–w–v
–another-maybe
–j-l-e-d–r-e-
-good-r-same-s-
–r—e——t-

Missing words: |better| |challenge| |crossword| |helpful| |intern| |know| |knows| |multiply| |newspaper| |position| |possibly| |puzzle| |would| |you| |your|

I knew in this other kind of ranking I would have gained some positions… :-)

see you on the next contest

Highest No. of 1st places:
1= Alex P. (18)
1= Alfonso Nieto-Castanon (18)
3 Abhisek Ukil (12)

Highest No. of 1st places in daylight;
1 Alex P. (18)
2 Abhisek Ukil (12)
3 Sergey Y. (11)

I have also got the ranking as suggested by Francesco, but it would possibly be a bit long for this comment… 97 unique participants. The top ten are:
Alfonso Nieto-Castanon (Place of the best submission: 1)
Nicholas Howe (6)
Magnus S (8)
Abhisek Ukil (9)
Dimitris Kaliakmanis (10)
Yitzchak (12)
Fel (13)
Michael C. (14)
Alan Chalker (17)
the cyclist (20)

The only Francesco in that ranking comes in at 56th place with the best submission on place 1074.

Does anybody else want to know her/his place in the player ranking?

I agree with Magnus and his suggestions. Another one that comes to mind is the ‘highest no. of 1st place’ prize. That would be a carrot many tweakers will seek and can be easily implemented using the All the Leaders stats.

Also like there is a ‘invite your friend’ page for Matlab web seminars, it would be nice to have one for the contest as well.

One thing to Matlab Contest GREAT Team:

we have the submissions classification: 1 rank for each submission,

is it possible to have also a player classification?

i.e. 1 rank = 1 player (only the best submission of this player)

Looking at the statistics page, the number of submissions is quite low during daylight (except for the last 15 minutes, of course). Maybe another prize for ‘Most hours lead’ or ‘Most hours lead during daylight’ would be a possibility to encourage more activity in daylight.

Alfonso would have won ‘Most hours lead’ as well this time with 29:16 (hh:mm) ahead of Alex P. (26:41) and Nicholas Howe (20:06).
Sergey Y would have won ‘Most hours lead during daylight’ with 17:18 with Michael Völker second at 15:07 and myself third at 12:36.

For the sake of posterity, I have been able to figure out the scoring formula and am posting it here as I traditionally do. As usual, it’s very similar to the recent contests:

score = k1*result + k2*e(k3*runtime) + k4*max(complexity-10,0) + k5*nodes

Where:

k1 = 0.01
k2 = 0.002
k3 = 4/30 (0.1333333)
k4 = 10
k5 = 0.01

Good luck to everyone in these final minutes of the contest!

Only moves can be used to make this vine have the best score possible. Longest possible vine will hit the bottom left corner of the board.

When I was newer to the contest it was always a thrill to see my name listed as a runner up even if I didn’t win a prize. So let me congratulate Alfonso, Robert Macrae, Kurt Janssens and Martin F. for being in the hunt and giving me something to shoot for! The contest is more fun the more people enter, and it looks as though one of those names is a new one.

For those looking to jump into the contest at this point, I plan to post to the newsgroup a summary of the techniques used in my Twilight code. I probably won’t get to it until this evening, though — gotta work on the Early Bird prize. ;-)

Here’s some suggestions for mini-contests:

-Best solver that always starts the vine at square 1
-Best solver that follows in the tradition of a ‘normal’ 15 square in that there ends up being only 1 empty square
-Best solver that always uses the full limit of moves
-Solver that has the most creative ‘artistic’ vine placement on select boards in the testsuite
-Solvers that when viewed in the MATLAB editor form ASCII art, yet are still able to run
-Most sophisticated / complex board that the leading solver is able to completely fill with a vine

Helen

As I indicated in the newsgroup thread, since I misunderstood how the moves were being scored, I decided to update the visualize abilities of the runcontest code to better illustrate what’s going on. I’ve uploaded it to the File Exchange since I thought it might be helpful to others: http://www.mathworks.com/matlabcentral/fileexchange/33600

The additions I made are:
-Add board # to the tile so you know which board you are looking at

-Color code the vine. Green parts are ‘good’ and scored, red parts are ‘bad’ and truncated by the grade code. An arrow shows the ‘start direction’ of the vine, while and X shows the end of it.

-Color code the moves. Blue moves are ‘good’ and scored, yellow moves are ‘bad’ and truncated by the grade code. A circle indicates the ‘source’ of the move.

Hopefully you all will find this helpful too.

The queue seems to be choking on submissions with non-english comments. I’ve removed the posts that were stuck and things seem to be processing correctly now.

Again, my apologies for the inconvenience.

Pity that none of the winning solutions really look like crosswords :) if the penalty was higher (or Inf), the whole competition would have a completely different nature.

I just found out that there was a no-bogus sub-competition :) If I knew that I would participate in that one, I got some pretty neat crosswords

http://img717.imageshack.us/i/resa.png/

Thanks again!

I really enjoyed seeing the code constantly improve with everyone’s contribution.
I’m sure the result is better than if everyone had worked separately on their own code, so this is a nice example of collaborative work :)

I had no strategy and just made any change I saw could improve the code; I don’t have the skills of those who made the winning solvers anyway. I’m just glad to see in the (so far) winning codes some ideas of mine remaining.
Oh and I think I hold a record for holding the first position for the longest with ‘Skin’s crawling’ :)

Thanks again to all!