The New Sudoku Players' Forum

by **m_b_metcalf** » Mon Sep 21, 2009 3:27 pm

Code: Select all: . 1 . . . . . 2 . . 3 . 4 . 5 . 1 . 4 . 6 . . . 5 . 7 . . . 5 . 3 . . . . . 8 . . . 1 . . . . . 1 . 8 . . . 2 . 9 . . . 8 . 3 . 8 . 3 . 4 . 7 . . 6 . . . . . 5 . ED=3.0/3.0/2.8

Regards,

Mike Metcalf

by **gsf** » Mon Sep 21, 2009 4:08 pm

Code: Select all: . 6 . . . . . 3 . . 4 . 5 . 7 . 8 . 7 . 8 . . . 2 . 6 . . . 4 . 5 . . . . . 9 . . . 4 . . . . . 2 . 8 . . . 1 . 6 . . . 3 . 8 . 2 . 1 . 4 . 9 . . 3 . . . . . 2 . ED=4.7/1.2/1.2

by **JPF** » Mon Sep 21, 2009 5:10 pm

Code: Select all: . 1 . | . . . | . 2 . . 3 . | 4 . 5 | . 1 . 4 . 5 | . . . | 3 . 6 -------+-------+------- . . . | 7 . 6 | . . . . . 8 | . . . | 2 . . . . . | 8 . 3 | . . . -------+-------+------- 1 . 2 | . . . | 7 . 5 . 6 . | 5 . 8 | . 4 . . 8 . | . . . | . 6 . ED= 5.2/1.2/1.2

JPF

by **gsf** » Mon Sep 21, 2009 5:11 pm

Code: Select all: . 6 . . . . . 4 . . 4 . 5 . 7 . 8 . 3 . 5 . . . 7 . 2 . . . 7 . 8 . . . . . 2 . . . 4 . . . . . 9 . 3 . . . 1 . 3 . . . 5 . 9 . 7 . 1 . 9 . 3 . . 2 . . . . . 1 . ED=4.8/1.2/1.2

by **JPF** » Mon Sep 21, 2009 5:30 pm

Code: Select all: . 1 . | . . . | . 2 . . 3 . | 4 . 5 | . 6 . 7 . 6 | . . . | 4 . 8 -------+-------+------- . . . | 6 . 1 | . . . . . 9 | . . . | 2 . . . . . | 8 . 3 | . . . -------+-------+------- 5 . 1 | . . . | 6 . 3 . 4 . | 3 . 9 | . 5 . . 7 . | . . . | . 1 . ED=6.0/1.2/1.2

JPF

by **gsf** » Mon Sep 21, 2009 5:56 pm

Code: Select all: . 2 . . . . . 6 . . 6 . 9 . 5 . 3 . 7 . 8 . . . 9 . 1 . . . 8 . 2 . . . . . 9 . . . 4 . . . . . 7 . 3 . . . 2 . 6 . . . 8 . 5 . 1 . 2 . 8 . 7 . . 9 . . . . . 1 . ED=5.9/1.2/1.2

by **g.r.emlin** » Mon Sep 21, 2009 6:00 pm

::: game 0090 status LIGHTNING 2009-09-21+18:00:31-0000 :::

Code: Select all: game 0090 2009-09-16+23:00:00-0000 LIGHTNING duration 4d18h dealer g.r.emlin pages 297-304 (open) 3 players ________________________________________________________________________________________________________________ 1.2/1.2/1.2 gsf 4.2/4.2/2.8 gsf 6.2/2.3/2.3 gsf 7.8/7.2/2.6 m_b_metcalf 1.5/1.5/1.5 gsf 4.4/4.4/2.8 JPF 6.5/2.0/2.0 gsf 7.9/7.2/2.6 gsf 1.7/1.5/1.5 JPF 4.5/4.5/2.6 gsf 6.6/6.6/3.0 m_b_metcalf 8.0/2.6/2.6 m_b_metcalf 2.0/2.0/2.0 JPF 4.6/3.2/2.6 gsf 6.7/6.7/2.6 gsf 8.2/8.2/2.6 gsf 2.3/2.3/2.3 gsf 4.7/1.2/1.2 gsf 6.8/6.8/2.6 m_b_metcalf 8.3/8.3/3.0 m_b_metcalf 2.5/2.5/2.5 gsf 4.8/1.2/1.2 gsf 6.9/6.9/2.6 JPF 8.4/8.4/3.4 m_b_metcalf 2.6/2.6/2.6 m_b_metcalf 5.0/5.0/3.0 gsf 7.0/7.0/2.6 gsf 8.5/8.5/2.6 gsf 2.8/2.8/2.8 gsf 5.2/1.2/1.2 JPF 7.1/7.1/7.1 m_b_metcalf 8.6/8.6/2.6 m_b_metcalf 3.0/3.0/2.8 m_b_metcalf 5.4/5.4/2.6 gsf 7.2/7.2/3.4 gsf 8.7/8.7/2.6 m_b_metcalf 3.2/3.2/3.2 gsf 5.6/4.4/2.6 m_b_metcalf 7.3/7.3/3.0 m_b_metcalf 8.8/8.8/2.6 gsf 3.4/3.4/2.8 gsf 5.7/4.4/2.6 gsf 7.4/7.4/2.6 gsf 8.9/8.9/2.6 m_b_metcalf 3.6/3.2/2.6 gsf 5.8/1.5/1.5 gsf 7.5/7.5/2.6 gsf 9.0/9.0/2.6 m_b_metcalf 3.8/2.3/2.3 gsf 5.9/1.2/1.2 gsf 7.6/7.6/2.6 gsf 9.1/9.1/2.6 gsf 4.0/4.0/2.6 m_b_metcalf 6.0/1.2/1.2 JPF 7.7/7.7/2.6 gsf 9.2/9.2/2.6 gsf ________________________________________________________________________________________________________________ open ratings : 4.9 5.1 6.1 8.1 9.3-11.4 submitter entries score response notes ----------- ------- ----- -------- ----- JPF 6 866 4h20m rare-1,pearl m_b_metcalf 16 4575 1h10m rare-0*2,rare-1*2,rare-2*2,diamond*2,pearl*7 gsf 34 5793 48m40s rare-0*3,rare-1*2,rare-2,diamond*3,pearl*15,highest-rating=269,last-entry=522 ----------- ------- ----- ----- 3 56 11234 rare-0*5,rare-1*5,rare-2*3,diamond*5,pearl*23

by **m_b_metcalf** » Mon Sep 21, 2009 6:14 pm

Code: Select all: . 1 . . . . . 2 . . 3 . 4 . 2 . 5 . 2 . 6 . . . 3 . 7 . . . 3 . 7 . . . . . 5 . . . 8 . . . . . 8 . 1 . . . 9 . 7 . . . 4 . 3 . 8 . 9 . 4 . 1 . . 6 . . . . . 8 . ED=4.7/2.3/2.3

Regards,

Mike Metcalf

by **gsf** » Mon Sep 21, 2009 6:40 pm

Code: Select all: . 9 . . . . . 5 . . 4 . 5 . 7 . 8 . 7 . 5 . . . 6 . 9 . . . 7 . 6 . . . . . 1 . . . 3 . . . . . 3 . 4 . . . 4 . 9 . . . 5 . 7 . 8 . 4 . 5 . 3 . . 3 . . . . . 6 . ED=8.1/1.2/1.2

by **JPF** » Tue Sep 22, 2009 5:35 pm

It's probably the right timing to revisit these tricky bonus values.

Some preliminary thoughts:
The bonuses for rare ratings like x/y/z are too complex and would need to know the list of all the rare ratings.
The real rare ratings are: 4.9; 8.1 and 10.7 +

Keep the rules as simple as possible.
Reduce the learning cost for a newcomer.

More time between two games: no need to rush, a bit of time to discuss what happened.

JPF

by **m_b_metcalf** » Tue Sep 22, 2009 6:21 pm

JPF wrote:It's probably the right timing to revisit these tricky bonus values.

Yes.

Some preliminary thoughts:
The bonuses for rare ratings like x/y/z are too complex and would need to know the list of all the rare ratings.

Yes, but I've written a little utility that tells me how often any given rating has been played. That made it easier to play ratings that were completely new, but with the passage of time there will be ever fewer. I think a brand new rating should get some bonus, less if it's trumped.

The real rare ratings are: 4.9; 8.1 and 10.7 +

Not quite: the first has been played fives times, the second 48!

Keep the rules as simple as possible.
Reduce the learning cost for a newcomer.

Absolutely. There's too steep a curve to climb now.

More time between two games: no need to rush, a bit of time to discuss what happened.

Quite.

Regards,

Mike Metcalf

by **gsf** » Tue Sep 22, 2009 6:42 pm

m_b_metcalf wrote:
JPF wrote:It's probably the right timing to revisit these tricky bonus values.
Yes.
Some preliminary thoughts:
The bonuses for rare ratings like x/y/z are too complex and would need to know the list of all the rare ratings.

Yes, but I've written a little utility that tells me how often any given rating has been played. That made it easier to play ratings that were completely new, but with the passage of time there will be ever fewer. I think a brand new rating should get some bonus, less if it's trumped.
The real rare ratings are: 4.9; 8.1 and 10.7 +
Not quite: the first has been played fives times, the second 48!
Keep the rules as simple as possible.
Reduce the learning cost for a newcomer.

Absolutely. There's too steep a curve to climb now.
More time between two games: no need to rush, a bit of time to discuss what happened.

Quite.

Regards,

Mike Metcalf

the referee config is listed in the first post after the rules
here are the settings related to these comments

Code: Select all: integer RARE_ED=2 # rare ER/EP/ED count (of all previous games) integer SCORE_HIGHEST=0 # highest ER/EP/ED rating adjustment integer SCORE_HIGHEST_PERCENT=10 # % of total at submission time for highest ER/EP/ED rating integer SCORE_LAST=0 # last entry adjustment integer SCORE_LAST_PERCENT=5 # % of total at submission time for last entry integer SCORE_RARE_ED=200 # rare ER/EP/ED rating bonus integer SCORE_TRUMPED_RARE_ED=50 # trumped rare ER/EP/ED rating bonus integer WAIT_START=16*hours # closed->start transition interval

the rare rating bonuses could be set to 0
that would reduce it to an announcement that a rare rating was hit
or
the rating histogram could be published on the results thread after each game (one ER/EP/ED count per line)
if we keep the rare bonuses, something like SCORE_RARE_ED=25 SCORE_TRUMPED_RARE_ED=10 might be more reasonable

WAIT_START could be increased by N*days: suggestions for N?

if the *_PERCENT scores are too complicated then the corresponding static bonuses could be used instead

by **tarek** » Tue Sep 22, 2009 8:30 pm

I have an idea about the rare ratings bonus :idea:

:

1. Compile a list of all ratings submitted in the patterns game (spanning all games with ER/EP/ED).
2. Remove all entries that were submitted in games that do not have an ED=8.5/8.5/8.5 submission (Remove patterns with no diamonds)
3. Remove all ratings that have ER=EP=ED (modification ER=EP=ED with a frequency of >= 5%)
4. For each ER keep the highest ED rating & remove the rest.

The resulting list is called (RARE RATINGS INDEX) & is updated at the final closure of each game (after the ::: more ::: entries).

A rare puzzle is a puzzle with a rating that will result in a CHANGE to the RARE RATINGS INDEX. (threfore the bonus can't be awarded prior to game closure)

If the above is too STRICT then a minor modification is to restrict the ratings on the RARE RATINGS INDEX to ratings with ER>4.0 & then instead of rare puzzles being the ones that can only cause a change to the RARE RATINGS INDEX, we can now consider puzzle as being rare if their rating is on the RARE RATING INDEX or if that rating will cause a change in the RARE RATINGS INDEX.

tarek

by **JPF** » Tue Sep 22, 2009 11:19 pm

gsf wrote:WAIT_START could be increased by N*days: suggestions for N?

Let's use N=2 for the moment and WAIT_START=16 hours+2 days to reach an agreement.

JPF

by **gsf** » Wed Sep 23, 2009 2:00 am

JPF wrote:
gsf wrote:WAIT_START could be increased by N*days: suggestions for N?

Let's use N=2 for the moment and WAIT_START=16 hours+2 days to reach an agreement.

ok

Code: Select all: WAIT_START=2*days+16*hours

The New Sudoku Players' Forum

Patterns Game