The New Sudoku Players' Forum

by **JPF** » Wed Jan 16, 2008 3:30 pm

Thanks gsf.

I think it's a nice improvement.
That should slow down a bit the game.

The remaining trouble is that people can't stay all the day (and night !) behind their computer to be able to react to new posts.
Note that this problem doesn't exist with only 2 players.
What can we do ?

What about g.r.emlin dealer ?

JPF

by **gsf** » Wed Jan 16, 2008 4:53 pm

JPF wrote:Thanks gsf.
I think it's a nice improvement.
That should slow down a bit the game.

it won't be until after the weekend that pearl trumps can get coded into the referee
I'll backup the current code in case its a bust

The remaining trouble is that people can't stay all the day (and night !) behind their computer to be able to react to new posts.
Note that this problem doesn't exist with only 2 players.
What can we do ?

lets see what happens with one change

What about g.r.emlin dealer ?

there's 88 candidates cached up
they match the rules w.r.t. symmetry, clues, rating
but I'm not sure about the quality w.r.t. game play
I have some heuristics but they could use improvement

we could simply make the dealer an observer
that might stir up some new players too

by **g.r.emlin** » Thu Jan 17, 2008 3:13 am

::: game 0012 idle -- with no further entries the game will close at 2008-01-17+04:00:00-0000 :::

Code: Select all: game 0012 2008-01-15+08:16:00-0000 duration 19h39m dealer TTHsieh index 840 (open) 4 players 1.2 1.5 1.7 2.0 2.3 2.5 2.6 2.8 3.0 3.2 3.4 3.6 3.8 4.0 4.2 4.4 4.5 4.6 4.69 4.7 5.0 5.2 5.4 5.6 5.7 5.8 5.9 6.2 6.5 6.6 6.69 6.7 6.8 6.9 7.0 7.1 7.2 7.3 7.4 7.5 7.6 7.7 7.8 7.9 8.0 8.2 8.3 8.4 8.5 8.6 8.7 8.8 8.9 9.0 9.1 9.2 submitter entries score response notes ----------- ------- ----- -------- ----- JPF 5 79 7m19s m_b_metcalf 2 127 20m59s weak-pearl gsf 25 749 1m48s weak-pearl*2 TTHsieh 24 1733 3m05s last-entry,highest-rating,weak-pearl*13,strong-pearl*6,dealer ----------- ------- 4 56

by **m_b_metcalf** » Thu Jan 17, 2008 4:45 am

::: comment :::
I'm unclear about the distinction between a weak and a strong pearl. Taking the SE 8.7 I posted on the previous page as an example, SS takes five steps to reduce the number of candidates (locked candidates, naked quad, swordfish), but solves no cell. The rules say

(a strong pearl has no simple sudoku constraint that leads to the first placement)

so why is that not strong (it's classified as weak in the score)?

Thanks,

Mike Metcalf (not complaining, just curious)

by **g.r.emlin** » Thu Jan 17, 2008 5:19 am

::: game 0012 closed (idle for 1d00h) -- preliminary results :::

Code: Select all: game 0012 2008-01-15+08:16:00-0000 duration 19h39m dealer TTHsieh index 840 (idle) 4 players 1.2 1.5 1.7 2.0 2.3 2.5 2.6 2.8 3.0 3.2 3.4 3.6 3.8 4.0 4.2 4.4 4.5 4.6 4.69 4.7 5.0 5.2 5.4 5.6 5.7 5.8 5.9 6.2 6.5 6.6 6.69 6.7 6.8 6.9 7.0 7.1 7.2 7.3 7.4 7.5 7.6 7.7 7.8 7.9 8.0 8.2 8.3 8.4 8.5 8.6 8.7 8.8 8.9 9.0 9.1 9.2 submitter entries score response notes ----------- ------- ----- -------- ----- JPF 5 79 7m19s m_b_metcalf 2 127 20m59s weak-pearl gsf 25 749 1m48s weak-pearl*2 TTHsieh 24 1733 3m05s last-entry,highest-rating,weak-pearl*13,strong-pearl*6,dealer ----------- ------- 4 56

by **gsf** » Thu Jan 17, 2008 5:19 am

::: pearl :::

Code: Select all: 000000000004020300050304020006070400080492010005030900030901050009060700000000000 6.5 4 000000000009020300040605070006080700070163050003040200060408010008010400000000000 8.7 11 000000000009010300040806070006040700070163050003090200060405010008020400000000000 5.0 3 000000000001070300040105060002040500070512080009060100020408070005030400000000000 3.6 4 000000000001020300040506010005070600070365080003090200010708040007010800000000000 8.6 99 000000000001020300050401070006090700080746020009050400060205080002030500000000000 8.2 11 000000000001020300020403060003060100070831020008050900050607010002090800000000000 4.0 4 000000000001020300040506020004010200030467080009080600020803060003090500000000000 7.4 11 000000000006090300040506020001080700070163050003040200060408070008010900000000000 8.9 99 000000000001020300030504010006010700050782090008040200060809030007030800000000000 3.8 4 000000000001060300030104050004020500020318090008040700050609040002070800000000000 8.5 99 000000000001020300040506090006080700070963050003040200010804070008050400000000000 9.0 99 000000000001090300040105060007020100080451020004060500060908050009070400000000000 7.6 11 000000000001020300080105060002090500070512080009060100020406070005070400000000000 8.4 99 000000000001020300020405060006050700080794020004010900060908010007030400000000000 6.8 4 000000000001090300040105060007020100080451020004030500060208050008010400000000000 7.8 11 000000000001020300040105060007030500080451020004060100060908050008010400000000000 4.6 99 000000000001020300030104050008060700020318090003040100050609020002070600000000000 7.5 11 000000000001020300020401060002050700080794020005010600050908010007030900000000000 3.4 7 000000000007040300080105090004050600050267030008010700040603010006020900000000000 5.6 3 000000000001020900040506010005080600070365080003090200010807040007010800000000000 2.6 4 000000000001020300040501090003010700080369050006070400060703040002090800000000000 3.2 4 000000000001020300040305060004030500020758090008090200060907020009080100000000000 7.3 11 000000000001020300020304050003040700070831020008070900050607040007090800000000000 6.6 11 000000000001020300020304050006030400080492010005070900030901080009060700000000000 3.0 4 000000000009020300040306070006080700070163050003040200060405080007030400000000000 2.8 6 000000000001020300040509020005070600070365080003090200020807040007010800000000000 8.3 11 000000000001020300030104050004060700020318090008040100050609040002070600000000000 7.1 11 000000000001020300040803060003010200070638050004090800010705080006080700000000000 4.5 11 000000000009020500040506070006040700070163050003090200090408010005070400000000000 4.4 7 000000000007020300040305070002030600010267080009050700030806040006090800000000000 7.2 99 000000000009020300030904080005010600040678090007040100050702030006030700000000000 2.3 3 000000000001020300070506090004010200030467080009080600010803060003090400000000000 2.0 3

by **gsf** » Thu Jan 17, 2008 6:48 am

m_b_metcalf wrote:::: comment :::
I'm unclear about the distinction between a weak and a strong pearl. Taking the SE 8.7 I posted on the previous page as an example, SS takes five steps to reduce the number of candidates (locked candidates, naked quad, swordfish), but solves no cell. The rules say
(a strong pearl has no simple sudoku constraint that leads to the first placement)
so why is that not strong (it's classified as weak in the score)?

the referee uses my solver -S -qss+G -f%#Pq to print the pearl index

--man for -qss wrote:A close approximation to the Simple Sudoku constraint order.
Differences will surface between SS coloring and X/Y cycles.

the 8.7 with pearl index 11 has some Y cycles and that's probably a bit stronger than ss coloring
it might make sense to move Y cycles after H4 (hidden quads) in the -qss order

by **TTHsieh** » Thu Jan 17, 2008 7:05 am

::: more :::

Code: Select all: 000000000001020300020401050003050600040762010008010700090803060007090200000000000 4.8

TTHsieh

by **TTHsieh** » Thu Jan 17, 2008 7:08 am

I found 39 ratings with pearl (* strong) in this round.

Code: Select all: 2.0 2.3 2.5 2.6 2.8 3.0 3.2 3.4 3.6 3.8 4.0 4.4 4.5 4.6* 5.0 5.6 6.6 6.69 6.7 6.8 6.9 7.0 7.1* 7.2* 7.3* 7.4 7.5* 7.6 7.7 7.8 8.2* 8.3* 8.4* 8.5* 8.6* 8.8* 8.9* 9.0* 9.1*

TTHsieh

by **JPF** » Thu Jan 17, 2008 10:28 am

::: pearl :::

pearls wrote:000000000001020300020403050002060500060758030007040800090506040006070100000000000 # 8.9 # 99
000000000001020300020104050002060100070815040008030900080706090007080400000000000 # 8.4 # 11
000000000001020300020403050002040600040271030008050700090507060007080100000000000 # 8.4 # 11
000000000001020300020403050002040600040271030008050700090705060007080100000000000 # 8.4 # 11
000000000001020300020403050002060700060251030008040500010506040009080100000000000 # 8.4 # 4
000000000001020300020403050002060700060251030008040500010605040009080100000000000 # 8.4 # 4
000000000001020300020405060002070600070869050008040900030607040007090100000000000 # 8.3 # 99
000000000001020300020403050002060400060758030008040500070506090006080100000000000 # 8.3 # 11
000000000001020300020304050002060400070859030008040500090507060007080100000000000 # 8.3 # 11
000000000001020300020403050002060400070859030008040500090705060007080100000000000 # 8.3 # 11
000000000001020300020304050006050700040278030002030900080105070004060100000000000 # 8.3 # 11
000000000001020300020301040005040600070268030002030900080704060007050100000000000 # 8.3 # 11
000000000001020300020405060002070400070863050008040600030607090007080100000000000 # 8.3 # 11
000000000001020300040105060002070100080216050009030600090807040006040800000000000 # 8.3 # 5
000000000001020300040503060004060100010478030009030700080607010002090400000000000 # 8.3 # 4
000000000001020300040305060004030700030481050009060800020608070008090100000000000 # 8.2 # 11
000000000001020300040305020004030600030471050008090700090207060007080100000000000 # 7.3 # 11
000000000001020300020403050002060700010879030008040500040907010007080600000000000 # 7.3 # 11
000000000001020300020403050002060500010758030008040100070905040009030600000000000 # 7.3 # 11
000000000001020300020403050002060500010758030007040100080905040009030600000000000 # 7.3 # 11
000000000001020300020304050002030600030271040008050700090507060007080100000000000 # 7.3 # 11
000000000001020300020403050002060500060247030008090700050706090007080100000000000 # 7.3 # 9
000000000001020300020103040002050600050247030008010400090405010007080900000000000 # 7.3 # 4
000000000001020300020403050002060700060258030009040500010506040008090100000000000 # 7.3 # 4
000000000001020300020403050002050600050267080009040500070805060008090100000000000 # 7.2 # 11
000000000001020300020405060002070500030869040008050600090603070003080100000000000 # 7.2 # 11
000000000001020300020403050002010600010256070008040500070502040003080900000000000 # 7.2 # 11
000000000001020300040103050002030600060257030008060500070506040006080100000000000 # 7.2 # 11
000000000001020300040503060004070500070452030008030700060709050007080100000000000 # 7.2 # 11
000000000001020300020403050002050600070268030009040500080705060007090100000000000 # 7.2 # 11
000000000001020300020403050002040600040271080009050700080507060007090100000000000 # 7.2 # 11
000000000001020300040503020004050600070681030009030800050807060008090100000000000 # 7.2 # 11
000000000001020300020405060002030100030261050007040600080906040009070800000000000 # 7.2 # 5
000000000001020300030405060002070800070261050009040600010607040003090100000000000 # 7.2 # 4
000000000001020300040105060007080900080763050003050600030608040002040700000000000 # 7.2 # 4
000000000001020300020104050005060100060217040008030700050906030007080600000000000 # 7.2 # 3
000000000001020300020104050005060100060217040008030700050906030007080400000000000 # 7.2 # 3
000000000001020300020405060002070400070368050003040600080607090007080100000000000 # 7.1 # 11
000000000001020300040103050002040600060257030008090500070504080006080900000000000 # 7.1 # 3
000000000001020300020301040002050600030247010008090400060409070009080500000000000 # 3.2 # 4
000000000001020300020104050002060100060517040008030700090706020007080400000000000 # 3.0 # 4

JPF

by **wintder** » Fri Jan 18, 2008 10:58 pm

JPF wrote:Thanks gsf.

I think it's a nice improvement.
That should slow down a bit the game.

The remaining trouble is that people can't stay all the day (and night !) behind their computer to be able to react to new posts.
Note that this problem doesn't exist with only 2 players.
What can we do ?

What about g.r.emlin dealer ?

JPF

This would be putting a lot on gsf and it is an early thought.

There should be a way to pm Gremlin with a list of "ready entries".

Gremlin accepts such lists and injects a submited puzzle after ten(?) minutes idle time on the board. (Give time for alternate entries.)

Gremlin would compare submission time of entry to current entries and "lose" conflicting submissions.

Gremlin would use the "current" submission rule and prefer earliest submission in a tie. The submissions would be pooled but sorted by entry time and rating. This done at the ten minute idle mark, not any earlier.
Redone every time so new entries get in.

I haven't thought it out deeply but it would be a way to help people who log on less often.

by **gsf** » Sat Jan 19, 2008 4:06 pm

pardon the short referee hiatus
retooling for pearl trumps
and mulling an imposed 10 min between submissions by the same player

by **gsf** » Tue Jan 22, 2008 8:23 am

the referee should now be able to handle pearl trumps
the first post has been updated but here's a recap of the changes:

(1) ratings are now ER=explainer-rating/pearl-index where explainer-rating is the SE 1.2.1 rating
and pearl-index is the (approximate) Simple Sudoku index of the hardest move
that leads to the first placement as listed by these options to the gsf solver (version 2008-01-18 or newer)

Code: Select all: -S -qss+G -f%#Pq

the periodic status will list the ER/largest-pearl-index for each ER

(2) trumps: a entry may have the same ER as a previous entry as long as it also has a larger pearl index
the new entry erases any scores of previous entries with the same ER
(strong pearls (pearl index 99) cannot be trumped)

(3) a strong pearl (pearl index 99) may not be the first entry for a given ER

(4) after making a submission a player may not make another submission until 10 minutes
have passed and another player has made a submission

we'll try this for a game or two and see if it helps level the playing field

by **g.r.emlin** » Tue Jan 22, 2008 8:34 am

::: game 0013 start at 2008-01-22+16:00:00-0000 :::

Code: Select all: dealer status candidate patterns (via private message to g.r.emlin) 1 JPF 1 coloin 3 m_b_metcalf 88 g.r.emlin past patterns 1 gsf 1 m_b_metcalf 1 ravel 1 wintder 2 Mauricio 2 g.r.emlin 15 JPF 20 TTHsieh

by **m_b_metcalf** » Tue Jan 22, 2008 3:26 pm

gsf wrote:(4) after making a submission a player may not make another submission until 10 minutes
have passed and another player has made a submission

Just for clarification: if a puzzle is submitted 'too soon', does that mean:

a) it is ignored;

b) it is queued; or

c) it is penalised?

Thanks for all your work.

Regards,

Mike Metcalf

The New Sudoku Players' Forum

Patterns Game 1.5