Patterns Game 1.5

Interactive on-site game threads go here

Postby 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
JPF
2017 Supporter
 
Posts: 6139
Joined: 06 December 2005
Location: Paris, France

Postby 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
gsf
2014 Supporter
 
Posts: 7306
Joined: 21 September 2005
Location: NJ USA

Postby 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
User avatar
g.r.emlin
Moderator
 
Posts: 11580
Joined: 07 December 2007
Location: www

Postby 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)
User avatar
m_b_metcalf
2017 Supporter
 
Posts: 13611
Joined: 15 May 2006
Location: Berlin

Postby 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
User avatar
g.r.emlin
Moderator
 
Posts: 11580
Joined: 07 December 2007
Location: www

Postby 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
gsf
2014 Supporter
 
Posts: 7306
Joined: 21 September 2005
Location: NJ USA

Postby 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
gsf
2014 Supporter
 
Posts: 7306
Joined: 21 September 2005
Location: NJ USA

Postby TTHsieh » Thu Jan 17, 2008 7:05 am

::: more :::
Code: Select all
000000000001020300020401050003050600040762010008010700090803060007090200000000000 4.8

TTHsieh
TTHsieh
 
Posts: 1390
Joined: 13 July 2007

Postby 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
TTHsieh
 
Posts: 1390
Joined: 13 July 2007

Postby 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
JPF
2017 Supporter
 
Posts: 6139
Joined: 06 December 2005
Location: Paris, France

Postby 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.:)
wintder
 
Posts: 297
Joined: 24 April 2007

Postby 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
gsf
2014 Supporter
 
Posts: 7306
Joined: 21 September 2005
Location: NJ USA

Postby 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
gsf
2014 Supporter
 
Posts: 7306
Joined: 21 September 2005
Location: NJ USA

Postby 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
User avatar
g.r.emlin
Moderator
 
Posts: 11580
Joined: 07 December 2007
Location: www

Postby 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
User avatar
m_b_metcalf
2017 Supporter
 
Posts: 13611
Joined: 15 May 2006
Location: Berlin

PreviousNext

Return to Interactive games