The New Sudoku Players' Forum

[g.r.emlin]

::: game 0120 closed LIGHTNING (idle for 1d00h) -- preliminary results :::

Code: Select all

game 0120 2010-09-25+22:00:00-0000 LIGHTNING duration 3d03h dealer JPF pages 649-658 (final) 6 players
______________________________________________________________________________________________________________

    1.2/1.2/1.2 champagne      4.4/4.4/4.4 champagne      6.7/6.7/6.7 champagne      8.2/8.2/8.2 champagne
    1.5/1.5/1.5 champagne      4.5/4.5/3.4 gsf            6.8/6.8/3.4 champagne      8.3/8.3/8.3 champagne
    1.7/1.5/1.5 champagne      4.6/4.6/3.4 champagne      6.9/6.9/2.6 JPF            8.4/8.4/8.4 champagne
    2.0/2.0/2.0 champagne      4.7/4.7/2.8 ronk           7.0/7.0/7.0 gsf            8.5/8.5/8.5 Mauricio
    2.3/2.3/2.3 champagne      4.8/1.2/1.2 champagne      7.1/7.1/7.1 gsf            8.6/8.6/8.4 Mauricio
    2.5/2.5/2.5 champagne      5.0/3.4/2.6 JPF            7.2/7.2/7.2 gsf            8.7/8.7/8.7 champagne
    2.6/2.6/2.6 champagne      5.2/2.5/2.5 JPF            7.3/7.3/7.3 gsf            8.8/8.8/8.8 Mauricio
    2.8/2.8/2.8 gsf            5.6/4.4/2.6 JPF            7.4/7.4/7.3 Mauricio       8.9/8.9/8.9 champagne
    3.0/3.0/3.0 gsf            5.7/3.6/2.6 JPF            7.5/7.5/7.5 champagne      9.0/9.0/9.0 champagne
    3.2/3.2/3.0 champagne      5.8/2.8/2.8 JPF            7.6/7.6/7.6 champagne      9.1/9.1/8.8 champagne
    3.4/3.4/3.4 gsf            5.9/4.0/3.0 ronk           7.7/7.7/7.6 ronk           9.2/9.2/8.8 champagne
    3.6/3.6/3.6 gsf            6.0/1.2/1.2 ronk           7.8/7.8/7.8 champagne      9.3/9.3/8.3 champagne
    3.8/3.8/3.8 gsf            6.2/6.2/6.2 champagne      7.9/7.9/7.9 champagne
    4.0/4.0/4.0 gsf            6.5/4.4/4.4 gsf            8.0/8.0/8.0 ronk
    4.2/4.2/4.2 champagne      6.6/6.6/6.6 champagne      8.1/1.2/1.2 ronk
______________________________________________________________________________________________________________

    open ratings : 4.9 5.1 5.4 6.1 9.4-11.4

       submitter  entries  score  response  notes
     -----------  -------  -----  --------  -----
     m_b_metcalf        0    150    56m07s  rare-0*3,rare-1*2,rare-2*2
        Mauricio        4    647    12m53s  rare-0*2,rare-1*2,rare-2,diamond*2,pearl*2
             JPF        6    761     1h32m  rare-0*2,pearl,dealer
             gsf       12   1053     9m08s  rare-0*6,rare-1*2,diamond*10,pearl
            ronk        6   1616    34m16s  rare-0*4,rare-1*4,rare-2,diamond,pearl*2,last-entry=356
       champagne       29   3960    13m44s  rare-1*2,rare-2,diamond*17,pearl*6,highest-rating=703
     -----------  -------  -----            -----
               6       57   8187            rare-0*17,rare-1*12,rare-2*5,diamond*30,pearl*12


[ronk]

::: more :::

Code: Select all

..9.2.5...6...5.4.3.....8.6.......3.2...7.9...4.......1.7.9...2.3.1...7...8...1..  ED=6.0/2.3/2.3

[champagne]

::: more :::

Code: Select all

..9.8.7...6...7.9.5.....4.3.......2.2...4.1...9.......4.3.1...8.8.6...7...6...3..  ED=9.2/9.2/9.0
..9.8.7...6...5.9.4.....3.2.......1.1...7.5...5.......9.7.6...3.4.3...7...2...4..  ED=9.1/9.1/9.0
..9.8.7...6...5.8.4.....3.2.......7.7...3.9...2.......3.1.9...4.5.1...3...6...8..  ED=6.8/6.8/3.6

[JPF]

::: more :::

Code: Select all

..1.2.3...4...5.1.6.....4.7.......8.2...3.6...7.......3.8.5...4.5.9...3...4...2..   ED=8.1/7.3/2.8
..1.2.3...4...5.6.7.....8.2.......7.5...3.1...9.......8.3.1...9.7.6...8...5...2..   ED=6.9/6.9/3.0
..1.2.3...4...3.5.6.....7.4.......8.2...7.6...5.......4.9.6...1.8.2...3...7...4..   ED=5.7/4.0/2.6
..1.2.3...4...5.6.6.....4.7.......8.2...3.1...9.......9.3.1...6.5.8...3...7...9..   ED=5.6/4.4/3.4
..1.2.3...4...5.6.7.....8.4.......7.2...9.1...3.......3.4.1...6.7.3...8...5...2..   ED=5.2/3.0/3.0


JPF

[JPF]

Some ::: comments :::

First of all, I would like to thank all of the players and viewers of this game I started three and a half years ago.
As of today the game got more than 1,010,000 views.
Special thanks to Mike who was the first follower of this new game and to gsf who developed his g.r.emlin creature allowing game continuity.

On several occasions we spoke about newcomers.
There are some natural hurdles to overcome to be a reasonable player, including:
• Algorithms
• Softwares
• Use of Sudoku Explainer
In addition, newcomers are facing with old timers skills and rhythm of the game as I mentioned before.
Take this last game: the pattern was given at 10 pm; 12 hours later, 50 ratings were already given on a total of 57.
Now there are professional players able to post puzzles days and nights!

I suggest some changes in the rules:

• Limited number of daily submissions for each player (could be 15/n ; n beeing the number of players )
• Automation of submissions should be prohibited
• Increase the “idle” period to 3 days.
It will allow late comers to stay in the loop.
• Allow a player to make a submission after a period of 2 days with no submission.
In many circumstances, it would allow to unlock the situation and restart the game.

JPF

[gsf]

::: comment :::

• Limited number of daily submissions for each player (could be 15/n ; n beeing the number of players )
• Automation of submissions should be prohibited
• Increase the “idle” period to 3 days.
It will allow late comers to stay in the loop.
• Allow a player to make a submission after a period of 2 days with no submission.
In many circumstances, it would allow to unlock the situation and restart the game.

I'll respond from easiest to hardest to do (w.r.t. the referee)

• increase the idle period to 3 days -- easy
• allow a player to make a "consecutive-entry" after 2 days idle -- doable with referee recoding
• limit #submissions per player per 24 hour period -- doable with referee recoding
  this could still allow a flurry of plays early in the game
  increasing the time between plays might take care of this (and would be easy to code)
  so maybe a combination: 1 hour between plays and 12 plays max per day
• prohibit automated submissions -- not possible to detect -- adding the previous changes will limit these anyway

we can still retain LIGHTNING games, but maybe decrease their frequency to say every nth game (5?)

[champagne]

::: comment :::

I must confess I am not at all interested in endless games.
I stopped that one more than one day before closure.

Others researchs are sharing scare ressources.


champagne

[JPF]

::: comment :::

gsf,
Thanks for your quick reaction.
I agree with your proposal which seems to be a fair compromise.

About the automated submissions prohibition, let's put it in the Honor Code of the game.

JPF

[ronk]

::: comment :::

Many wouldn't participate in bowling leagues or golfing leagues if a handicap system didn't somewhat level the playing field.
How about a handicap system for this game?

[gsf]

::: comment :::

Thanks for your quick reaction.
I agree with your proposal which seems to be a fair compromise.

About the automated submissions prohibition, let's put it in the Honor Code of the game.

well we do need to compromise between different levels of play
we already have 10 min between moves vs. lightning
maybe we should experiment with another level starting with the suggestions mentioned
that way we might keep interest up for all levels of play
maybe a few more posts to reach consensus?
like what frequency for lightning vs 10min moves vs n moves per day
with enough players the n moves per day might bring about interesting strategies

re: robo submissions -- I do that (interspersed with manual moves)
as can be seen from the past few months, my current strategy is not that good
I'm attempting to capture what I would do -- its a challenging problem

to keep it fair I only make submissions during my waking hours
and also added a randomized 15-30 min wait between moves
for recent games my cache of playable puzzles has been weak
its usually empty before the game ends

so my honor code view is to mimic what I would do

[champagne]

::: comment :::

first of all some facts on game 120 results.



column 2: total number of submissions
column 3: still there at the end of the game
column 3: ratio c3/c2

Code: Select all

champagne  37  29    0.78
gsf        20  12    0.60
ronk       18   6    0.33
m_b_metcal 14   0    0.00
Mauricio   11   4    0.36
JPF         9   6    0.67

          109  57   

Clearly different strategies.
I can add that my first submission is the tenth. I was sleeping when the game was launched.
As I am a poor old guy, my bladder wakes me up 2 or 3 times a night so I started some hours later.

I like the idea of ronk's to put a handicap.
It could be a first time lag for skilled players before the first submisssion.
No need to add coding, it should be a fair acceptation of the handicap.

This should be enough to give to newcomers the time to find diamonds or to catch rarity bonus.

BTW, an extension of the rarity would be in favour of newcomers.

I also would be in favour of a limitation in the number of games per month (say 2).

whatever will be the final choice, it should prevent to reduce the number of submissions to something very close to the final number of entries.

champagne

[joel64]

::: comment :::

For me the handicap already exists with the difference in power of the computers used. To create a new handicap would be unjust.

[m_b_metcalf]

::: comment :::

I like the idea of ronk's to put a handicap.
It could be a first time lag for skilled players before the first submisssion.
No need to add coding, it should be a fair acceptation of the handicap.

One way to implement this is simply to add a rule that the winner of one game cannot make a submission to the following game for, say, 12 hours after its start.

I share champagne's view that games should not go on endlessly -- stopping after 24 hours without a submission is perfect.

For me the handicap already exists with the difference in power of the computers used. To create a new handicap would be unjust.

joël: I think you may be misunderstaning this -- a handicap here is meant in the sense that it's used in golf, to give a disadvantage to the stronger player.

In general, I've become resigned to playing a lousy end game. I've steadily improved my programs since Game 1, but others have done so faster. A constantly moving target but a stimulus nevertheless. Well done jpf for the initial idea. And, in fact, one of my programs is currently being considered for inclusion in the next SPEC suite!

Regards,

Mike Metcalf

[champagne]

::: comment :::

One way to implement this is simply to add a rule that the winner of one game cannot make a submission to the following game for, say, 12 hours after its start.

To go in the same way, we could work on the last year results and consider the total number of game wons.

One parameter is for sure the delay before the first submission;

Another possibility is to lock at the start the low part of the game for skilled players;

eg of a rule:

any player having won "n" (n maximum 5, the last game counts for 3) during the last year has to wait n*2 hours before the first submission and n*4 hours before submitting an ER below 7.0

champagne

ps: the referee could give the status of handicaps at the start of a game, the rule should be fairly applied without new coding.

[joel64]

::: comment :::

With my weak English, I will try to explain my fears on the application of the rule of handicap.

For a player using an average computer his only chance against a strong configuration is to answer very quickly.

Its penalization is more important in the event of period of waiting of tender. During this time the strong configuration will be able to make a treatment and to send a great number of results.

Joël