Gomoku Statistics: Black vs. White - How effective is the forbidden zone? Gomoku, Connect6

36 replies. Last post: 2010-07-13

Reply to this topic Return to forum

Gomoku Statistics: Black vs. White - How effective is the forbidden zone?
  • Julius Sneezer at 2009-09-13

    There must have been many Gomoku games played here over the years, and I wonder if the statistics (does anyone know how to find them?) show that the forbidden zone does in fact even out the advantage held by black. At renju.net I found that in tournament play, over about 1000games, Renju with move restrictions for black, additionally played with swap, came out almost exactly even for black and white. The forbidden zone is much simpler and more elegant, but is it as fair?

  • ondik at 2009-09-21

    No it is not. The opening rule here at little golem is very bad and imbalanced and may be one of reasons why there are no good players except russians playing in here. Just look at games in first league of gomoku championship. In my opinion the starting player has at least 80 % advantage.

  • ypercube at 2009-09-21

    It would be if some statistics offered more evidence in support of this theory

    (I don't disagree ondik, Black proabably has advantage - but 80% ? - it sounds too much.)

  • Julius Sneezer at 2009-09-21

    When i was looking at these stats at renju.net, i also looked at stats from a single Gomoku tournament (the exact rules used didn't seem to be specified) and it was 80% first player/black.

    I wonder if the 'renju -ish' option for the second player of swapping the first 3 moves, which is possible in Gomoku at Kurnik and Richard's PBeM, would help much.

  • Thomas at 2009-10-07

    Suggestion for new Gomoku rules.

    Rule details I'm not sure about are in [].

    The game is played on 19 * 19 board, initially empty.

    The first player places two black and one white stone on the board, then the second player decides which player is White and which is Black. [maybe one black stone must be on the center point]

    From then, players alternate placing a stone of their color onto an empty point, with White starting. Together with placing a stone, a player can prolong the game if it has not already been prolonged.

    If a player forms an orthogonal or diagonal unbroken row of 5 [or more] of their stones, that player wins.

    If there are 100 stones [or any other number] on the board and no player has a winning row and the game has not been prolonged, the game is draw.

    If the board is full of stones and no player has a winning row, the player who has prolonged the game loses.

  • dushoff at 2009-10-07

    Has anybody done the Connect6 statistics? Is Connect6 balanced?

  • Arty Sandler at 2009-10-07

    Connect6 is pretty balanced according to the statistics I saw on few sites.

    For example:

    http://brainking.com/en/GameRules?tp=82

    http://www.iggamecenter.com/stats/game19.html

    On both statistics there are slightly more wins by the first player than by the second player (50.5% vs 49%) but it's not any close to Gomoku's 80% vs 20%.

  • Thomas at 2009-10-07

    I have counted it for league 1 and 2 of the gomoku championships 10 to 13 (12 sections)

    278 games won by black (61.8%)

    168 games won by white (37.3%)

    4 drawn games (0.9%)

    in every section there was more games won by black than by white (at least 4 more)

  • euhuang ★ at 2009-10-13

    I got a message from Professor I-Chen Wu's NCTU Connect6 Lab. This lab has been analyzing Connect6 games on Little Golem. Here is its statistics so far:

    Total game count* : 17739

    B-Win : 9185 (51.77%)

    W-Win : 8554 (48.23%)

    \* Counts only those games with at least one of the two players ranked above 1800.

    For your reference.

  • ypercube at 2009-10-13

    I'd like to see statistics from other ratings as well.

  • euhuang ★ at 2009-10-13

    @ypercube: The reasons why the above NCTU Connect6 Lab's statistics doesn't count games with both players ranked below 1800 are two fold:

    (1) to avoid non-sense games which might possibly disturb statistics

    (2) the total game number is simply too large to collect & analyze by NCTU C6 Lab itself. (Maybe the Little Golem webmaster can easily provide the total statistics.)

    (By the way, the above statistics show results of LG C6 games up to 2009/9/10.)

  • Julius Sneezer at 2009-10-13

    One thing to point out is that, as I understand it, the Gomoku discrepancy only becomes overly apparent with very strong players (ie in Japan, 'dan-level'). This discussion may not immediately affect players whose present rating is below about 1800.

    Even though LG Gomoku *may* be sufficient for beginners (like myself) it may not offer much to strong players, or to players who dedicate themselves to getting stronger (it additionally helps to have strong players around for that).

    Simply saying “play Connect6 instead” to me isn't sufficient as they are vastly different, though related, games (“play Connect6 also” is a more reasonable proposition); and currently Connect6 is still too new to have the resources that other n-in-a-row games have online, such as Nosovsky's writings.

  • JerNYC at 2009-10-29

    I don't understand all the fuss. I play at all the major board game sites and LG's forbidden zone is definitely challenging for player one. I'm certainly no “beginner” and ypercube can vouch for me. I have a 1900+ rating on two sites. I haven't played here enough to get my rating up on LG. But, even the Pro-Go Moku opening move restrictions on the other sites still allow for a decent playing field. It really depends on what kind of player you are. Did the statistical analysis take into account who the players were? Are we talking about players of equal strength? Is there one particular player that wins most of the GM championships on LG or is it evenly spread among several top players?

    What I want to know is why LG allows overlines in their GM. It should be a strict five. I'm also not that crazy about the graphics of the board and pieces. The pieces are very chunky and the board gets crowded in a hurry which plays tricks on the eyes.

  • Wang Zhiguo at 2009-11-29

    Thomas (new Gomoku rules) is “Swap1” rule ,reference: http://www.playok.com/

    and A better rule is “Swap2”, Because the first player places Unknown opening to opponent ,so it is not fair.

    Chinese-style gomoku rule is “Swap-F"(Swap after the first move)

    The first player place one black stone on the board, then the second player decides which player is White and which is Black.[Like “Go” opening]

    so it is much simpler and more elegant, more fair!

    game room : http://www.freewzq.com/cn/gameWzq.htm#TableID=Room_9&TableIndex=1

    game record :http://www.freewzq.com/cn/ASP/gameRecord_WZQ_list.asp

  • Richard Malaschitz ★ at 2009-11-30

    I suggest these new rules for Gomoku:

    Board:

    1. Board is 15x15

    Opening rules:

    2. First player put on board 3,4 or 5 moves. First move is in central point.

    3. Second player can use SWAP rule

    Win rule:

    4. Player win when create 5 or more stones in row

    No draw rule:

    5. PASS move. When player pass his move, rest of board will be filled with opponents stones and game is finished. If there will be no 5-in-row, player who passes is winner.

  • FatPhil at 2009-11-30

    I really like the no draw rule. I play it in a couple of card games, and it works really well.

    I see nothing negative about any of the other rules either. However, is '3, 4, or 5' too much flexibility? Why not just fix one of those?

    I of course defer to those who actually know more about the game.

  • Julius Sneezer at 2009-11-30

    i agree about the 'no draw' rule, i think it's brilliant. i was hesitant about '3,4,5' but after imagining how it would play out i think it is very good also.

  • Julius Sneezer at 2009-11-30

    (also, because the specific rules are unique, without changing the essentials of the game, it would be 'cheat-resistant').

  • Anatolij at 2009-12-03

    Richard What is your opinion about the false rating Kung Fu Panda O_O ?

  • I_EI_EI_EI_EI_E at 2010-04-29

    At renju.net I found that in tournament play, over about 1000games, Renju with move restrictions for black, additionally played with swap, came out almost exactly even for black and white.

    ---——————————-

    That is only half right.

    Swap made balance between black and white.

    But “restrictions for black” is worse,it made not balance.So renju`s board is 15*15,because 19*19 is very good for white at later-half game.

  • I_EI_EI_EI_EI_E at 2010-04-29

    At renju.net I found that in tournament play…

    ---—————————-

    I collect some data about this game.There are 425 games Group A and Group Q:

    black win : 167 .

    white win : 192 .

    The average-move of not-draw game is : 45.

    and :

    black win before 45 move : 108 .

    white win before 45 move : 90 .

    black win after 45 move : 54 .

    white win after 45 move : 102 .

    black win after 60 move : 22 .

    white win after 60 move : 53 .

    At later-half game , Black and white party winning percentage ratio of less than 1:2

    So,renju is not a balance game.At early-half game,that is black good.At later-half game.

    Because in renju , it “restrictions for black”,black and white sides use different rules to win,How funny it is!

  • I_EI_EI_EI_EI_E at 2010-04-29

    I suggest these new rules for Gomoku:

    Board:

    1. Board is 15x15

    ---———————–

    The smaller the board, the less variation。

    Opening rules:

    2. First player put on board 3,4 or 5 moves. First move is in central point.

    3. Second player can use SWAP rule

    ---—————————-

    “Muse central point” = less variation.

    Win rule:

    4. Player win when create 5 or more stones in row

    ---—————

    That is good. 6 include 5 , so 6 ought to win.

  • I_EI_EI_EI_EI_E at 2010-04-29

    The first player place one black stone on the board, then the second player decides which player is White and which is Black.

    ---———————-

    Yes,This rule is fair and much simpler.

    19 * 19 is good , it would more variation.

  • Julius Sneezer at 2010-04-30

    these days i think we should have renju rules (with restrictions for black) as an option / variant for gomoku. i would like the challenge of learning something new and a bit complex. alot of people play this version of the game all over the world but there aren't many opportunities to play it on english sites.

  • I_EI_EI_EI_EI_E at 2010-05-01

    to plasticonion :

    Why dont you think,should have a renju++ ( both with restrictions for black and white ).

    At least,renju++ is fair.

    Renju only restrictions for black , it unfair , and how funny it is.

  • Julius Sneezer at 2010-05-01

    ni hao, it's pretty much accepted that the player who moves first (black) has a very large advantage in free gomoku. move restrictions for black in renju are intended to level the playing field, as is swap. it seems to be the case, according to statistics, that gomoku with swap is more fair than free gomoku, but renju with swap is more fair than gomoku with swap, and as fair as connect6.

  • Julius Sneezer at 2010-05-01

    PS rejecting renju on the basis of its complicated rules is a mistake. 'a minute to learn, a lifetime to master' -type rules are a thing of the past. mathematicians admire such rules, but people in general society respect games with complex rules, like chess and go. they are impressed by people who can play complicated games, and players who learn such games feel like they have accomplished something.

  • I_EI_EI_EI_EI_E at 2010-05-02

    renju has no real fair.

    in renju , black is better than white before 45 moves , and white is better than black after 45 moves.

    if there is a rule , no swap , and :

    i play 5 games by black ,you play 5 games by white.

    Do you think that is fair?

    yes it is fair,but it is not real fair.

    in fact , “move restrictions for black " is a unsuccessful design. designer what limit black advantage ,but black sure win.

    and this design be continued

    how funny it is!

    PS : in all chess game rule,black and white sides use same rules to win,but renju it is not.

    so, renju`s board only 15 * 15,how small and how funny it is!

  • I_EI_EI_EI_EI_E at 2010-05-02

    I collect some data about gomodu.

    There are 27736 games from “PLAYOK match(organiser is rewolwer (reg.))“,from 2007-10-15 to 2008-4-14.

    (with out games that with out unuseful datas such as less than 3 moves and tie games )

    black win : 14238 .

    white win : 13498 .

    (About 1:1 -> 1.05:1)

    black win before 30 moves : 6107.

    white win before 60 moves : 6077.

    (About 1:1 -> 1.004:1)

    black win before 60 moves : 12611.

    white win before 60 moves : 11939.

    (About 1:1 -> 1.05:1)

    black win after 60 moves : 1627.

    white win after 60 moves : 1559.

    (About 1:1 -> 1.04:1)

    black win after 90 moves : 1289.

    white win after 90 moves : 1263.

    (About 1: -> 1.02:1)

    black win between 30 and 60 moves : 6504.

    white win between 30 and 60 moves : 5862.

    (About 1:1 -> 1.10:1)

    black win between 60 and 90 moves : 1289.

    white win between 60 and 90 moves : 1263.

    (About 1:1 -> 1.02:1)

    All indicators close to 1:1 .

    And let's look renju:

    Such as,after 60 move,black and white party winning percentage ratio less than 1:2(please see above article).

    PS:

    gomoku with swap is more fair than free gomoku

    ---————–

    did you means : gomoku with swap is more fair than free gomoku with out swap …? This is not comparable.

    If you are willing to , please tell me gomoku with swap and free gomoku with swap comparison between , thx.

  • I_EI_EI_EI_EI_E at 2010-05-02

    gomoku with swap and free gomoku with swap all has real fair.

    but renju has no real fair,bucause in renju`s rule , black and white sides use different rules to win,how funny it is!

  • I_EI_EI_EI_EI_E at 2010-05-03

    did you said “free gomoku” means “free-style gomoku( 6 in a line is win )“?

    if not , i am sorry :)

  • Julius Sneezer at 2010-05-03

    thanks for the statistics ni hao, they are interesting, but as i posted earlier in this thread, the first move advantage is said to be more pronounced between stronger players. i do not argue that renju should replace gomoku at little golem, but that it should be added as a variant, so players have it as an option. nobody wants to dedicate time to becoming a stronger player in a game if this merely renders the outcome of the game more and more predictable.

  • I_EI_EI_EI_EI_E at 2010-05-04

    thanks for reply.

    gomoku`s board is 19 * 19,and renju`s board is 15 * 15.

    so gomoku`s variants more than renju`s variants.

    why renju`s board must be 15 * 15?

    because renju only restrictions for black , that is un-real-fair.

    and ,if you like renju,why you dont like renju++( both with restrictions for black and white )?

    At least , renju++ is real fair.

  • Anatolij at 2010-05-04

    According to the rules of fair play, we suggest, at least for major league

    play without clones. For my part, I tidy up their clones from the higher

    League-bonus and badboy, I also support other strong players.

  • I_EI_EI_EI_EI_E at 2010-05-04

    I am sorry for clones use of site loopholes.

    At first, I have access to TOP20 A Game spectrum, I saw two players a draw deliberately, and all of a draw like this, this makes me very angry, while LG-gomoku rules did not really feel a draw, but uncertain, want to strong players play , but if the points slowly, feeling very far away,also see a lot of players in 2000 points, the following points, very few high scores and not often, some of them have not play , so there is no rising opportunities, this new user to the past master is a torture. So think of this method, many accounts, won more points, into the Championship League: 1.

    I hope that LG will become better and better.

  • FatPhil at 2010-07-13

    Just for reference, don't think that your change of name yesterday will cover your history of sock-puppeting and early resignations.

Return to forum

Reply to this topic