Maintenance for the week of December 23:
• NA megaservers for maintenance – December 23, 4:00AM EST (9:00 UTC) - 9:00AM EST (14:00 UTC)
• EU megaservers for maintenance – December 23, 9:00 UTC (4:00AM EST) - 14:00 UTC (9:00AM EST)

ToT custom tournament

Danzio
Danzio
✭✭
Hey guys, have a question to discuss and collect your opinions. I'll be holding a ToT mini tournament in my guild and there's a question about how to choose winners from round robin groups in case their amount of wins are equal. I mean, if we have 2 winners from each group, but there are 3 or more people with 2 wins in top of table or something like that.
Some options to handle this are:
- choose total amount of prestige as 2nd parameter: could not probably work, because some players strategy is build about patron collecting and not prestige gaining;
- chose duration of game as 2nd parameter (which player have minimum total number of rounds to win is better): it is difficult to count rounds and play the game simultaneously, also it could have same problems as prestige checking;
- use an additional game to determine better player: well, it will take additional time, and average time of 1 game is nearly 20-30 minutes now. Also it cannot handle 3+ equal-wins players case.

So I would like to discuss it with any of you, maybe you can suggest some great ideas how to handle this question!
PC EU | Proud kwama in Kwama-and-Sujamma | Doing raids, craft and stuff | Healing bosses, taunting group, damaging AOE-mechanics | CP 1900+
MAIN | Halintar | Altmer Templar (healer) | Worshipper of Auri-El
And many other dudes (just... too many)
  • tsaescishoeshiner
    tsaescishoeshiner
    ✭✭✭✭✭
    ✭✭
    Round Robin ties are somewhat easy to break, given that everyone plays everyone.

    The players in the tie will have played each other, so you can knock out whoever in the tie has LOST to someone else in the tie.

    Failing that, you can add up the scores of the players they lost to. Whoever lost to the players with the lowest scores is ranked lower. Or, whoever lost to players who beat the other players.

    Otherwise, have a race to see who can beat an NPC the fastest! My record is 1:56.
    PC-NA
    in-game: @tsaescishoeshiner
  • dmnqwk
    dmnqwk
    ✭✭✭✭✭
    What sort of group size are you considering?

    Groups of 3 (3 matches total)
    Whether 1 player or 2 advance from the group, there's no need to worry about fairness past 'drawing names of tied positions'. This is because they've not been in a position of awesome luck, then suddenly lose their last match to end up in a playoff.

    Groups of 4 (6 matches total)
    In a top 1 only scenario for advancement, there could be a 2, 3 or 4 player tie for this. That would be hell.
    DONT MAKE GROUPS OF FOUR because groups of four suggest you'd be better off straight elimination (and provide byes where necessary, such as if you have 13 players in the tourney).
    In a top 2 advancing - see above (because you could still have the 3 or 4 tied player scenarios affecting it, while a tied for 2nd place would be a single elimination match).

    Groups of 5 (10 matches total)
    If you have someone go 5/0 then you're safe, because it only leaves a possibiliy for a tied 2nd place between 2 players.
    If the top person finishes 4/1 then a 2nd could finish 4/1, but the winner would be whoever beat the other. If you allow a 2nd advancement from this group you'll probably face a situation of 2/3 2/3 2/3 0/5 for group and risk a triple playoff scenario (see Groups oof 3 explanation for tiebreaker)
    If nobody goes better than 3/2, you could get a three-way tie for 1st but if they're doing this there's no clear leader and see Groups of 3 explanation

    Basically I'd suggest this:
    1. DO NOT USE GROUPS OF FOUR they just extend the duration of the tournament.
    2. 1 advancement per group is best, but for Groups of 5+ a 2nd is fine if you accept the extra headaches.
    3. Tiebreakers should be Versus, then drawing of straws (for groups of 3-5). The reasoning is if you remove a win from a player and their percentage of wins falls below 50%, there's no real need to worry it comes down to luck.

    Obviously you might include groups of 6, but then you open yourself up to some nasty situations of ties!

    Groups of 6 (15 matches total)
    Too much explanation but, like groups of 4, you risk every single person going 3/3. This is why even groups suck!
  • Danzio
    Danzio
    ✭✭
    Thanks guys for your ideas! Currently, we found a way to overcome this problem, using Sonneborn-Berger score. Shortly: we can sum up scores of players, to which tied ones lost. This means player who lost to more successful opponents advances.
    We currently have only group of 5, so I think there will be no big deal with issues as @dmnqwk suggested. As for group of 3 - there is a possibility to have a triple deadlock (each one lost to each one), with 1 winning point on each participant. This could be quite difficult to handle.
    Anyway, thank you for your ideas!
    PC EU | Proud kwama in Kwama-and-Sujamma | Doing raids, craft and stuff | Healing bosses, taunting group, damaging AOE-mechanics | CP 1900+
    MAIN | Halintar | Altmer Templar (healer) | Worshipper of Auri-El
    And many other dudes (just... too many)
Sign In or Register to comment.