Possible UNFairness in TMS Tiebreaker system

Discussion in 'TCG News & Gossip Discussion' started by SteveP, Feb 1, 2004.

8 league13 468 60
Thread Status:
Not open for further replies.
  1. SteveP

    SteveP Active Member

    Let me pose a hypothetical. Your school plays football in a tough conference. At the end of the season, the top 2 teams have identical win-loss records, both conference-wise and non-conference-wise. These two teams played each other once during the season. But, because the conference is large, all teams DON'T get the chance to play each other during the season.

    So, how do you determine who the conference champ is?

    1. The champ is determined by who won the regular season match-up.

    2. The champ is determined by whose opponents were tougher.

    3. Have a playoff.

    I ask this question because this came up today at the Lakewood Colorado City Championship. The top 2 players had 15 points each. Both players played each other. However, the TMS Tiebreaker system gave the top position to the player who actually lost to the other player he was tied with.

    In Decipher Swiss Tournaments, if there are only two top players with identical records, and those players played each other, the winner is determined by who won the match-up, not the tiebreaker. The tiebreaker is ONLY used to determine the winner IF AND ONLY IF there's a 3-way tie (or more) at the top,OR the two players never played each other (very unlikely in swiss-style tournaments).

    Did we give the Lakewood Colorado CC Medalion (and VIP Package award) to the wrong player?

    I was surprised when the printout showed the match-up winner in second place instead of first. That difference was only 3 percentage points.

    Any thoughts?
     
    Last edited: Feb 1, 2004
  2. NoPoke

    NoPoke New Member

    Yep.. exact same thing happend yesterday. However it seemed to be determined by the 2nd place player having slightly easier opponents in the first two rounds.

    Now to add further difficulty to the use of resistance alone, the player who came second had as one of his opponents a player who droped from the tournament. So dropping potentially penalises opponents. :(

    Note that we are running manually and might not be emulating the intended activities of the tournament software. Its easy to emulate its actual behaviour :(
     
  3. TheCrossFormatKid

    TheCrossFormatKid New Member

    In my opinion, this dilema could be solved by cutting to a top8 at the end of swiss rounds. Pure swiss isn't nearly as fair as swiss with t8.
     
  4. marril2k

    marril2k Member

    I'm not sure if it's fair to call what TMS generates as "OP-Match Win%"... I've run a few test runs with Reporter and TMS running along side each other, using manual pairing on one based off what the other generates. While the points may match up, for some reason the "tiebrk" are different. Thoughts?

    -marril
     
  5. NoPoke

    NoPoke New Member

    Reporter uses a different system. It starts off similar: since win percentage and opponents resistance are closely related. The handling of byes and drops being the significant differences.

    Reporter has a 33% minimum for op win %.

    Both TMS and Reporter aim to establish which player had the tougher ride to break the tie. Neither Reporters' nor TMSs' approach is without problems.

    hmmm I think I'll try a side by side comparison just to make sure that both are behaving as I expect/understand.
     
    Last edited: Feb 2, 2004
  6. JohnnyBlaze

    JohnnyBlaze New Member

    Steve - Contact Nintendo on this. Its gotta go to the player that straight out won and now has 18 points over the player that lost but still had a higher % but still has 15 points. That is crazy what happened :eek:
     
  7. mysterioustrainer

    mysterioustrainer New Member

    Just for the record, I was the one who didn't pull off the top win. Oh well, second place wasn't that bad.
     
  8. Adv1sor

    Adv1sor New Member

    If both players had the same number of points then you have to go with the tie breaker percentages. The fact that they played each other doesn’t matter.

    You can have the same kind of problem cutting to a top eight. I saw a tournament where one player entered the top eight undefeated while another squeezed in with two losses. The player with two losses ended up winning the tournament. Is that fair? Depends on your definition I guess.
     
  9. SteveP

    SteveP Active Member

    JohnnyBlaze, no need to contact Nintendo. Both players finished 5-1, 15 points.

    My thoughts and opinions are intended NOT to slam the tie-breaker system. It works okay, although slightly different than the DCI tie-breaker system (as pointed out by NoPoke). I'm just trying to argue the merits of a "tie-breaker over-ride" when there's only 2 top players that are tied (or for that matter, when ANY two players are tied in the final standings). I'm of the opinion that head-to-head match-up SHOULD take precedence in these situations, NOT the tie-breaker.

    Top 8 is NOT an option for many tournaments. It extends the tournament 3 extra rounds (2-3 more hours). Plus, as has been pointed out, the EXACT SAME issue comes up when cutting off the top eight players and using the tiebreaker.
     
    Last edited: Feb 2, 2004
  10. GreatFox

    GreatFox New Member

    When that happenes.... its always best to do a sudden death match betwwen the two. Especially when it comes down to who will recieve the VIP Package.
     
  11. Adv1sor

    Adv1sor New Member

    I like that idea Great Fox so long as this is pointed out before the tournament starts. Otherwise it may look as though you are waiting to see who wins before deciding what to do.
     
  12. NoPoke

    NoPoke New Member

    hmm playoff aren't without their problems either.

    1) It adds an extra round
    2) where A beat B in the main tourney and the tie-breaker indicates A should win what do you do when B beats A in the playoff? It seems just as unfair as having B as the winner in the main tournament on tie-breaker.
     
  13. Glumanda

    Glumanda Rulings Compendium Translator

    Can someone explain, how the tiebreaker is calculated in the TMS software? I tested it including ties and drops and couldn´t figure out the calculation method.
     
  14. SteveP

    SteveP Active Member

    GreatFox, I TOTALLY disagree with a playoff when using software that calculates tiebreakers. In my example at the COlorado tournament, what happens if those top two players (who played each other in the swiss rounds) played in a playoff and the OTHER player won (the one who lost in the swiss rounds). PLUS, how do you report that ONE-GAME match if the tournament is sanctioned? You CAN'T just decide to have a game that impacts the results in a sanctioned tournament and NOT report that game!

    Now, if you're doing a tournament manually, on paper, and you don't know how to calculate tiebreakers, then I suppose you MUST do something to determine who gets the Medalion and VIP Package award. In that situation, I'd think head-to-head match results would determine the winner. If that STILL doesn't resolve it, then a playoff might be the only option (although a VERY BAD option IMO). Hopefully, that won't come up and TOs will be able to use tiebreakers if needed to determine the winner.

    One final idea. At the Lakewood tournament, we ran minimum+1 rounds as was suggested for last year's SBZs. Maybe that caused the problem of using the tiebreaker to determine the Champ. If we'd gone one less round, maybe we'd have had an outright winner based on game points. Nevertheless, when you're running one, big age-modified tournament, and you must pick your age-group champs from the final standings, you could STILL have potential tiebreaker champion issues.

    BTW, we had decided before printing out the final standings that we'd need to run a playoff game if the top 2 were also tied on tiebreaker points (I highly doubt PUI would've sent an addition Medalion in support of dual champs). Luckily, the tiebreaker worked.
     
    Last edited: Feb 2, 2004
  15. JohnnyBlaze

    JohnnyBlaze New Member

    Steve - Based on the above wording, I thought that the scenario was that they both had 15 points and then played against each other. Then the winner of that match was given 2nd place due to % points.

    Now that I am reading your other posts I get the true scenrio.

    So in this instance it would of been ok to run a playoff?
     
  16. DaytonGymLeader

    DaytonGymLeader New Member

    Actually, I haven't been able to figure out HOW TMS does it's Tie Breaker Points. Several emails to PUI, Jimmer, Dave, and the writers of the software have yielded no answers, mainly beacuse I haven't gotten any response. I guess I need to get back to the drawing board and do some more checking.

    Believe it or not, the way that we used to do it under DCI was Match Points, Tie Breakers, First Initial of the Player's Last Name. Now how fair is it to lose a T8 spot when you're in 9th place, tied with 8th in both Match Points and Tie Breakers with a last name starts with Z?

    If something like that comes up, I would lean more towards match play in a playoff for those two folks only. However, this is the system we've been left to deal with. Until the "correct" answer comes down on high, we, as TOs need to explain how these situations will play out before the events start.
     
  17. ScythKing

    ScythKing Member

    FWIW - I always run a top 8. Just for this reason - it's just too dang hard explaining tiebreakers to folks that feel their performance was better.
     
  18. yoshi1001

    yoshi1001 New Member

    The problem with all this is "better" is a relative term. The only way to truly determine who is better is to play hundreds of matches. In an effert to be more reasonable, people have invented tiebreakers to predict what the outcomes would be.

    But you all knew that.

    Anyway, I would say head-to-head should come first. I mean, the way the NFL does it can't be all bad (as long as we don't go to instant replay).
     
  19. NoPoke

    NoPoke New Member

    Yoshi its not about trying to determine who is better.

    Tie breakers aren't about trying to predict the outcome of hundreds of games

    The DCI decided not to use the result from the head-to-head in determining the tiebreaker. They know more about tcgs than the NFL!

    Since you don't seem to be clear on the purpose of a tiebreaker .Here's a little snippet from Reporter....
    The last part is key to SteveP's original question.

    If you run a swiss tournament under reporter than reporter will use tiebreakers to seperate the top two players. TMS does the same. Both pieces of software WILL use tiebreakers to decide who wins. I'm not aware that a TO has discretion to overrule the software and reverse the final standing between the number 1 and 2 players. So SteveP has no option but to say that the player who won WON. Now he could just shrug his shoulders and declare that life is tough, but instead he choose to highlight the issue. Its hard not to feel that the 2nd placed player got a raw deal.

    I'd really like to see a response from PUI on this.
     
    Last edited: Feb 2, 2004
  20. GreatFox

    GreatFox New Member

    Oh... I see. I first thought that both players had tied Point and MatchWin% wise. If they both had the same Points and MatchWin%(tiebraker) then you would have to do a sudden death or full match betwwen the two (your choice).

    I only mentioned using Sudden Death because that happened at our league once. We cut to top 8, however; both 8th and 9th place had the same score and MatchWin%. So who would countinue to top 8. In this case, a sudden death match was made between the two. The victor would continue on. Of course, since they were the only ones playing and it was only used to determain who continued, that match was not reported as it was not really nessacery to report.

    But in your situation above... you have no choice but to go by what TMS displayed.
     
Thread Status:
Not open for further replies.

Share This Page