Tournament Tie Breakers: How My Previous Opponents Losing Their Round 9 Games Cost me a Top 32 Spot in the Clash

    • bobinchese wrote:

      I'm not saying tiebreakers mean nothing, even one match means strength of schedule is slightly higher. What I'm saying is that a system where we have a play in round for top 8 is a better system because a head to head match between two contenders for top 8 is far more significant than OMW%.
      the idea for the change is for less bad feelings correct? Well how would yoh feel if you were in 8th With tiebreakers and lose to that single elimination round to who finished 9th... regardless someone will have bad feelings .

      Now what happens when 8th place is X-2... how many people get to participate in this single elimination round? Because that would be a long addition to the tournament for people on contention
    • bobinchese wrote:

      I'm not saying tiebreakers mean nothing, even one match means strength of schedule is slightly higher. What I'm saying is that a system where we have a play in round for top 8 is a better system because a head to head match between two contenders for top 8 is far more significant than OMW%.
      the idea for the change is for less bad feelings correct? Well how would yoh feel if you were in 8th With tiebreakers and lose to that single elimination round to who finished 9th... regardless someone will have bad feelings .

      Now what happens when 8th place is X-2... how many people get to participate in this single elimination round? Because that would be a long addition to the tournament for people in contention
    • I can at least say that a playoff for top 8 would make the clash even worse for Europeans. It already lasted until 5am Monday morning... It means top 8 draft would have to be at a different time (as the other top 8's would have to wait), something that makes it a lot harder to make time for the tournament.
    • Norwind wrote:

      I can at least say that a playoff for top 8 would make the clash even worse for Europeans.
      If you read what I suggested - which is basically a play off for the last few top 8 spots instead of doing it based on breakers - it doesn't add time because you already took a round off the length of the event. So the event ends up being shorter for 95+% of people playing it and the same length for the dozen or some or did well.
    • One issue with having one fewer Swiss round and one more Single Elim round is that getting or not getting byes for that first Single Elim round could make some people a bit salty depending how the numbers work out; but probably no more than just missing out on tiebreaks already does.
    • AnomalyCobra wrote:

      Turning a 9 round swiss into 8...
      Which would leave 2 undefeated and roughly 7-11 1 loss players. And have the bottom few of them play single elimination for the last spots of top 8?
      Basically yea. That is essentially what happens the last round of a lot of swiss events anyways. This way the event is just shorter for most people while still having a big event to play in, with no one missing a top cut on breakers.

      The post was edited 1 time, last by JeffHoogland ().

    • i don't mind this in theory but a couple of issues..

      First the game has to recognize how many single elimination games are needed for a clear top 8. Like if there are 1 loss players up to 12th place, that means there would need to be 4 single elimination matches ranging from 5th place to 12th place. Then in a following tournament there might only be enough 1 loss players to fill 9th place, which would mean only 8th and 9th need to play a game. I'm not sure how easy that would be to code.

      Second is this, for the people that lost the tiebreaker single elimination games, do they finish ahead of the 6-2 players since they end at 7-2?
    • Eraia wrote:

      Skirovik wrote:

      THIS IS NOT RNG.
      How is it not RNG? I am assigned opponents at random and THEY are assigned their opponents at random, thus the strength of those opponents is rng. Pure and simple. 100%.
      Hex tournaments are actually what is known as "Swiss" style - this means that after the first round, winners are paired up against winners, and losers against losers (with occasionally someone being paired up or down a tier if the numbers don't work out nicely). This means that if you keep winning, you'll end up playing against other people who also won (which will thus improve your T1 tiebreaks).
    • Showsni wrote:

      This means that if you keep winning, you'll end up playing against other people who also won (which will thus improve your T1 tiebreaks)
      Except... tiebreakers are always against people who won the same number of rounds as you(or else it wouldn't be a tie at all), so would've had the same 'improvement' to their opponents. Which again means it all comes down to rng.
      Gamer. Streamer. Photographer. Writer. Anime Lover. Possessor of Stuffed Animals.

      Also... I'm terrible at this game.
    • Eraia wrote:

      Except... tiebreakers are always against people who won the same number of rounds as you(or else it wouldn't be a tie at all), so would've had the same 'improvement' to their opponents. Which again means it all comes down to rng.
      Not quite. Among people who go x-2, if you lost two matches early and win the rest, you will have played against people who are x-2 when they play you. If you loose two matches at the end, you will have played against other people who are x-0 when they play you. So on average depending on what happens after they play you, the second person might be slightly more likely to have better tie-breakers.
    • And downmatching and upmatching also happens while obviously only to resolve uneven results its probably a pain for both parties involved like for me in clash nr 2 when i was downmatched against a deck that was inferior and i just feel bad because i knew it would cost me tiebreakers either way and wasn't a very fun or fair match.
    • True downmatching and upmatching is really also making it not really fair and I don't get it why it's even possible, you should face off someone with the same ranking range I always got matched up against way stronger opponents. Like I played Havoc last round ruined his chance to get 32th spot and had no chance at all to win.

      The post was edited 1 time, last by Treasure ().

    • Treasure wrote:

      True downmatching and upmatching is really also making it not really fair and I don't get it why it's even possible, you should face off someone with the same ranking range I always got matched up against way stronger opponents. Like I played Havoc last round ruined his chance to get 32th spot and had no chance at all to win.
      there aren't always an even amount of players with a ranking... unless uit force the tournament to start with a specific just of players there will always be situations where three are an odd number of players at a ranking leading to down matching
    • Why keep ignoring the fact that matches against easier opponents are matches you are more likely to win? The guy with the weaker tiebreakers has a result nearer the top of their range than the one with stronger tiebreakers: 1% is still a small fraction of the margin of variance of course

      Suppose every player in the tournament had an exact % score of how good they were.
      Player A has better tiebreakers than player B (on same score).

      The tournament result tells us that A has a score somewhere in the range of 53%-67%* and player B is in the ranger 49%-65%.
      Player B might be better but that's not the way to bet -it's not a random choice (let alone one of equal probability).

      If they had swapped opponents it is more likely (compared to the chance for B) that A would have won an extra match and more likely that B would have lost one (thus moving them out of tiebreaker territory)


      * these numbers are totally arbitrary to make my point clearer and I have made no kind of effort to calculate what kind of range and overlap is appropriate for the variance inherent in the game mitigated by the number of rounds but it is clear that the closer the tiebreaker the greater the overlap and that the player with the higher tiebreaker will always have the higher range.

      I'd actually expect the ranges to be much closer to each other than my example and for there to be considerable overlap across different win-records: 32nd place with poor luck can very easily be a better player than 7th place with good luck in a single tournament.

      The more tournaments you play, the closer your results will be to those deserved by your ability: current tiebreaker system does get you there faster than straight RNG -if it doesn't get you there much faster it's worth trying to improve on but throwing it out for RNG is definitely a step back in that regard
      -if you want to justify straight RNG for the purpose of predictability I think that would have limited effect but it is reasonable to subjectively consider that a trade worth making.