Resolving issues with non-elim formats

Hi all,

I literally logged on just to record this idea because it's been sitting around for ages and I wanted to put it out there.

One of the big issues with non-elim formats (RR, Swiss, group stages, etc.) is redundant matches. These are matches where it doesn't matter what the outcome is for one or both players, it will not affect their overall tournament progression, whether they've clinched a spot in the next round, or whether there's no possibility that they'll win/advance. Such matches tend to be plagued with inactivity because they're somewhat meaningless. The question then becomes, how can a host create a format that is non-elim, while minimising these meaningless matches? To this, I'd like to propose an idea

• Non-elim stage, followed by a single elim stage

This gives everyone a decent chance to compete and position themselves for the elim stage. The elim stage is necessary because it's just a really efficient means of bringing the tournament to a conclusion without redundant matches

• Everyone advances to the elim stage

The most surefire way to create redundant matches in the non-elim stage is to make it so that only a portion of all players advance to the elim stage. Doing so means that there will inevitably be players that find themselves in lame duck scenarios, and they're a high risk to go afk

• The elim stage is seeded based on non-elim results AND higher seeds have an advantage in terms of asymmetric win requirements

For a long time I was thinking of ways to progress from non-elim to elim stages while keeping matches meaningful without resorting to bad solutions such as double elim- seeding alone just didn't seem enough to me. Then it occurred to me- make it so that players with a significant seeding advantage need to win fewer battles to win a round. An example would be if a 1st seed faced a 16th seed- the 1st seed might need to win only two battles to advance, but the 16th seed might need to win three.

Another way of describing it would be to say that the higher seed starts a Bo3 set already up 1-0

To me, this strikes an ideal balance. Almost every match in the non-elim stage becomes meaningful since getting a higher seed produces a significant advantage. Meanwhile, you're not really put out by this arrangement because you don't need to do significantly more prep, unlike say, double elim, which would require you to prepare twice as many teams. There are of course a few questions this raises:

• Does this seeding advantage persist through the entirety of the elim stage?

One thought that occurred to me is that you could make it so that this seeding advantage does not persist throughout the entire elim stage, such that by the time you reach say, semis, players swap back to being on a level playing field. I think either approach is fine (I personally like the idea), but if this is implemented, it obviously shouldn't be done so early that seeding advantage becomes meaningless. Consider reserving it for semis or finals, if you do this at all

• If a lower seed beats a higher seed, do they claim the higher seed?

To me the answer is no. Although this could be a really elegant method of curtailing the seeding advantage heading into the finals, I'm concerned that players that miss out on a top seed might tank to snag a low seed- if beating a single high seed means that they gain high seed advantage, then they might see that as preferable to risking say 3 rounds at a seeding disadvantage

• How much of a seed difference do you need for asymmetric requirements to kick in?

I think there's a fine balance to be struck so that seeding advantage is meaningful, but not so strict as to be unfair. For instance, it doesn't seem reasonable to me to give the 4th seed an advantage over the 5th seed, since they're already so close. By contrast, if you need to be say, eight seeds above an opponent to gain an advantage, then it's barely relevant. I think 2-4 is probably the ideal number.

• What about tied non-elim records?

This plays into the question above, but obviously, if there's a tie, there should be no seeding advantage, even if there are enough players tied that there would otherwise be a seeding advantage in place. For instance, if there are 2 players tied for first after the non-elim stage and 5 players with the next best record, the first 2 players would both be treated as #1 seeds, while the next 5 players would all be treated as #3 seeds, and players thereafter treated as #8 seeds. Alternatively, you could treat the first 2 players as #1 seeds, the next 5 players as #2 seeds, then the rest as #3 seeds and so on, but you can see that in this example seeding advantage probably wouldn't mean much, as there would be insufficient differences

Anyway, that's my idea, let me know your thoughts!

 

If the issue in solo tours is people not playing enough games then how about having a bracket to rank everybody: for instance both semis losers could play each others to determine who ranks 3rd and who ranks 4th. This also doesn't really pose any issue regarding the length of the tournament since all the brackets become independent quite fast. The other option would be adding a loser bracket but then you still run the risk of only playing two games in the whole tournament

 

Playing a minimum number of games was very much my priority here. One of my main complaints about single elim is that for half of all participants they're expected to play a single set and then sit out, which is probably the main thing I had in mind when making this.

There's no two ways around it- my proposal does extend the length of tours. Personally, I don't mind longer tournaments, and I was envisaging longer tournaments, possibly spanning the length of one of the old seasons. That said, even if you were to implement a group stage, that could work without extending the duration of the tour too much. With 4 player groups, you could potentially do the non-elim stage within 2 weeks, and you'd only be squeezing in one game in excess of the 1 game per week standard. That's perhaps not an ideal solution, but it is definitely workable imo

I also didn't consider team tours at all, so I'm not really expecting this to be applicable there

Another thought that just occurred to me is that I suspect there's a way to adapt to implement this in a way that resembles a "soft" double elim, which might be more time-efficient than my initial two stage approach. That said, I haven't thought through the specifics, so I don't have a concrete proposal