As always when the number of teams in the League changes, it requires a shifting in the way the divisions and postseason work. With an all time high of twenty teams slated to compete in the 2015 Season, this year will be no exception. Read on to find out how it will work!
With there being twenty teams, the two natural configurations that we considered were having two divisions of ten teams or four divisions of five teams. The most challenging issue with this though was that there are eleven teams in the Eastern Time Zone (EST), five in the Central Time Zone (CST), and four in the Pacific Time Zone (PST). We also had a major goal of making the number of EST vs PST matches as low as possible; the three hour time difference had consistently been causing problems for teams in recent seasons, with such matches either starting too early for those in the West or ending too late for those in the East, something to which there was no real solution short of making the time control of such matches undesirably short.
Given these issues, there really was no way to divide the twenty teams into either two or four divisions without there being a clear problem in this regard. After considering options and talking with teams, we decided to go with the two division format, placing all the EST teams, other than Boston, in the Eastern Division with all the remaining teams in the Western Division. While having Boston in the “West” was not the most desirable option from a geographical or historical standpoint, with their team actually semi-preferring (due to players working late) the later starts that come with being in the Western Division, this option was acceptable to them and, as such, this course of action definitely seemed to be the best thing in terms of making things work logistically for all the teams collectively.
The way the schedule will work is that each team will play each team from their division once and one of the teams in their division twice (there will be no inter-division play) - teams’ repeat opponent being based on preseason seeding (1 vs 2, 3 vs 4, 5 vs 6, 7 vs 8, and 9 vs 10) which can be found here.
While we generally have preferred in the past to have some inter-division play to give teams a wider variety of opposition, with the League having grown to its current size, there will be almost no impact from teams only playing within their division, as they will only play one team more than once. And doing it in this way will completely eliminate there being any EST vs PST matches prior to the Championship (other than those involving Boston) which, once more, is what seems to be in the League’s best interest.
The playoff format will be very similar to the NFL’s current system, with the top six teams in each division qualifying for the postseason. A high number to be sure, but beyond just qualifying for it, seeding will be extremely important as any playoff match prior to the Championship will give the higher seeded team draw odds, and there are also byes for some of the higher seeded teams.
To be more specific, in the Wildcard Round the third seed will play the sixth seed while the fourth seed will play the fifth seed (in both divisions) with the top two seeds in each division getting byes. Then, in the Quarterfinals, the first seed will take on the lower seeded Wildcard winner while the second seed will play the higher seeded Wildcard winner (again in both divisions). After that, in the Semifinals, the two Quarterfinal winners from each division will meet, and then the Championship will consist of the two Semifinal winners.
As mentioned, seeding will be very important as any playoff match prior to the Championship will work as follows: if the higher seed in a match is seeded two or more higher than the lower seed, they will receive draw odds and color choice for the match in question. If the difference in seeding is only one, then the higher seed picks between draw odds and color choice with the lower seed getting the other advantage.
Finally, in the Championship, there will as usual be no draw odds, with the team with the better collective record to that point (playoffs included) will get color choice for the match, and if the match ends up tied, it will be decided by the gauntlet blitz tiebreaker.
With the new record size of the USCL, we look forward to 2015 Season to be the biggest one yet, as we are thrilled to continually add new players and areas to the experience in the hopes of continual growth and more and more of America’s top players able to compete.
Hm, would five divisions of four teams make sense? It fits excellently into the 10-week regular season: each week, a different pair of divisions “play against” each other, while the others plays amongst themselves; in this way, each team plays their division opponents twice and one team from each of the other four divisions. Playoffs can work by taking the top two teams from each division and the top third place team, with the division winners getting a first-round bye and normal single-elimination stuff happening with the resulting 8 teams. The four PST teams can be in their own division, so each team will only have EST v PST game, but the main disadvantage to this structure is that those games are spread throughout the season rather than being concentrated in a single week. Advantages include more interaction between divisions (instead of just the championship) and a somewhat more natural selection of opponents however. Also this is very similar to how NFL runs (though obviously 24 teams is more convenient for this as you can have 2 conferences of three 4-team divisions). Maybe something to consider for next year.
An interesting suggestion which I did not consider - as mentioned we only really thought about two or four divisions, not five. Like most things, there pretty much always will be some advantages/disadvantages to each configuration. Under yours, teams would play three teams twice rather than just one team twice - is this a good thing? Not sure, very much a matter of opinion, some would say playing your own division mates more is better for determining a fair final order in each division, others would say a variety of opposition, if fairly balanced from team to team, is better.
Under your system it seems like 11 teams would make the playoffs which I kind of like. 12/20 is a bit much, ideally closer to 50% is better I think which is main reason why we made seeding very important by having draw odds all the way until the Championship - since it’s much easier than usual to make the postseason, wanted to make sure the difference between each seed was reasonably significant. That said, the “top third place team” making playoffs definitely seems shaky in my view. After all, if teams play six of ten matches vs their own division, clearly two third place teams play completely different schedules - certainly might be far easier to get a good record as third place in one division compared to another (if the fourth place team in your division completely tanks or if in some other division the top team totally dominates). In the current system teams play an identical set of opponents as those they are competing with for playoff spots, other than one match, the one repeat opponent which I think from a fairness standpoint is better. Add that to the fact that a certain division will have three of four teams qualifying for postseason - just seems like too much. I personally have generally favored larger divisions since it tends to make the playoff races more exciting and also because in a situation where two of four teams in a division qualify for the playoffs, one team tanking again has a huge impact on that division - naturally one team completely dominating/completely tanking has much less effect on a ten team division than a four team division.
No idea if we made best choice for how to break things down - as always it comes down to judgement calls + opinions how things should be. Definitely nothing is set in stone, if we keep this same set of teams next season, we might well reassess based on way things went this year to see if something else might be better. If we expand to 22 teams that will definitely be a challenge of how to break things down - two divisions of 11 teams is a problem since whenever there is an odd number of teams in a division, every week at least one team in a division must play outside the division which will make scheduling quite tough. It will end up that nearly all teams will have to play at least one match outside their division meaning they would not play every team in their own division. And this is definitely not ideal as in any normal system you play vs every team in your own division at least as much as you play any team outside of it. If we can go to 24 teams though, that would likely be quite ideal as it would give us many options. Time will tell!
I think that there should be a blitz playoff with the winner deciding which configuration to go with. After all everything else in chess seems to be decided this way.
I know this perhaps biased by our team last year, but I feel as though the “gauntlet” blitz tiebreaker effectively gives draw odds to the team with the best player. This is a bit unfair, since often times the more balanced team has been the more successful one as well. I think the blitz tiebreaker should simply be a “match” of itself: Each Board 4 would play, Board 3, etc. If this too was a draw, the top boards would then play each other until there’s a knockout (switching colors with each draw).
Yes, this is certainly a very valid concern. The blitz tiebreaker, while very exciting for the fans, is definitely somewhat skewed in this regard. Hard to say exactly how much so, but I think saying that it gives draw odds to the team with the best player is definitely overstating things. After all if you go back to the early years of the USCL when we had the tiebreaker four years in a row in the Championship Match, in the majority of the years the team that was “favored” for that very reason did not end up winning.
My general view is more that in any regular USCL match, each board contributes an equal 25% to the overall result while in the blitz tiebreaker these %s are clearly completely skewed - no idea what they actually become, but obviously Board 1s % is far higher and Board 4s is far lower. Is that wrong? Depends who you ask I guess, but I personally would not be at all opposed to having a new system in which the % each player contributes during the tiebreaker is far closer to even from player to player than what I currently think it is, even if it’s not exactly equal.
That said, this is definitely the kind of thing where to properly assess its merit, it would probably need many, many years of data to really gauge which obviously we don’t really have. Definitely open to suggestions in general if people have them - you are certainly not the first person to make this point, but Greg and I sort of decided that it might be hasty to change things based on just one year’s result (especially again since the prior years did not suggest this), particularly since this upcoming season, the blitz tiebreaker will only be potentially used in the Championship, not in the Semifinals or anywhere else like last season.
Mr. Katz idea of 5 divisions with 4 teams is more akin to NFL model (8 divisions of 4 teams each) rather than NHL model (4 divisions of 7 to 8 teams each). NFL model seems best for playoffs among 4 major sports as it rewards good teams with bye and forces lower seeds on road (similar to draw odds in chess). A model similar to NFL makes the regular season matter, unlike in NBA or NHL. 12 of 20 teams in playoffs is 60% qualification and dilutes regular season as majority make it. Also, with new system of 2 divisions, it forces Boston team in West which can be 3 time zones away for some matches. Very unlucky for them and could damage morale/participation. I like Mr. Katz idea.
A division structure of the following under a 5 division 4 team model could look like this:
1) Northeast Division - Boston, New England, Connecticut, Manhattan
2) Mid Atlantic Division - New York, New Jersey, Philadelphia, Baltimore
3) South - Carolina, Atlanta, Miami, St. Louis
4) Southwest (with apologies to Minnesota) - Minnesota, Rio Grande, Dallas, Lubbock
5) Pacific - Arizona, Seattle, San Diego, San Francisco.
The set up would keep regional rivalries and ensure at least one top historic team in each division. No perfect solution exists but only teams in different time zones than other division teammates would be St. Louis (Central) and Arizona (Mountain).
For blitz tiebreak, perhaps having a best of 5 game playoff match between 1 vs. 1, 2 vs. 2, 3 vs. 3, 4 vs. 4. The board 1 has 40% importance, board 2 30% importance and so on. The danger under current scenario is you could have boards 2 to 4 all lose, only to have super GM board 1 save day and beat 4 players in a row.
These are all lively points and enjoy following USCL. Keep up the good work Greg and Arun!