tangentially significant conundrums

Considering OWL Season 2 Formats

possible adjustments after the inaugural season

September 8, 2018 tsc

With Season 1 of the Overwatch League in the books and the expansion cities for Season 2 being announced, no time seems better than now to begin speculating on what Season 2 might look like.

Divisions and Teams

Status Quo

Keeping the divisions the same as today would look like the following:

It’s at this point that the divisions begin to look a bit large, though they would serve as a natural formation point for conferences in the future.

Splitting the Divisions

The Atlantic and Pacific divisions could also be converted to conferences immediately, with each having two divisions of five teams each. One possible split would look as follows:

These divisions feature some geographical absurdities, like Shanghai and Guangzhou not being in the same division despite being very close, and Philadelphia being much closer to most of the North Atlantic teams than any of the South Atlantic teams, but it’s par for the course given that Season 1 featured the ridiculousness of Dallas and Houston being in different divisions despite being closer to each other than to any other team.

The divisions also end up being fairly well-balanced, with each having three teams from Season 1 and two expansion teams.

Regular Season

For Season 1, each of the 12 teams played 40 sets for a total of 240 sets played during the regular season. There’s a myriad of changes that could be made to the regular season, so I will focus on what I consider some of the most likely possibilities.

Simple Round-Robin

The simplest season would have each team play the 19 other teams once. However, I don’t expect this is very likely since it would represent a fairly significant decrease in total sets played during the regular season.

Double round-robin, where each team plays 38 sets total, is also an option but would represent a very significant increase in the number of sets played during the season.

Intra-division Focus

Similar to traditional sports leagues, Season 2 could also have teams play the teams in their division more frequently.

With the Atlantic and Pacific divisions being retained, this would result in each team playing the nine other teams in its division twice and the ten teams in the other division once for a total of 28 sets for the season. This seems to be the most likely possibility as it doesn’t significantly shrink or expand the season (Season 1 having 240 total regular sets and this scenario adds up to 280 sets).

With the Atlantic and Pacific divisions being split into conferences with two divisions each, there are several options. One option would have each team would play the four other teams in its division twice and each team in its conference’s other division once for a base number of 13 sets. Each team would also play an odd number of teams from the other conference to bring the total up to an even number, most likely five or nine for totals of 18 or 22 sets, respectively. The main consideration with this scenario is ensuring that the selection of teams played from the other conference does not significantly impact strength of schedule.

Another option with two conferences is having teams play other teams in the same division three times, teams in the other division of the same conference twice, and teams in the other conference once, for a total of 32 sets. While each team would still be playing less sets compared to the previous season, the league overall would have a third more sets in the regular season, which would present significant logistical challenges given no other changes.

Postseason

With 20 teams in the league, it’s likely that eight teams will make playoffs, as keeping the playoff size at six would make playoffs seem exceptionally small and having 12 teams makes it likely that teams with a losing record will make the playoffs.

A ten-team playoff is another possibility, albeit less likely. It would require a play-in series for the bottom four teams to narrow things down to the top eight, so they are similar to the eight-team formats I will focus on.

Atlantic vs. Pacific

This format is most similar to most traditional sports with two conferences: the top four teams in each conference would battle in a bracket to determine the conference champion, and then the two champions would face off in the final for the title of league champion.

The main downside of such a format is that it presumes that both conferences are roughly the same strength—in a situation where one conference is much stronger than another, there’s the potential for stronger teams to miss the playoff in favor of weaker teams from the other conference, as well as the league final to be a much lesser match than one of the conference playoff games.

Full League Showdown

Instead of having conference playoffs leading to a league final, there could also be a straight eight-team playoff, where the best eight teams are entered regardless of conference, which would produce a more competitive playoff overall.

To ensure the divisions remain relevant, the top seeds could be reserved for the top team in each division or conference, as they were in Season 1. In addition, one of the best aspects of later stage title match format could be taken, in which the conference leaders would be allowed to select their first-round opponent from among the other teams qualifying for the playoffs. This may prove to be a more valuable benefit than a first-round bye was in Season 1.

Single or Double Elimination

One potential prospect is changing the structure of the playoffs from single-elimination to double-elimination. The main advantages of double-elimination tournaments are that they add more legitimacy to the final results (as teams can lose once to get eliminated because of a fluke, but are very unlikely to lose twice to get eliminated while still being the best team), as well as allowing a greater chance for teams to adapt during the course of the bracket (one of the most famous examples being the Evil Geniuses victory at The International 2015 against CDEC despite being swept in the upper bracket final by the same team).

Certainly, in Season 1, it is interesting to consider whether the New York Excelsior (by far the most dominant team in the league over most of the season) and Los Angeles Valiant (hot coming into the playoffs with the Stage 4 title win over New York) could have come back after their crushing losses and adapted sufficiently to make a grand final, or whether they truly could not adapt to the playoff meta and would have been soundly defeated again. (Of course, there is the argument to be made that if the teams could adapt, they would have done so in the time between the first series and the second series.)

The main downside of a double-elimination bracket is its complexity; most sports use single-elimination brackets which are conceptually easier to understand. It would also take significantly more time to run a double-elimination bracket, as almost double the number of games have to be played due to the addition of the lower bracket. Finally, and perhaps most fatally, the format of the grand final of a double-elimination bracket would be quite difficult to pin down; if the earlier bracket games were best-of-three series of best-of-five sets, the grand final being the same would render the upper bracket team at a disadvantage (where it could lose only one best-of-three series and still not win the tournament), whereas being two series would make it exceedingly difficult to schedule and play due to potential significant variances in length (between two and six best-of-three series depending on what happens). The preferred option in such cases, a best-of-five series, would also be quite lengthy and difficult to schedule.

For these reasons, it’s rather unlikely that a double-elimination playoff would ever come to fruition, but if the difficulties were possible to overcome, it would be a welcomed improvement.

Set Format

The format of each set was very standard for Season 1; in the regular season, a minimum of four games played with one on each standard game mode (Control, Escort, Assault, and Hybrid), with an additional tiebreaker map if necessary. The playoffs were not much different, with all four of the standard game modes represented in the first four maps of the best-of-five set, and the set ending only when a team had won three maps (with unlimited tiebreaker maps until then).

The maps played were predetermined. Thus, given the rich variety of maps available in Overwatch, an easy way to add strategic depth is to allow teams to select the maps for the set (via some standard pick-ban process as demonstrated in other esports), either some time prior to the match to encourage preparation on the specific maps selected or on the day of the match to allow for the possibility of surprise picks.

In addition, regular season sets could be reduced in length, allowing for more of them to be played and giving additional meaning to playoff series. The simplest would be a best-of-three set with each map being of a different game mode. The game modes for each set could be set up to give home-team advantage, where the home team either bans one game mode from being played or selects two of the three game modes to be played (with the away team selecting the other), giving concrete meaning to the home-away designation.

Final Thoughts

With the league expanding so much for Season 2, there’s potential for significant changes to be made for the season format. If anything, Blizzard has proven to be adaptable with the Overwatch League, so fans should a greatly enhanced experience in the next season.

In any case, it will be interesting to see what exactly comes back from the drawing board for Season 2.