OWL by the numbers: New York Excelsior's troubles adapting to new meta

ESPN Stats and Information

The New York Excelsior ended the regular season on a bit of a down note. It finished with the third-best record in Stage 4 (7-3) but had as many losses as it had in Stages 1-3 combined (27-3). It made its fourth straight stage playoffs (no other team has more than two) but failed to win the stage playoffs for the first time since Stage 1 (losing to London in the Stage 1 finals).

Many believed New York's mediocrity in Stage 4 was due in part to Excelsior's already clinching its spot in the season playoffs and being more concerned about preparing for that than a "meaningless" stage match. However, after being swept 3-0 by the Philadelphia Fusion on Wednesday, one has to wonder what meta it was preparing for.

Single support with triple-tank?

Single support with triple-tank was played 1.8 percent of the time in Stages 1-3. In the playoffs that percentage has jumped to 30.8 percent, the highest pick rate of any team composition outside of 2-2-2 (two tanks, two supports, two DPS) in any stage this season.

When Excelsior faced the Fusion on Wednesday, it ran single support with triple-tank only 2.5 percent of the time (1 minute and 10 seconds). The playoff team with the next-lowest pick rate of that composition was the Los Angeles Valiant with 29.4 percent (who also lost its first quarterfinals match).

The composition with the most time played in the playoffs thus far has been Mercy, Roadhog, Orisa, D.Va, Widowmaker and Hanzo -- for a total of 1 hour and 49 minutes. New York accounted for none of that playtime.

New Hanzo

One reason New York did not play the above-mentioned composition is its overall lack of Hanzo play.

Hanzo's rework between Stage 4 and the playoffs has catapulted him into the meta like never before. During the regular season, Hanzo had a league-wide pick rate of 3.4 percent, ninth among DPS heroes. In the playoffs, that pick rate has gone all the way up to 49.3 percent, second among DPS heroes.

New York again decided to counter the meta by picking Hanzo 6.8 percent of the time (3 minutes and 7 seconds). What is even more surprising is that Excelsior DPS Hae-Seong "Libero" Kim played more Hanzo than anyone else during the regular season (2 hours and 2 minutes). The Excelsior instead opted to play Tracer 80.6 percent of the time, more than double the league average during the playoffs (29.4).

Confusing tank line

The Excelsior tank line on Wednesday was almost exclusively limited to dive tanks D.Va (100 percent pick rate) and Winston (95.4 percent pick rate). The 95.4 percent pick rate of Winston was greater than the London Spitfire (47.3 percent) and Fusion (39.8 percent) combined.

As a result, New York played only 4.6 percent of its time on Orisa and 2.5 percent on Roadhog. The league average on those heroes during the playoffs has been 40.6 percent and 36.2 percent, respectively.

New York's reason for not playing more Orisa and Roadhog is likely not a lack of options. During the regular season, main tank Dong-Gyu "Mano" Kim played the second-most Orisa (5 hours and 48 minutes) and off tank Tae-Hong "MekO" Kim played the second-most Roadhog (3 hours) in OWL.

New York seems to have the pieces in place to play to the meta, so why not do it?

Solving the meta

The easy solution when looking at the above information might be for New York to switch over to single-support with triple-tank. They could move Libero to Hanzo and have Mano and MekO play more Orisa and Roadhog. Done deal, right? Well this leaves NYXL with one glaring question -- what does league MVP Seong-Hyun "JJoNak" Bang play?

During the regular season, JJoNak played Zenyatta 93 percent of the time. In the playoffs, Zenyatta's overall pick rate has dropped to 48.3 percent, lowest in any stage this season and well below his regular-season average of 84.5. With Zenyatta no longer a must-pick on every map, where does this leave JJoNak?

The only non-support hero JJoNak played during the regular season for more than 30 minutes was Roadhog (33 minutes). If New York wants to play the meta it will likely need JJoNak to show he can flex onto Roadhog. If not, New York will simply have to win as it did on the previous metas, by playing around JJoNak's Zenyatta.