Day 2 Statistical Preview: Ironside v. Revolver

July 5, 2013 by Sean Childers and Jeremy Weiss in News with 0 comments

Yes, it’s early.  But with a currently 3-0 Revolver looking strong in Day 1 action, the winner of their Friday’s 2v3 against Ironside (3-0) will have the inside track to play in the Sunday finals.  Ultiworld and USA Ultimate are experimenting with live tracking of both teams using the Ultiapps statistical tracking system.  Though we only have a few hundred plays on each team – too small for rigorous or perfect analysis – there are some statistical storylines to keep an eye on during our live stream.

Each team’s statistical notes from day 1 are summarized, and then we present a ranked table of our newest metric – Estimated Contribution.

Revolver Day 1: 

Josh Wiseman and Cassidy Rasmussen led the vertical attack for Revolver – each contributed about 30 yards (combined throwing and receiving) per each point when they played. But it was another Revolver stud, Mac Taylor, that most controlled the disc, sporting a usage rate near 25% (touching the disc on nearly one-fourth of his team’s throws while on the field) and assisting on over 33% of the points in which he played.  Devon Anderson (8 goals) and Beau Kittredge (6 goals) were the heaviest goal scorers.

From an efficiency standpoint, most Revolver players completed over 90% of their throws. Rasmussen is probably the hottest offensive player: He was incredibly involved, racked up tons of yardage, and rarely threw it away.

Spatially, Revolver worked the disc especially strong along the flick-sideline – perhaps unsurprising given Revolver’s strong throwers and athletes.  But it will be interesting to see if Ironside can play more effective force-side defense than Revolver’s day one competition, or if they instead try to throw junky or middle forces.

Ironside Day 1:

Looking through the Ironside day 1 statistics, it’s tough to spot any weaknesses.  Some of their better-known players (George Stubbs, Danny Clark, Will Neff) were average on some of our metrics – but more because other players like Josh Markette and Adrian Banerji dominated downfield yardage.  One difference between the Ironside day 1 offense and Revolver’s was the relatively even level of disc sharing amongst the team, with only a couple players taking more than 20% of the team’s touches when they were on the field.  Matthew McDonnell and Alex Kapinos were a bit more involved in scoring goals, but the overall distribution of endzone involvement was also tighter than Revolver’s.  Efficiency-wise, our yardage-per-turnover metric liked Banerji, Markette, and Christian Foster the most.

The estimated contribution metric (a somewhat complicated process) looks at each team’s spatial probability (what is the team’s probability of winning the point from a given location on the field) and estimates how much each player’s on-disc involvement (throws, catches, and blocks) affected the team’s probability of scoring the point, conditioned on how much each player played.   There are at least two interesting things to note from day 1’s tables: First, the tightness in the values in the Revolver table suggests that — while different players took on very different roles for the team today (see above) — a lot of them were about equally successful in terms of their contributions to the team’s winning probability.  Second, some of the Ironside players that weren’t identified in the above metrics are more highly favored in terms estimate contribution (Malecek, Stubbs).

**Note:  These numbers can’t be compared across teams.  The calculation of the score is based on each team’s individual spatial characteristics, so a good team will penalize an average player while a good player on a bad team will look great.  It’s best used as a ranking methodology intra-team.

CORRECTION: The Ironside table has been updated.

Revolver Expected Contribution Table:

 Ironside Expected Contribution Table: