Playoff Predictions: S2 Pre-Quarterfinals

As discussed on Episode 308 of the Parity Podcast, here are my picks for the first week of playoffs (listen to the show for my full bracket predictions). I'll be updating my picks each week as new results and stats come in.

Team names are accurate at the time of posting, but who the heck knows what they'll be by the time you read this. Enjoy!

---

[Pre-QF] Wu-hoo! (Heather) (25) vs. TooTrains TootFurious (Al) (21)
[Pre-QF] S2: E9 Bedlam in the Big Top (Rachel) 20 vs. Tied Pod Challenge (Jon) (21) *universe*
[Pre-QF] Live Free or Tie Hard (Travis) (29) vs. #TimesUp (Kindha) (18)
[Pre-QF] Shantay, You Stay (Laura) (16) vs. Pad Tie (Nat) (12)
[Pre-QF] Higher Seed (Bossy) (24) vs. Even (Adam) (17)

The higher seeds are the favourites in all the matchups this week except the 8 vs. 9 game. It's a toss-up but Jon's team has the slight edge.

Travis has the strongest offensive team heading into playoffs followed closely by Heather. I expect both teams to make semis. Laura has by far the strongest team defensively, which should take them deep into the playoffs, but their offensive conversion has only been average and poses a liability.

The biggest underdogs at this stage are Kindha and Adam whose rosters will need to overperform in the next few weeks to stay competitive.

---

For the methodology nerds: I generate an offensive and a defensive rating for each team using an average of their players' offensive/defensive ratings (50% weighting) and offensive/defensive efficiency (50% weighting). Full player ratings are available here. I divide each team's offensive rating by their opponent's defensive rating to generate a score multiplier (which I normalize based on the average league offensive/defensive differential) and then multiply that by the league average number of goals per game to predict how many goals each team will score in each game.

The biggest caveat to this methodology (other than the inherent limitations of the underlying player stats/ratings) is my assumption that every player will play an equal number of points in every game. Even one missing player can make a big difference since the margins between teams are relatively small.

For example, if Brian Perry didn't play for Heather tonight, that prediction changes from a 25-21 win over Al's team to a 24-24 toss-up. Players of Brian's calibre are typically replaced with subs. But if a team is missing its bottom players and doesn't replace them they can get significantly stronger. For example, if Al's bottom man and woman don't show up tonight, I predict a 23-23 game that could tip either way.