What Score-Adjustment Isn't Telling You by @BrianK_PI - PensInitiative | Pittsburgh Penguins Blog | Rumors | News

The Latest

Post Top Ad

Tuesday, April 7, 2015

What Score-Adjustment Isn't Telling You by @BrianK_PI

All stats from War on Ice and at even strength. Stats are current through April 5th.

As "fancy" stats, "enhanced" stats, "analytics" - whatever you choose to call the more in depth look at hockey data - have progressed, they've helped us gain a better understanding of what is happening on the ice. It's a simple premise - shots for are good, and shots against are bad. You can score when you attempt shots, and you can get scored upon when the other team attempts shots. An elite possession team can have possession rates 5-10 percentage points better than their opponents; no matter how good a team is, they're never going to consistently have a shooting percentage 5-10 points better than their opponents. The easiest way to score more goals is to generate more shots.

As time has progressed, the stats have progressed to better ways of looking at the data with better correlations. Corsi, all shot attempts, turned into Fenwick, all unblocked shot attempts. An emphasis on score effects led to a closer look at the Corsi and Fenwick close stats, the Corsi/Fenwick events that take place with a tie score or within 1 before the third period. Of course, close possession rates ignore large portions of the on-ice action for stats needing large sample sizes, and score-adjusted possession rates proved to be a better way of assessing the data by taking all action into account.


According to the possession data at even strength, the 2014-15 Pittsburgh Penguins should be doing very well. They're 7th in both score-adjusted CF% (53.0%) and FF% (52.9%), though 10th in GF% (52.2%) and 13th in scoring chances for percentage (51.7%). Of course, over the past few months they've struggled to beat some of the worst teams in the league - they're 14-13-3 in their past 30 games despite a schedule that currently averaged 86.5 points over the course of the season. But they're fourth in score-adjusted CF% (54.0%) and third in score-adjusted FF% (54.0%) during the 2015 calendar year, along with a SCF% of 53.1% (6th). That hasn't translated to the scoresheet though as they're just 23rd in GF% at 48.3% and they've taken a hit in the standings. They've certainly had bad luck, but does the problem run deeper that that?


Above - Pens CF% vs League Average (minus Buffalo) with League-Wide Ranking
Below - Pens SCF% vs League Average (minus Buffalo) with League-Wide Ranking
While score-adjusted possession gives a team's cumulative ability across all scores, it doesn't break out how a team has done at each score differential. And taking a closer look at the Penguins' possession numbers reveals something surprising. Pittsburgh has been an elite team at 5 on 5 when tied, leading by 1, or leading by 2, but as soon as they lose the lead they turn into a completely different team, ranking in the bottom half of the league while trailing by 1 or 2. What further compounds the problem is that the Penguins are one of the least disciplined teams in the league. They have more difficulty driving possession while down, and the propensity at which they take penalties gives their opponents extra opportunities when tied or trailing to overcome Pittsburgh's elite possession, let alone further putting the Pens at a disadvantage when their opponents are in the lead. Most troubling of all is the fact that Pittsburgh is TWO entire penalties per 60 minutes above the league average while leading by a single goal. The stupid and careless penalties are a problem that have plagued this team before, and they're plaguing this team now. The problem could be even more severe if the Pens weren't also one of the best penalty killing teams in the league.

Pens Penalties per 60 Minutes vs League Average
The penalty issues and the possession difficulties when trailing are a horrible combination that leads to Pittsburgh not being down, but usually out, if they fall behind. They are 0-18-5 when trailing after two periods this season and are the only team who hasn't come back to win a game. Even the Buffalo Sabres have managed to win in that scenario and that franchise is actively trying to ice the worst team in recent memory. There's no reason why as strong of a possession team as the Penguins are when tied or leading should be that poor when trailing. The high rate at which the team takes penalties works to start undoing the good they are doing, but whether it's coaching putting them in a bad situation while they're losing, a poor team mindset when they fall behind on the scoreboard, a combination of the two or something else entirely, the Pens transform into a bad team when behind and it's killing them. They do have some serious depth issue, they are yet again seeing players out of the lineup, and they're certainly icing a less-talented roster at the bottom of the lineup then more balanced teams, but that's no excuse. Certainly not after showing what they're capable at different score differentials.

So yes, the Penguins have been a really good possession team this season, and on the whole their score adjusted numbers look great, but a closer look shows a team that's one part Jekyll, one part Hyde. The team has struggled against the better teams, and if they make the playoffs, they'll find themselves out in a hurry if they play like that if/when they find themselves down. The good news is that taking penalties is something that's entirely (mostly, if you believe if conspiracies) within their control, and a team that can play that successful of a possession game tied or ahead should be able to adapt. But something will need to change, or changes will certainly follow after the season.

No comments:

Post a Comment

Post Top Ad