Sometimes statistical analysis isn’t just about developing new, complicated formulas to quantify performance; it can also be about interpreting existing data differently. An example of this might be how we interpret power play success rate.
Currently, power play success is calculated by simply dividing the number of PPG scored by the number of power play opportunities. Using this calculation, the NHL’s more successful teams in power play success rates converted at roughly 19% and up last season. However, there are inherent flaws in using this simple formula to calculate power play success rate.
First, not all power plays are created equally. For example; teams generally aren’t successful when earning a PP near the end of a period. Teams generally lose any momentum generated during the first part of a PP when it is interrupted by the end of the period. Then they have to start the next period’s already shortened advantage by taking the faceoff at center ice rather than in the offensive zone.
Another reason that can adversely affect PP conversion percentage while not truly representing how well the PP is performing is that teams don’t always get the full term of their PP. In fact, using some numbers gleaned from ESPN’s in-game play-by-play reports, 43 of 107 power play opportunities that the Atlantic Division teams have enjoyed to date (40.2%), didn’t last the full terms for different reasons. So why is it that not all power plays last the full duration?
One reason is overlapping penalties. For example; if Team A goes on a 2 minute PP at the 5 minute mark of the 1st period but commits a 2 minute penalty of their own at the 6 minute mark, Team A only receives 1 minute of actual power play time. On the flip side, Team B also gets a 1 minute power play when their original infraction expires following 1 minute of 4-on-4 play. If neither team records a goal on their 1 minute power plays, then both teams sit at 0 – 1 (0%) on the PP.
A second reason PP opportunities get cut short are when a PP goal is scored. Is it perhaps more impressive when a team converts a PP goal just seconds into the advantage rather than scoring with only a few seconds remaining on the penalty? Maybe not in an individual instance but if a team can score PP goals in less time than another team then maybe that can lead to more on-ice success.
Another flaw in only utilizing success rate percentage is in the fundamental difference between 2 minute and 5 minute power plays. With a 2 minute power play, the man advantage concludes either when the team on the PP scores or at the end of two minutes; assuming no other penalties occur. On a 5 minute man advantage, the team on the PP can score as many goals as they can during that time. In theory, this team can finish with 2 or 3 goals on one PP. Statistically, how can that work?
Simply saying a team converted on 1 – 5 PP attempts (20%) doesn’t necessarily tell the whole story. What if that team had 2 full 5 minute advantages, 2 full 2 minute advantages and scored 1:30 into their 5th PP, also a 2 minute advantage? A 20% success rate is pretty good but does it tell the whole story? One PP goal scored when a team has 15:30 of PP time may not compare so favorably against a team that can score three PP goals with that same amount of PP time. It is often referred to as “power play efficiency” after all. Scoring more goals than your opponent is the object of the game isn’t it? Three is greater than one and scoring three in the same amount of time as another team scores one is the definition of more efficiency.
Now please realize that this isn’t the first time this has been brought up but perhaps the better way to gauge the success rate of PP’s is to calculate how many minutes on the PP it takes for a team to score a PPG. Using ESPN’s game play-by-play reports again, I’ve taken the opportunity to do those calculations for the current season for each of the teams in the Atlantic Division. Before we see those results, let’s see where each of the teams rate under the traditional means of PP calculation through October the twelfths games:
Team PPG PPO PP%
Flyers 9 31 29%
NYR 7 31 22.6%
Pens 5 23 21.7%
Devils 5 24 20.8%
NYI 4 20 20%
By contrast, here is how the Atlantic Division teams rank by virtue of PPG-per-minute:
Team PPG/PP Minutes
Flyers 1 – 5:15
Pens 1 – 6:24
NYR 1 – 6:24
Devils 1 – 7:57
NYI 1 – 9:06
To put these numbers in some perspective, I factored simply using minutes on the power play. I didn’t differentiate between 1 or 2 man advantages. Something else to consider; the Penguins, before last Saturday night’s contest against Toronto, had scored 1 PPG for every 10:02 of PP time. After scoring 3 PPG in just 5:57 minutes of PP time, their rating went all the way to 1 PPG for every 5:12 of PP time; a dramatic improvement to say the least. Of course that number dropped 1:12 per PPG after an 0 – 6:00 showing against the Senators on Monday night.
These types of jumps are to be expected since most of the teams in the Atlantic have only played a handful of games. One great or terrible PP performance can inflate or deflate a team’s performance significantly.
Despite these factors, both of these rankings are very similar whether using straight conversion percentage or PPG-per-PP minute to calculate success. The only difference is that the Rangers and Penguins flip-flopped spots. The hypothesis here is that we will see some fluctuations as the season goes on; perhaps even some big-time anomalies too.
I will continue to maintain the ratings for the Atlantic Division while also attempting to go back and grab the stats for the rest of the NHL. It will be interesting to see how the numbers compare at the conclusion of the regular season.
If anyone wants to email me directly to comment or ask questions here at Inside Hockey, I can be reached at: firstname.lastname@example.org. Thanks for reading.