### The Adjusted Power Play Percentages

A post on NBC Pro Hockey Talk reminded me what bothers me about how power play percentages are calculated.

The current system takes power play goals scored, divided by the opportunities. Makes sense, except when you take into account the fact that not all power play opportunities are a full two minutes. If a power play is interrupted by another penalty, it may be only a minute and a half, or only 10 seconds. Either way, they all still count as one power play, or one opportunity. To me, it doesn’t seem like the most accurate method of measuring a team’s power play effectiveness.

Power play percentages should be measured as an expression of goals per two minutes of power play time, not goals per power play opportunity, because not all power play opportunities are a full two minutes, so the percentages derived from them are unreliable.

NHL.com thankfully has a terrific statistical database. I was able to use the power play data for each team as a figure of time instead of opportunities. It also breaks the power plays up into the different scenarios: 5-4, 5-3, and 4-3, which I think is also relevant.

The chart below lists the adjusted data:

Note: Chart currently sorted by 5 on 4 adjusted percentages. Google is struggling with the ability to sort columns.

LEGEND: PP Opp= Number of Power Play Opportunities  PP Adj Time= Time on the Power Play, with seconds adjusted as a percentage of a minute (for the purpose of calculation)  PPG= Power Play Goals  STD %= Standard Power Play percentage, using the current method  PP% 2:Opps= Goals scoring efficiency per 2 minutes of Power Play time  Min/G= How much Power Play time is needed for a goal

A few things.

1. I had to express the time not in minutes and seconds, but as a figure of a ratio of minutes. Thus, New Jersey’s power play time of 406.50 minutes is the same as 406 min 30 seconds. This was done for more accurate calculations.
2. The colors indicate the different power play scenarios: Blue is overall, green is 5 on 4, orange is 5 on 3, and yellow is 4 on 3.
3. The first noticeable element is that the (what I’m referring to as) Adjusted Power Play Percentage is higher for every team. This is because as I mentioned, teams are currently credited for a power play regardless of duration. If a team had far more abbreviated power plays, their Adjusted Power Play Percentage will be much higher than their standard percentage.
4. There actually isn’t a huge difference in the rankings between the two methods, but there are a few. The best example of this discrepancy is with the New York Rangers and Philadelphia Flyers. Using the standard method, the Flyers had a power play percentage of 16.6% (I thought it was higher too). The Rangers’ was 16.9%, slightly higher. But looking at the Adjusted Power Play Percentage, the Flyers’ is 19.73% and the Rangers’ is 19.65%. This is because, while the Flyers had 295 power play opportunities to the Rangers’ 290, they actually had less minutes of power play time: 496.62 minutes (496:37) to 498.73 minutes (498:44). Thus, the Flyers were technically more efficient than the Rangers on the power play, even though the standard method of measurement (Goals/Opportunities) gives them a higher percentage.
5. The Minutes / Goals scored statistic is kind of fun. This is a measurement of how much power play time was needed for a team to score.
6. 5-3 percentage is incredible, and it points out the flaw (if you want to call it that) of this method of measurement. Technically, it is impossible for a team to have over 100% efficiency on the power play. But in the Adjusted method, it is not. This is because the teams that were over 100% essentially never (or rarely) needed the full two minutes to convert a goal on the 5-3 power play. But this also explains why 5-3 power play percentage using the standard means of measurement seems so low: because they are pretty much never a full two minutes. The adjusted method blatantly points out the teams that were efficient on 5-3 power plays.
7. There were far less 4-3 power plays than 5-3, thus, the data for them is even more shallow. For example, last season four teams failed to convert on any 4-3 opportunities. Thus, their time needed to score on a 4-3 power play was infinity.
8. The 5 on 4 Adjusted Percentage is the most important and telling statistic when measuring a team’s power play efficiency. This is the measurement of the most common power play and its efficiency per full two minutes. This method is more interesting than the standard here, if not simply because the range was greater: Vancouver’s 5-4 power play converted at 31% while Florida’s came in at just under 12%. The Canucks only needed 6.44 minutes (6:28) of 5-4 power play time to score. The Panthers needed 16.78 minutes (16:47). The Canucks would have almost been able to score three power play Goals before the Panthers were able to put in one.
9. The summer is too long. It always leads to something like this.
2 Responses to “The Adjusted Power Play Percentages”
1. Player-X says:

Interesting Stuff, seems like placings stay mostly the same as in Standard rankings, but one can more clearly see the discrepancies between team’s proficiencies.

Couple of questions:
(1) Is there some trick to having the chart sort by individual categories? I can’t get it to do so.
(2) How do the stats take into account that when a team scores a goal, they do not use the full two minutes? Is “actual” power-play-time-used included in the calculations? I guess that is the minutes per goal column, yes? Maybe a shortcut stat would be a percentage from time used/time allotted due to scoring.

I was surprised at the Kings rank in 5-3, 10th was way better than I would have guessed. Anecdotally, my impression was that they had been dreadful, when in fact the Kings were pretty good 5-3. Even more surprising was Vancouver’s near total failure, ranking only above Edmonton and scoring only one 5-3 goal last year. Talk about counter-intuitive…

It would also be interesting to see a breakdown of PP points by position on each team. Do the top teams correlate, as in percentages of points by defensemen, for example?

• crownedroyal says:

Dammit. I thought that link would take you to a sortable chart but I can’t get it to work. Sorry about that. If you’d like, I can email it to you as an attachment. If the email you registered with is accurate I can send it there.

This method kind of does that. If a team only takes 20 seconds of a power play to score, then will then have more “Goals per 2 minutes of power play time” which is what I refer to as the “Adjusted Power Play Percentage”.

I was also somewhat surprised at the Kings 5-3 efficiency. It definitely didn’t seem that great last season.

That is not a bad idea, but just off the cuff, I doubt there is a correlation between effective blue-liners and power play success, just thinking about Weber, Yandle, Chara, those teams are in the bottom 10. Of course it’s not always a bad thing, just not sure that there would be an obvious correlation.