I posted an STR analysis about half way through the season which can be found here. STR, or Special Team Rank, is a stat that tries to capture how good a team plays when on the power play or the penalty kill. It is calculated by simply adding the penalty kill and power play percentages together. An average team should add up to 100%. At the midway point in the season, the Devils ranked 6th in the league in STR. Now here is how the season ended.
|Team||Raw PP%||Raw PK%||STR%|
That's right. The Devils ranked 2nd in the entire league. Though some called our power play a paper tiger, it ended the year at a very respectable 19.5%. When I posted my STR review earlier this year, FrankG929 suggested that I also account for short handed goals against and short handed goals for. I did this by subtracting SHGA from PPG and subtracting SHG from PPGA (Power Play Goals Against). The adjusted STR rankings are below.
|Team||Raw PP%||Raw PK%||Adj. PP%||Adj. PK%||Adj. STR|
The Devils remain in second place, but are a lot closer to Pittsburg. Below is a chart of how teams were affected by adjusting the STR SHGs. The numbers are the charge for the team in the before column. Look at the STL example if it is confusing. They had the third best, but after the adjustment they are in fourth: a change of minus 1 spot.
Ottawa and Edmonton stand out here as the pretenders in STR. Their ability to give up short handed goals is a lot stronger than their ability to score them. Calgary, Dallas, and Carolina are the most improved. Their SHG +/- are 5, 3, and 3 respectively. Now what does STR imply about winning games? When I did this analysis at the midway point of the season, a weak correlation said that the Devils should be doing a lot better than they were doing based on their special team performance. In this analysis, a stronger correlation agrees even more.
Point % = 1.3455 (adj. STR%) - 78.352
When comparing STR% to Point %, it is clear the Devils underperform. Being to the right of the line is bad, while being to the left of the line is good. Can you spot the Devils? With this linear fit, we would expect the Devils to end the season with 109 points. That is
just slightly higher than the 88 points they finished with. Also, the correlation is 40.3%. As we would expect, STR% would not be a perfectly correlation with Point% as special teams is just one part of hockey, but statistically it does account for 40.3% of the variance, so it is clearly significant.
How much do the Devil's underperform compared to their STR? By a lot. The chart below shows how teams Point% compared to what the adj. STR% fit would predict. I call it the 'Who needs Special Teams?' chart.
Clearly Anaheim doesn't need them to win while The Devils most certainly do. That's right. The Devils finished second to last. The differential is huge! Interestingly, Buffalo blows us out of the water, but their lack of success is not at all surprising to me. Anaheim has never struck me as a special teams dynamo, which my data confirms, but clearly they don't care. Their 5v5 hockey is more than good enough to compensate for lackluster special teams.
I want to end on the tone of my last comment. Anaheim shows how a team doesn't need special teams success when they have such great 5v5 play. The Devils have the opposite problem. Besides truly having stellar PK and PP abilities, their 5v5 play just hurts them too much. The only team with fewer 5v5 goals this season? You guessed it: Buffalo. Now some of the negative variance is caused by the Devil's horrendous shootout record, but 5v5 hockey is clearly something the Devils need to work on. Hopefully my analyses of STR in the future show an improvement in 'Who needs Special Teams?' graph...as long as it comes through a higher Point% and not a lower adj. STR%.
Feedback would be much appreciated. Thanks for taking the time to read.