One of the more underreported stories (including by me) related to the Oilers this season has been the steep decline in the power play since a brilliant run to start the year.
The chart above shows the Oilers’ cumulative power play efficiency over the course of the season, meaning that the first point is from game one, the second from games one and two, the third from games one to three and so on. At game five, the Oilers reached their season peak; this was followed by a steep decline and then a long, slow drop-off. Currently, they are in their worst run of the season, a 2-for-29 stretch since their 8-2 win over Calgary.
Lies, Damned Lies, and Statistics
The question is what it means, since we can present the data in a number of ways. For example:
- “The Oilers scored 10 power play goals on 28 opportunities in their first five games. They’ve scored 23 in the last 131 opportunities – a 17.6 percent clip and below NHL average. This is a sub-average group that got hot for five games.”
- “The Oilers have gone 2-for-29 over their last nine games. Prior to that they went 31-for-130 – a 23.8% clip and one that would have led the league last year! This is a great group in a cold stretch.”
This is what Mark Twain meant when he talked about lies, damned lies, and statistics – the data can be interpreted in different ways, and it’s important to express it properly or else one arrives at an erroneous conclusion.
So what is the answer? Based on power play efficiency alone, the safest answer is probably that the Oilers are worth their season number – it’s the largest sample, and doesn’t suffer from any arbitrary cut-off. However, there’s more to the story than just goals per opportunities .
The Magic Power Play
Power play efficiency captures only a small piece of the picture: goals. Nobody would argue that goals are the object of the power play, and thus the most vital piece, but they’re also relatively rare events. The Oilers have 30 goals and 165 shots in the most common power play situation – 5-on-4 play – this year, and over that small of a sample the percentages can mislead. Last season, Jordan Eberle scored 34 goals thanks in no small part to an 18.9 shooting percentage – this year he’s on a 23-goal pace over the same number of games despite a shots increase because his shooting percentage has slipped to 10.4 percent. Whether one is optimistic or pessimistic on Eberle, there’s no arguing that shooting percentage has had a massive influence for good and for bad on his goal-scoring totals.
Last year, the Oilers had a really good shooting percentage number – they were one of two teams in the league with a shooting percentage above 16 percent in 5-on-4 situations. Nashville was the other. Looking at recent years, I found that the trend for high shooting percentage teams was to see a massive drop in shooting percentage the next year:
|2008-09 San Jose||15||13.1||-12.67%|
For Nashville, that proved true – they’re down to a 12.6 shooting percentage in 5-on-4 situations this year, and have seen the power play fall to 19th this season after leading the league last year. Edmonton, however, has actually increased their shooting percentage – in 5-on-4 situations, they’re at 18.2 percent this year. In terms of shots/minute, though, they are the second-worst team in the league.
Looking at league-wide trends, that seems unlikely to continue. The 2008-09/2009-10 Washington Capitals were the last team to boast a 15+ 5-on-4 shooting percentage over two consecutive seasons; they fell to the bottom of the league in year three. This is the real concern. Either one believes the Oilers have found a way to sustain a high shooting percentage 5-on-4, based on ~120 games of them doing so, or one believes that they’re riding percentages that can’t be sustained, based on what has happened to other NHL teams. I don’t have a crystal ball, but I lean toward the latter explanation: I can’t help thinking that eventually the Oilers’ inability to generate shots on the power play is going to cost them.
But then I thought that last year, too.
Recently around the Nation Network
- Flames Nation: Calgary’s race with Edmonton for the bottom
- Canucks Army: Duncan Keith – Sexist or just a jerk?
- NHL Numbers: PDO numbers by NHL team, April 22
- Wanye: Photoshop and draft updates
- Brownlee: The way I see it
- Lowetide: Good arrows from OKC
- Strudwick: The Nuge and the beach
- Lowetide: Going mobile
- Gregor: Nugent-Hopkins needs surgery
- Follow Jonathan Willis on Twitter!