It’s all about special teams… right?

Conventional Wisdom suggests that games are won and lost by special teams: if you can kill penalties and score on the power play, you will be a successful team in the NHL.

And sure, that’s easy to accept, but is it causality (special teams play leads to wins) or just correlation (good teams play good special teams)?

While I am no statistician – and believe you me, I am both mindful and thankful for that every minute of my life – I think I’ve devised a simple little experiment which will allow us to test this Conventional Wisdom. Here’s how it works:

Create three lists: the first list will rank the teams by 5-on-5 goals for / goals against. I’ll call this our “Even Strength” ratio. Yes, we’re leaving out 4-on-4 (and 3-on-3) situations, but those are special situations that will be difficult to factor into our list. It’s simpler (and I believe better) to just ignore them. The second list will rank teams by their “Special Teams Rating” – Power Play Percentage plus Penalty Kill Percentage. The rule-of-thumb is that teams over 100 are playoff teams and teams under 100 are not. The third list will have the teams ranked by current number of points.

The test here will be to see if one of our first two lists matches the third list better than the other. To test this, I’ll determine the difference in ranks between the 1-and-3 and 2-and-3 and find the standard deviation of those number sets.

Are you still with me? Great. Here’s our results:

Standard Deviation between Points per Game and 5-on-5 Goals For / Goals Against: 4.93

Standard Deviation between Points per Game and Special Teams Rating: 8.54

For the record, I hadn’t calculated the numbers before writing this post. But I think this shows that the “Special Teams Rating” that many talk about – looking at penalty kill and power play – mean less to a team than we may like to think. Sure, the Thrashers have a dreadful penalty kill and it hurts them greatly, but the Thrashers are also something of a reverse case. They are 6 places from their 5v5 spot (ranked 26th in record, 24th in Special Teams Rating, but 20th in 5v5 play).

What immediately becomes visible, however, are cases for the following teams:

• Columbus: Ranked 14th, but 5th while 5-on-5 and 26th in Special Teams. Columbus shows us that a team which has terrible special teams – mostly due to a woeful power play – can make up for it in 5-on-5. Another example of that is…
• New Jersey: Ranked 4th, the Devils have a Special Teams Rank of 16 – which suggests they’d finish at the bottom of the playoff contenders. Their strong play while 5-on-5 (ranked 2nd) has elevated them to a Cup contender.
• Minnesota: Despite having the 2nd-best Special Teams Rating (with the help of an astounding 87.4% penalty kill), the Wild are mired in 21st thanks to their 24th-ranked 5-on-5 play.

Of course, there are always exceptions to the rule. Take the following teams as an example:

• New York Rangers: A 24th-ranked 5-on-5 ratio doesn’t hold the Rangers back from being ranked 13th overall. The Rangers have an 11th-ranked Special Teams Ratio. Is that the key to their success? If so, you would expect the Rangers to have more opportunities on the Power Play, but they come in at 15th in Power Play Opportunities and 18th in Shorthanded Situations. In this case, the Rangers are simply an outlier.
• Philadelphia: The Flyers follow the same outlier case as the Rangers, but find themselves shorthanded the 2nd most number of times in the NHL. Their 82.9% Penalty Kill would be expected to sway their overall position more. (The Flyers are ranked 7th, but are only 14th in 5-on-5. They have the 4th-best STR.)

The conclusion, therefore, can be stated very simply: In general, most teams will trend toward winning or losing more closely with their full-strength performance, not special teams performance. There are cases where having strong special teams can offset poor performance at even-strength – or vice-versa – but these cases are found primarily in teams that take (or draw) and exceptional number of penalties.

For Thrashers fans, if you think I’m kidding, I simply suggest you look at these two numbers. These are the penalty kill percentages for two teams:

Thrashers: 75.2%

Red Wings: 77.6%