2013 Tampa Bay Lightning numbers: Through game 15

This week I'll take a closer look at power plays and penalty kills and their effects on the Lightning's season so far, as well as introduce you to some work that has just come out.

Unless otherwise indicated, all stats are for 5v5 play and for players with at least 5 games. Data is from behindthenet.ca unless indicated.

1. Basic stats:

Let's begin with some of the basic stats that we track every week. This is through Tuesday afternoon (before the Toronto game).

Fenwick close 44.14 (29th)

Fenwick tied 43.39 (28th)

Corsi On range 0.00 (Steven Stamkos) to -31.82 (Adam Hall)

Relative Corsi range 18.6 (Steven Stamkos) to -23.9 (Adam Hall)

Corsi QoC range 1.608 (B.J. Crombeen) to 0.403 (Brian Lee)

iPDO: 1097 (Tom Pyatt) to 927 (Brian Lee)

tPDO: 1031 (ESSh% .12/ ESSv% .911)

QS (data from hockey-reference.com and theahl.com):

Anders Lindback: 4/11 (36.4%); 1 wasted, 4 bail outs

Mathieu Garon: 1/4 (25%); 0 wasted, 0 bail outs

Riku Helenius: 8/18 (44.4%); 0 wasted; 2 bail outs

Cedrick Desjardins has only played one game for the Syracuse Crunch, and it was not a QS.

2. Special Teams Efficiency

Special Teams efficiency is a way of getting an idea of how much power plays and penalty kills are affecting the outcome of games. It works in principle much like PDO. Every power play is also a penalty kill, so the sum of power play rates plus the sum of penalty kill rates across the league will add up to 1.00 (or 100%). Thus, for each team, PP% + PK% should regress towards 1. Derek Zona calls this stat Special Teams Efficiency (STE).

Over the past several years, STE has correlated quite well with standings outcomes, so it's possible to use this as a shorthand look at whether teams are getting or losing more goals than average in these situations. It's sort of like looking at whether special teams are having a disproportionate effect on season results. In an 82-game season, we'd expect teams to fall somewhere between 90 and 110, the majority between 95 and 105. In a shortened season we expect to see a wider range than normal.

Data from nhl.com

What this indicates is that man-up/man-down situations are netting the Lightning more goals than the "average" team. As of Feb. 5, however, the Lightning's STE was 119, so that is coming down.

The reason that one might be concerned with a very high STE is that no one is guaranteed the "right" number of power plays each game, which makes it kind of tough to rely on those "extra" goals. I think, though, that there's no reason to be alarmed at this number right now. We might keep track of things to see if there's any precipitous drop in special teams performance, however.

3. Penalty differential

Across the league, we've seen more power plays over last season. In 2011-12, officials awarded 3.30 power plays per team per game. Right now, the average is 3.97. It was 4.36 as of February 2. Perhaps this is a function of the officials not getting their training camp and preseason. Perhaps it's the new rules interpretations. Who knows? Either way, guys are being sent to the Terrarium of Iniquity more frequently than they were last season.

Which got me wondering how that's working out for the Lightning.

Over the last five years, the Lightning have averaged 3.95 power plays opportunities per game and 4.02 times shorthanded per game. It seems that, looked at in the big picture, the Lightning are benefitting only very slightly from the jump in penalties called, because they've been above league averages over that time anyway.

4. Another reason we don't worry about shot quality

Every once in a while, someone will raise the issue of shot quality. The assumption in hockey has always been that where you shoot from matters more than how much you shoot, and it's okay to allow a high number of shots as long as they're from the perimeter and not the dangerous areas like the slot. And that's the way hockey teams have played for decades. It's one of the basic strategies of the game.

It may be true. The problem is that no one has been able to show that taking mostly good shots (or giving up mostly bad ones) is better than taking a lot of shots no matter where they come from. All the data we've been able to gather over the past decade or so shows that regardless of shot distance or location, you're better off having a team that shoots more than their opponents.

There are various explanations for this, but the most compelling one is the rebound/redirection issue. A bad-angle shot can turn into a second, third, or fourth chance pretty quickly, and the chances that a rebound will go in are higher than the chances that an initial shot will.

But here's another wrinkle in the shot quality debate. Apparently the data on NHL.com for where shots come from is incredibly inaccurate. As in only about 12.5% of the shots Chris Boyle (Habs Eyes on the Prize) tracked were coded accurately on NHL.com. Which means that anything we thought we knew about shot quality based on this data is wholly worthless. So now we're kind of starting over on the issue of whether it matters where shots are taken from.

And in other news, take a look at the amount of work it is to collect this data. Now you know why I don't do it.

And finally.....

5. Zone entries

This is something I'm very excited about. Eric_T. (from Broad Street Hockey and NHLNumbers.com) and a group of collaborators have been tracking the difference between dumping vs. carrying the puck into the offensive zone. And they found that--well, I'll let him tell it.

shot differential at 5-on-5 appears to be largely determined by neutral zone play, and that retaining possession as a team enters the offensive zone is particularly important, generating more than twice as much offense as a dump-and-chase play.

This project is still in the relatively early stages, but so far the signs tend to bear out the initial findings that Ryan Shannon deserved a new contract how you get up the ice matters an enormous amount. And it vindicates all those times you yelled in frustration at the dump and chase. It'll be interesting to see how things develop in this project, especially as they're addressing the question of turnovers.