clock menu more-arrow no yes mobile

Filed under:

2013 Tampa Bay Lightning statistics: Games 1-5

While the Lightning are off to a great start, underlying numbers point towards a correction in the future.


Well, the Lightning are off to a blistering start, with a record of 5-1-0, sitting briefly at the top of the Eastern Conference on Sunday evening. (They're in second place as I type this on Wednesday morning.) It's very tempting to project this success down the line, but--well, let's just say I've been burned before. When looking at the underlying numbers from the Lightning's first five games, I noticed signs that may indicate trouble ahead.

One of the best indicators of whether a player is benefiting from the randomness inherent in hockey (or, if you prefer, whether they're getting lucky) is PDO. This is the team's shooting percentage while the player is on the ice plus the team's save percentage while they're on the ice (check out this post I did last season for a better explination of PDO). PDO regresses over time towards 1000. High PDOs come down; low ones come up. Over an 82-game season some players and lines can sustain higher or lower PDOs than others. Steven Stamkos generally runs around 1020; Nate Thompson runs around 940 to 960. But over time the bumps in this stat generally even out.

Of all 21 Lightning skaters, 9 have PDOs of 1100 or higher; of the 18 who have more than 2 games, 9 have PDOs of 1100 or higher. There are no players with PDOs under 1000, and the range for all skaters is 1021-1250. For those aren't sure what that means, that's incredibly high and indicates that the team has been experiencing a lot of good bounces. Things aren't going to stay this good, although it's impossible to know just when or how the correction will come.

Take Victor Hedman. After five games, his stats looked like this:

Corsi On


Relative Corsi


Off. Zone Starts


On-Ice Shooting %


On-Ice Save %


Corsi QoC


Corsi QoT


This indicates that while Hedman is on the ice, the team is not only taking far more shots than when he isn't on the ice, but also that 20% of those shots are goals. At the same time, he hasn't been on the ice for any 5-on-5 goals against. And while that's impressive and the kind of stat that can make you smile, it's completely unreasonable that this is a situation that can last. At some point, those numbers are going to come down, and likely quite a bit.

For instance, his on-ice shooting percentage last year was 9.76%, which came while playing more with Steven Stamkos (who shot 19.8%) and Martin St. Louis (shooting 13.8%) than any other forwards. (his regular defensive pair was Eric Brewer.) So far this season, his top linemates have been Sami Salo (0%), Nate Thompson (0%), and Stamkos (20%).

There are other signs that some of the underlying issues the Lightning experienced last season are not banished altogether, despite the hot start. First, the team isn't controlling the puck very often in most situations. They take fewer than 50% of the shots on net in all situations except when they're down by 2 or more goals. What that means is that except at those times when teams are just dumping everything they can on the net in desperation, the Lightning are being outshot consistently. They're 24th in the league in Fenwick Tied* (46.67%) and 26th in Fenwick Close (46.15%).

This is further reinforced by the fact that only 3 skaters have positive Corsi scores. Three. And one of those (Marc-Andre Bergeron) has only played in two games. That means that everyone not named Victor Hedman and Steven Stamkos is getting outshot, some by only a little overall, but outshot nonetheless. Over time, getting outshot, even with very accurate shooters on the roster, makes it much harder to win games.

One final statistic I'll mention: Quality Starts. Again, a Quality Start is when a goaltender gives his team a good chance to win (based on save percentage). That's a save percentage of above .917 (currently) or fewer than 3 goals against while facing 25 shots or fewer. Anders Lindback has 2 Quality Start and Mathieu Garon has one. That's a 40% QS rate for Lindback, which is slightly low. We'd like to see it be above 45%. The very best goalies seem to run QS rates around 55-60%.

None of this is to say that the Lightning are going to start losing games in bunches or slide into the depths of the Southeast Division. It's just a reminder that things aren't quite as solid as their win-loss record would indicate. It's unlikely that they can continue to score at quite this rate. However, it's also likely that Anders Lindback's numbers will fluctuate wildly from game to game.

One last caveat about advanced statistics, however. In this shortened season, we're likely to see a great deal more variance in all shot, scoring, and save metrics. Things simply won't have time to even out like they would in an 82-game season, and even in an 82-game season, variance plays a huge role in the outcome of games (somewhere around 40% of the standings appears to be attributable to randomness.) It is entirely possible that some of the Lightning will be able to maintain higher than normal shooting rates by simple luck alone. It is possible that the Lightning's lack of puck control will hurt them less than it would if there were three more months to get through.

Then again, randomness is (naturally) unpredictable, and it could very well come back to bite the team next game, next week or next month.

*Fenwick scores are Shots For as a percentage of all shots taken. They're broken down into game situation because teams that are behind tend to shoot more to try to catch up (something called the score effect). The further behind a team is, the more they tend to shoot. The larger the gap in the score is, the more shot ratios are skewed. Thus, checking how teams control the puck when the score is tied or within one goal (close) helps to reduce the effect of desperation on shot ratios.