I do not claim to have a perfect or complete understanding of the study of NHL statistics, (referred to in the "business" as "analytics") but certain parts of it are not very hard to understand. For instance, it is simple common sense that if a team is outshooting their opponent they are likely possessing the puck more than the opposition: a team can't shoot a puck they do not possess. The opposite is equally obvious: a lack of shots likely means a team is not possessing the puck as much. This is measured statistically by something referred to as "Corsi." Corsi is the number of attempted shots, shots on net, and blocked shots a team collects in the course of a game, reflected as Corsi/60, or for an individual player, as icorsi/60. Analytics people can analyze Corsi for and against (how many shots for a team and how many shots against a team) and while a player is on the ice and while he is off of it, and compare the two. Looking at the Panthers team Corsi/60 for this past season, I expected to see an awful finish. After all, a team that possesses the puck more than an opponent typically wins games, and that was something this team did very little of in 2013-14.
Imagine my surprise than to see that the team was 17th in the NHL in even strength 5 on 5 Corsi For/60 (How many shots the Panthers were taking at even strength per 60 minutes). To put this in the most basic of ways: the Panthers were better than almost half the league when it came to shot differential. I would have expected to see something far worse from the second worst team in the league. Even more startling was that the Panthers even strength 5 on 5 Corsi Against/60 was 8th in the league, just two positions below Boston's (how many shots were being taken against the Panthers in 60 minutes of ice time). Now to some degree this could be expected: teams that are losing late in a game are typically permitted by their opponent to take more shots as the team with the lead sits back and attempts to give up only low quality outside shots while not taking any big chances in the offensive zone. Still, the Panthers at 17 in even strength Corsi/60 are in good company, with Columbus at 19, Detroit at 22, Montreal at 18, and Tampa at 16, all of whom were playoff teams. Even more interesting was that in 5 on 5 situations when the score was close, the Panthers still had a Corsi For % of 49.7%, good for 17th in the league, and tied with Pittsburgh. Anaheim was 15th in the league with 49.8% and defense oriented Nashville was 19th with a 49.4% (Buffalo and Toronto were worst in the league with a 41.1% and 42.1% respectively). This piqued my interest and it sent me looking for where exactly I could find the indicators that would show me how this team ended up in 29th place.
It did not take too much effort to find. The Panthers were:
- dead last in the league in save percentage
- second to last in goals against per 60 minutes (GA/60)
- last in goals against while on the penalty kill
- second to last in shots against while on the penalty kill
- second to last in save percentage on the penalty kill
- last in power play goals per 60 minutes
- second to last in power play shots for per 60 minutes
- 25th in the league in shooting percentage
Bad numbers, to be sure, but one thing caught my eye: while having some of the worst save percentage and special teams numbers in the league, the Panthers were 21st in the league in shots against per 60 minutes (SA/60) at even strength ("Shots Against" are actual shots that get through and are on net, they are quite literally, shots on net against). So nine teams in the NHL gave up more shots against per 60 minutes than the lowly Cats, but this team finished dead last in save percentage. The "stats guys" will tell you on all of their websites and blogs that defensemen have no impact on SV%. I can't say I necessarily understand this, but o.k., if that is the case than Panthers goalies were dead last in the league while there were nine teams who allowed more shots on goal per game.
To be frank, I would have expected the team with the worst save percentage in the NHL to also be one of the two or three worst in giving up shots per game as well. This was absolutely the case with respect to the penalty kill, where the Panthers were second to last in shots against, but what about even strength? Toronto, Ottawa, Buffalo, Edmonton, Colorado, Philadelphia, Carolina, and Montreal all gave up more shots against per game than the Panthers did at even strength. But only Florida and Edmonton transferred from the ten-worst in the league in shots against list (Florida at 21st in shots against, Edmonton at 27th) to the list of worst 5 teams in the league in save percentage (Panthers 30th and Edmonton 26th). Toronto, Ottawa, and Buffalo were the worst three teams in the league in shots against per game, but finished 9th 19th, and 14th (respectively) in SV%. This led me to take a closer look at our goalies. Could the 2013-14 Panther's goaltending have been one of the biggest problems on this lackluster squad?
Five goalies played 6 or more games for the Panthers this season. Tim Thomas played 40 games, followed by Scott Clemmensen with 17, Roberto Luongo with 14, Jacob Markstrom with 12, and Dan Ellis with 6. So how much impact did they have on the dismal season? Numbers may lie, but it still appears the answer is "rather substantial." Looking at goalies who played similar numbers of games on other teams, it is apparent that the best the Cats got this year from a goalie was average, and the worst was well below that.
We will start with Tim Thomas and his 40 games. Thomas faced 1206 shots and posted a .909 SV% and 2.87 GAA. This averages to 30.15 shots per game. Three other goalies I looked at from this season also played 40 games: Evgeni Nabokov, for the Islanders, Carter Hutton, in Nashville, and Karri Ramo for Calgary. Nabokov faced 1085 shots, for an average of 27.125 shots per game and posted a .905 SV% and 2.74 GAA. Ramo faced 1091 shots for an average of 27.27 shots per game put up a .911 SV% and 2.65 GAA. Hutton played 40 games in Nashville, facing an average of 25.15 shots per game and posting a 2.62 GAA and .910 SV%. Thomas was comparable to these other goalies. To take a more in-depth look I also looked at several other goalies who played 36-48 games this season, putting them within 8 starts of Thomas in his time with the team. The first of these is Braden Holtby, who played 48 games for the Capitals, faced 1475 shots (an average of 30.72 per game) and posted a .915 SV% and 2.85 GAA. Cory Schneider played 45 games for the Devils and faced an average of 24.60 shots per game on his way to a .921 SV% and 1.97 GAA. Eddie Lack played 41 games for Vancouver and faced an average of 25.65 shots per game. He posted a 2.41 GAA and .912 SV%. Martin Brodeur played 39 games for the stingy Devils, facing an average of 24.89 shots per game, with a 2.51 GAA and a .901 SV%. James Reimer played 36 games for the Maple Leafs and saw an average of 30.41 shots against. Reimer posted a .911 SV% and 3.29 GAA. Carolina's Anton Khudobin played 36 games and faced an average of 29.88 shots per game with a .926 SV% and 2.30 GAA. Robin Lehner of the Senators also played 36 games and faced an average of 31.69 shots per game with a .913 SV% and a 3.06 GAA.
Looking at these other goalies, all from teams that did not make the playoffs, it is apparent that Tim Thomas was somewhat effective for the Panthers, if not merely average. He saw a similar number of shots to all but Schneider and Brodeur (the Devils gave up the fewest number of shots per 60 minutes of any team in the NHL this season) and posted comparable numbers in the SV% and GAA departments with the majority of them. He was not facing an extraordinary number of shots compared to these other goaltenders, but none of them (other than the Devils tandem and their few shots per game) were going to be considered all-stars either. Also worth noting is that certain teams are willing to allow a great number of low percentage chance shots from certain areas of the ice - Randy Carlyle is known for this with the Maple Leafs. Thomas would end up going 16-20-3 with the Cats, but injuries hampered him and resulted in his backups playing 29 games.
Scott Clemmensen played 17 games this season. He saw an average of 26.70 shots per game. His SV% was an ugly .896, while his GAA snuck up to a 3.09. He finished with a record of 6-7. He is also 36 years old, and well past his prime. By comparison, Philipp Grubauer of the Capitals also played 17 games this season. He faced an average of 27.58 shots, but posted a 2.38 GAA and .925 SV%, both far better than Clemmer. Justin Peters in Carolina played 21 games and saw an average of 29.80 shots while putting up a .919 SV% and 2.50 GAA, again, far better. Ben Scrivens played 21 games for the Oilers this season and faced an average of 35.28 shots against. His SV% for the Oilers was .916 and his GAA was 3.01. From the playoff bound Penguins, Jeff Zatkoff played 20 games and averaged 29.10 shots on his way to a .912 SV% and 2.61 GAA. Niklas Backstrom of the playoff contending Wild played 21 games, and like Clemmer, struggled while facing an average of 26 shots against per game. He posted a 3.02 GAA and .899 SV%.
Clemmensen did not face a large average number of shots per game, but he certainly struggled, more so than many of these other goalies who played a similar number of games, saw more shots, and, other than two, played on non-contending teams. To be fair, Clemmer did face some games with an extraordinary large number of shots, such as his last start against the Islanders where he faced 40, but his average number of shots against is still low.
Jacob Markstrom was perceived to be the goalie of the future in Florida, but in 12 games this season, he faced an average of 23.75 shots against, and had a horrifying 3.52 GAA and .874 SV%. By comparison, Jaroslav Halak played 12 games for a not so good Capitals team this season, facing an average of 31.91 shots per game but also posting a 2.31 GAA and .930 SV%. Michal Neuvirth played 13 games for the Caps and faced an average of 32.23 shots against per game. Neuvirth had a GAA of 2.82 and SV% of .914. Joey MacDonald, in Calgary, played 11 games and faced an average of 24 shots against. He had a GAA of 2.90 and SV% of .890. Petr Mrazek of the playoff bound Red Wings played 9 games and saw an average of only 19.77 shots per game. He posted a 1.74 GAA and a .927 SV%.
Markstrom may end up being a good goalie one day, but it sure was not happening in Florida, or anytime soon. In 4 games with Vancouver he had a .868 SV% and a 3.00 GAA while facing an average of 19 shots against per game. He faced a below average number of shots per game and still put up numbers that show he is nowhere near ready. Markstrom went 1-6 with the Panthers. For those paying attention, that is 13 losses for Markstrom and Clemmensen. Things get even worse when you add Dan Ellis and his 0-5 record to this total. The sieve-like Ellis faced an average of 27.5 shots per game for the Cats, and posted a 4.81 GAA and .836 Sv%, (well below average) while facing a rather normal number of shots per game for an NHL goaltender. That is 18 losses for these three goalies. Let's take that one step further though and look at the all important goal support question for these three Panthers backups.
Looking only at games where these 3 goalies played 30 minutes or more: In the 6 games Ellis appeared in and went 0-5, the Panthers scored 2 or more goals in 4 games and 3 or more goals in 2 of those. Markstrom had 9 losses where he played 30 minutes or more. In those 9 losses the Panthers scored 2 or more goals in 5 of them, and 3 or more goals in 3 of those games. Clemmensen had 9 losses where he played 30 minutes or more. In those games the Panthers scored 2 or more goals in 6 of them and 3 or more in 3 of those games. Certainly there was an impact there, but how much is hard to tell, as it is impossible to tell how the goalie's play effected team play in those games, if at all. The Panthers were shutout in one game that Ellis played 30 minutes or more in (Ellis played 39 minutes in that game in relief of Luongo), and blanked twice in games that Markstrom played 30 minutes or more in. In the two games Markstrom lost where the Panthers were shutout, the numbers are rather telling regarding team impact. The Kings shut the Panthers out on October 13, 2013. In that game Markstrom faced 29 shots against and posted a .897 SV%. The Panthers only put 20 shots on Ben Scrivens as the Kings played their stifling defense to a tee. On November 1, 2013, the St. Louis Blues shut the Panthers out 4-0. In that game Markstrom faced only 23 shots and posted a .826 SV%, giving up 3 goals on 7 shots in the 2nd period. It is especially apparent from the Blues game that once the opposition gained a lead they sat back and let the Panthers come at them. The Panthers put 31 shots on Brian Elliott that night. It is readily apparent that Markstrom's play had a significant impact on the team in that game.
Dale Tallon looks that much smarter for having moved Markstrom for Roberto Luongo when looking at the numbers. Luongo was facing an average of 27.54 shots in 42 games for the Canucks this season and he put up a .917 SV% and 2.38 GAA. In 14 games for the Panthers he faced an average of 30.85 shots, roughly 4 more shots per game, and his SV% bumped up to a .924, while his GAA slipped slightly to a 2.46. Plenty of NHL analysts claim that Lou plays better when he faces more shots. If that's true, the Panthers are definitely a good fit. His record was 6-7-1-1, similar to Clemmensen's. Lou had a far better SV% than Tim Thomas did, and a better GAA, while facing a similar number of shots per game (Lou faced roughly close to one more shot per game).
The goalie position alone did not cost the Panthers this season. Poor team defense, and abysmal special teams were crippling. However, a league worst SV% does add an important clue to diagnosing what else went wrong. Simply put, other teams faced an equal or greater number of average shots per game and got better goaltending. The result for those teams was a better record than the Panthers. This was true of every team finishing behind the Panthers in shots against per 60 minutes, except for Buffalo (28th in the league in shots against per 60 minutes but 14th in SV%).
Good goaltending can bail bad defense out, and good goaltending can "steal" games on occasion. The Panthers need desperately to repair their broken special teams and improve team defense, but a season's worth of average to above average goaltending will provide this team with a big boost next season. This is not the type of boost that will propel them into the playoffs, but one that, if combined with the repairs mentioned above, could put the team in contention for a wildcard rather quickly ("contention" being the key word). Do not misread this to say that a new goalie will get the team into the playoffs, it's not that simple, but it is likely to curb some of the losing and stop the cheap and easy goals that a team lacking scoring ability has trouble making up. In its most simplistic form: a team cannot have the worst SV% in the league, or be in the bottom 5, and be competitive.
A full season of a healthy Luongo is a great rock to build upon. With that problem temporarily resolved, the team can move to working on the special teams and defense. Tallon pulled a real coup in obtaining Bobby-Lou services. But, if he is injured, all bets are off, with the current lack of a decent backup or a ready alternative waiting in the wings. Putting that aside, however, it appears one of the team's biggest statistical issues has been put to rest, and that bodes well for putting together more victories in the future.