In tremendously simplistic terms, an NHL regular season can be broken into four groups of 20-games for purposes of analysis. The usefulness of a 20-game block is, admittedly, somewhat abstract, but it also presents some valuable information. NHL players and teams can be tremendously streaky, and with the 21st Century emphasis on video coaches, studying opponents and the game, opponents can start identifying trends in a team to focus upon in upcoming games. There is no question that opponents in today’s NHL are better prepared to play the Florida Panthers than ever before, with scouting reports and video study identifying how to try and counteract the Cats game plan.
With the concept of “trends” fresh in our heads, I have decided to take a look at the Panthers trends in 10-game blocks, and report the data at the 20-game marks. By doing so, we can attempt to identify who is truly improving, playing well, or not-contributing. We can also see if the team has, or can, address its weaknesses. Let’s start with an example:
The power play. Through the first 10-games of the season, the Panthers power play was one of the worst in the league (again). The statistics at the 10-game mark showed that the Cats power-outage with the extra man was below league average, with a 12.50% success rate. That placed the Panthers 24th in the league in terms of power play percentage.
But at LBC, we do not merely use statistics, we also watch the games, and pay attention to our coaches and player’s post-game comments. There is no question (from watching games) that the power play actually did improve over the course of the first 10-games. It may have been ineffective in the extreme at putting points on the board, but at least the team started to gain entrance to the offensive zone with possession, and maintain possession. From whence we came, that was an improvement.
But after an 0-7 power play performance against Boston, in the tenth game of the season, (in a game that saw the Panthers surrender a short-handed goal), Coach Gerard Gallant called out his man-advantage as inadequate. We now can move forward into the second unit of 10-games (in our 20-game review) and we can see that in the second group of 10-games, the power play is still below average, and went 0-4 in a loss to the Flyers. Over the second 10-games, the power play barely improved at all.
That is how we are seeking to use this analysis, to identify whether the team is trending in good or bad directions. Lets continue to explore the team statistics:
Stat: After 10: After 20: Trend:
Goals-for 26 52 Same
Goals-for-game 2.60 (19th) 2.60 (14th) Same
Shooting % 8.3% 8.5% About same
Goals-Against 24 53 worsening
Goals-against-per-game 2.40 (10th) 2.65 (18th) worsening
Save % .910 (14th) .911 (19th) worsening
Shots-for-per-game 31.3 (7th) 30.5 (11th) worsening
Shots-against-per-game 26.8 (4th) 29.7 (12th) worsening
Power-play % 12.5% (24th) 14.5% (23rd) improved
Penalty-Kill % 78.6% (22nd) 81.4% (19th) improved
Penalty Minutes 89 (10th) 161 (8th) improved
Penalty-min-per game 7th 8th About same
PDO 99.8 99.7 Same
After 10 games, the Panthers were at league average in goals-for and above league average in goals against. The team was below league average on the power play and penalty kill, save percentage and shooting percentage.
After 20 games the Cats were still average in goals-for and against, well below average on the power play, slightly below average on the penalty kill, and below average in both save % and shooting %. This does not make me feel comfortable….
The Players:
Lets start our look at the players with possession stats, and all possession starts with face-offs. Six Panthers took 28 or more face-offs through the first 10 games, while Jonathan Marchessault would stop taking face-offs over the next 10:
Name: After 10: After 20:
Jonathan Marchessault 28-24 (53.8%) No longer statistically applicable
Aleksander Barkov 72-66 (52.2%) 142-12 (48.3%)
Derek MacKenzie 53-59 (47.3%) 102-101 (50.2%)
Vincent Trocheck 94-105 (47.2%) 206-202 (50.5%)
Jared McCann 13-15 (46.4%) 19-25 (43.2%)
Denis Malgin 17-36 (32.1%) 40-70 (36.4%)
That should make you sort of wonder why Nick Bjugstad went to wing over rookie Denis Malgin. Vincent Trocheck and Derek MacKenzie lead the team in face-off wins, with Malgin trailing badly.
The Panthers advanced possession stats are actually a pundit’s dream, with positive shot differential and low PDO. Yet, amongst the players, there were some surprising possession results. Here are the Corsi-For % stats for the team:
Name: After 10: After 20:
Vincent Trocheck 58.4% 55.6%
Reilly Smith 57.6% 56.3%
Jaromir Jagr 57.5% 55.9%
Jason Demers 56.6% 53.1%
Jonathan Marchessault 56.4% 54.4%
Colton Sceviour 55.5% 51.4%
Aleksander Barkov 55.2% 54.6%
Mark Pysyk 54.7% 51.1%
Jared McCann 54.4% 51.2%
Alex Petrovic 54.3% 53.8%
Shane Harper 54.1% 49.3%
Aaron Ekblad 53.2% 51.1%
Michael Matheson 52.9% 51.4%
Keith Yandle 52.8% 50.5%
Denis Malgin 51.1% 50.8%
Derek MacKenzie 50.0% 45.6%
Greg McKegg 48.0% 48.1%
Kyle Rau 47.6% 46.5%
The surprises are obviously with respect to Ekblad and Yandle being so low on that list. Both of these fantastic, elite defensemen are expected to drive possession for the team, and instead, were near the bottom in this possession metric. Their respective Corsi-For Relative % numbers offer a similar surprising result, with Yandle at -2.7 and Ekblad at -2.1 (after 10). The other surprise is Barkov. The young Finn was expected to dominate, and is typically a two-way, possession beast. Instead, he was well behind his line mates and the second line with a Corsi-For relative % of 1.2 (after 10).
As for the drops in the 20-game percentages for most of the players, that should be expected as their time-on-ice goes up. At the 10-game mark, the stats were still skewed in some respects by the lack of numbers. Lets go ahead and plug in some zone start numbers to see who the coaching staff was trying to shelter:
Name: O-Zone Start rate after 10: After 20:
Denis Malgin 71.7% 65.5%
Jared McCann 71.7% 66.3%
Keith Yandle 62.6% 55.6%
Aaron Ekblad 60.8% 56.1%
Jaromir Jagr 59.8% 52.7%
Shane Harper 59.2% 50%
Aleksander Barkov 58.3% 53.5%
Colton Sceviour 58.3% 48.5%
Greg McKegg 56.1% 48.8%
Jon Marchessault 54.9% 53.1%
Jason Demers 53.3% 49.7%
Alex Petrovic 52.9% 49.2%
Mark Pysyk 52.4% 48.3%
Derek MacKenzie 52% 44.5%
Reilly Smith 51.5% 50.7%
Mike Matheson 50.6% 49.7%
Kyle Rau 50% 48.8%
Vincent Trocheck 49.1% 47.6%
A cursory look at the numbers after 10 games showed that the kids on the 3rd line were being massively sheltered, along with the top pair defense (from that time period, when Yandle and Ekblad were paired). That makes those Corsi numbers a little more concerning over the first 10, and there was not any significant change over 20 to suggest that the low Corsi numbers are dissipating for Yandle and Ekblad.
We can certainly see that MacKenzie and his line mates are getting a lot of defensive zone time. More striking is the use we see for the Trocheck line. That line is being used in all situations and are seeing more ice time over the second 10 games than the presumptive first line of Barkov, Jagr, and Marchessault.
Moving to goaltending, the Cats goaltending tandem was decent over the first 10, but slid towards the 20-game mark. Roberto Luongo went 3-4 in the first 10 games, with James Reimer posting a 1-1-1 record. Over 20 games, Luongo went 7-6-0, while Reimer went 3-3-1.
Name: GAA-1st 10 GAA-20 SV%- 1st 10 SV%-20
Roberto Luongo 2.30 2.31 .912 .921
James Reimer 2.61 3.05 .907 .897
Reimer saw an average of 28 shots per game, Luongo faced an average of 26 per game over their first 10 games. By the 20-game mark Lu was facing an average of 29.1 shots against while Reimer saw an average of 30.42 per-game. The Panthers are giving up more shots-against as the season progresses. While Luongo got better, Reimer posted poor numbers. The goaltending may not have been out of this world, but putting aside the numbers and watching the games, the duo was hung out to dry by the defense on numerous occasions.
Looking at some miscellaneous categories, over the first 10-games of the season, Vincent Trocheck and Aaron Ekblad led the team with 22 hits each, followed by Derek MacKenzie with 21 and Shane Harper and Alex Petrovic each with 17. After 20-games, MacKenzie leads the team with 49 hits, followed by Trocheck with 48, Ekblad with 41, and Petrovic with 31. Aaron Ekblad led the team after the first 10-games with 13 blocked shots, followed by Mark Pysykl with 11 and Michael Matheson with 10. Colton Sceviour led the forward corps after 10-games with 9 blocks. After 20-games, Pysyk leads the team with 27 blocked shots, followed by Yandle with 25, Matheson with 22 and Demers with 20.
Finally, we move to the all important goals and assists categories, where we saw some rather unexpected results after the 10-game mark
Name: Goals/10: Goals/20: Assists/10: Assists/20:
Marchessault 6 9 5 7
Sceviour 5 5 2 4
Trocheck 4 6 7
Matheson 2 2 2 5
Barkov 2 2 3 9
Harper 2 2 1
Ekblad 1 5
Smith 1 4 2 3
Jagr 1 2 3 7
Pysyk 1 1 2 2
Malgin 1 4 2 4
Yandle 0 1 4 10
Petrovic 0 1 3 4
Demers 0 2 3 6
MacKenzie 0 2 2 3
McKegg 0 0 2 2
Jokinen 0 1 2 3
There are, or course, some absences on the list over the initial 10-games (most notably Jared McCann), and some top-6 players showing concerning scoring problems. Consider that over the first 10-games, Colton Sceviour played only 146 minutes, and Jonathan Marchessault 181 minutes, vs. Vincent Trocheck’s 208 minutes and Aaron Ekblad’s 236 minutes, or Aleksander Barkov’s 193 minutes, and they still managed to out-produce all of their other teammates. Similarly, Reilly Smith played 194 minutes over the first 10-games. The player with the fewest minutes (other than Kyle Rau, who played only a couple of games over the first 10) was Jared McCann with 100 minutes, so we can’t be that surprised that he failed to produce offensively.
Jumping ahead to the 20-game mark, we still don’t have sufficient production from the team’s “top players.” Barkov, Jagr, Smith- none have been producing a great amount of offense. Say what you will about Marchessault not being Jonathan Huberdeau, but he has led the team in scoring and points at both the 10 and 20-game marks, and he has not played as much as the Trocheck line either. Malgin has produced some offense, but he has suffered somewhat of late through rotating line mates. Ekblad has begun to produce goals, which is essential for the team. Nonetheless, he also led the team in ice time over 20-games, and got a healthy dose of offensive zone starts: translation: he’d better produce.
Finally, one more thing to include in our analysis. Over the first 10-games, Marchessault went +6, while Vincent Trocheck went -3. Barkov was +3 over the first 10, while Aaron Ekblad was -3. Petrovic was +5 over the first 10, and his partner, Mark Pysyk was +6. Plus/minus is a very flawed statistic, but it can be (at a minimum) an eyebrow raiser: Some of these players are on the ice for an awful lot of goals-for, and others were on the ice for an awful lot of goals-against, whatever those reasons may be. After 20-games Marchessault dropped to +3; Trocheck dropped to -4; Barkov climbed to +4; Ekblad dropped to an incredible -7; Petrovic climbed to +8, and Mark Pysyk dropped to +4. Just something to consider.
The Panthers possession numbers were 4th in the league, with a 53.2% Corsi, which is outstanding. Yet, the team is barely above .500, and is worsening (according to the numbers) with respect to shooting percentage and save percentage. So, how do we reconcile that information? Last season, former LBC writer Shane O’Donnell posted an article examining some factors that were leading to just this sort of outcome early last season in the NHL. Several of the factors he identified that explain great Corsi and less-than-great outcomes are present here: The Panthers special teams are abysmal (again), and they are not getting sufficient goaltending from James Reimer. Both of those factors can negate great Corsi results.
Yet, watching the Panthers it is also evident that the team is “jobbing” the system. In the Philadelphia game, the Panthers fired more than 30 shots against the Flyers, while giving up less than 30 against. That is a positive shot differential that Corsi stats pick up on. Yet, a large number of those shots were from low percentage scoring areas of the ice. Even in the win over the Rangers, the Blueshirts averaged closer shots to the Panthers net than the Cats registered against the Rangers. We are seeing that a great deal this season, as the Panthers fire a great number of shots from the defense, but not a great number of high probability shots from the slot. Kevin Allen of USA Today wrote about the players “jobbing” the Corsi system in October. In that article, he stated that players know about Corsi and how it effects their contracts, and some will fire shots from anywhere just to gain positive Corsi. The Panthers are effectively doing something similar.
James Reimer takes an inordinate amount of blame in the numbers, but aside from the Toronto game- where he was awful, it appears he is more the victim of very poor team defense, than a single scapegoat for the team’s problems. I am not alone in pointing to the team’s defense. Gerard Gallant has been pointing out their failures repeatedly. Watching the games shows that the defensive failures are costing the team goals-against far more than the goalie play. But the team’s star forwards are also failing.
The lack of goal production from the likes of Barkov, Smith, Jokinen, and Jagr is problematic. Instead, the team has been forced to rely more heavily on Trocheck, and timely contributions from defensemen and bottom-6 forwards. The team’s struggles against bigger teams have also been evident.
What comes next is anyone’s guess, but the team’s conversion to a purely analytical approach has proven to be a bumpy road. We will look again at the trends at the 40-game mark, and hope that the team has put its troubles behind. For the time being though, this is not the start we envisioned, and comparisons to the 2015-16 Panthers start are not a source of comfort, as that team needed a 12-game winning streak to pull them to success. We can’t count on that happening again.