Filppula Unlikely to Change Much for Flyers

The biggest move the Philadelphia Flyers made on the trade deadline this spring was moving Mark Streit to the Tampa Bay Lightning for Valtteri Filppula and two minor picks. My initial reaction is that it was not much of a return. Filppula was a player who Tampa couldn’t give away last season, and the picks were of practically zero marginal value.

The other perspective is that Filppula is a solid 2-way center, who better balances the Flyers lineup. Much of this idea is based on the fact that, at the time of trade, Filppula had 35 even-strength points, a better pace than nearly every player on the Flyers who couldn’t seem to score at 5v5 this season.

The season is now over, and Filppula has 20 games as a Flyer under his belt. What has he done so far? Measured individually, not much.

Filppula’s Individual Performance

The other day I wrote about which Flyers finished strong. These numbers were from the trade deadline, March 1, to the end of the season, representing 20 games:

FW finish

All four of these measures reflect poorly on Filppula’s performance as a Flyer. His possession numbers (relative corsi percentage and relative expected goals) are among the worst on the team. Furthermore, his production, the precise thing some said he would bring, was in the bottom third of the team as well.

I’ll throw him a bone and say it’s just 20 games. Indeed, players like Simmonds, Voracek and Konecny were comparably bad in this time period (Voracek, in fact, spent a fair amount of time on a line with Filppula). I’ll simply give him a pass on this bad start in Philly.

What concerns me, however, are his long term trends. Filppula is now 33-years old. This is the age when most hockey players start to noticeably decline. Really, it would be unusual if Filppula DIDN’T decline at this age.  Finishing the season with 42 points, up from 31 last year, might imply that he is not declining along normal NHL patterns. I don’t think that’s accurate.

filp excel

Taking a long term view over 5 seasons we are seeing a common pattern of decline. The path is bumpy, but the general direction of his possession (relative Corsi) and scoring chance differential (relative expected goals) is downwards.

The real anomaly in the chart were his points production in the first half of this season with Tampa. I call it an anomaly because it’s an unpredictable spike that runs counter to what his possession numbers were doing at the same time. That is a major red flag that indicates unsustainable performance.

Takeaways:

  • Filppula’s stats, both points and “advanced” shot metrics, in his first 20 games with the Flyers are very poor.
  • His multi-season trends in advanced stats point to the usual decline a 33-year old NHL player faces.

Indirect Effects of the Filppula Trade

But a-ha! The real benefit of the Filppula trade isn’t that he will tear up the scoresheet, it’s that he brings balance to the Flyers’ lineup! Or so they say.

Exhibit A for this argument is that in the 20 games Filppula has been a Flyer, Couturier had 18 points! Exhibit B is that the Flyers’ overall 5v5 goal difference also got better during this period. Let’s take these ideas one by one.

First off, this logical leap is at high risk of a “post hoc error.” To put it simply, just because results are different AFTER an event occurred (ie, a trade), doesn’t mean that the results are different BECAUSE the event occurred.

Take Couturier for example. Yes, his points rose sharply when Filppula arrived. Looking at his advanced stats, however, there was little change in his performance when Filppula arrived.

coots rolling relCF-marked

coots rolling offensive chances - marked

His relative Corsi started going up BEFORE Filppula arrived, and his game-to-game chance creation (expected goals per 60 minutes) hasn’t really changed from where it’s been the last 2 seasons. The difference was the puck started going in on those chances. Did the presence of Filppula CAUSE that to happen?

I know some have taken the position that the Filppula trade motivated Couturier, or gave him better a more comfortable role as “third line center.” On the former, I can’t read Couturier’s mind, so I can’t rule it out. On the latter, it’s not like Couturier is seeing easier matchups.

Unfortunately my go-to stats site (www.corsica.hockey) can’t provide quality of competition stats as of certain dates, but it is true that Couturier has the most difficult quality of competition over the last two seasons overall of all Flyers. This hasn’t changed since Filppula’s arrival.

Looking at individual recent games, Couturier is still getting the toughest matchups. Against Carolina, he was playing against Staal-Aho-Teravainen, their best line. Against the Blue Jackets, it was Wennberg-Foligno-Saad, their top unit. Against NJ, it was the Taylor Hall line. When he scored 3 points against the Islanders, his primary matchup was against John Tavares.

It simply is not true that Couturier is getting better matchups now the Filppula allows the Flyers to re-configure their lines. If Couturier was getting the hardest matchups both before and after the trade, what are we even really saying to call Filppula the “2C” and Couturier the “3C”? It’s a distinction without a meaningful difference. For these reasons, I am skeptical of the argument that Couturier’s boom in late-season scoring was because of Filppula’s arrival.

Turning to Exhibit B, the “advanced”, shot-metric stats also undermine the argument that the Flyers have been better at 5v5 since Filppula arrived.

flyers roll cf-marked

phi roll xGF-marked

Put simply, the Flyers possession as measured by Corsi percentage got WORSE when Filppula joined the team. Their differential of scoring chances as measured by expected goals is solid, but largely similar to where it’s been the last few months.

So why did I mark the Flyers 10 game winning streak on those charts? For one thing, the 10 game rolling average lags, taking into account previous games as of any particular date. Second, there is a clear disconnect between the win streak and the advanced statistics. The win streak, however, can be explained by just one stat.

phi roll pdo-marked

PDO is shooting percentage plus save percentage. Generally, that number tends to hover around 100%. It’s not a coincidence that win streaks are accompanied by a temporarily high PDO when your goalie is making more saves than normal, and more of your shots are going than normal. The things is, PDOs don’t tend to stay significantly above 100 for very long, and a wins dependent on high PDO while other advanced metrics are low usually are just temporary mirages.

This bit of statistical wisdom was proven correct in regards to the Flyers win streak earlier this year. They simply weren’t as good as a 10-game win streak would normally imply, and that success proved to be unsustainable. By the same token, their improved goal difference at 5v5 since Filppula joined the team is largely attributable to a PDO jump as well, and thus likely isn’t a sustainable, repeatable model of future success either.

Takeaways:

  • The Filppula trade may or may not have mentally motivated Couturier to play better, but it didn’t result in Couturier having better possession, creating more chances, or getting easier matchups.
  • The Flyers’ shot metrics did not meaningfully improve after Filppula arrived, and their change in 5v5 goal difference after the trade is more likely because of a temporary jump in PDO.

Conclusion

My view of the numbers show 4 things, which I’ll repeat here.

  • Filppula’s stats, both points and “advanced” shot metric, in his first 20 games with the Flyers are very poor.
  • His multi-season trends in advanced stats point to the usual decline a 33-year old NHL player faces.
  • The Filppula trade may or may not have mentally motivated Couturier to play better, but it didn’t result in Couturier having better possession, creating more chances, or getting easier matchups.
  • The Flyers’ shot metrics did not meaningfully improve after Filppula arrived, and their change in 5v5 goal difference after the trade is more likely because of a temporary jump in PDO.

For me, the bottom line on Filppula is that he’s an “okay” player. He’s not terrible, and I’ll grant him he’s better than his individual stats show in his first 20 games as a Flyer. Even granting him that, he appears to be trending down as most 33-year old NHLers do, and he will likely continue to be a negative player on relative shot differential and relative scoring chances as he’s been the last 3 seasons. For these reasons, pegging Filppula as a 40-point player is an overly optimistic projection.

Additionally, his “indirect” benefits to the Flyers are not as promising as they seem. Advanced stats and an examination of Couturier’s matchups cast doubt on the story that Filppula has brought a meaningful benefit to the Flyers’ and Couturier’s play. Ultimately, to say otherwise is to argue that the advanced shot metrics don’t matter, and that Filppula’s weak individual stats for 20 games somehow caused the Flyers’ PDO to jump in late season games (when the Flyers were already de facto eliminated from the playoffs).

I’ll hope for the best from Filppula for the full year we have him next year, and I’m sure he will make some nice plays here and there. On the whole, however, he’s unlikely to be a terribly effective “2C” for the Flyers, or someone who meaningfully improves the team.

Photo Source: Eric Hartline – USA TODAY Sports

One comment

  • I don’t disagree that Filpulla may not be the solution, but in a vacuum, we traded Steit for him. Is he better than Streit, with out a doubt, is he better than PEB. It is a clear win for the Flyers

Leave a Reply