The statistical argument against benching Sanheim

Updated: January 23, 2018 at 12:20 pm by Matt Brauckmann

If you’re like a majority of Flyer fans in this city, you might be scratching your head at the reason why rookie Travis Sanheim had been benched and subsequently sent back down to Lehigh Valley when there is clearly an opening with the big club.

There’s a lot that goes into this decision, a lot more than Sanheim’s stat line but when you dig into the more advanced stats, it becomes very clear that Sanheim should be in the lineup over Brandon Manning.

Why Manning Is In the Lineup

There’s been a lot of speculation on Twitter about why Brandon Manning is even in the lineup to begin with, but when push comes to shove, the Flyers and Dave Hakstol have put a lot of value in the veteran presence that Manning brings to the blue line. Manning played his first NHL game during the 2011-12 season and has been nearly full-time for the Flyers since the 2015-16 season. That may not seem like much but considering how young the Flyer defense core is, having a 27-year old on the blue line probably holds a lot of weight in Dave Hakstol’s eyes.

Then there is also the fact that Brandon Manning really hasn’t been that bad in his time with the Flyers.

He finished as a -12 last season but that was a season in which a majority of the Flyers finished in the negative of that category. He was a plus in every other season he’s played in the NHL. That’s something that a lot of teams will look for especially in a 5th or 6th defensemen.

But therein lies the problem.

Manning is a 5th or 6th defenseman. He isn’t going to generate offense, but for the most part, he is going to do his job for the 15-16 minutes a night that he should be on the ice. Manning isn’t a guy who is a power play defenseman and he isn’t a guy that should be playing 18-20 minutes a night against another teams top line. But for some reason, that’s where Dave Hakstol has decided to use him.

That being said, the Flyers have the perfect guy that fits the role of someone who could play on the 2nd power play unit and eat up minutes against the 2nd or 3rd line of opposing teams. Travis Sanheim.

Advanced Metrics Primer

Before we dive into Sanheim’s advanced stats, I want to go over some of the terminologies that will be used here, so here’s a handy little guide combined with my own explanation below:

  • CF – Corsi For – Shot attempts for at even-strength.
  • CA – Corsi Against – Shot attempts against at even-strength.
  • CF % – Corsi for percentage – CF / (CF – CA) The percentage of shot attempts generated to shot attempts allowed while a player is on the ice.
  • CF % rel – Relative Corsi for percentage – The difference between the teams Corsi for while the player is on the ice as opposed to when he is off the ice. if a team generates a corsi of +2 per 60 minutes with the player on the ice and that drops to -3 corsi per 60 minutes when he isn’t, his relative corsi is +5 corsi per 60. This stat basically takes the rest of the team into account.

You might be asking why these stats are important. Essentially the NHL does not keep track of time on attack or time spent in the defensive zone, so this is the closest we can get to putting a number on that. If a player has a CF % of less than 50%, he’s most likely spending more time in his defensive zone than his offensive zone. If a player is over 50%, he’s probably spending more time in his offensive zone. Generally speaking, a CF % above 55% is considered excellent and anywhere below 45% is considered bad.

There are of course other things that factor into this such as zone starts and who the player is playing with, but that’s where things can get really in-depth and complicated.

For Sanheim’s case, we will look at Corsi, but we will also look at two other stats:

  • oiSV% – On-Ice Save Percentage – Just as it sounds. A team’s save-percentage while said player is on the ice. League average is .920.
  • PDO – Percentage Determined Outcome – SV % + SH % –   The sum of on-ice save percentage and on-ice shooting percentage. This stat is aptly referred to as the “luck stat” by statisticians. The league average is 100.

Ok, now that we’re through that, let’s dig into the metrics.

The Case For Keeping Sanheim Around

Now that you’ve got all the advanced stats you need under your belt, let’s take a look at Sanheim’s case. He had a stat line of 1 goal, 4 assists, and was a -10 in his 35 games played this season. By now, it is well established that +/- isn’t really a great stat to look at, especially for defenders. It varies from year to year based on how the team is performing and generally isn’t indicative of how a player is performing.

We’ll delve right into the statistics on this one.

Player CF % CF % rel oiSV % PDO
Travis Sanheim 54.20% +7.5 .890 94.5
Brandon Manning 48.90% -0.2 .936 101.5
Andrew MacDonald 44.00% -5.3 .951 103.6
Shayne Gostisbehere 50.90% +3.0 .920 101.1
Ivan Provorov 49.00% +0.1 .921 102.2

I threw in MacDonald, Gostisbehere, and Provorov as a measure of comparison here, but our main comparison is Sanheim and Manning here.

Right off the bat, you can see that Sanheim’s CF % is way up there. In fact, it is the best on the team and almost a full percentage point better than Sean Couturier who is easily the team’s best defensive forward. Why does this matter? Take Couturier as a prime example. Couturier had a high CF % almost every year but his scoring and +/- didn’t reflect well because he was playing with two linemates who were not very good and Couturier wasn’t put in big situations. Now he finds himself in those roles and is one of the Flyers most prominent players.

That takes care of CF %, so let’s talk about CF % rel. As you can see, that number is also way up there. Sanheim’s +7.5 CF % rel is not only a team-best but is 8th among all NHL defensemen. To give you an idea, that puts him right alongside guys like Eric Karlsson, Zac Werenski, Marc Giordano, and Seth Jones

This means he generates some serious offensive zone possession when he is on the ice compared to the rest of the team. As you would expect, Andrew MacDonald is deeply in the negative here, and that directly reflects what you see in his on-ice play. This stat is important for defensemen because it gives a good measure of how a guy can set up plays and drive the offense. Not surprisingly, Gostisbehere is right around where I would expect an offensive-defenseman to be, and Manning is right around where I’d expect a stay at home defenseman to be.

The only outlier in Sanheim’s stat line is the .890 oiSV%, which is a stat that is relatively out of his control. When I see a save-percentage that low for a player it means one of two things:

  1. The defender is making one-off catastrophic mistakes/turnovers night in and night out that hang the goalie out to dry.
  2. The goaltending is consistently subpar when that player is on the ice.

It’s easy to go with option number one, but think about it, is that really the problem? No. Watch Sanheim’s game and you’ll quickly notice that his one quirk is getting knocked off the puck and losing board battles. He isn’t allowing awful turnovers or making catastrophic mistakes in front of his own net and with possession metrics as good as his, option number 1 is quickly eliminated.

Things get even more interesting when you look at the Flyers as a whole because some of the most controversial guys on the team get the biggest boost from their netminder. Take a look at the oiSV% for some of the top line guys and Sanheim:

Player oiSV%
Travis Sanheim .890
Sean Couturier .909
Claude Giroux .909
Wayne Simmonds .917

The oiSV% is below average for those. Now take a look at the oiSV% for some of the “fringe” guys:

Player oiSV%
Michael Raffl .952
Andrew MacDonald .947
Valtteri Filppula .944
Brandon Manning .936
Jori Lehtera .936

The last thing to take into account here is the PDO statistic. I mentioned above that it is commonly referred to as the “luck” statistic, and Sanheim has just about none of it. The Flyers shooting percentage and save percentage is low when Sanheim is on the ice, resulting in a team worst PDO of 94.5. This is most likely a result of Sanheim being on the ice with forwards who can’t finish their chances (oiSH % for Sanheim is a team-low 5.5%) and some bad-luck with goaltending.

What Does This Mean for Sanheim?

With everything laid out above it is clear that despite some rookie struggles, Sanheim has played relatively well for the Flyers. Has Brandon Manning been awful? Not by any means, in fact, as I mentioned before, he is an adequate 3rd line defender and has done a decent job for the Flyers. That being said, the Flyers now have a better option and a guy who can really create scoring chances. They need to manage that the right way.

Sanheim’s situation is eerily reminiscent of Ivan Provorov’s, but has just not been handled the same way. Think back to last year when Provorov got torched by the Blackhawks. He didn’t get sent down, benched, or have his ice time cut, he was thrown right back out there for 19+ minutes in the next game, and two months later came back and netted two goals against the very same team who embarrassed him. That’s the type of opportunity a guy like Sanheim needs.

It’s extremely hard to defend the decisions this team is making with Sanheim. Hextall has preached time and time again about these guys developing with the big club and actually playing, but that isn’t what they’ve done here. Instead, Sanheim played 30+ games and then was benched for a guy who is clearly inferior. It makes things even worse when you see a guy like Andrew MacDonald who is out there getting 20+ minutes a night.

The Flyers finally did some good by getting him out of the press box and onto the ice in Lehigh Valley, but he should no doubt be skating with the big club and developing. There’s only so much devemopment that can be had in the AHL. Putting him back with the Orange and Black would greatly benefit both sides.


Bookmark and Share

, ,

Leave a Reply