clock menu more-arrow no yes

Filed under:

Shayne Gostisbehere’s return nears, who should he replace?

New, comments

Ghost has missed the team’s previous ten games.

New York Rangers v Philadelphia Flyers Photo by Len Redkoles/NHLI via Getty Images

Flyers defenseman Shayne Gostisbehere will be returning soon. Possibly as soon as the team’s next game, Thursday night against the New Jersey Devils. Should he do so, head coach Alain Vigneault will have a decision to make on which of his six current defensemen should come out of the lineup. Let’s take a crack at it ourselves.

First off, yes, we are going to talk about PDO. It’s a stat that has fallen out of favor a bit, mainly because there are better, more specific ways to make an attempt at evaluating luck. However, it will always be a good way to quickly find a shortlist of players whose struggles or successes are likely being exaggerated.

If you’re not already familiar, a player’s PDO is their on-ice shooting percentage plus their on-ice save percentage. An average PDO is considered 100, and players north of that figure are considered lucky. Now, luck can be a touchy word, so let’s go with this instead; players on each side of the benchmark can expect regression towards it, the more games they play.

What’s the point?

Teams win games by scoring goals. A wild concept, surely. What plus-minus is supposed to do is give you a general look at which players are on the ice for more goals for than they are against, but it does an absolutely terrible job at doing so. A much more accurate way to look at “pluses” and “minuses” would be a player’s 5-on-5 (or even strength) goal differential, as it removes situations where teams have an extra skater, but both are influenced by two things that spike and dip rather unpredictably: shooting percentage, and save percentage.

Their unpredictability is what makes them important — previous goal results don’t lead to future goal results. We know that shot-based metrics like Corsi and expected goals do a better job at predicting future goals, so when a team goes on a PDO bender all while getting out-shot by a large margin, we can expect them to win fewer games moving forward. The same can be applied at a player level, in terms of their team out-scoring opponents while they are on the ice.

So to figure out who Gostisbehere should be replacing in the lineup when he returns, let’s use the previous ten games to look at Flyers defensemen recent on-ice results.

One defender stands out with an 85% goals for percentage at 5-on-5 — Robert Hagg. He’s been on the ice for six goals for and just one against, while no other defenseman that has appeared in all ten games has been on the ice for even one more goal for than against. Only Mark Friedman and Justin Braun — each with five games played, and a plus-one goal differential — are above 50% with Hagg.

But other defenders on the team had a much more positive impact in the games, despite their ratio of goals for to goals against being weaker.

On-Ice 5v5 Stats Last 10 Games

Player CF% GF% xGF% On-Ice SH% On-Ice SV% PDO
Player CF% GF% xGF% On-Ice SH% On-Ice SV% PDO
Ivan Provorov 55.78 46.15 53.59 6.00 90.41 96.41
Matt Niskanen 56.79 50.00 56.36 6.45 91.30 97.75
Philippe Myers 55.08 50.00 48.33 9.88 90.00 99.88
Travis Sanheim 55.48 50.00 52.33 7.79 92.00 99.79
Robert Hagg 41.41 85.71 44.28 10.91 98.84 109.75
Justin Braun 40.46 66.67 41.64 6.45 97.44 103.89

When one of the top two defensive pairs were on the ice in the Flyers’ previous ten contests, they owned over 55% of the shots taken. However with Hagg on the ice they owned just 41%. Weighted shots, or expected goals, have the split look less dramatic, but still present. So why have more goals wound up in the back of the Flyers’ net while the top four defensemen have been on the ice? Goaltending.

While Flyers netminders had stopped over 98% of the shots they had faced with Hagg on the ice, they only stopped anywhere between 90-92% of the shots they faced with the four. Taking a look at rate stats, we can also see that the Flyers have allowed more shots against with Hagg on the ice than any of the four per minute, and only Philippe Myers has been on the ice for more expected goals against per minute. They aren’t a better shot suppression team with Hagg, they never have been over multiple seasons now.

Stats aside, even just watching these past few games it’d be a hard sell to suggest that Ivan Provorov, their only minus defenseman, has been worse defensively than Hagg. An impossible sell, really.

So if we’re to think of the Big Four as safe, that leaves just Hagg and Braun. Both branded as reliable, stay-at-home defensemen, only one of them has the results this season to back it up. One would imagine that if a defenseman was actually good at keeping shots to the outside, thus reducing their danger, that their team would give up fewer shots and chances against with them on the ice. And with Braun, the Flyers have.

All of which leads us to this: When Gostisbehere makes his return to the lineup, it should be in place of Hagg. His team-best goals for percentage and plus-minus rating of late isn’t an accurate representation of how well the Flyers had control while he was on the ice. Nor is it indicative of their future goal results with him out there.

In closing, things like on-ice shooting percentage and save percentage become even more important as the trade deadline approaches. Don’t get fooled by point totals driven by unsustainable shooting percentages, or you’ll just be disappointed by seasons end. Individually, players can run high for the majority of the season, even the entirety of a season some cases. Understanding what is likely versus what is unlikely to continue can save a team from making a deadline day mistake.

Data via Natural Stat Trick