clock menu more-arrow no yes

Filed under:

Exploring if Ivan Provorov is as good as advertised

New, comments

We had lofty expectations for the third-year pro heading into the season, but were we expecting too much?

Heather Barry - SB Nation ©

The Flyers’ unequivocal number one defenseman for the next decade. A franchise player. A future Norris Trophy winner. All things that have been said about Ivan Provorov. And while a number of Flyers have struggled through the 2018-2019 campaign, maybe none have been more disappointing than he has been based on expectations. A season like Provorov has had will, of course, lead to questions. Tough questions at that. Questions that may lead you to facing answers that you don’t like. And that... that’s exactly what we’ll be doing today.

Data used courtesy of Corsica.hockey and Natural Stat Trick. All shot metrics are from a 5-on-5 game state and are adjusted via Corsica’s method.

Great via the eye test, not via the stat sheet

Let me begin by saying that when I say he’s been great by the eye test, I’m not talking about this season. I’m talking about his full body of work over his first 250 NHL games; you’d be hard-pressed to find someone who believes Provorov has had a good season this year. Provorov, then 19, broke into the NHL and quickly became the Flyers’ lead back, averaging just under 22 minutes per game at the conclusion of his rookie season. His ability to outwork opponents in the dirty areas of the ice, break the puck out under pressure, and avoid making the big mistakes had fans and analysts alike falling in love with his game.

The thing is, he had both a Corsi-For percentage (CF%) and Expected Goals-For percentage (xGF%) below 50%, and negative relative stats. Given that he was playing top pair minutes as a 19-year old, it was easy to look past the numbers. The mere fact that he was able to hold his own in such a role and not fall flat on his face was a victory in itself. There were other factors beyond just his own play of course, like playing with Andrew MacDonald for the majority of the season, but even away from him it wasn’t like he was posting otherworldly numbers. He was good, but he was nowhere near the best — not that he needed to be in his rookie campaign.

One year later, in his sophomore season, we saw improvement in just about every facet of his game. Visually, he looked like a stronger player who was more adjusted to the pace of the NHL. Statistically, he posted positive RelT numbers in both CF and xG, improved his scoring rates, and had a positive on-ice goal differential for the first time. He was certainly good, and better than he was the year prior, but a 49.71 CF% and 50.9 xGF% still had him a far cry away from Erik Karlsson and company.

So where was the disconnect between his perceived level of play and what the data suggests over his first two seasons of his career? For starters, he excelled in specific areas of the game that can be eye-catching. Per Corey Sznajder’s manually tracked data, Provorov was one of the best defensemen in the league at exiting the defensive zone with control, and was not far behind in his ability to create zone entries with possession over the two-year stretch. Plus, he scored 17 goals in his sophomore season, a figure that no defenseman in the league could surpass. What’s more eye-catching than goals?

Now, you could say, well it doesn’t necessarily matter that he’s not driving play as long as he’s doing those specific things so well. However, the problem with that is that Corsi continues to be best predictor of future success at the public level. And while the stats don’t say he’s a bad player, they definitely don’t say that he’s been as good as we like to believe he’s been. And in the present, not even the eye test way of player evaluation backs him this time around.

What might be the single most frustrating thing about Provorov’s season has been the mental mistakes that he had previously avoided making. There have been countless times this season where his decision making with the puck has just simply been off, and outside of a few good games, he hasn’t been the same calm, smart puck moving defenseman that he looked like in seasons prior. His 47.15 CF% is the lowest mark of his early career, although relative to his team he’s outdone his rookie season, and his 49.12 xGF% slots in between his rookie and sophomore seasons. By every way of player evaluation, Provorov’s play has regressed since last season, which is not exactly breaking news to those who have caught a number of Flyers games this year.

While obviously not a good thing, it doesn’t mean that he can’t get back to where he was and improve further. Development is rarely ever linear, plus, there are a few things out of his control that have had an impact on everything we’ve discussed to this point. The first of which being the defensive pairings that he’s been a part of.

Heather Barry - SB Nation ©

Partners have been less than ideal, but not awful

When rookie Provorov blew people away via the eye test but lagged behind in play-driving, a lot of blame was placed on the MacDonald effect, and rightfully so. It’s no secret that MacDonald has been a Corsi drag on his teammates dating back to his Islanders days, and it was no different in the 2016-2017 season when he played alongside Provorov. In over 800 minutes together that season, the two posted a 47.46 CF% and a dreadful 44.63 xGF%. Away from MacDonald, Provorov’s CF% jumped above 52 and his xGF% improved marginally.

Along with MacDonald, Provorov has had three other primary partners through the first three seasons of his career.

Provorov’s Pairings Stats

Partner TOI CF% CF% Rel xGF% xGF% Rel
Partner TOI CF% CF% Rel xGF% xGF% Rel
MacDonald 1245.35 47.33 -2.14 46.20 -3.71
Gostisbehere 1044.78 52.12 3.12 53.11 4.20
Sanheim 861.30 46.80 -1.95 50.11 0.62
Hagg 511.31 46.48 -2.99 47.87 -2.45

Let’s make this simple: half of his partners have shown to be at the very least above-average NHL defenseman, while the other half haven’t. And when you’re locked in as the team’s number one defenseman, your partner probably shouldn’t ever be a replacement level skater. Considering how much time he’s spent with skaters who qualify as such, the “partner” argument is the strongest case to be made in Provorov’s favor. On the flip side, when you break it down by season, it becomes clear that the only time Provorov has won both the shot quantity and shot quality battle was when he played alongside Shayne Gostisbehere last season.

If you want to blame his poor shot metrics on MacDonald in year one, you can’t discount the role Gostisbehere played in his strong shot metrics in year two. In 2017-2018 Gostisbehere delivered a near Norris-caliber season as the pair posted a 53.11 CF% and a 55.29 xGF% in just over 750 minutes together. Away from Ghost, Provorov was a 46% Corsi player in just under 600 minutes with MacDonald or Robert Hagg. His true talent level? Likely somewhere in the middle, and that’s where his partner for the second half of this season comes into play.

While Travis Sanheim has the potential to become a great defenseman in this league, he’s not there yet and has been nowhere near as good as ‘17-18 Gostisbehere — few have. He’s also been nowhere near as bad as Provorov’s other partners, making him our best bet at evaluating Provorov with simply a good partner.

So, how have they done? The duo of Provorov and Sanheim hasn’t been good by Corsi, 46.67%, but neither has the rest of the team. Still, their -1.34 CF% relative is exactly that — negative. The good news is that weighted shots are much kinder to them; the two have a 50.62 xGF% and a +2.35 xGF% relative in their 816 minutes together. So they’ve been good, but they haven’t been as good as you’d imagine a pair with a true-talent number one defenseman and a true talent top-four defenseman would be.

Away from both extremes, Provorov looks to be a solid defender. But partners aren’t the only factor in his on-ice numbers, and the situations he plays in can give us an even further look at what he’s being asked to do.

His usage doesn’t absolve him

Another way of justifying his sub-par numbers is to point at the players that he’s matched up against on a nightly basis. It inherently makes sense; if a defender is constantly having to play against the likes of Sidney Crosby and Alex Ovechkin, they’re not likely to have a good time. But, if a player is out there with his own team’s top stars as well, the balance of power begins to even out.

While Provorov had already been facing teams’ best skaters through the first two years of his career, year three has brought his toughest challenge yet. Back in December we explored Provorov’s usage this season, and whether any defenseman had seen success under such circumstances. In the piece, Jake concluded that his usage at the time was resulting in a “near impossible endeavor in which to succeed.” But in the just under four months that followed, Provorov’s harsh deployment softened slightly, and no longer is he among the select few defenseman playing in the most challenging situations across the league. It’s not that his usage has been easy by any means, but now the data points to one of his teammates having faced tougher usage than him over the course of the full season.

Provorov & Gudas 5v5 Usage Factors

Player TOI% QoT CF% QoT xGF% QoT TOI% QoC CF% QoC xGF% QoC ZSR
Player TOI% QoT CF% QoT xGF% QoT TOI% QoC CF% QoC xGF% QoC ZSR
Ivan Provorov 28.38 49.08 49.28 29.46 50.15 50.07 43.03
Radko Gudas 27.41 46.62 47.66 28.62 49.85 49.70 42.36
Note: It appears that Corsica’s skater stats page does not currently include data from games 80 and 81 of the Flyers’ season.

What the above table tells us is that Radko Gudas has faced slightly lesser competition (QoC) and played with significantly worse teammates (QoT) than Provorov this season. It also tells us that he’s had very similar deployment in terms of zone starts, but even then he’s had bit fewer in the offensive zone on average in comparison to Provorov. We know that time on ice isn’t the best indicator of a player’s true talent level — especially not on this team with how they had ordered their defensive depth chart earlier this season — but it is one of three ways, along with Corsi and Expected Goals, to gauge the talent that someone plays both with and against. Luckily in our case, all three are in agreement in regards to Gudas versus Provorov.

Per HockeyViz, Provorov tends to see more time with the top three forwards on the depth chart than the average defender. While Gudas on the other hand spends a higher percentage of his time in the company of the seventh, eighth, and ninth forwards on the team, also based on the average defender. That, along with the defensemen they play with, leads to the relatively large gap between all three of their QoT data points, the biggest of which being based on their teammates’ Corsi. Provorov’s teammates on average have a higher CF%, xGF%, and play higher in the Flyers’ lineup, giving Provorov an advantage over Gudas, even though Provorov has faced marginally tougher competition.

There’s something to be said about facing the league’s top stars each game, but it’s the quality of the players you play with that is of greater contextual importance. This is mostly because who you play with is far and away more of a constant than who you play against. Using the Flyers’ game against the Toronto Maple Leafs last week as an example, at 5-on-5 Provorov played 21:58, with 94.31% of his minutes coming alongside Sanheim. However, he only spent 44.92% of his time on the ice with his most common opponent, Morgan Rielly. Furthermore, when you compare the amount of time he spent with the forwards on both teams, it presents more of the same. Sean Couturier, his most common forward teammate, was on the ice with him 51.21% of the time, while his most common forward opponent, Andreas Johnsson, was on the ice with him just 36.49% of the time. This will almost always be the case, as a coach has full control over who a player plays with, while they do not have full control over who they play against.

Over a full season, a player will tend to play with the same skaters, or skaters with a similar talent level, but face differing levels of competition from shift to shift. No amount of attempted line matching can override that, and that’s one reason why Gudas’ QoC numbers are as close to Provorov’s as they are, despite it feeling like Provorov is always up against top talent. Even so, while his opponents have been slightly stronger, other defensemen in the league like Mark Giordano, Aaron Ekblad, and Hampus Lindholm to name a few, have faced similar, or tougher competition and have all had better seasons than Provorov.

The whole reason that we’re even comparing the two defenders is to show that Gudas has been able to have success while being dealt a tougher hand than Provorov, who has struggled through the majority season. Despite the circumstances, Gudas boasts a 49.38 CF%, +2.76% RelT, and a 51.18 xGF%, +3.52 RelT, figures well north of Provorov’s from earlier. And if Gudas can thrive in such a situation, why can’t Provorov do the same? That right there is when the question, “Is Provorov actually as good as we think he is?” was born.

Heather Barry - SB Nation ©

Closing thoughts

Ivan Provorov is a good player. Ivan Provorov is not a truly great player — yet. The question at hand is one that I don’t think has a definitive answer right now, but his statistical profile doesn’t exactly paint a pretty picture, even after you account for all the factors that are out of his control. The only time Provorov has driven play anywhere near an elite level for an extended period of time was alongside Gostisbehere a season ago, while Ghost was in the midst of the best single season by a Flyers’ defenseman since Chris Pronger’s arrival in 2009. If that’s the type of performance it takes for Provorov to not be teetering on the wrong side of the shot and scoring chance battle, that’s a problem.

As an upcoming restricted free agent, Flyers’ General Manager Chuck Fletcher will be tasked with signing Provorov to his first non-entry level contract in the near future. Last July we looked at what his next contract may be, and came up with three main, long-term options; six years with a $6.1M average annual value, seven years with a $6.7M AAV, or eight years with a $7.3M AAV. Now, while his ‘18-19 season certainly won’t help his agent in negotiations, it probably won’t hurt him that much either, thanks to Provorov’s reputation around the league. If I’m the Flyers I’m pushing for our six-year projection from last Summer and looking to avoid a cap hit north of $7M. Any higher and you’re betting on him becoming a better player than he currently is, and while it obviously wouldn’t be the worst bet to make, there’s always the risk that he never truly hits that next level.

Maybe all that needs to happen for him to quiet any doubt is a full season alongside a good, consistent partner. Whether that’s Sanheim, Gostisbehere, Myers, or someone who’s not even currently under contract with the Flyers remains to be seen — but to say Provorov being the player that we expect him to be moving forward is crucial would be an understatement.