11
Photo Credit: Dennis Wierzbicki-USA TODAY Sports

Benning the Numbers (Part 2)

In Part 1 of this series of articles, I pulled a variety of key statistics on Matt Benning that I use to characterize and assess players, particularly defenders.

Those numbers should have reconciled well for you vis a vis the eye test. We have a player who helps the team a lot on the attack, but his efforts are partly offset by his defensive shortcomings. He gives up a lot on defense, and is not a good penalty kill guy.

Even so, overall the team does better than break even with Matt Benning on the ice, and that should generally be considered a success for any 3rd pair guy. His strength on PP2 is a bonus.

Advertisement - Continue Reading Below

But wait! There’s a fly in this ointment (there always is).

Benning isn’t alone on the ice. While his on-ice numbers do a decent job of capturing his influence on the team and his teammates, what about the influence of those teammates on him?

Let’s take a look.

Fayning Interest

Advertisement - Continue Reading Below

Like penalty differential, another now-longstanding addition to my statistical overview of defenders is to look at their results with defensive partners. We can call this the “Fayne” lesson.

Back in the day, analytics types ripped almost every defensive trade/signing made by the Oilers, and with good reason. Mark Fayne was an exception – one of the few defensive signings that were Analytics Approved(tm).

<stats rant>

People often use that circumstance to incorrectly blame the Fayne signing on analytics. Nuh uh. In reality, Fayne was signed by a distinctly non-analytics GM, and the signing was universally lauded by old school SEEN ‘IM GOOD MOAR BIG types. Because Fayne was big, he was a veteran, and he had solid results on a defensively well coached team.

Advertisement - Continue Reading Below

So if you want to try and use Mark Fayne as an example of some sort of analytics mistake which invalidates the entire field, you should also acknowledge that he was a scouting/seen him good mistake and invalidate those approaches as well, no?

</stats rant>

In any case, while I doubt the analytics naysayers learned anything, one big difference is that analytics folks after the fact asked the question as to why the numbers were mistaken on Fayne. As it turned out, Fayne’s numbers were goosed by the fact that he was playing within the paint-dryingly locked-down New Jersey system, which was a good fit for him. More importantly, he did this while partnered with the (at the time) outstanding Andy Greene. That made him look a lot more capable than he was.

Context matters, eh?

Joshua Tree

Embed from Getty Images

Advertisement - Continue Reading Below

Since then, no analysis, especially of a defender, is complete without some sort of U2 (WOWY – with or without you) analysis.

Here are Benning’s results with and without his 5 most common defensive partners.

Partner EVTOI with CF% with CF% wo CF% wo Benning GF% with GF% wo GF% wo Benning
Klefbom 537 50.4 52.7 51.7 47.7 60.6 41.3
Sekera 211 49.8 53.4 45.8 25.0 60.0 21.4
Auvitu 115 52.7 50.4 51.3 61.5 65.2 63.0
Russell 100 45.8 51.9 47.8 57.0 51.0 43.8
Nurse 96 58.0 50.5 50.0 83.0 47.0 52.7

A few things stand out.

One is that, overall, while there’s variation here and there, Benning’s results were not goosed by his partners. He did well with almost everyone, and the others, for the most part, did not do better away from him. (There is a statistic that summarizes this WOWY effect into a single number, called a “RelTM” statistic. Benning sits at a healthy +1.41 for shot attempts and an excellent +9.15 for goals by that measure)

The second thing that pops out of these numbers for me is that I sure hope the Oilers try Benning with Nurse next season. While the ice time is too small to draw any sort of robust conclusion, in the equivalent of 6 to 8 games of ice time together, they do appear to work well. Really well. It’s worth a try.

And thirdly, I’m still scratching my head over why the Oilers let Yohann Auvitu go. Yes, he was quite bad defensively, and sheltered like a Sarah McLachlan song. But man, the Oilers outscored like a bat out of Heck with him on the ice.

In any case – we can be pretty comfortable that this isn’t a Fayne situation. Benning’s decent numbers follow him through his pairings, they aren’t because he’s always partnered with someone amazing.

The McDavid Effect

Nov 9, 2017; Newark, NJ, USA; Edmonton Oilers center Leon Draisaitl (29) celebrates with center Connor McDavid (97) and defenseman Matt Benning (83) after scoring the game winning goal in overtime at Prudential Center. Mandatory Credit: Ed Mulholland-USA TODAY Sports

Speaking of someone amazing … the bro with the sickest flow … by now we all know that a critical part of evaluating any Oilers player is to see to what extent his numbers were goosed by McDavid. It’s not a matter of if, it’s a matter of how much!

In Benning’s case, we see something interesting:

  with McD without McD
5v5 TOI (mins) 349 780
5v5 CF% 52.8 50.4
5v5 GF% 58.7 45.1
5v5 Points 13 3
All Situations Points 15 7

Did you spot what I mean by interesting?

The uninteresting part is the individual scoring. Without McDavid (like Caggiula), Benning’s scoring dries right up. If we overemphasize Benning’s points in contract discussions, we’ll get in trouble. No surprise there.

That said, since Benning didn’t spend an unusual amount of time with McDavid, if we estimate what he’d look like with a “normal” top line player in place of McDavid, Benning would likely be closer to 15 points on the season. That’s noticeably weaker, but still OK for a third pairing guy.

The surprise to me is in the shot and goal metrics. Benning is one of the few players to hold his own when McDeity is not on the ice! His shot metrics are above breakeven. And while his GF% falls below breakeven, in both cases, Benning does considerably better than most of his teammates do without Connor – about 0.8% better on the shot metrics (vs 49.6%) and a full 4.5% better on the goals (vs 41.6%).

Given what we usually see in this regard, that’s a significant plus.

What We Learned

OK! So in part one, the numbers pointed to Benning as a capable third pairing defender with decent attacking ability mostly but not entirely offset by defensive shortcomings.

After pulling apart the WOWY and McDavid effects, well, for once it turns out that Benning is a depth guy who is actually still pretty capable when he’s away from McDavid.

This is good! A big problem the Oilers have is how terrible the team is when McDavid is not on the ice. It is more likely that Matt Benning is part of the solution rather than part of the problem.

In Part 3, I’ll use this information to pull together some comparables as a basis for evaluating Benning’s new contract.



  • Just a Fan

    “One is that, overall, while there’s variation here and there, Bennett’s results were not goosed by his partners”

    Who is this “Bennett” you speak of and why do his partners want to goose him?

  • Big Nuggets

    But how many offensive zone starts did he get without McDavid? Was it a lot? and could those numbers be matched by a defender that was too busy taking defensive zone starts? There is always another layer to the context of these stats.

    • There is always another layer, and it’s easy to get overwhelmed for sure.

      I think by the time you’re looking at WOWY on zone starts you are usually digging past the useful limits of the data.

      In any case, I’ve generally found zone starts have relatively modest impact on overall numbers unless they are at an extreme.

      Benning’s starts with McDavid (60%) were not far off his overall ZS numbers, so the uplift suggests the overall impact would be close to negligible.

      It’s a good thought though, and cogently pointing out shortcomings in the data or the analysis always makes for fair game.

      • Big Nuggets

        Its a good article. My comment was not meant as a criticism, just worrying that people will draw the wrong conclusions.

        The discrepancy between cf with Russell and with Nurse is interesting but I would guess it has to do with how they were used by Tmac more than Nurse and Benning making magic together. If TMac is playing a Nurse Benning duo I bet he is putting them in positions to score, whereas with Russell the may have been given more defensive assignments, or perhaps they went with Nurse Benning combo against weaker teams. Admittedly I am just guessing, I’m not an analytics expert.

  • WhoreableGuy

    Unbelievable the amount of dislike that Auvitu got last season. He was brought in for cheap ($700,000) as a 6 or 7 D, did his job well and could even fill a forward spot as well. He wasn’t there taking ice time away from anyone that deserved it. I wish the Oilers tried to keep him.

    • crabman

      I’m not sure the team didn’t try to keep him. I didn’t see any articles that said as much, of course I could be wring.
      I think it was more that he wasn’t getting many games or playing time in 2 consecutive years in the NHL and thinks it would be better for him and his career to go back ti Europe.
      I think Auvitu could have been a good asset on a better defencive team that had a stronger top4 that ate up most of the minutes. If he was left to play little minutes and sheltered, while getting pp time he could carve out a real niche, but that’s a pretty specific role.

  • crabman

    I’ve said it before and I’ll say it again, I liked Benning as a bottom pair defenceman this year and didn’t see him as part of the problem. I’m happy to see articles really digging into the topic and validating my eye test.
    I think a lot of the time people get more caught up in what a player can’t do and lose sight of the bigger picture. And once the pitchforks sre out everyone who was on the fence seem to jump on board with tue player trashing.

  • Bumbo

    It always amazes me how much the numbers tilt when McDavid is on the ice. A fun article would be an entire team comparison of when he’s on and when’s he is off.