• Checking In On Eberle

    by  • March 22, 2013 • Hockey • 4 Comments

    If you follow me on Twitter, you’re probably aware that I’ve been having a bit of fun with how utterly right I have been about Jordan Eberle’s point production this season. Way back last summer, I had this to say:

    I’ve said elsewhere that I expect Eberle to get somewhere between 55 and 60 points and that still seems like a pretty decent projection to me. I wouldn’t be stunned if played a full year or close to it, and didn’t break 50 points though. I’d be a lot more surprised if he broke 70 points without a significant improvement in his Corsi or shot numbers.

    My views were not universally well received. Eberle is currently on a 57 point per 82 games pace this year. Two kind of caveats: first, it’s hard to get too worked up about stats of any type this year because this isn’t exactly a normal NHL season. I made this point before the start of the season and figured I’d get people saying I was trying to back away from what I said. As it so happens, I have no need to back away from what I said (subject to my next point) and one would have figured that having played with Eberle would benefit him but the point still stands – this isn’t a normal NHL year and whatever happens should be taken with a grain of salt.

    The second point is that, as with high school math, how you got to your answer matters. Process trumps results and all that. With that said, I thought it might be interesting to re-visit what I said and the extent to which it holds up.

    If you assume a TOI increase to bring him to 45th amongst NHL forwards, that he is a Tanguay when it comes to on-ice S% and average IPP and that he plays all 82 games, you come up with 82 games * 14.82 minutes * 10.8% on-ice S% * 28.23 S/60 * points on 72% of goals = 44.46 5v5 points. He had 56 5v5 points last year. I’m not at all sold that he is a Tanguay – everyone looks like a Tanguay when the pucks are going in – so I sort of suspect that that’s a pretty generous expectation for him at 5v5. It also knocks 12 points off his season from last year.

    I’m going to look at this as if we’ve played a full season and Eberle’s played 82 games. The TOI is pretty close – he’s played 15.05 5v5 minutes a night so far. The on-ice S% is absurdly low – the Oilers are shooting 5.98% with Super Sniper™ Jordan Eberle on the ice. The S/60 is up substantially – I’ll talk about below – to 34.5 S/60. He’s had points on 73.3% of the goals that he’s been on the ice for.

    If you run the math on that, it’s 82 games * 15.05 minutes * 5.98% shooting percentage * 34.5 S/60 * points on 73.3% of goals = 31 points. 31 points! After he had 56 at 5v5 last year! So I look pretty good with two of the numbers I came up with so far – the 15.05 minutes and the percentage of goals on which Eberle would get points (his individual points percentage or IPP).

    Unfortunately, those are the things a monkey could have predicted – all I did was just pick the mid-point for the top 90 F in the NHL in terms of 5v5 and assume that Eberle would be that. Pretty easy. As for the IPP thing, that’s pretty easy to predict too – very few guys seem to continue getting points on 85% of the goals scored when they’re on the ice in the long term.

    Let’s be honest: predicting that Eberle’s on-ice shooting percentage would go down was pretty much shooting a sluggish fish in a small barrel that didn’t have a lot of water in it. The fact that the Oilers are shooting 6% with him on the ice is well below what we’d expect and, if you’ve watched all of the games, you’ve noticed a ton of chances for Eberle and RNH (with Eberle on the ice) that were shot wide or saw a puck jump over a stick. In 25 or 30 game stretches, that sort of thing happens.

    The interesting thing, for me, is the thing that I’ve been most wrong on: the shooting rate when Eberle’s been on the ice. Last year, it was 28.3 S/60; this year, so far, it’s been 34.6. That’s a huge leap. There’s something going if you look at the shot attempts though. Here’s Eberle’s Corsi data broken down for the past two seasons:

    The Corsi% is, in fact, up, and that’s great, subject to a point I’m about to make. The mix of shot attempts when Eberle is on the ice is different this year though. Last year, 51.7% of the shot attempts with Eberle on the ice made it through to the net at 5v5. In 2010-11, it was 51.8% This year, it’s been 59.1% so far. The best argument that the pro-Eberle segment had was that his shot rate would increase and they have; I wonder whether or not this is something that tends to carry on though or whether, over the course of a long season, you run into stretches where shots are blocked and miss the net.

    The other point I wanted to make was in the nature of that old Vic Ferrari question, who’s zooming who? Here is the Corsi data for Eberle with and without Taylor Hall so far in his career.

    Now, in fairness to Eberle, Hall tends to play with Hemsky when he’s not playing with Eberle and Eberle’s most common LW other than Hall is…probably Ryan Smyth, but still. The Hemsky/Hall pairing has historically played tougher minutes than when he’s together with Eberle and yet Eberle has a massive falloff without Hall while Hall’s is much less noticeable.

    In summary, Jordan Eberle’s not going to have a lower on-ice S% than Zac Rinaldo forever, so the points number is going to go up. If the S/60 number at 5v5 is real, then he’s going to crush what I gave as a points range for him. I’m not sure that it is though, because of the curious rise in shots getting through to the net. If you assume that that’s statistical noise – and I suppose it’s something I can look into – then we’d expect something like 30.2 SF/60 at 5v5 for Eberle going forward. That decline would eat up a healthy chunk of the positive regression in on-ice S%. Worth watching as the season goes forward.

    Email Tyler Dellow at tyler@mc79hockey.com


    4 Responses to Checking In On Eberle

    1. dawgbone
      March 22, 2013 at

      I mentioned this in the C&B top 25.

      I’m curious if the shots increase is tied at all to the fact that the shooting % is down. I mean when you score a goal, the shots stop occurring and you start the play 114 feet away from the opposition’s net. All the work you just did to gain the zone has to be replicated after you’ve already played part of a shift. In other cases you leave the ice completely.

      When the puck doesn’t go in the net, the puck remains in the zone. Maybe you get a rebound opportunity or regain possession of the puck and get another shot several seconds later. Maybe the goalie covers the puck and you get a face-off with the puck still in the opposition zone.

      Seems to make sense at a glance anyways.

      • dawgbone
        March 22, 2013 at

        or start it 89 feet away… whichever math makes more sense.

    2. Jake
      March 23, 2013 at

      Regarding the Corsi data:

      In addition to the “Eberle-less Hall gets Hemsky, while Hall-less Eberle gets Smyth” point that you make, couldn’t some of the difference between the two also be due to the fact that a good chunk of Eberle’s time apart from Hall has been while Hall is injured/suspended? With Hall out of the lineup, the opposition is likely able to key in on Eberle line (leaving them without a Zetterberg/Datsyuk conundrum, if you will). The extra 250 mins of 5v5 ice time for Eberle seems to support this.

      (Not that approx. 15-20 games worth of ES data should cause a 3.5% drop relative to Eberle’s other ~1000 5v5 minutes. However, if that is causing a drop, as well as a lower quality of linemate is causing a drop, then the 3.4% difference perhaps isn’t as significant as it seems.)

    3. Pingback: Spectors Hockey | NHL Blog Beat – Saturday, March 23, 2013.

    Leave a Reply

    Your email address will not be published. Required fields are marked *