• Applied Scoring Chances

    by Tyler Dellow • January 7, 2012 • NHL • 8 Comments

    Jason Gregor raises a good point in the comments to my last post:

    The Oilers have 32 PP goals 5on4 at the midway point of the season, last year they had 41. Wouldn’t you say they are significantly better on the PP. I would argue that the combination of RNH and Eberle have made them much more dangerous.

    Is it possible to account for the amount of “easier” goals this year, basically from better passes, than last? I think that is the major factor on their PP. They, mostly started by RNH and Eberle, move the puck around so much more crisp and clean that they actually have an advantage with the extra man now.

    I’m not sure if it is even possible to calculate that, but to me that is the biggest difference on the PP. Rarely does RNH stand facing the boards. For years we’ve seen their PP waste time with guys having their back to the play.

    Just a thought if it is possible to look at that. I’m assuming we’d have to break down every PP goal, but the cross seam pass from RNH to Eberle is the most effective new addition to the PP that I’ve seen.

    Gregor’s not the only one to think this – I’ve felt myself that the PP has looked more dangerous this year and I’ve noticed other people who I respect making the same comment. Here’s the thing though: the chances don’t say so. When I wrote that post, I used David Staples’ chances to do a quick comparison of the ES chance ratio for this year and last – it was pretty close, like .443 last year to .46 this year, or something like that. Vic Ferarri has pointed out before that chances tend to correlate pretty tightly with shot counts and it seems to bear out – the shot count hasn’t really improved much and the ES chance ratio needle has barely moved.

    I didn’t check the PP chances at the time but, as I said, I felt Gregor was making a sensible point, so I went and looked. According to Derek Zona, who maintains records of Dennis’ counts (we really should here), the Oilers had 261 5v4 chances last year. So far this year, with 39 games logged…130 chances. They’re actually getting a bit more 5v4 time this year too, 6.4 5v4 minutes per game versus 5.7 5v4 minutes last year.

    So what does the chance data tell us? It says what the shot data says – this team isn’t any more effective than last year’s team. You’re left to bet on the quality of the chances having improved dramatically, which would suggest a sustainable high S%. Everything we know about S% tells us that that’s very unlikely to be the case.

    About Tyler Dellow

    8 Responses to Applied Scoring Chances

    1. Coach Pb
      January 7, 2012 at

      Came to the same conclusion. RNH or not, they aren’t going to succeed with these shot rates.

    2. dave
      January 8, 2012 at

      Interesting points and I dont disagree with anything you say – but there is something that caught my eye

      you said ‘everything we know about s%….’ got me thinking – there appears to be 3 attitudes to stats that will tell us something meaningful about a team or player – roughly they might be –

      1. Stats are generally rubbish (they dont tell the real picture of whats going on)
      2. There are loads of stats that tell us meaningful things – but you have to pick the right ones
      3. We dont yet have the ‘golden’ stat that truly evaluates a player/team

      Now I’m personally 2 or 3 – but whose to say that 1 isn’t correct – I just dont know…..

    3. marconiusE
      January 8, 2012 at

      Do you know if shooting% stays relatively consistent from professional level hockey all the way down to the beer leagues?

      • Tyler Dellow
        January 8, 2012 at

        There’s no reason to think that it would. As you get further away from the NHL, the skill level varies more.

        There’s an analogy to be made to baseball here. BABIP is pretty random at the Major League level; pitchers don’t have a great deal of ability to control balls in play. As you get further from the majors, this is less true.

        • marconiusE
          January 8, 2012 at

          but wouldn’t the lack of skill in goaltending as you went down through the leagues keep the shooting percentage higher?

          (please ignore the double post below)

    4. Hambone
      January 8, 2012 at

      How about more articles about how incompetent Lowe and Tambellini are? I liked those.

      • marconiusE
        January 8, 2012 at

        but wouldn’t the lack of skill in goaltending as you went down through the leagues keep the shooting percentage higher?

    5. January 9, 2012 at

      It could be as simple as Hall and Eberle (16 of the PP goals) are simply better shooters than we have employed on the PP before – more specifically they have become better shooters in their sophomore season.

      Eberle has gone from ~%11 to ~%18 and hall 11.8 to 13.5. Considering those two are shooting a lot of PP shots for the Oilers, if you believe their real ability is closer to the sophomore numbers versus the rookie numbers, that could account for real improvement for the Oilers (although Eberle’s shooting % is very high, it’s the second highest amongst the top 30).

      It probably wouldn’t be a huge difference, and you’d prefer higher shot rates for sure but would explain real improvement even if a small amount.

    Leave a Reply

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