• Isolating The Problem With Taylor Hall

    by  • January 22, 2014 • Uncategorized • 16 Comments

    Diving into doing something analytical is the sort of thing that can spin you off in a direction that you weren’t anticipating. Take Taylor Hall. It has not escaped my notice that he’s getting killed in the Corsis this year. This is a new thing. Here’s what Hall’s career looks like to date.

    For three years, given the context of the Oilers (nuclear wasteland), his numbers are really good. This year, he’s completely off the rails. The Oilers are giving up an extra six SAA per sixty with him on the ice and generating 11 fewer SAF. The Corsi% is at 42.7% instead of 50.4%.

    This has been kind of hidden because Hall is putting up points. It’s weird but if this was a decade ago, nobody would be talking about Hall having a bad year. There’d be talk about his +/- being the fault of the goalies and the defence and the Oilers needing to support Hall better. The thing is though, this stuff matters. The bad Corsi% matters. Let’s look at the goals for and against and shots for and against with Hall on the ice for 2010-13 and 2013-14.

    You can see that, through his first three years, Hall was +1 at 5v5 with the goalies in and the Oilers got outshot by 12 over 2480.13 minutes. Again, given the context he was in, that’s not bad. This year, he’s -9 at 5v5 with the goalies in, and the Oilers have been outshot by 80 with him on the ice. Yes, the save percentage when he is on the ice is down, but some of that is covered off by an uptick in the shooting percentage when he’s on the ice. The real problem is the getting badly outshot thing, which is a new problem. It is a serious problem though, regardless of what the boxcars say.

    I wondered why he seemed to be posting this atrocious Corsi% in game after game against teams that he’d done well against before this year. I probed into two games, one against the Blues from 2012 and one from 2013. I was able to create this table, which I’ve posted previously, of how the Oilers generated shots with Hall on the ice against the St. Louis Blues with Taylor Hall on the ice on February 29, 2012 and December 21, 2013. As I mentioned, we can see that the shots off zone entries in which the puck was carried in were way down.

    There’s an obvious complement to this, which is how the Blues created shots against the Oilers with Hall on the ice at 5v5 in that game. I’ve put those tables to the left. It looks like the Blues were just killing the Oilers by creating shots off controlled entries. I took a bit of a closer look at it though and I think it’s a bit misleading – the Oilers had one really bad shift where they got slaughtered and gave up four shot attempts. In terms of the number of times that the Blues carried the puck into the Oilers end and created at least one shot attempt, things were pretty similar to the halcyon days when Hall was cock of the Corsi walk.

    This got me thinking about a concept that I fooled around with last summer – breaking the game down into shifts and looking at results on a per shift basis. Let’s look at the defensive end of things first.

    Raw numbers are a bit confusing, given the wildly different shift volumes, so let’s express that table in terms of percentages.

    You can see that this year looks pretty similar to the past three seasons, in terms of the volume of shifts that end up with no shot attempts against. The reason that we see a bit of an uptick in terms of the SAA/60 when Hall is on the ice looks to me to be more shifts where the opposition have hemmed them in. 6.9% of Hall’s 5v5 shifts this year have seen 3 or more SAA; prior to this year, it was 5.5%. It’s not a huge thing in the grand scheme of things but I think that that’s the answer to that.

    OK. That’s the answer to that. Now let’s delve into the offensive zone. First the raw numbers:

    Raw numbers stink, let’s look at percentages.



    There’s your headline right there. In his first three years in the NHL, the Oilers got a shot attempt on 49.7% of Taylor Hall’s 5v5 shifts. This year, they generate a shot attempt on 39.7% of Taylor Hall’s 5v5 shifts. I…I don’t know what to say. 20% of shifts on which the Oilers generated at least one shot attempt for with Hall on the ice from 2010-13 have just evaporated. This does not seem like A Positive Development.

    I did a rolling ten game graph of Hall’s rate of shifts on which he got at least one shot attempt for. I think it kind of tells the story here. He’s kind of rolling along, 2013-14 starts and he goes off the cliff. Like an 18 wheeler.

    It’s notable, I think, that the rate at which Hall generated multi-SAF shifts once the first SAF is generated is essentially unchanged. Here’s that information, in handy table form.

    Once the first SAF is generated, it’s Showtime, as it’s generally been in his career. It’s important to think about the data here – I suspect the vast majority of shifts with multiple SAF saw the multiple SAF recorded while the puck was still in the offensive zone. “Shot attempt, puck recovery, shot attempt” rather than “Shot attempt, puck recovered by other team, puck cleared, puck taken back, zone entry, shot attempt.” The latter seems a lot closer to me to how the game actually works than the former.

    That points us in a very specific direction: something that the Oilers do is making it much tougher for them to generate the first shot attempt this year. Once they get that first shot attempt, the party’s on. But there’s something that makes it harder for them to get that first attempt. Perhaps by coincidence, perhaps not, that fits nicely with the data I generated from that Blues game, in which the Oilers rate at generating shots off controlled zone entries cratered from the February 29, 2012 game to the December 21, 2013 game.

    I think it raises two pretty critical questions. First, is what I observed when contrasting the two Blues games a real thing, in terms of the Oilers suddenly having become much worse at generating shot attempts off controlled second entries? Second, if that is the case, why is it? Does something that the Oilers are doing leaving their own end or entering the offensive zone result in attacks with less support for the puck? Have the coaches asked them to do something that’s resulted in fewer shot attempts per carry? I appreciate the importance of questions about hit totals as much as the next guy (mostly because I enjoy exasperated responses and a circus), but these would be pretty fine questions for someone to raise with the coach in a media availability.

    A final point: lost in the wreckage of the Oilers season is that one of last year’s problems, the team’s complete inability to do anything when Hall wasn’t on the ice, has gotten much better this year – last year they were around a 42% Corsi% without Hall on the ice, this year they’re at 46% or so. If Taylor Hall was Taylor Halling, things would probably look even better than that, as the non-Hall lines got more offensive zone starts and fewer defensive ones. That leads to more favourable zone starts for Hall and there’s probably a bit of a virtuous cycle that kicks in: good things beget good things. The shot attempt generation problem with Hall on the ice seems like a thing that, if sorted out and fixed, could pay some pretty immediate dividends.

    Or, you know, MOAR FACEWASHING. Either way.

    Email Tyler Dellow at tyler@mc79hockey.com


    16 Responses to Isolating The Problem With Taylor Hall

    1. T.Rade M El
      January 22, 2014 at

      Trading Hall plus a pick or packaged player(s) could likely land the top pairing D that plays 25-30 mins a game the team must have to be legit contenders. . It’s what a good GM would do.

    2. David Goodhart
      January 22, 2014 at

      Isn’t the answer available from the recent Staples interview with MacT? The GM said he wanted Hall in particular to play an offensive game heavier on the cycle with more prolonged attacking zone possession instead of attacking with speed for one shot and out. It may be that Hall has been told to avoid the quick shot on zone entry to avoid quick puck retrieval and zone clearance by the defenders. With results the opposite of intended.

    3. godot10
      January 22, 2014 at

      The rest of the Oilers really aren’t doing any better under Eakins than Krueger if you take out all the easy games against the Eastern Conference, who the Oilers didn’t play last year, certainly with the margin of error.

    4. Bruce McCurdy
      January 22, 2014 at

      Morbid fascination compels me to ask about the shift with 8 SAA. Do you know which game that was, & was it in garbage time or more significant? Was it an exceptionally long shift?

      Holy Hannah, IIRC even Chorwick & Strudney only gave up 11 SAA in the 225 inglorious seconds that comprised The Shift.

    5. LoweAsWeGo
      January 22, 2014 at

      Taylor Hall entered the league reckless, was babied for the first few years (tantrums indulged, guys jumping off the bench every time he accidentally fell down, doesn’t try to make his linemates better), and is now a spoilt child and spoilt prospect. Oilers fans and players were too sensitive because the expectations were too high. That sensitivity derailed the team from making progress by winning, and everyone got caught up defending Taylor Hall’s honor and wasted energy on a plug like Hordichuck or a cheapshotting coward like Eager.

      Take away the expectations and entitlements. Success will take root just like in Pittsburgh, Chicago, Los Angeles, and most importantly like in Oklahoma City.

      Taylor Hall needs a complete rebuild as a person, and there’s no stability or leadership within the team to help that. He’ll probably injure himself and be the next Eric Lindros. We should have drafted some goalies and defencemen early, instead of chasing flashy teenage scorers. The whole roster is in jeopardy because the team can’t outscore the opposition, and the only players that resign must really enjoy boozing after a big night of losing.

      Katz should take the hundreds of millions in tax money we’re gifting him, and pay for a whole roster swap with some sunbelt team. All the way up to and including Kevin Lowe.

      The Oilers should forget about being the best team in the league. Focus on moving from 30th up to 29th past Buffalo, and get some positive momentum going — otherwise every season becomes a failure when we don’t win the cup, make the playoffs or any other distant goal. Achieve a modest goal, and then set a modestly higher goal. It’s that simple.

    6. OilLeak
      January 22, 2014 at

      Could coaches be keying in on Hall (more so then before) realizing that he is usually the catalyst for the Oilers offense thus suppressing his shot attempt totals?

      Take Hall’s shot attempts away to keep the Oilers offense at bay. Could be an old wives tale.

    7. Tyler Dellow
      January 22, 2014 at

      • Bruce McCurdy
        January 22, 2014 at

        Thanks, Tyler.

      • January 23, 2014 at

        Perfect song choice here.

      • Curcro
        January 25, 2014 at

        There’s your problem too much Jones.

        • Woodguy
          January 25, 2014 at

          Jones flying the zone early in the middle of that clip is so Ryan Jones.

    8. speeds
      January 22, 2014 at


      Is there anything to the idea that Hall’s Corsi drop might be injury related? How were his stats prior to his injury? The games after his return? Anything there?

    9. Skipper
      January 22, 2014 at

      I think your missing the fact that he played with nuge and eberle primarily those first 3 years.

      This year those guys have been split up, hurting his corsi but improving the teams corsi.

      The guy that said to trade hall needs to give his head a shake.

      • Simon Lamarche
        January 23, 2014 at

        He actually spent more time with Nugent-Hopkins this year than last (63% instead of 60%).

        He did spend less time with Eberle this year (46% instead of 66%).

        If it is indeed Eberle who made his line go from a 42% to a 50% shot differential last year, Eberle is underpaid and LoweAsWeGo is right, the Oilers should trade Hall.

        I don’t buy it though.

        (5v5 stats from http://stats.hockeyanalysis.com)

    10. Murat
      January 23, 2014 at

      If I’m reading right, zone entries for both teams are happening with the same frequency and in the same ways in the past, and Edmonton’s d-zone looks almost exactly the same as in the past, and if Hall’s line generates one shot they get more at the same rate as in the past. The only difference is that more often, Hall’s line doesn’t get a shot at all.

      With this in mind, what makes sense?

      1) Injury. We do know that Hall has missed at least one morning skate due to a sore shoulder.
      I just don’t see it. To me, if an injury was holding Hall back, we would either see it in reduced foot speed or his abilities to win the puck along the wall. I don’t think he looks slower and his rate of multiple shot attempt shifts is unchanged. If he (or his line) is still recovering pucks, can we really say an injury is to blame?

      2) Tactics. MacT has said he’s pleased with Hall’s decision making with respect to dumping the puck when the carry option isn’t there.
      On first glance, this could be it. Hall’s “carry” number goes down in STL game two. That said, wouldn’t his “dump” number go up as a result? Perhaps he is dumping it more, causing his carry number to decrease, but his line isn’t recovering the puck very well?

      I am struggling to see an obvious answer. They’re still entering the zone but when they do, they are not getting the shots. Could this really be a tactical thing? Maybe those BS drop passes they’re slowly phasing out actually led to more shots off the rush. Maybe someone is hurt and they’re worse at cycling after retrieving a dump.

      I don’t see it, but I also don’t see Hall as reckless or a baby or whatever else is being suggested.

    11. matt
      January 27, 2014 at

      Any chance of posting the scatter plot of the rolling average graph, rather than a connecting line?

    Leave a Reply

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