Dear NHL: Stop the madness (Updated)

If there’s one thing that anybody who regularly watches the Edmonton Oilers knows, it’s this: there is some incredibly unwatchable hockey played after a team is eliminated from playoff contention. And there’s absolutely no reason it has to be that way.

The Gold System

At the 2012 Sloan Sports Analytics Conference, a guy named Adam Gold proposed a simple change to the way that the draft order is calculated for teams outside of the playoffs. Rather than a weighted lottery system, which still encourages teams to lose for better draft position, Gold suggested that once teams are eliminated from the playoffs a second race – one for the first overall pick – begins. Teams eliminated earlier would have more time to run up wins, giving them a better shot at the top draft pick; that means that the draft order would be similar but each team would always have an incentive to win, right down to Game 82. 

Gold’s presentation can be found here, and Rob Pettapiece did a detailed analysis for NHL Numbers last summer, but that’s the gist of it.

C’mon!

NHL teams get really funny once they have no incentive to win. It’s not a coincidence that Ales Hemsky, Ryan Nugent-Hopkins and Lennart Petrell suddenly decided that it didn’t make sense to keep playing through injury once the Oilers were eliminated from playoff contention. It’s also not a coincidence that Calgary’s line combinations have been getting pretty hilarious in recent games; here’s the forward lineup from their game yesterday against Nashville:

  • Baertschi – Reinhart – Cervenka
  • Hudler – Horak – Aliu
  • Bancks – Byron – McGrattan
  • Hanowski – Street – Jones

Eight of those guys have combined for 387 AHL games. This season. That’s not even the worst part: in those 387 games, they’ve combined for only 159 points. The other four forwards are enforcer Brian McGrattan, rookies Ben Hanowski (four games of professional hockey experience) and Roman Cervenka, and actual NHL player Jiri Hudler.

It’s embarrassing to see, and virtually every team in the situation does it (including the Oilers), because the NHL’s current system encourages it. It doesn’t have to be that way.

Stop the madness.

Update

There was a lot of concern in the comments section that truly bad teams wouldn’t be able to improve under this system, so I went back and calculated a rough version of how this would have worked last year. I started counting points after a team could no longer overtake the eighth seed in their conference by winning all of their games; ideally we’d have goal differential and regulation/overtime wins as a tie-breaker but I didn’t go into that much detail here – instead I used actual NHL standings and full-season goal differential as tie-breakers because that information was much more readily available. Here are the first overall pick standings from last year using that system: 

Team W L OT Points Gold Actual
Columbus 7 6 0 14 1 2
Minnesota 4 2 1 9 2 7
Edmonton 3 4 2 8 3 1
Montreal 2 1 3 7 4 3
Toronto 2 2 1 5 5 5
Anaheim 2 3 1 5 6 6
Tampa Bay 2 1 1 5 7 10
Carolina 2 2 0 4 8 8
Calgary 2 0 0 4 9 14
Winnipeg 1 1 1 3 10 9
N.Y. Islanders 1 3 0 2 11 4
Buffalo 0 0 1 1 12 12
Colorado 0 2 0 0 13 11
Dallas 0 1 0 0 14 13

In some ways, this is actually a fairer system, since the worst team in the league (Columbus) ended up picking first overall, rather than a significantly better Oilers team. The far right column shows where teams actually picked, while the column labeled "Gold" shows where they picked based on this system. Keep in mind, too: these teams had no incentive to win, and it’s probably safe to assume that with the incentive to win some of the worst teams on this list could have been more competitive than they are.

Even as-is, though, this doesn’t look that bad. The worst teams still make the earliest picks, and only the New York Islanders (dropping from fourth to 11th) saw their draft position fall more than two spots. This system serves the same purpose as the lottery (making bad teams better through draft picks) but does it without encouraging them to lose. 

Recently around the Nation Network

    • RexLibris

      Not to blow my own horn here, but that is why I wrote this: http://flamesnation.ca/2013/4/22/for-the-calgary-flames-a-tutorial-on-tanking

      And I second your call for Barkov, with a compensatory option in Monahan. Either way, when it comes around to Jay Feaster’s time to draft I hope he has a choice between Nichushkin and Nurse.

      I’d like to watch him tell the hockey world how either of those players are going to be the best players in their draft years ten years from now.

      • Romulus' Apotheosis

        Barkov, Monahan, Lindholm.

        Where we are now, we’ll get one for sure. That’s good news!

        The Flames are probably tanking at tanking because they know their off-the-grid pick will be available whenever they want him.

    • Ducey

      Luck. And home cooking.

      They are much better at home.
      They are outshot badly every game yet win. The game against Detroit, Detroit scored on their own net twice.

      • Romulus' Apotheosis

        I should have used “competitive” more circumspectly… but the point remains Calgary is having a hard time losing. They lost last night only 4-3 afterall.

        I imagine Feaster is pulling his hair out as we speak: “what do I have to do to lose around here anyway!?”

        BTW if we lose the next 3 we have a very good shot at 4th.

        BARKOV!

        • Rocket

          Barkov is listed as 2nd best draft prospect by central scouting so I doubt The Oilers get him.

          Still, never underestimate this teams’ ability to fall to 2nd overall pick.

          As Jordan Eberle just said: “Three years of this sucks.”

          • Romulus' Apotheosis

            Unless I’m mistaken central scouting ranks in two lists: NA and EURO.

            Barkov is 1st on the Euro list. But most lists have him behind Jones and the East Coast 2.

  • I’m fine with the lottery system the way it WAS before they made even this change to it.

    The fact is the draft works the way its suppossed to. Bad teams fail long enough to gather excellent players, climb out of basement. There is nothing the NHL can do to stop bad management so the Islanders wallow for particularly long times, but eventually the cream rises back to the top.

    I dont care for getting cutesy with extra point calculations and side races that would be more akin to a side pot in Poker than something that benefits truely bad teams.

    The lottery isnt a prize for teams that can win. It’s a measure to prevent teams that cant from NEVER being able to climb out of the basement.

    As it stands right now the 9th team in the West could select 1st Overall in the NHL entry draft, and that would be a terrible injustice to the clubs that cant compete as it is.

  • From the Pettapiece analysis:

    What’s encouraging about these results is they imply that this new system would maintain the entire reason to have a draft (distribute new talent to encourage parity) while removing its current worst feature (Fail for Nail, Stop Winnin’ For MacKinnon, etc.). Half the time, the first pick would go to one of the four worst teams — CBJ, MIN, NYI, TBL. Almost never does a 9th or 10th-place team receive it. So the bad teams still pick earlier, but now they’d actually have to earn that pick. I can get behind that.

    The idea is that the order isn’t much different, but teams now have to win rather than lose.

    • G Money

      The only flaw with that system is that expansion teams get crushed. The 1992-93 Senators were 10-70-4 and the Sharks were 11-71-2. They could have feasibly been the worst teams in the league but draft 10th or later. We know the NHL is looking at expansion and that is why they won’t switch systems.

      • Expansion temas have changed – you look at teams like CBJ and MIN, and while they were bad they were nothing like those early 1990’s teams.

        Also: I find it hard to believe current owners much care about the fate of expansion teams.

        Edit to add: And anyway, that’s easy enough to fix with an expansion exemption, if they were so inclined.

        • G Money

          Just because CBJ and MIN weren’t colossally awful doesn’t mean the next round won’t be. There is also the possibility for any team to be historically bad as well.

          The owners care about the fate of expansion teams because expansion fees are big money but revenue sharing takes some back. An expansion exemption likely wouldn’t work because if an expansion teams ended up around 20th they wouldn’t have a clear way to determine their spot in the draft.

  • Ducey

    The only problem is that the truly lousy teams will never be able to get better. They miss the payoffs and then can’t win enough games to score a really great draft pick.

    There may be teams that tank but the reality is that most crummy teams stink becuase of poor management, injuries, poverty, bad luck (Tampa Bay is 28th and is -1 in goal differential, WTF?) or some combination of those. They can’t just turn it on in a second season.

    All you are going to do is reward teams who are usually pretty good but hit a rough enough patch of injuries that they miss the playoffs. They get the #1 and then bounce back into the top 10. Meanwhile the NYI and FLA’s never can get better.

    I like the system as it is.

  • I don’t see how it would solve all that ails, for every Calgary where management pulls the chute, there is the Edmonton Oilers where the players have pulled the chute.

    And I don’t think the players give a flying you know what where the team drafts.

  • The only problem that the pathetically bad will stay pathetically bad. Perhaps Number 17 drafts first and then you go back to the last place team drafing 1 in subsequent rounds.

    That and it would have meant the flames would have had a bunch of high draft picks over the past years. And that is just a bad idea. If we could some how eliminate the flames from the formula it could work.

    By the way what ever happened to Team Fist?

    • No, the draft order wouldn’t change that much.

      The worst teams have been eliminated for some time now – so they’d have 10 games to run up wins. Team 17 probably only gets one game, because they only get to count wins toward their draft position after they’re mathematically eliminated.

      • G Money

        Your analysis (actually, your conclusions) are incorrect I think. You’ve shown that the draft order would not change very much, but that’s with teams that were playing under the old system. You would expect that the draft order almost certainly would look VERY different if the teams did not have the same incentive to tank.

        You should expect that in the big picture, this system would favour the teams just-out-of-the-playoffs rather than the truly putrid teams. Over the longer run, you would likely see the better draft picks going to the mediocre 17-25 teams rather than the truly bad 26-30 group.

        And in this case, I think the solution is worse than the problem. The tradeoff you’re making is that
        – teams playing meaningless games try harder to win them, but
        – mediocre teams now have a chance to return quickly to playoff contention, while truly bad teams have a harder time getting out of the basement

        And as others have pointed out, it doesn’t make a difference if your players have checked out – many of them, especially pending UFAs, wouldn’t care where the team drafts.

        Might be better to simply fine teams that do not dress a minimum number of their starting lineup from prior to elimination.

          • G Money

            Um, no, not sure how you got that from my post.

            My thesis is that under the current system, all teams (whether decent teams just missing the playoff cutline, or the truly putrid) go in the tank.

            The end result is that all teams are now truly putrid (and you see the feces fest like last nights Cgy-Nsh game).

            But since the lousy teams have less talent and are likely already lower down in the standings, things don’t change much standings wise for the rest of the season. So you end up with a high correlation between the lousiness of the team and the higher (i.e. better) draft picks. Just as it should be.

            Under an “incentive to finish higher” system, teams just missing the playoff cutline – who have vastly more talent than the truly putrid – can make up a ton of ground simply by winning whatever is left. After all, if the lousy teams had the ability to win just by trying harder, they wouldn’t be out of the playoffs early, would they?

            So … you reward the mediocre teams with better draft picks, and punish the truly bad teams (who need the draft picks to improve) with lousier draft picks.

            Alternatively, if player apathy is the number one driver of lousy finishes, then this new system won’t make a difference at all.

            So, in summary … the system you’re proposing either punishes lousy teams with lousier draft picks and rewards mediocre teams with better ones, OR it has no effect on draft position.

            All for the questionable benefit that the teams might play a bit better in meaningless games.

            Like I said – better to fine teams that ice lousy lineups, if that is what you want to improve, rather than create a system with perverse incentives and negative long-term draft consequences.

          • Rambelaya

            @ GMoney

            I think the presumption of the Gold model is that the truly bad teams will be out of the playoffs significantly earlier than the mediocre bad teams, and have more change to accumulate wins.

            This isn’t a “count wins from some arbitrary day equally for everyone”, it’s about the lousy teams being eliminated earlier, and getting a head start on getting a few wins.

            Let’s look at 2 cases:

            1) – 2 closely matched teams are in 9th/10th, battling for 8th. Theoretically they should get similar draft picks. Assuming neither makes it into 8th, then the 10th place team would be mathematically eliminated maybe a day or a couple days before the 9th place team. That doesn’t give either team much chance for wins, but should theoretically allow them to get at least 1 more win in the remaining games than the 9th place team. Either way, because they were both eliminated closely together, their wins would be reflective of their ranking.

            2) a 9th and 15th place team. The 15th place team is horribly worse, and struggles for every win (see Colorado this year). But Colorado has been mathematically eliminated since when? – beginning of April? Maybe earlier? So every win after that counts towards their draft ranking. The 9th place team is still playing for “playoffs” until maybe the last game of the season, and may not even have a chance to get 1 more win after being mathematically eliminated. This means that Colorado will have a significantly better draft ranking since they’ve been able to accumulate at least a handful of wins between when they were eliminated and when the 9th place team was eliminated.

            I’m not saying whether it would work out better or worse in the long run for any given position, but I can see why lousier teams aren’t necessarily at an immediate disadvantage. They simply have so much more time to collect wins.

          • G Money

            @ Rambelaya,

            I get the concept of the early start for lousy teams helping to offset the inherent unfairness of the exercise.

            The point is – if the bias doesn’t work, it means that a lousy team lost its draft position, which defeats the point of the draft positioning system (which is to let lousy teams improve).

            If the bias does work, then you get a system that doesn’t have any meaningful effect. If teams know that they can’t change the draft position, will that change their behaviour (“shut it down”) from what is happening now? Doubt it.

            Which leads to a secondary question … what nasty unintended consequences will you inject?

            One person already pointed a possibility out, which is that teams who recognize they have no chance will tank as early as possible in the season to maximize their chances. So far from improving the quality of play, you will make it *worse*. Tanking from the beginning of the season sounds a lot worse than tanking once you’re eliminated, yes?

            Secondly, what about teams that shut players down for injury? Players play through injury because the pain and risk are worth the possibility of making the playoffs. Once that’s gone, they shut it down – as you or I guarantee would in the same sort of situation.

            Should the Oilers be in a position to ask Hemsky and RNH to play out the season to improve the Oilers draft position (a situation utterly meaningless to Hemsky given he’s probably quite aware that he’s at high risk of being traded)? Is that a fair position to put teams OR players in?

            The whole purpose of the regular season (for players) is to determine who gets in the playoffs, and secondarily positioning. That’s why once you’re eliminated, the games automatically become meaningless for the players. Anything you inject that benefits or hurts the team based on results after the games become inherently meaningless seems questionable to me in terms of its motivation value.

            So bottom line … if you do this, what you’ve done is implemented a complicated system with either *no* effects, or *bad* effects, and likely all kinds of negative unintended consequences, all in exchange for the hope that a few more meaningless games will be a little less meaningless.

            Not worth it.

            By the way, this discussion in some ways is a repeat of what has happened repeatedly in the past. Teams would tank to get first overall. So they introduced the lottery to ‘fix’ it. Teams still tank, this time to get a lottery pick. This year they introduced the new lottery that allows all non-playoff teams to participate (though that might be an Oiler-specific rule). Teams still tank. It’s the nature of the beast.

  • Quicksilver ballet

    With too many teams and not enough talent to go around, it was only a matter of time. Developmental teams in almost half of the NHL markets, at big league ticket prices. Milking each market for every possible dollar has become more important than the product on the ice. The league should start drafting 19-20 yr olds and contract by 4 teams.

    The Coyote situation has gone silent. Could it be, the league is finally going to pull the plug on that market?

    • vetinari

      “Fists” are for winners… put them away until (hopefully) next year…

      As for the quality of hockey for eliminated teams, it does suck. The other option, which I don’t propose, is relegation to the AHL. Wouldn’t that be a kick in the nards and encourage teams to not finish last? Sure, you won the draft but you have to finish in the top of the AHL the year following to make it back to the NHL.

      However, I do like the idea of somehow giving a “bonus” to the non-playoff teams based on their wins/points after being officially eliminated. Maybe each point improves their chances by an additional one percentage point or something like that?