• An Educated Guess About Realignment And The Playoff Format

    by  • February 27, 2013 • Hockey • 6 Comments

    The details of the latest NHL realignment proposal leaked today, resulting in much puzzlement about one aspect: only the top three teams in each division are guaranteed a playoff spot, with the fourth place team in each division vulnerable to missing the playoffs if the fifth place team in the other division has a better record than they do. It’s being described differently elsewhere but that’s essentially what this is, with the added variable that the worst team to qualify for the playoffs will face the best team in the first round, regardless of division.

    Many people seem to be questioning this, wondering why the league would dilute the divisional principle that it seems to be pursuing. I was kind of wondering this too, so I went back and looked at the data for the last seven years. What I did was pretty straightforward: I just reorganized the league into the new divisions and looked at what the standings would look like. There are 14 possible cases of a fifth placed team in one division being better than a fourth place in the the other division – two conferences and seven years. (Obviously, you can’t have both fifth placed teams better than both fourth place teams in their conference.)

    It turns out that this has actually happened relatively frequently: five times out of a possible fourteen. In 2008-09, the Hurricanes finished fifth in the Atlantic, with 97 points; Montreal finished fourth in the Central Division with 93 points. In 2010-11, Montreal finished fifth in the Central Division with 96 points; the Rangers were fourth in the Atlantic with 93 points.

    Over in the Western Conference, in 2005-06, Vancouver would bump out Winnipeg (things are a little more fictional here, with Atlanta having played a very different schedule but it’s hard to imagine they would have piled up more points with a Mid-West Division schedule than they did in the Southeast). In 2010-11, the Kings displace the Blues, 98 to 87. In 2011-12, Calgary knocks out Dallas, 90-89.

    The schedules that teams will play in the proposed conferences are actually pretty equal – a Pacific Division team would have 71 games that are the same as those played by a Mid-West Division team and a Central team’s schedule will include 73 games that are identical to those played by an Atlantic team. It’s not unreasonable to compare between the two divisions and, if a fifth place team posts a better record than the fourth place team in the other division, let them into the playoffs.

    (Only two third place teams out of 28 would have made the playoffs in the last 7 years that didn’t actually make it; the 2008-09 Minnesota Wild and the 2008-09 Florida Panthers. Presumably, their records would have been inflated enough with the extra games against terrible teams that finished behind them to at least make it look a little more respectable.)

    I would also think, although I’m not entirely certain of this, that a fifth place team in one division that is able to accumulate more points than the fourth place team in the other division is almost certainly a stronger team, given that they will have had to play extra games against teams that managed to finish ahead of them in the standings. If you finish fifth in a division and finish ahead of the fourth place team in the other division in terms of points, you’ve managed to do so despite playing those teams that you couldn’t overcome in the standings an extra 7-8 times. I would think, although I haven’t done any math, that there’s a pretty strong implication that you’re the better team there.

    One thing that this doesn’t do is address the issue of the odds being better of making the playoffs in a 14 team conference than in a 16 team conference. Elliotte Friedman referenced this in his piece about the realignment:

    So, let’s say the Oilers finish fourth in the Pacific next year, but get to the playoffs as the “remaining team” with the most points. And say Chicago wins the Midwest (ahead of St. Louis and Nashville) and has fewer points than Pacific winner Vancouver. The Oilers technically become a Midwest Division club. They would play the Blackhawks in the first round and, with a win, the Blues or Predators in the second.

    It wouldn’t be a surprise if the league wasn’t 100 per cent thrilled with this solution, because it might hurt the development of some divisional rivalries. But, it also increases the possibility of, say, Vancouver vs. Chicago. What it also proves is that players will accept the possibility of worse travel if it means a better chance at the playoffs, as suspected when the first re-alignment attempt was nixed last season.

    Taking what Friedman says here as correct, I think that the concerns of the players have been somewhat misrepresented by the media covering this. This realignment proposal does not mean a better chance at the playoffs: 8/14 or 8/16 is 8/14 or 8/16. That issue isn’t being addressed. It is, to my eye, a fairer system, because it ensures that a team like the 2010-11 LA Kings, with 98 points, makes the playoffs instead of the St. Louis Blues.

    The one fewer playoff appearance per 14 years that every team in the East will suffer struck me as something of a red herring anyway; the NHL isn’t likely to be a 30 team league that much longer and if you’re a player whose career lasts long enough to lose such a playoff appearance, you’re probably very wealthy and aren’t that worried about the impact a single playoff appearance where you scored some goals could have on your career earnings. The fact that this is fairer, in terms of allocating playoff spaces based on results, does strike me as something that the players would worry about – while owners can live with missing the playoffs some years with a stronger team and making it in other years with a weaker team, the players have a shorter window to worry about and don’t have time for universe to find its own way back to equilibrium.

    (Aside: while we should celebrate the NHLPA’s apparent refusal to accept a system in which it was more likely that a truly terrible team would make the playoffs, part of me wants to have seen just what Bettman’s piling of small sample size events (SO/OT wins), incentives to play for OT (guaranteed point for each side) and small divisions could have done for the playoffs. We could well have seen some truly terrible teams make it by virtue of luck and a soft division. Like, um, the 2007-08 Edmonton Oilers, who won an astounding 15 games in the shootout and would have finished fourth in the Pacific that year but would lose a tiebreaker to Chicago, who would cross over.)

    There’s a principle here that’s somewhat intuitive, I think: as the divisions or conferences or whatever we call them get smaller, the likelihood of a “bad” conference in which a genuinely terrible team gets rewarded increases. To take it to an extreme, if you had 15 divisions of two teams, with the first place finisher in each making the playoffs and one wildcard for the best second place team, you’d probably end up with some conferences that had two objectively terrible teams, one of which gets to make the playoffs simply by virtue of being paired with a slightly worse team. Four divisions of seven or eight teams isn’t that bad, obviously, but it does up the risk of a team that isn’t in the top sixteen getting that fourth seed.

    The whole thing is kind of an interesting example of the tension between the commercial and sporting realities of running a sports league. If it was a pure sport thing, you might not have any divisions and play a 58 game home and home schedule, with the best 16 making the playoffs. Turns out the market will absorb more hockey than that, so it’s 82 games. Also, it’s way cheaper in terms of travel and makes more sense TV wise if teams play more games in or around their time zones, so you get conferences.

    As soon as you step away from having one table with all of the teams in the league in it though, you run into sporting fairness problems – how do you ensure that the best 16 teams make the playoffs? The NHL’s solution has been not to worry about that. It’s no secret that the Eastern Conference has been generally weaker than the Western Conference since approximately forever. According to Jeff Sagarin, since 2005-06, the Eastern Conference has had 6, 7, 4, 9, 5, 8 and 6 of the top 16 teams in the NHL. They still got 56 playoff spots over that time, despite having just 45 teams who were amongst the 16 best in the NHL. As an aside, you could say that the East deserves to be punished for its decade plus inability to match up to the West by having an additional two teams to contend with, one of whom has been a pretty perennial playoff contender. If you aren’t going to take building a decent hockey team seriously, why should you be rewarded with the same opportunity at a playoff spot as the half of the continent in which this pursuit is taken seriously?

    I would guess that the reason that the league has proposed what it has is that it is trying to find a way to balance commercial and sporting realities. Divisional playoff games, in a setup in which divisional regular season games are most common, tend to pay off in the form of valuable high profile games the following year between two teams that don’t care for one another This is, I think, particularly true where two teams meet up a few times within a five year period or so. By my count, under the proposed setup, 10 of 56 first round series in the past seven years would have been re-matches of prior first round series from the past seven years. The existing format provided 26 re-matches in 105 total playoff series. I would expect the proposed format to produce a significantly higher number of rematches – if I could play out the playoffs under the proposed new system, the rematch number would skyrocket as teams met in the second rounds, both in rematches and in series that rendered subsequent first round series rematches.

    Whether you like what the NHL proposes to do or not is a matter of taste, I think. As someone who’s concerned about getting roughly the right teams in the playoffs, I like the introduction of a mechanism whereby a fifth placed team could make it in place of a fourth placed team with a worse record. I think it’s best to view this change to the playoff format as one that’s intended to ensure more entertaining matchups by giving us teams with a history together more often, which has been tempered with a recognition that it’s important that the best teams make the playoffs. That seems tough to argue against to me.

    (Fun fact: Calgary would have appeared in the playoffs six times in the last seven seasons under the proposed re-alignment, instead of four. Pity for them.)

    Email Tyler Dellow at tyler@mc79hockey.com


    6 Responses to An Educated Guess About Realignment And The Playoff Format

    1. palefire
      February 27, 2013 at

      Please correct me if I’m mistaken, but I believe the reports I saw suggested that each team would play home-and-home series against the teams in the other three divisions, and either 5 or 6 games against the teams in their own division. That leaves only 56 games in common between the schedules of two teams in different divisions (plus two games against each other), which I think is way too unbalanced to make a straight-across points comparison. I’d have no complaints if the only thing that counted for the wildcard berth was, say, all your extra-divisional games plus your final home and away games against each team in your own division. But that’s complicated, and we already know that 3 points for a regulation win and 2 points for an overtime/shootout win is too complicated for the geniuses at NHL HQ.

      • Tyler Dellow
        February 27, 2013 at

        They’re playing 3 against out of division conference teams. So Edmonton will play 3*7=21 against the Mid-West. That means that 21+32+(3*6=18 divisional games) are the same. It’s about the same as now, north of 90%.

    2. Pingback: Spectors Hockey | NHL Blog Beat – February 27, 2013.

    3. don__cherrys_conscience
      February 28, 2013 at

      Tyler, from a lifelong hockey fan living in NHL Siberia (Columbus, Ohio), just wanted to let you know that your work is great. Your posts have a depth that many NHL markets in the U.S. (not just the media in Columbus) never get to regarding understanding why things happen in a game.

      Just an observation regarding the realignment plan: I see the league “mimicing” the NFL, the 800 lb gorilla of sports, in stressing the importance of regular season divisional games. Currently, when I go the league’s web site, and look up the standings, I always sort by conference. Last’s season L.A. Kings were a perfect example of the “just get in the top 8 and qualify” mentality promoted on a lot of U.S. broadcasts, especially by the national broadcaster, NBC.

      The NFL stresses winning the divison as the only sure route to qualifying for the postseason, and not depending on tiebreakers as a wild card (note how the NHL picked up the term “wild card” from the NFL, everybody knows what that means). I am a long-time NFL season-ticket holder, and the first thing I look at each April, when the schedule is finalized, are the dates of the three divisional home games. NFL team executives openly talk about building rosters to regularly beat their divisional foes. Some divisions even have their own identities…for example, the AFC North is known for reams having open-air stadiums, playing in cold weather, and having hard-hitting defenses and strong rushing attacks.

      If this plan is approved, and the divisions stayed aligned this way beyond the first 3-year agreement, it will be interesting to see if clubs

    4. don_cherrys_conscience
      February 28, 2013 at

      build rosters based on the makeup of their divisional opponents.

    5. Tach
      March 1, 2013 at

      If you acknowledge fairness as a key component, I don’t know why they don’t totally do away with the West/East axis entirely. Top 3 teams in each Division/Conference in, plus the next 4 best teams from any division. The “Wild Card” teams are seeded into the divisions to give the top teams the most favourable match up (ie the 16th place team into the 1st place team’s division, the 15th place team into the next best division leader’s division). I haven’t worked out Tyler’s math, but I must imagine that in under this system, the odds of the best 16 teams not being in the playoffs is incredibly small.

      Could you end up with a Boston/LA first round match-up? Sure. But that is not logistically or logically more troublesome than a Nashville/Vancouver first round match-up once you allow for the 4 wild card teams in any event.

      Plus it avoids the inherent assurance that the Eastern teams basically don’t leave their time zone for the first 3 rounds of the playoffs, which is far more offensive to me as a Westerner than the regular season travel issues.

    Leave a Reply

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