The final playoff spot in the Eastern Conference currently belongs to the Toronto Maple Leafs. They hold that berth by a single point, and if they can cling to that toehold for 15 more games they’ll be heading to the playoffs for only the second time in the last dozen years.
The stakes are high and the margins narrow, and success or failure will almost certainly come down to the four-game homestand which closes out Toronto’s regular season. If there’s a sudden chill in the air, it’s because the Leafs’ opponents in those games are excellent, and if there’s one thing Toronto has been lousy at this season it’s showing up against excellent foes.
Of the five Canadian teams still in the playoff race, Toronto is uniquely bad when facing a top-end rival.
If the trend above holds, Toronto could see its playoff hopes die in the season’s final days, and even if the team were to make the postseason it would then need to find a way to turn that record around. Edmonton and Montreal have similar problems, while the picture is rosier for Ottawa and Calgary.
The question is whether this is really a valid way of looking at these teams’ seasons.
There’s a wide range in quality between the NHL’s best teams and its worst teams. A night in Washington or Pittsburgh comes with a much lower expectation of success than a similar evening spent in Glendale or Denver. But is matching up well against the powerhouses a better indication of future playoff success than beating the also-rans?
An analysis of the last three NHL seasons suggests that breaking down a team’s performance this way does not improve predictability, but the details are interesting.
A gambler calling playoff series based on each team’s regular season goal differential would have been correct 60 percent of the time, choosing the right winner in 27 of 45 playoff series. Another gambler, following the same methods but using just goal differential against top 10 teams, would also have correctly chosen the winner of 27 series.
That’s awfully close, and a look at the correlation between regular season goal differential and playoff wins over those three years may help explain matters (a perfect correlation equals 1.0):
Regular season goal differential (all teams) vs. playoff wins: 0.35
Regular season goal differential (top 10 teams) vs. playoff wins: 0.34
Regular season goal differential (bottom 10 teams) vs. playoff wins: 0.15
The first two numbers are virtually identical, as we would expect from two approaches with the exact same success rate at predicting playoff series. The third number, which shows how well teams did against bottom-feeders, is much weaker, but still seems to have some value.
Put another way, a win over Washington does tell us more about a team than a win over Colorado, but the win over Colorado still tells us something.
We might illustrate this point with two versions of the Los Angeles Kings.
On the left, we see the 2013-14 Kings. They destroyed bad teams, running up a massive plus-38 goal differential against bottom-10 clubs. They weren’t nearly so good against quality teams; the average playoff team goes minus-four against a top-10 club, so minus-one is nothing to be excited about.
On the right, we see the 2015-16 Kings. They were much better against top teams. Over this three-year span, only one club did better than that plus-12 goal differential against the toughs. They were below average against bottom-feeders, but then bottom-feeders don’t make the playoffs.
The team on the left won the Stanley Cup. The team on the right won one playoff game.
Those are just two examples, and certainly the reverse happens too. Chicago last season got knocked out in the first round after devastating bottom-10 teams during the regular season (plus-47) but getting its teeth kicked in by the other two-thirds of the NHL (minus-21). However, it should be clear that a team which gets its points in the regular season by picking off the weakest members of the herd can suddenly be a real threat to the tougher marks in the playoffs.
That takes us back to the Maple Leafs.
Toronto has made the playoffs once in the past 11 years, thanks to a hot start and a 48-game schedule in 2012-13 which didn’t quite give the team enough time to play its way out of the postseason. The last time the Leafs were playoff-caliber over 82 games was 2004 — Pat Quinn was coach, Mats Sundin was captain, and the salary cap was just a glimmer in Gary Bettman’s eye.
Whether or not the Mike Babcock-coached, Austin Matthews-led 2016-17 version of the team reaches the same benchmark as that 2003-04 squad remains to be seen, and we make no guarantees here. What we do know is that it would be foolish to write them off either now or during a hypothetical playoff run just because they had problems against specific opponents over the first 80 percent of the season.