Started From The Bottom Now We Where?: Follow-up seasons from last-place teams

The Leafs, as you all know, finished in last place last season. No one was especially bothered by this, especially not after it snagged us Auston Matthews. But a new season is soon to be upon us, and most fans (although not all) seem to be expecting the Leafs to be better this year.

Let's see about that.

I looked at the last eight teams to finish in last place, and how they did the following season.

N.S. stands for "next season", and refers to the year after the team finished 30th. Records from the 2012-13 season are marked with a * to indicate the total has been prorated to 82 games. Columbus gets a big boost from the prorate, but it's consistent with their movement in the standings.

If you'd like to go a bit denser, I've put the full table with all the team stats in a Google doc. L.P. stands for "last place", and refers to the team stats from the year the team finished 30th, as distinguished from "N.S." stats. For every stat, the leaguewide rank in that stat for that year is in brackets. Have a look at some terrible numbers.

To be clear from the outset: this is a very limited sample size. The idea is to look at last place teams and try to give context for what happens with them; it's not meant to be statistically predictive, because it won't be. The primary takeaway from this chart is that bad teams are bad.

That said, the average last-place team improved by about 10.9 points, or 8.3 if you remove the short season where Columbus had a huge jump. The bottom-feeders tended to jump by 4.6 wins (3.6 without Columbus.) This is a little misleading, because four of the teams improved at least 12 points, and four of them improved less than 4. Still, if you just want a back-of-the-envelope prediction for the Leafs, this would make for a 2016-17 finish between 77 and 80 points, and 32-34 wins.

The Leafs, though, are an unusual last-place team for several reasons. For one, they have the highest point total and are tied for the highest win total of any last place team since 2008. They also are the only last-place team in our sample to be above 50% in 5v5, score-adjusted CF% (the Leafs were 13th in the NHL, with 50.23%.) The Leafs also had easily the best-ranked penalty kill in the last-place sample, although special teams numbers bounce around quite a bit from year to year. They were undone by somewhat below-average goaltending and a miserable shooting percentage, but insofar as there are good last-place teams, the Leafs were one. (We're grading on a curve here.)

I'm going to look specifically for comparables for the Leafs. We'll start with the last good last-place team.

The 2007-08 Tampa Bay Lightning

The team in question was Tampa Bay, and if you look at the spreadsheet, you'll immediately see why. The Lightning were middle-of-the-pack in goals per game, penalty killing, possession, and shooting percentage, and their goaltending was just absolutely terrible. Their .9010 is the third-worst 5v5 save percentage since 2007. The 29th place team in 2007-08, Toronto (deep sigh), was a full .006 better. Let's take a second away from the numbers and look at what was going on there.

The Tampa Bay Lightning won a Cup in 2004 with Nikolai Khabibulin as their starter, ably backed up by a rangy, impressive-looking goalie named John Grahame. (Grahame actually had a better save percentage in the Cup year than Khabi, putting up .913 in 29 games to Khabibulin's .910 in 55.) After the 2005 lockout, Khabibulin signed a big free agent deal with Chicago, and the Lightning gave Grahame the reins for 05-06. Grahame turned out to be unable to sustain his impressive performance as a backup, putting up a miserable .889 in 05-06, backed up by a similarly subpar Sean Burke (.895.) The Lightning were at the beginning of a goaltending wilderness in which they, at different times, gave the reins to Johan Holmqvist, Marc Denis, Kari Ramo, Marc Denis, and finally Mike Smith. Until Smith in 08-09, none of these goalies put up .900 in a season. (And it was touch and go after that, too: between 2007 and 2013, the Lightning had a different goalie lead them in games played every year. The Ben Bishop trade only gets better the more you think about it.)

At any rate, 2007-08 was the depths, as four goalies--Holmqvist, Denis, Ramo, and Smith--combined to form a platoon that played like a fart noise. As a result, a quite decent team finished last in the NHL. Vincent Lecavalier and Martin St. Louis both had great years (92 and 83 points), respectively, and they were followed by Brad Richards and Vinny Prospal, but it was nowhere near enough. The Lightning tried desperately to rectify the situation at the deadline--Richards and Holmqvist were dealt to the Dallas Stars at the deadline for Mike Smith, Jussi Jokinen, Jeff Halpern, and a fourth--but that didn't fix things.

The last-place Lightning got two pieces of good news: they drafted Steven Stamkos in June, and their goaltending stabilized the next season as Mike Smith put up .916. And...they got five points worse. They were only saved the indignity of back to back 30th-place finishes by the total implosion of the New York Islanders.

What the hell happened?

As you can see from the spreadsheet, the Lightning went from being a decent team with bad goaltending to a bad team with average goaltending. Several factors played into this:

1. Vincent Lecavalier suffered a shoulder injury on a dirty hit by Matt Cooke (what a surprise) in the 81st game of the 2007-08 season, and was never the same again. He declined from being a superstar to merely good, dropping from 92 to 67 points. There's nothing wrong with being a 67-point player, but a 25-point drop is going to hurt.
2. Vinny Prospal was declining with age.
3. Brad Richards was gone.
4. Dan Boyle, the Bolts' 1D, was told by ownership that he could either waive his NTC or he would be put on waivers and claimed. Boyle went to San Jose.
5. On a related note, Jay Feaster, then the Bolts' GM, resigned rather than keep dealing with the owners interfering with his management of the team.
6. Coach John Tortorella, who led the team to its Cup, was fired for not having good enough goaltending. While I have mixed feelings on John Tortorella as a coach (i.e. I don't think he's very good), the Lightning absolutely would have been better off keeping him...
7. ...because his replacement was Barry Melrose.

I could go into what a trainwreck of a coach Barry Melrose is, but I think it's common wisdom at this point and this article would have an additional 2000 words. Suffice it to say, he was awful and he helped make the team awful.

Reassuringly for the Leafs, once you look more closely, there's little or no commonality between them and the last good bad team. All of the Leafs' key players are 27 or younger, and (knocks on wood) as long as JVR recovers fully, all of them should be in good health come the start of the season. The only major players the Leafs have shed in the past year are Dion Phaneuf and P.A. Parenteau; I think Parenteau is going to be missed more than people realize, but neither is even close to the level of a Richards/Boyle loss. Leafs' ownership continues to trust the process. Lou is, by his own account, bleedin' blue. Mike Babcock is not Barry Melrose.

So if the Lightning aren't the right comparable, let's look at the two teams who had the biggest jumps: the Columbus Blue Jackets in 2013, and the Sabres this year. Are we in line for a similar increase?

The 2011-12 Columbus Blue Jackets

They had a PDO spike in the short season. Woo.

The 2014-15 Buffalo Sabres

This is more interesting. Buffalo quite brazenly tanked for a pick in the McDavid-Eichel draft and I don't at all blame them for doing it.

People get on the Sabres for dismantling their team, but the truth is, their team was already total garbage, as you can see from the table. The only two teams to finish under 55 points since the 2005 lockout are Buffalo in 2013-14 and Buffalo in 2014-15; given that they were already one of the worst teams this century, it made no sense in summer 2014 not to plan for a tank.

Having said that, you have to acknowledge that they did tank, and they put on one of the worst non-expansion teams in the history of the NHL. Probably the saddest thing is that they actually had quite respectable goaltending (from Michal Neuvirth and Anders Lindback; our boy Jhonas Enroth had a poor year.) But they were the worst possession team since the stat has been tracked, by a huge margin. According to Corsica, the worst team in the NHL typically has a score-adjusted 5v5 CF% in the 42-44% range. The Sabres are the only team since 2007 to finish under 40%, and they wrapped at 36.16%. This is equivalent to how the Washington Generals do in basketball.

The tank successfully concluded as the Sabres snagged the No. 2 overall pick. Buffalo promptly disposed of their team and replaced it with a different, better team. The changes started behind the bench; Tim Murray fired Ted Nolan and upgraded to Dan Bylsma.

As for the on-ice personnel: of the Sabres five leading scorers in 2015-16, three were new to the team (Ryan O'Reilly, Jack Eichel, Evander Kane), one had an ELC tryout the year before (Sam Reinhart) and one more than doubled his point production (Rasmus Ristolainen.) Either at the 2015 deadline or in the offseason, they added Zach Bogosian and Cody Franson on defence. Their special teams zoomed from dead last to top-twelve. And as you'd guess from all this: their 5v5 CF% jumped more than ten points, all the way up to a wait-actually-that's-still-kind-of-bad 46.75%.

As you've probably already figured out, the Leafs are very unlikely to jump 27 points in a year, the way Buffalo did, because they're a hell of a lot better than Buffalo was. Buffalo had so much room to grow because they started so poorly. (Incidentally, while they were outside our sample, this is also true of the last team to make the playoffs after finishing 30th, the 2007 Philadelphia Flyers. Philly went from having catastrophic goaltending to having quite good goaltending, and that fixed a lot of problems for them.)

To return to the Sabres-Leafs comparison: if the Leafs managed to push up their CF% by 10.6 percentage points, they would be the best possession team in the history of the stat. They aren't going to do that. The Leafs were very narrowly bad enough to finish last; they don't really equate to the Sabres.

So we're left with the Panthers, the Oilers, and the Islanders in our comparable hunt. Allow me to briefly dispense with two of them.

The 2012-13 Florida Panthers

The Panthers were a merely subpar possession team that was really, really bad at almost everything else. They had what looked like a good powerplay, but the next year it fell apart. They had absolutely terrible 5v5 shooting percentages, but this was not an isolated incident. The worst team in 5v5 shooting percentage from 2010-2015 was Florida (at 6.88%; no other team was under 7%.) At that point, some of their shooting is more than mere bad luck; they just didn't have many good forwards. Remember when they had that first line of Versteeg-Weiss-Fleischmann in 2011-12 and it somehow worked?

The shooting percentage point may be a cautionary tale for Leafs' fans, given that our squad just finished 30th in 5v5 shooting %. Some of that is bad luck (cough cough NAZEM KADRI cough) but some of it is also that we had a very weak forward group; we probably deserved to be a below-average shooting team.

The 2010-11 Edmonton Oilers

They were bad at everything. Then their goaltending, penalty kill, and 5v5 shooting got modest bumps, and their powerplay exploded. That was enough to push them up to 74 points despite the fact their possession was only slightly improved. In subsequent years, they struggled to maintain the special teams percentages and stagnated. ¯\_(ツ)_/¯

The 2008-09 New York Islanders

So here's my best effort at a Leafs' comparable. The Isles had no staggering improvements, but got a little bit better in several areas. They scored a little more, they got slightly better goaltending, and their shooting percentage went from abysmal to mediocre. They had a meaningful incease in possession of 2.6 percentage points, though this still only brought them to bad. As a result, they jumped eighteen points in the standings.

In 2008-09, the Isles had very limited offence. Their leading scorer was Mark Streit, whom you may recall is a defenceman; other than that, no player broke 40 points (Kyle Okposo peaked at 39.) They had a couple of aging veterans producing for them in Bill Guerin and Doug Weight.

The next season, they added first overall draft pick John Tavares, who put up 24G-30A-54P as a rookie, leading the team. Guerin departed and Weight got hurt, but Okposo rose with the offensive tide to hit 52 points, and Matt Moulson had his first 30-goal season. The Isles had continuity in coaching, as Scott Gordon continued to helm the team (though he would thereafter be replace by Jack Capuano.) Their goaltending was modestly better under the aged Dwayne Roloson, after a miserable year under Joey MacDonald.

No comparable is perfect, and especially so for the Leafs, who are the best worst team since 2008. The Leafs are going to add three star rookies, not just one and a half (Tavares and Josh Bailey), and they're starting from a higher place than the Islanders. The rookies will likely have more than a few growing pains, especially on the defensive end, but they also have some hope of boosting the moribund offence. The offence should be better in any event; Leo Komarov probably isn't scoring 19 goals again, but Kadri likely won't shoot under 7% and JVR and Bozak will hopefully recover to full health.

To get a little more technical about it: if the Leafs' 5v5 shooting percentage, their 5v4 shooting percentage, and their 5v5 save percentage were all to jump to league average, and they had exactly the same shot totals, they could expect a cumulative improvement of about 45 goals. Considering their goal differential was -48 last season, that would bring them within a stone's throw of being league average in goal differential, and presumably in standings points--probably leading to a point total in the high 80s.

All things considered, I think an improvement in line with the Isles' 18 points makes reasonable sense. This would leave the Leafs at 87 points. Last season that would have meant they were on par with the Minnesota Wild--who made the playoffs in the weaker Western Conference--but it would have left the Leafs six points short of the final Wild Card spot in the East.

Of course, an above-average PDO spike is never totally out of the question, and if the Leafs were to benefit from one, they might well push into the 90s. In which case they could do something none of the eight teams we looked at did: go from last place to the playoffs.