Advanced Stats Say Sabres May Have Been NHL's Worst Team

There's no sugar coating it, so I'll be blunt: A deeper look into the Buffalo Sabres' statistics during the 2013 NHL season say they might have been the worst team in the NHL. Not among the worst. The worst.
Yes, there were teams who finished lower in the standings, but in order to do a proper analysis, you have to throw out the team's finishing spot in the Eastern Conference. Because of an extremely flawed scoring system, shootouts, luck, opponents' injuries, goaltending variances and small sample sizes, teams can appear better or worse than they actually are when simply looking at the standings.
One example of this might be last year's Florida Panthers, who played in an abysmal division and had 18 – yes, 18 – shootout and overtime losses. The reality was that the Panthers were a mediocre-to-bad team by every other metric includings shots for vs. shots against, possession and goal differential – where they finished minus-24.
In the past three seasons, only three teams have made the playoffs with negative goal differentials, but the Panthers were one. Were they a good team? No. They won 38 of 82. But they were a three seed and thought of as the NHL's surprise, goodwill story.
So you can see that record isn't always the best way to determine who the best teams are. What is?
1) Puck Possession
The teams with the puck the most are the best. Crazy, right? Team possession is determined by using shots taken at the opponents' goal vs. shots at your team's goal. Team possession statistics are an excellent predictor for which teams will make the playoffs and even win the Stanley Cup. Last season, the Kings were an eighth seed, but they were fourth in the NHL in possession. The Devils were 10th in the NHL. This season, 12 of the 16 playoff teams are in the top half of the league in possession and there's a good chance a few of them would have fallen out had there been an 82-game season (looking at you, Toronto.)
The Sabres were the NHL's worst possession team. In terms of total shots toward the opponents' goal in close situations (between a 2 goal lead or deficit) the Sabres took a mere 43.73%. That is incredibly bad. Almost historically. In fact, since they started keeping the stat in 07-08, only two teams have ever been lower. For a little more context, L.A. took a league best 56.98 percent of shots toward the goal. Chicago, New Jersey, Boston, St. Louis, NYR and MTL followed. No coincidence that, outside of NJ, those are some of the league's best.
The Sabres were even worse when the “going got tough.” WGR's Paul Hamilton has talked for years about Buffalo's struggles when they get up in a game. He's 100% correct. With a 1-goal lead, the Sabres are by far the worst team in the league with a total of 35.87% of the shots heading toward their opponents' goal. The next worse is 40.42. And with a 2-goal lead? Also dead last in the NHL with an astonishing 29.51%. The league's best? Chicago at 55.00%
Buffalo was also the league's worst possession team when trailing by one goal.
On an individual level, the Sabres only had two players who played more than 20 games who took more attempts at the goal than against: Marcus Foligno and Christian Ehrhoff.
2) Goal differential
Remember the Panthers thing? Hardly any teams ever make the post-season without a positive goal differential and when they do, they are usually a fluke or snuck in. Well, the Sabres didn't sneak in, but their losing ways were no fluke. Buffalo finished the year with a goal differential of minus-18, which ranked 13th in the Eastern Conference. Pittsburgh was by far No. 1 at plus-46.
So if they were only 13th worst in the Eastern Conference, how could they be the worst team in the NHL in terms of goal differential. Two ways: A) Luck B) Goaltending.
Start with A. does a week-by-week update of PDO, a statistic that compares a team's on-ice shooting percentage to their opponents' on-ice shooting percentage to determine whether a team was having good shooting luck for or against. A micro example of how PDO works would be Tyler Ennis. Last season, he had a crazy-high 18.3 shooting percentage and the NHL's highest PDO. This season, he regressed to a shooting % under 10 % because, in reality, he wasn't an 18.3% shooter. Wayne Gretzky is an 18% shooter.
It works for teams, too. Sometimes you have pucks bounce in that shouldn't, sometimes you face backup goalies for six straight games and get goals you otherwise wouldn't. This statistic keeps track of that.
And the Sabres were 12th in PDO – or as stat guys might say, the 12th luckiest team in the NHL. They didn't have a negative goal differential because they were getting screwed by bad bounces, in other words. They had a negative goal differential because they were a bad hockey team. In terms of luck, too, the Sabres won an East-best seven shootouts.
Now for B. The Sabres had good goaltending. That makes them look even worse as a whole.
Buffalo's team Even-Strength save percentage (which is the best measure for evaluating goaltending) was .923. League average is .919. The best in the NHL was .933 and worst .890. The New Jersey Devils were a good team that got terrible goaltending with a .906 EV save %. The Sabres actually received good goaltending from Ryan Miller and Jhonas Enroth and still ended up with a minus-18 differential.
3) Special teams
Power play and Penalty Killer aren't a be-all-end-all of evaluating teams. But a good PP like the Capitals have can cover up some shortcomings and a good PK can bail a team out. There is a lot of luck involved at being good on special teams – that's why it's No. 3. But there's also teams who are consistently good at 'teams and some that always seem to be bad (ahem, Boston).
The Sabres were bad on both PP and PK. They ranked 29th in PP at 14.1% and 26th in PK at 79.2%. If you added up the two percentages for every team, the Sabres would be last in total special teams, with Winnipeg a close trailer by 0.2%.
Was Anyone Worse?
The Florida Panthers were probably worse. Their minus-59 goal differential is pretty incredible. It's hard to say a team that was 41 goals worse could be better, but there's even an argument to be made that if the Sabres and Panthers played 1,000 more games with their same team, the Panthers would eventually come out on top. They had better possession and special teams, but got unbelievably bad goaltending with a .899 EV save %. The Panthers were also by far the league's unluckiest team in terms of PDO. And, of course, the Panthers also walked away with a great chance at getting the No. 1 overall pick.
What does it mean?
The Sabres were not a good hockey team in 2013. At the trade deadline, they started to move in the right direction in looking at a complete overhaul, but they should not be anywhere close to done. Buffalo's club is filled with players who struggle to dominate puck possession and who perform even worse under pressure.
Management, whomever that might include, shouldn't settle for moving a few players. They should evaluate which players are the best in terms of puck possession and in terms of mental toughness. There was a noticeable difference when Kevin Porter, Brian Flynn, Mark Pysyk and Adam Pardy were on the ice compared to their teammates. Maybe they should be back. But every player who isn't helping the team get more pucks to the net is part of the problem, not part of the solution.
Also if the Sabres are able to work out a contract extension with Ryan Miller, it should be considered. He's one of the best EV goalies since the lockout. The Sabres will never end up like the Panthers or Devils with him in net. But, of course, if you dominate the puck like Chicago, even average goalies can turn into good ones.
Next week...we'll look at individual Sabres players. Who should stay, who should go statistically speaking.