The terms are everywhere these playoffs, part of hockey's new analytics lexicon.
Corsi. Fenwick. Zone starts. Score effects. Percentage Driven Outcome. Scoring chance share. Expected goals.
This is Big Data, NHL style, like baseball's Moneyball but with the complicating factor that hockey is more heavily influenced by randomness and variance, with so much depending on how the puck caroms around the ice. One bounce off a shin pad or a post can change a game. Or a series.
Can any of it be predicted by the numbers?
Many NHL teams are trying. Even laggards such as Edmonton, Florida, Toronto and Washington made high-profile additions in their stats departments last year, attempting to catch up to what other franchises had been doing for years. (Hockey remains years behind baseball and basketball.)
The most useful and popular statistics in the NHL all focus on measures of puck possession, since the team that possesses the puck more often than its opponent – even by small margins – generally has more success.
It's more complicated than that because the measures vary with the impact of great goaltending, penalties, special teams and whatever the hockey gods have to say.
"[Puck possession] is one of the things that's proven to be more repeatable," explains Matt Fenwick, an engineer from Edmonton who was in on the ground floor when these stats started being accumulated and assessed, around 2008. "What this is about is the search for real skill."
The first two rounds of this year's playoffs have offered some great examples of the strengths and limitations of hockey's new numbers. What follows are breakdowns of four key teams, their stories and how their playoff paths were – or weren't – foreseeable from the start.
The pretender: Calgary Flames
What analytics said: The Flames had no chance. They were by far the weakest possession team to make the postseason this year, with an average score-adjusted Corsi of 44.3 per cent. At even strength, they were the third-weakest NHL team in terms of controlling play and of keeping the puck out of their zone.
The other 15 teams that made the playoffs averaged better than 52-per-cent possession, led by two teams that earned semi-final berths in the Stanley Cup playoffs: Chicago and Tampa Bay.
What happened: Calgary, an outlier much of the season, possessed statistically unmeasureable qualities such as heart and grit (if not the puck), and for that, coach Bob Hartley will likely win coach of the year.
As well, the Flames, who generated a low number of scoring chances, managed to score on an abnormally high percentage of their shots, and had a remarkable number of late-game comebacks with their goalie pulled. Their success defied the odds, in other words, and that tends to be unsustainable. They also relied on some players with poor possession numbers, such as Deryk Engelland and Lance Bouma, who played substantial minutes at even strength.
So while Calgary’s season was a huge leap forward for a young team in the midst of a rebuild, it invited considerable skepticism from the stats crowd. Missing captain Mark Giordano – one of the better possession players in the game – due to injury, the Flames still managed to advance to the second round of playoffs in part because their first-round opponent, Vancouver, was an even weaker possession team.
Calgary’s second-round matchup wasn’t as forgiving. Powerhouse Anaheim outscored the Flames 19-9 and outshot them by 45 to win the series in five games. The Ducks possession rating in the series was 58 per cent, a level of domination that was to be expected given the two teams’ regular-season stats.
What will happen next: Many teams in similar situations have typically faltered the following season. The Toronto Maple Leafs (in 2013) and Colorado Avalanche (in 2014) were also poor possession teams that had good records one year but imploded the next, missing the playoffs. They are both still searching for answers.
“We’ve seen it over and over,” said Fenwick, a long-time Flames fan who remains optimistic they can counteract this potential regression. “It’s very likely that their even-strength shooting percentage will regress. That is going to cost them some goals.”
That doesn’t guarantee Calgary will bottom out. Management could make substantial changes in the off-season, replacing poor-possession players with strong ones. Their young players could also improve dramatically.
Without those changes, however, it’s highly improbable they’ll repeat this season’s 97 points and make the playoffs in 2016.
The two-man show: Montreal Canadiens
What analytics said: Montreal was good, not great, with two major strengths – one of the era’s best goaltenders, Carey Price, and an outstanding top defenceman, P.K. Subban. But the rest of the roster didn’t contribute enough to produce a championship-calibre team.
The Habs play a grinding, simplified style that doesn’t lend itself well to good possession numbers or high-calibre scoring chances. The Canadiens were the second-weakest possession team to make the playoffs, with numbers that were superior to Calgary (49.1 per cent when blocked shots are factored out) but trailing everyone else.
So with great goaltending, they could win a round or two, but a championship was highly unlikely.
What happened: Price is expected to be named the league’s MVP after elevating the Habs to a remarkable 110-point regular season, putting them second in the NHL despite finishing 20th in scoring. Price’s league-leading .933 save percentage during the year was one of the best in recent NHL history, and it offet other deficiencies.
The Canadiens had the weakest offence among the teams that made the postseason. They scored only 16 goals in their first nine playoff games, and lost the first three games in Round 2 to the Tampa Bay Lightning. Montreal battled back to extend the series, but as expected, Tampa’s superior offence ultimately won out.
What will happen next: In the playoffs, Montreal showed it could be a better all-around team. The Habs surprised many by controlling the play well against both Ottawa and Tampa Bay, posting a 52-per-cent possession rating that was driven heavily by a strong counterattack and taking more shots per minute than they did during the year.
Coaching systems play a key role in this regard. Canadiens coach Michel Therrien gets criticized by the analytics crowd because his team struggles to score, and also because of his history. He coached a star-studded Penguins team back in 2008-09 that had serious issues with puck possession. When he was let go late in that season, his replacement, Dan Bylsma, changed the system and the team’s possession numbers improved sharply. The Penguins went on to win the Cup.
It’s widely believed that scenario could repeat itself in Montreal if a) the Canadiens were to change coaches and add one of the star candidates available, such as Mike Babcock, and b) give largers roles to young players such as Alex Galchenyuk and Nathan Beaulieu.
At the very least, the Habs appear to have more talent than their goal totals under Therrien indicate.
The challenger: New York Rangers
What analytics said: These Rangers were slightly weaker than the team that lost to the Kings in the 2014 Stanley Cup Finals. New York lost underrated defenceman Anton Stralman via free agency to Tampa Bay, and that played a role in their possession numbers sliding back closer to league average.
The Rangers still boast a talented cast of forwards and one of the best goaltenders in the game, Henrik Lundqvist. That combination made them the NHL’s top team (with 113 points) in the regular season.
But they also led the NHL in Percentage Driven Outcome, a remarkably predictive stat that measures how fortunate teams are over the course of a season.
While not a powerhouse in terms of advanced statistics, New York is well-coached (Alain Vigneault is up for coach of the year) and was deemed a credible threat to win the weaker Eastern Conference when the playoffs began.
What happened: The Rangers barely made it through to the final four. They did not play particularly well against the injury-depleted Pittsburgh Penguins in Round 1, slipping by with a series of one-goal wins. It then took a remarkable comeback in the second round to topple the Washington Capitals in overtime of Game 7 on Wednesday.
The NHL’s third-highest-scoring team during the season, the Rangers have been stymied by great goaltending in the playoffs, producing an average of only two goals per game. And no team has better illustrated how thin the margins can be in the playoffs: Eight of New York’s 12 games ended in the same score – 2-1.
When games are that close, the hockey gods can override the stats. And clearly, the gods liked the Rangers.
What will happen next: New York is the weakest possession team remaining in the postseason (50.6 per cent in 2014-15), and it has had difficulty generating offence despite solid scoring-chance numbers.
The Rangers have also lost two of their top players at driving play – Mats Zuccarello and Dan Boyle – to concussions in these playoffs. Whether or not they can return and make a contribution will be a key factor in Round 3 against the Lightning.
Another issue is that the Rangers spend a lot of time in their own end when one of their top defencemen, Dan Girardi, is on the ice, which could be bad news against the high-powered teams still in the hunt.
While traditional measures such as points and goal differential indicate New York should be the favourite to win the Stanley Cup this year, hockey’s newer statistics suggest the opposite: The Rangers are not dominant team at controlling play, and could struggle to score goals against the remaining contenders.
Despite having the top seed, the analytics say the Rangers are likely the underdogs the rest of the way – only a historic performance by Lundqvist would give them a chance to win.
The powerhouse: Chicago Blackhawks
What analytics said: The most dominant team of the analytics era is still very good despite the aging core of its roster.
The Blackhawks have played more playoff games (104) than any other team since 2009, for very good reason: They – along with the Los Angeles Kings – have been the NHL’s puck-possession leaders for years.
Chicago has a great coach and a handful of elite players at both ends of the rink. Captain Jonathan Toews, defenceman Duncan Keith and winger Marian Hossa are all excellent according to these new metrics, and their reasonable contracts have allowed general manager Stan Bowman to surround them with talented teammates.
Chicago was terrific this season from the start, with a 58-per-cent rate of controlling play in the first 20 games. With the Kings missing the postseason, the Hawks entered the playoffs as the top team statistically, albeit with a difficult playoff path.
What happened: The Blackhawks made short work of that difficult path, getting through to the conference finals – for the fifth time in the last seven years – in only 10 games.
Chicago has actually been outshot in these playoffs heading into Round 3 against Anaheim, but that stat can be deceiving. Some of it comes down to score effects – the Hawks had the lead in most games, and teams with a lead tend to sit back and allow more opportunities. The rest can be attributed to facing two strong teams, Nashville and Minnesota.
In terms of score-adjusted possession, the Blackhawks were the better team, with a 52.5-per-cent rating in the first two rounds. Their share of scoring chances generated at even strength, meanwhile, was also strong.
Their biggest weakness has been killing penalties: They’ve allowed 50 chances per 60 minutes of play at five-on-four, the worst rate of any remaining team.
What will happen next: With seven key players left from the teams that won Stanley Cups in 2010 and 2013, the Blackhawks are the favourites. Everyone believes Toews, Patrick Kane et al can win another title because they’ve done it before.
But the current Blackhawks are not as historically dominant as those earlier teams, which ranked among the best-ever since these stats began being compiled. They’re still good, but they’re more vulnerable to an upset.
More importantly: Who will win the Stanley Cup?
Teams that excel according to these new numbers tend to win Stanley Cups. Five of the last seven Cup-winners were dominant possession teams, at better than 55 per cent. No poor-possession teams have won since the statistics were introduced, and that trend will continue this year.
By any measure, the West is again the stronger of the NHL’s two conferences, making either Anaheim or Chicago the favourite in the final. The Blackhawks are ranked slightly ahead of the Ducks because of their experience.
The Ducks are an interesting case because, statistically, they improved as the year went along, adding key depth players such as defenceman Simon Després before the trade deadline to boost them into the NHL’s elite over the final 20 games (53.1-per-cent possession). They have a big, physical team, and they’ll be well rested after needing only nine games to get through two rounds.
Hockey’s new numbers gave Chicago, Anaheim and Tampa Bay high odds of progressing this far, and at this point there’s very little separating them. Which one wins the Cup will come down to who stays healthy and gets the best goaltending.
But it never hurts to have the hockey gods on your side.