Faces and Places Shiffrin heats up sponsor market First Look podcast: Opening Day and more Raveling ‘an information reservoir’ F1 players accelerate growth in U.S. Instagram expands its student program Plugged In: Amy Trask Venue lockers deliver merch, food SunTrust Park brew steeped in the game Teams to get millions in relocation fees
SBJ/January 22 - 28, 2007/This Weeks News
NHL cap creates a new class of scouts
Published January 22, 2007
A pile of e-mails awaited Barry Hanrahan, the Philadelphia Flyers’ assistant general manager, when he returned to the office after a five-game road trip in January, but he paused for only one.
Titled “Flyers — Statistical Model Project,” it was seven pages long and offered recommendations for measuring the value of hockey players. Though it came from a fan, Hanrahan pored through it with the same curiosity he brought to the most recent Flyers statistics.
“You never know what you’ll find,” he said, recalling the e-mail. “We get some pretty interesting things, and one way or another, each has an effect on how I can improve player evaluations.”
Before the NHL lockout, evaluating talent on the ice was a simple task handled almost entirely by front-office experts such as Hanrahan. But the salary cap, part of the collective-bargaining agreement signed in July 2005, has changed that, putting a much bigger premium on accurately evaluating players and getting a better value for a franchise’s money.
A small industry of consultants has emerged as a result.
The consultants range from average fans and statisticians to former NHL general managers and entrepreneurs. As the league gathers in Dallas this week to watch hockey’s best players in the NHL All-Star Game, they will be crunching numbers and designing statistical models to identify the talented but undervalued players who stayed home.
Using statistics to isolate talent is nothing new in sports. Oakland A’s general manager Billy Beane popularized the practice in baseball, and the NBA has had no shortage of efforts to determine the true value of a basketball player. But hockey has generally been immune to such number crunching.
The lockout and the salary cap changed that. Suddenly, teams that could once buy their way out of poor player acquisitions could no longer do so. Teams with payrolls of $60 million or more now had to fit under a $44 million ceiling. That placed a premium on finding talented, underpaid players.
Former GM Mike Smith (left) is part of
Coleman Analytics, which focuses on
players who produce under pressure.
One of the most successful consulting groups has been Coleman Analytics. Richard Coleman, a San Francisco-area statistician, and Mike Smith, former Chicago Blackhawks general manager, launched the venture this season, and already it has secured contracts with five NHL teams, Smith said. The cost ranges from $40,000 to $90,000 based on the number of years on the contract.
The group isolates player performance using a variety of categories but focuses on tying it to pressure situations. The resulting model allows teams to identify players who can help hold a one-goal lead in the final two minutes of a game, or score a tying goal in a 2-1 game.
The information allows managers and coaches to know who to put on the ice at key moments and gives them an additional factor in determining who to sign, Smith said.
“If you had $800,000 to sign a player and you want them to fill a role on a team, you can use our info to assess — based on past performance — who fits that bill,” he said. “You don’t rely solely on it, but it can be confirmation of what you think you know.”
One general manager working with Smith agreed. Speaking on the condition of anonymity to prevent competitors from knowing his team uses the system, he said, “I don’t think the system they’ve put together is foolproof or by following it you’ll be a Stanley Cup winner, but it’s another way to get there.”
Rick Gray founded Rainy Day Sports believing he could offer the same service. A marketer by day, he spent his nights plotting every shot taken in the NHL last year. The result gave him the likelihood a player would score based on location. He compares it to on-base percentage in baseball as opposed to batting average.
“We’re hoping it lets GMs find a way to exploit the gap between high-performing and low-paid players,” said Gray, who has solicited 18 teams but landed no contracts. The cost for working with him for a season would approach $25,000, Gray said.
RealGM, which is launching an NHL service after years of working with as many as nine NBA teams, has focused its efforts on creating a patented Web site that allows teams to isolate potential trades based on selected performance and salary cap criteria.
The benefits of the consultants and their statistical models could have applications beyond finding talent, general managers and agents say. They also can help in negotiations.
The general manager working with Coleman Analytics said new criteria for evaluating talent could hurt a player’s leverage in contract talks. In the course of 82 games, he said, a forward who scores 40 goals might deserve $3.5 million, but if half those goals come in “low-pressure situations” like a 4-0 lead, then that could drop to $2 million.
“No one’s used it like that yet,” one GM said, “but it’s certainly there.”
Still, many teams have been hesitant to embrace the new firms. Tom Lynn, assistant general manager with the Minnesota Wild, thinks they need to become more multidimensional before they are worthwhile.
“It’s like cell phones in the first generation,” Lynn said. “Until they can come up with one that can do multiple things, they won’t be really useful.”
Consultants say those attitudes will change with time, much the same way they did in baseball following Beane’s success with the A’s.
“Hockey still hasn’t developed a system for dealing with the cap,” Smith said. “That’s why we think the timing of this is good. Fast-forward 10 years and we believe this will be commonplace.”