Eye Test vs Analytics and How it Affects the NHL Trade Deadline

The NBA saw a trade deadline this February like they’d never seen before.

The Phoenix Suns were the talk of the league, pulling the plug on the season and overhauling nearly a third of their roster to begin a massive rebuild — something the NHL is more than familiar with.

What NBA fans found to be a wild, out of control deadline day is a yearly occurrence in the NHL. Every season, teams make a choice at the trade deadline — and either go all in to push for the Stanley Cup, or pull the plug and start gathering assets to retool the roster over the summer. This is far from out of the ordinary; with big names dealt at the deadline every year, hockey is a sport where rosters get shaken up on an annual basis in pursuit of Lord Stanley’s trophy.

This season, though, the league has seen a massive shift towards analytics-based rosters; with some teams building on possession numbers and complementary production levels, the remainder of the league’s front offices are left wondering whether the hype is worth it.

The anti-analytics crew have one defense they often give — the ‘eye test’.

Although most hockey analysts who support their claims with advanced numbers spend time watching the game itself, a good chunk of the modern-day analytics personnel boast that they never watch the game at all. They’re a departure from the traditional hockey insider; they don’t buy jerseys to support their favorite teams, and they don’t purchase tickets. They don’t sit in the press box, either; rather, they come to conclusions based on the stats from each and every game, comparing ice time and possession metrics to determine which players are really worth what we think they are.

This has made the trade deadline even more interesting than normal.

Players like Mats Zuccarello have never been the league’s idea of a team superstar; standing just over five foot seven on a good day, Zuccarello is a fast but unrefined hockey player. His passing is a little erratic, and his positioning is questionable to the naked eye; his team wins games, but they don’t always look very pretty doing so.

As an emerging offensive producer, though, Zuccarello is hard to overlook — size aside, he’s able to pull his weight both in scoring and ice time. He skates a heavy schedule, but manages to consistently put the puck in the back of the net — whether he’s shooting at a high or low percentage, suggesting that he’s a strong producer whether luck is on his side or not. Advanced analytics examine this kind of data, and the advanced numbers crew will be the first ones to tell you — as a pending free agent, a player like Mats Zuccarello deserves a hefty raise.

Analytics don’t tell the whole story, though, and the other side of the argument is quick to point this out.

Players who are underproducing in one market may be lacking more than just raw talent, argue the traditionalists. A player who is given a fresh start somewhere else — especially if his teammates believe in him and give him the spaces and opportunities to pick up his production — is going to see a jump in his numbers, largely based on what can only be described as ‘intangibles’. You see it time and time again; when thrust into a leadership role in Arizona, waiver pickup Mark Arcobello fired off five goals in seven games, and snake-bitten Mike Ribeiro has seen his production go back up on the Nashville Predators less than a season after getting bought out of a multi-year contract.

It’s ironic, therefore, that there’s an analytics conference going on in the midst of the hockey trade deadline — and that many of the things brought up in trade talks are part of the arguments being made at the Sloan Sports Conference Panel.

With Calgary’s Brian Burke arguing that hockey is too ‘physical and random’ to really use analytics to determine player value, you can see why players like Jared Cowen, Chris Stewart and Erik Cole are even brought up in trade talks. Cole brings veteran experience and consistency to a team that may need it, while Stewart has been struggling to perform on par with his scouting reports (on a pretty slapshod team) and could likely benefit from a new set of linemates. These ‘intangibles’ drive player value up — while raw data for Stewart would send his trade value through the floor.

The same goes for two of the centers rumors have focused on — Antoine Vermette and Tyler Bozak. Based on their advanced data, both are league-average second or third line centers; Bozak gives much better offensive support, while Vermette is more well-rounded and capable of producing in any situation. The fact that they’ve both been used as top line centers on teams that need more support than they can lend, though, have driven their trade value down; Vermette is considered to be overrated relative to his asking price, while Bozak is considered too expensive for his relative production.

Yet Vermette is a two-way veteran center with a low cap hit, something many teams covet — and while Bozak is misused as a top line center, he’s certainly not overpaid for a second line guy with similar production.

The trades made at the deadline are never made without pushback from the NHL community, but this year should be something especially unique — because the relative value from team to team is now being measured in two completely different ways.

Don’t miss an episode of The Natural Hat Trick! Subscribe on iTunes

To Top