ZedJ

Load management in hockey: Why it isnt happening, why it should and why it probably wont

In the summer of 2018, the Toronto Raptors made a big bet, trading away DeMar DeRozan, the face of the franchise, for a potential MVP candidate who had only suited up in nine games the year prior. One year later the risk was well worth the reward as Kawhi Leonard carried the team to its first-ever NBA championship. 

It’s what happened in between those two moments though that’s of pivotal importance to the end result and of major interest to sports at large: Load management. If not for the Raptors’ program that kept Leonard out of 22 regular-season games there likely would not have been a championship. 

Leonard averaged 39 minutes throughout the playoffs, a mark he reached just four times during the regular season. It was that extra rest during the season that likely allowed him to ramp things up come playoff time. In a story earlier this year by Jovan Buha and Sam Amick, it was noted that Leonard’s new team, the Los Angeles Clippers, didn’t believe it was a coincidence that the number of games missed during the regular season was close to the number of games he played in the playoffs.

Advertisement

“By shaving off nearly 750 minutes from his regular-season workload (Leonard’s 34.0 minutes spread over 22 games), he was able to log a league-leading 939 minutes in the 2019 postseason,” Buha and Amick wrote.

It’s not only extra rest, but it’s also a precaution to lower the chance of injury and re-injury.

“Me missing games isn’t just to keep me fresh,” Leonard told reporters after a late-March game last season. “It’s obviously making sure I don’t re-injure something that I was out for last year.”

The key to the Raptors’ championship run was that they needed Leonard healthy and they needed him at his best. During the regular season, the team’s net rating was plus-7.9 with Leonard and plus-3.6 without, according to NBA.com. In the playoffs, they were even better at plus-8.3 with Leonard on the floor compared to minus-5.5 without. That’s a major discrepancy and one the Raptors likely saw coming, knowing they were deep enough to handle his absence during the regular season, but that his impact would be felt much greater in the postseason against tougher competition.

Load management was arguably a key factor in that and the Raptors’ success. It was about getting Leonard to be his best when it mattered most. While it comes with its own issues, it still begs the question as to why the concept hasn’t expanded to some degree outside the NBA, specifically a brutally tough and belligerent sport played on ice.

Why not hockey?

Hockey is a gruelling sport. Over an 82-game season the bumps and bruises add up, and that piles up even further the deeper a team goes into the postseason. With the intensity ramping up in the playoffs, the more lenient rulebook, and the potential for heavier usage in unlimited overtime periods, it’s no surprise the post-mortem injury list would make some people squeamish. Issues that may have been minor during the regular season might be aggravated further in a win-or-go-home contest, and while the culture says to “play through it,” it likely makes more sense to limit that need in the first place by ensuring teams have the best, freshest, healthiest versions of their players in time for the grind.

Advertisement

Teams do that to some extent, resting a swath of players in the final game of the season – a usually meaningless contest – but it’s extremely rare to see it handled the way NBA teams do. Earlier this year it seemed like Boston was heading down that path with Patrice Bergeron, but that now looks to be a one-off to increase rehabilitation time, rather than a full-season trend. Without many other examples, it seems as if the NHL at large has room to take things a step further. I mean, we already see it with goalies where the league has shifted away from 65-plus game workhorses to more even splits. Managing player workload seems like the next logical step. 

But there’s a major difference between the NHL and the NBA, and it’s likely at the forefront of why we’ve seen load management manifest in one league and not the other: Parity. The NHL has it, the NBA doesn’t. 

In the NBA, elite teams can afford to take the risk of resting star players, even very early in the season, as it usually doesn’t affect the big picture. The risk is worth the reward for some teams where the playoffs are a foregone conclusion. In the NHL, playoff races are much tighter. Even for the teams at the top, a division win could be the difference between a Round 1 minnow and a legitimate contender. There’s also the matter of games being inherently closer in the NHL, where very few games ever top 70 percent in win probability, a high mark that’s much more common in the NBA.

In the NHL, there’s a cultural tendency to be more risk-averse and with regards to load management, parity appears to be a legitimate reason for it. No coach wants to explain a loss after a star player rested, especially if there’s even the slightest chance that it might mean the loss of a playoff berth. A team is not in until there’s an “x” beside their name. 

That’s fair, but it overstates two things: How large of an effect a player has on a single game, and how likely a team with a substantiative edge in the standings is to miss the postseason. It’s those two things that might have teams missing the mark when it comes to resting players down the stretch: For some teams that have a comfortable place in the standings, it’s probably not a big deal. Really, it all comes down to timing.

Let’s start at the game level. For the rest of the NHL season there will be about 25 games where the favourite has a 70-percent chance or better of winning. That’s about as many as there has been in the season to date. Compare that to the NBA where there are as many games with odds that high in an average week according to 538’s game probabilities. It’s 219 games over the rest of the season, more than half of the remaining games. In the NHL, the average favourite wins 58 percent of the time. In the NBA, it’s 71 percent and that’s shown plainly in comparing the distribution of game probabilities in both leagues. In the NBA, there are going to be 124 games with a probability greater than the largest in the NHL: 79.9 percent in a game between the Lightning and Red Wings in March.

In terms of parity, the two leagues could not be any more different. From that perspective, I understand the hesitation in sitting a star player in the NHL where there are significantly fewer gimmies but there’s another key difference and that’s the effect of a single player on those probabilities. If Leonard – the NBA’s fifth most impactful player on win probability – is out for the Clippers, the team’s chances drop 10 percentage points against an average team. For the NHL, that would be the Avalanche’s Nathan MacKinnon, where the impact is “just” four percent in comparison.

Advertisement

That figure for Leonard comes courtesy of Jacob Goldstein of BBall Index, the creator of Player Impact Plus-Minus (PIPM) which aims to measure the value of a single player. According to his data, the largest impact on a single game belongs to James Harden, where the Rockets win probability would drop nearly 13 percentage points. For hockey, that honour belongs, of course, to Connor McDavid, but the effect is 4.6 percent. According to Goldstein’s work, you’d have to go all the way down to the 24th most impactful player, Toronto’s Pascal Siakam, to get to McDavid’s effect for hockey. When looking at the two league’s top 10, it’s a stark contrast.

The data above isn’t a strict translation from a player’s value as it depends on the minute distribution in each player’s absence. Giannis Antetokounmpo carries the highest value in the NBA but plays for a deep Milwaukee team hindering his effect on win probability. It’s the same reason neither of Nikita Kucherov or Sidney Crosby make the list and it’s why the question of resting players is team-dependent. Some teams – usually the ones who would be in a position to do so in the first place – can afford to do it more than others.

Now, three-or-four percent is nothing to scoff at by any means, but it obviously pales in comparison to what NBA teams would have to deal with if a player rests. It slightly levels the playing field in the two very different contexts. The games in the NBA may be more lopsided to create more leeway for rest, but the players also mitigate that by having a larger impact.

Wonder why the Penguins were able to continue rolling without Crosby (and a slew of other players) or St. Louis without Vladimir Tarasenko? It’s because in hockey, a single player doesn’t make as large of a dent on single-game outcomes. If Pittsburgh is playing Detroit at home, the team would have a 75.1 percent chance with Crosby, but if he takes the night off that only drops to 72.7 percent. It would be very hard to pin a loss on his absence alone. As best as we can estimate right now the best players in hockey are worth around three-to-five wins above replacement and dividing that by 82 games means around 0.037 to 0.061 wins lost per game. It would take a lot of absences for that loss to be truly felt.

And that’s where a team’s current standing comes into the equation.

The Hurricanes looked like a very safe playoff bet on January 17 with playoff chances hovering above 80 percent, but then they lost Norris Trophy hopeful Dougie Hamilton, worth 3.2 wins, for what could be the season. If he misses five games or 10 games, that’s manageable. If Carolina was further from the cutoff line that’s manageable too. But at 82 percent and with Hamilton potentially missing most of the remaining season, the impact is huge and caused the Hurricanes’ playoff chances to drop 14 percentage points to 68 percent. 

That’s a significant loss, but it’s worth noting the team was still more likely than not to make the playoffs, despite losing one of their most impactful players. That helps inform how much leeway a much safer team has in terms of purposefully holding impact players out of games for the purpose of rest.

Advertisement

Take Tampa Bay for example. The Lightning have surged of late and are now second in the league with a 39-15-5 record (before Monday’s game), 13 points up on the ninth-place team in the East with a strong group that shouldn’t let up down the stretch. The team already has 83 points in the bank with 23 games left to play, and with the playoff cutoff in the East projected to be around 97 points, they only need 14 points to get there – a paltry 55-point pace the rest of the way. For all intents and purposes, they’re in the playoffs barring some sort of cataclysmic misfortune thanks to the points they’ve already accrued, the level of talent the team has, and how little season there is left.

Here’s how misfortunate they would have to be.

Right now, the team is without Nikita Kucherov, Steven Stamkos, Anthony Cirelli, Ryan McDonagh and Jan Rutta. That’s one of the best players in the game, an elite center, a Selke-calibre center, a top-pairing shutdown defenseman, and a solid top-four guy. It’s nearly 12 wins out of the lineup, enough to severely limit a majority of teams to a significant degree. Most of them should be back soon, but for argument’s sake, let’s say they don’t come back until the playoffs. If that were the case, Tampa Bay’s chances of missing the playoffs drop from 99.992 percent to… 99.99 percent. 

That’s literally nothing.

Add Brayden Point, Ondrej Palat, Victor Hedman, and Mikhail Sergachev to the injury list and the team loses another 10.7 wins, bringing the total out of the lineup to almost 23, dropping the team’s expected win percentage all the way down to .410, a true-talent 79-point team that would rank among the league’s very worst. That’s still well above the 51-point pace the team needs to likely make the playoffs though and unsurprisingly their playoff chances remain strong at 97.8 percent despite all the misfortunes.

Remember the classic Simpsons episode, “Homer at the Bat” where Mr. Burns builds a company softball team out of MLB ringers and laughs at the unlikelihood of nine separate misfortunes plaguing his team? It would take even more than that to stop the Lightning who could still lose Alex Killorn, Tyler Johnson, Kevin Shattenkirk, and Erik Cernak – literally the only remaining difference-makers on the team – from their roster and still have an 89 percent chance at making the playoffs. They would be a true-talent 65-point team, a team worse than the Red Wings on paper and it wouldn’t even matter. That’s how safe they are: The entire team can go on vacation and basically let the AHL team (plus Andrei Vasilevskiy) take over and it would probably be fine.

Though obviously not to the same degree, it’s a similar trend for the other five teams with playoff chances at 99 percent or higher: Boston, Colorado, Pittsburgh, St. Louis, and Washington. These teams haven’t technically “clinched,” but it would be mathematically improbable for any to miss at this point. Some teams are safer than others, but the same message holds true: It would take a lot of misfortunes for things to go awry at this point.

Advertisement

Data as of February 14

What the chart above shows is how much leeway each team has when it comes to their playoff chances, depending on how many wins are out of their lineup over the remainder of the season. For every team here, there’s a lot of wiggle room to rest players and still have a highly probable shot at the playoffs. With an extremely conservative cutoff point of 99 percent playoff probability that wiggle room ranges from three wins out of the lineup (Pittsburgh) to 19 wins (Tampa Bay). It’s a margin that only increases further as the end of the season gets closer and as teams inch closer to the playoff cutoff line.

Even the least safe team above, Pittsburgh, would have to employ an extremely aggressive rest strategy to reach its already very conservative cutoff point of three wins. That equates to Sidney Crosby, Evgeni Malkin, Bryan Rust, Kris Letang, Patric Hornqvist, Jason Zucker, and Jared McCann missing 13 games, half of their remaining season – something no one in their right mind would ever suggest. No one is expecting or asking for these players to take a lengthy vacation until the playoffs begin, but it shows that a couple of vacation days here and there very likely won’t hurt.

Tampa Bay has 23 games left. If the Lightning gave each of their five best skaters four-to-five games off apiece, it would equate to around one win lost in total, a loss that wouldn’t even register in the team’s playoff chances based on the above chart. In fact, when each player is taken out of a specific game, the grand total in expected wins lost is even less than a win due to the team’s exceptional depth. It would be virtually inconsequential in the grand scheme of things.

The Lightning would be favoured in three fewer games, yes, but those were coin flip games regardless. In any sense, those games are probably not ones where players should be rested anyway since they’re higher leverage tilts. It’s the games against the Red Wings, Devils, and Senators of the world where it’s much more ideal. You’ll notice too that the effect of rest drops the higher a team’s win probability is in a given game – the law of diminishing returns.

But even this scenario that I’ve laid out is likely a bit on the outlandish side given the current load management environment in the league. Still, it exemplifies what’s possible for teams that are as good as clinched. For the greater good of being as fresh as possible for the playoffs, it feels like a worthy trade-off. There may not be much clarity or certainty in how big of an effect resting of this nature would have for a hockey player, but it’s hard to argue it wouldn’t be positive. When the games remaining have such low stakes for some teams, what’s the harm? An extra five games of rest in games that don’t matter can be a huge difference-maker in games that do.

That’s all well and good, but it would be dishonest to stop there without acknowledging the one extremely obvious drawback to all of this: It absolutely, unequivocally sucks if you’re a fan who paid hard-earned money to attend a game and the stars decided to take a night off. Hockey is the ultimate team game, but there’s no denying it’s the Kucherovs, Crosbys, and MacKinnons of the world that are putting butts in seats – even if the league itself has struggled to market its stars. There will be savvier fans that will acknowledge it may be a necessary evil to increase the odds of a parade a few months later ever so slightly, but that won’t make it any less disappointing if it were to actually happen.

Hockey, like all sports, is an entertainment product first and foremost and that should be the number one priority. Resting star players is in favour of that, but only in the long term in an abstract way by virtue of attempting to increase the odds of a team winning it all. It’s also a way of keeping guys fresh for the entire season.

Advertisement

Our own John Hollinger, who previously worked with the Memphis Grizzlies, had an interesting anecdote on that angle in a story earlier this year. While working with the Grizzlies, the team had a 4-in-5 very early in the season and had to rest both Mike Conley and Marc Gasol for a game. The whole article has some great points on load management, but this passage, in particular, offers a thought-provoking perspective on the fan angle from a team perspective.

“I’m sure there were people in Minnesota who bought a ticket thinking they’d see Mike and Marc play and they were probably upset. (Not that upset, since they got to see the T’wolves beat our asses by 36, but still). But there were also fans who bought tickets for our final 76 games, and they thought they were seeing Mike and Marc, too. We gave them a much better chance of that happening; indeed, both players finished the season healthy and played 69 and 74 games, respectively, that season – including a few strategic rests in between.”

Despite the health benefits to players, the NBA ended up cracking down on the concept of load management a few seasons ago when the issue was running rampant across the league.

Back in 2017, the league made strides to curb some of the issues with a few policy changes such as a prohibition on resting healthy players during “high-profile, nationally televised games” as well as a guideline to not allow rest of multiple players for the same game or any healthy player on the road. In a further step this year, the league cracked down on teams even using the “load management” terminology, forcing teams to use more creative language like “knee injury management” and deeming Leonard himself “not a healthy player under the league’s resting policy.” The Clippers were even fined $50,000 for saying the wrong things.

Right now it’s obviously not a problem in the NHL considering the very idea isn’t entertained until the final week of the season, but I don’t doubt there will be a day in the future where load management will be more commonplace. When that time comes the NHL could adopt similar policies.

That’s the main issue, but there are some others as well starting with the culture of hockey. The NHL is still a very results-oriented business and there’s no doubt a large and vocal old-school contingent might voice its displeasure the second a decision to rest a player goes awry with a loss, especially if it’s in a very winnable game. Hockey is the “toughest” sport where players play through the pain – a completely different vibe from the very idea of load management. 

Back in the good old days, the thought of taking a night off to recover wouldn’t even be thought of, and that sentiment is still thoroughly ingrained in hockey culture. In that vein, there’s the possibility that players don’t even want to take games off. When a culture is built upon toughness, the idea of needing extra rest can be misconstrued as “soft,” adding another hurdle to cross. There are also the players that chase statistical milestones that may not want to let rest get in the way of that.

Advertisement

At the same time, there’s also the idea to some – right or wrong – that championships are the only thing that matters and if load management helps gain an edge, maybe players won’t be dismissive of it after all.

Finally, there’s the fact that the playoff race isn’t the only race. While Boston, Colorado, Pittsburgh, St. Louis, Tampa Bay and Washington look set to make the postseason, their seeding is still very much up in the air, especially considering there are two teams from each division listed here. In the league’s current parity landscape, there’s sometimes a sizeable difference between playing the third seed in a division versus playing a wildcard team, and those races are far from settled. Resting is still an attractive option, but it’s to the extent of the gap in the division race and lowers the margins teams can play with where the projected gap between two teams is two or three points as opposed to 15.

Load management will probably one day find its place in the NHL. As analytics and sports science becomes more prominent around the NHL, the positive value that rest brings will likely become too difficult to ignore. The Raptors and Leonard showed that last season and though that’s an extreme case, there’s room in the NHL to emulate the NBA’s more subtle programs. The NHL schedule is long and the game itself is rugged – it’s a grind, one that only becomes more gruelling the longer a season goes. A few extra days of rest here and there seems like a benefit a lot of teams and players.

There’s a hidden benefit to that too and it’s teams being able to see better what they have in their organizational depth. A few games missed by a star gives players chances to move up the lineup and call-ups to show what they have at the NHL level. It’ll help better evaluate what players are made of.

Load management isn’t for every team or every player, but for some it can be a huge benefit to a player’s and team’s bottom line. It’s about saving a player’s best for when it matters most, and for some teams the risk is well worth the reward, especially for deep teams where that risk is minimal at best while the potential reward is significantly larger in comparison.

What it all comes down to at the end of the day though is what’s best for the team. What the fans want matters as they pay money to see an entertainment product and they deserve to see the best that can be offered, but that best can be a long playoff run, maybe even a championship that will be remembered forever. Raptors fans might have been frustrated by the 22 games Leonard missed at the time, but those games have long been forgotten, replaced by the memory of an epic championship run that will last a lifetime. They have load management to thank for that.

Data via Evolving Hockey, NBA.com, and Jacob Goldstein of BBall Index

Blake Murphy contributed to this story

(Top photo of Brad Marchand and Nikita Kucherov: Mike Carlson / NHLI via Getty Images)

ncG1vNJzZmismJqutbTLnquim16YvK57kG9ncXFiZXxzfJFpZmlqX2aFcLjOmptmpZGjrqixzJ6lrWWZo3qpu8KknLJlp53GbrXTZqCspqRitaK8z56loqaXYsSpxYyiq2armKTCrbCMmqWdZaedxm6102anq6eSlq%2BtxYywpqesXw%3D%3D

Christie Applegate

Update: 2024-04-26