Buy low or declining asset? An analytical look at Phillip Danault

On Friday night, the Canadiens brought back a player who had slipped through their fingers in the summer of 2021. Montréal was desperately looking for a centerman, and Phillip Danault represented one of the few candidates realistically available via trade. Much has been said about the fact that the Quebecer has only recorded five points since the beginning of the season, but what do the more advanced statistics say?

If you follow the hockey analytics community, you may have come across a post by JFresh on the social media platform X, reacting to the trade. He views Danault’s return to Montréal positively.

His model, HockeyStats.com, remains enthusiastic about Danault’s performances. It ranks him in the top 35% of the league in WAR (wins above replacement). This metric attempts to quantify a player’s contribution in terms of wins. The model takes into account the player’s last three seasons, while weighting recent seasons more heavily. The only aspect of Danault’s game that appears to be in decline is his finishing ability. However, we know that finishing is highly volatile, subject to luck and before definitively saying that a player is below league average in this category, we need a large sample size (multiple seasons).

When we look under the hood, JFresh’s enthusiasm is justified. When Danault is on the ice at five-on-five, the Kings post a 51.5% shot attempt share, a 52% shot share and a 57% goal share. They register only a 48% expected goal share, according to Natural Stat Trick. However, Danault’s usage can partially explain this last metric. According to Hockeystatcards.com, Danault is the Kings’ centerman who has faced the highest quality of competition this season. Add to that the fact that the centerman plays on average with teammates whose offensive performance ranks outside the top 60% of the NHL. Despite this, the Kings rank in the 67th percentile for shot attempts per sixty minutes with Danault on the ice at even strength. This suggests that the Kings at least partially control play and spend time in the offensive zone when he is deployed at even strength.

So, is everything OK? Did the Habs just find a hidden gem who has simply been unlucky over a small stretch of games? Well, not so fast.

Danault is still searching for his first goal of the season. Natural Stat Trick credits him with five individual expected goals. This statistic is agnostic to shooter quality. It estimates what an average shooter would produce in similar circumstances. This is probably the right approach. We mentioned it earlier, shooting percentage is unstable and sometimes driven by randomness. Players generally regress toward their career averages and toward the league’s mean. Over his career, Danault has indeed been roughly a league-average finisher. Since 2016–2017, his actual goal totals and the expected goals credited to him by Natural Stat Trick have tracked fairly closely. Two seasons are exceptions: 2025–2026 and 2024–2025. However, thanks to certain research efforts (such as those by Eric Tulsky), we know that a player’s finishing ability tends to decline with age. We also know that when players decline, they can decline quickly. Have the Kings identified a rapidly declining asset and managed to move him, along with his entire cap hit for the next two seasons, before it was too late?

Public models struggle to properly capture rush chances. Public expected-goals models generally cannot distinguish whether a shot comes from extended offensive-zone pressure or from a two-on-one rush, for example. Rush chances are more likely to generate goals. Is it possible that the expected goals credited to Danault overestimate his true offensive contribution, because he may be below average in transition? Private data compiled by Sportlogiq and shared on RDS.ca this week suggest that Danault does not even rank among the NHL’s top 400 players in terms of scoring chances on the rush. However, that same model is more optimistic than Natural Stat Trick’s and estimates that the Kings generate a majority of expected goals with Danault on the ice. It is also probably normal for a defensive center to generate few counterattack chances. Creating those opportunities requires taking risks.

Advanced statistics are not so simple after all… First, Danault’s performance appears to be stable, with the exception of one statistic that we know tends to regress toward career averages. Second, we also know that the skill measured by this statistic declines with age.

If I were to venture a prediction, it seems likely to me that the Canadiens made the right decision by acquiring Danault. He fits the profile of a strong buy-low candidate. We also know that this model worked in the past: using a defensive center earning $5.5 million (or $3.1 million at the time) to neutralize the opposing team’s highest-paid players was part of what allowed Montréal to reach the Stanley Cup Final. Danault will free up the Canadiens’ best offensive players by allowing St-Louis to deploy his top units for offensive-zone starts against lesser competition. And remember Eric Tulsky, whose work I mentioned earlier? The team he now runs, the Carolina Hurricanes, were rumored to be interested in Phillip Danault. Needless to say, he has access to far more information than we do.