clock menu more-arrow no yes

Filed under:

When Will the Advanced Stats Revolution Come to Junior Hockey?

New, comments

While the NHL seems to be embracing statistics, CHL leagues are not currently publishing or even tracking many useful stats.

Ken Andersen/Getty Images

Last summer, the Wild drafted Reid Duke in the sixth round, a player with skills that would have indicated that he should be drafted several rounds earlier. He was drafted to the WHL with a 1st round pick in the 2011 Bantam draft fifth overall. While not all the players drafted in the first round of a bantam draft will show NHL potential, many do, so this still is a good indication that he’s a player who has significant potential. With Duke’s skill set, he probably could have been picked in the late third to early fourth round of the NHL draft without it being a reach.

But he was still available for the Wild at 169th because he was on the Lethbridge Hurricanes, the 2013-14 season’s worst team in the WHL. His 40 points in 62 games means he had .645 points per game, but was second on the team in that metric. Any team looking at his stats would combine that fact with the fact that he was -45, and would naturally be unwilling to risk drafting him, even though he has obvious talent and possesses a goal scoring skill set.

After playing only one game with the Hurricanes this season, Duke was traded to the Brandon Wheat Kings. Before being injured in mid-February, he was an essential part of their deep forward pool and played second or third line minutes. This year, he produced at a .911 points per game pace and was ninth on the team in points per game. In these two Reid Duke seasons, his situations were radically different.

With Lethbridge in 2013-14, his team had a goal differential of -187, while the Brandon Wheat Kings had a goal differential of +121 during the 2014-15 regular season. It’s easier to produce more points on a team that has greater scoring prowess. It’s much easier, even in the WHL, for good defensemen to shut down a single great line if a team has only one scoring line. All stats that are purely based on point production are going to make Duke look like an improved player during his first post-draft season, but that’s not necessarily the case. Duke had greater production in a second or third line role than he did in the previous season, but he probably produced more offense with less ice time because the Wheat Kings had more skilled players throughout their roster.

With the stats currently published by the WHL, OHL, and QMJHL, there are limited ways to analyze Duke’s season to see how much improvement he made this year. It would be helpful to know his percentage of offensive zone starts, his ice time, etc, but the WHL doesn’t track these stats. Maybe he just had an extraordinarily good shooting percentage in 2014-15 or maybe he actually shot more because his teammates made it easier to spend time in the offensive zone. As a team, the Wheat Kings' estimated Fenwick close for 2014-15 was 55.403%, best in the WHL. The Hurricanes had a 40.548% Fenwick close when Duke was with that team in 2013-14. These team based stats show that his 2014-15 team shot more and/or prevented more shots than his 2013-14 team, but don't necessarily demonstrate that Duke himself shot more or prevented more shots.

Because of the limitations to the stats tracked by the leagues of the CHL, robust calculations, like those done about individual NHL players, cannot be made to assist in examining the performance of individual players in the CHL. While the leagues track and publish shots on goal by team, only the QMJHL publishes individual player’s shot information, and none of the leagues track and publish ice time for the individual skaters. All of the advanced stats studies that can be done are purely based on goals and assists totals.

Additionally, the QMJHL has changed their website to make it less friendly for anyone attempting to use web scraping tools to build stats websites. They also threatened to serve CHL Stats with a cease and desist order. While most websites that publish NHL advanced stats have ads that generate (small) revenue, CHL Stats does not have any ads, so it is doubtful that the QMJHL would have actually taken legal action against CHL Stats. But it does suggest that currently the feelings regarding the advanced stats movement are rather hostile, at least on the part of the QMJHL.

As long as the WHL, OHL and QMJHL are not tracking ice time and individual shot tallies, the advanced stats revolution is not going to come to junior hockey. But an increased acceptance of advanced stats in the NHL can only help motivate the leagues of the CHL to track and publish more information. Proponents of advanced stats occasionally disregard "the eye test" method, but when it comes to junior hockey, there is not a catalogue of stats that allow for anything other than watching the games. For now, anyone wanting to try to examine prospects for improvement has to do so the old fashioned way.