Djurgården and Slot Entries

Djurgården and Slot Entries Good analytics borrow, great analytics steal DANIEL W 25 NOV 2023 The best advice I can give to anyone trying to get into hockey analytics1 and looking for inspiration is to look at similar sports that have advanced their respective analytics fields much more than hockey has. Those would be basketball and football. A while ago I saw this article from Stats Perform talking about actions in and entries into the penalty box: The edge of the box is a liminal space between hope (outside) and expectation (inside). Once an attacking player gets on the ball inside the penalty area their chance of scoring is significantly increased, and so is the potential sanction for a defender who makes a mistake. Preventing your opponents reaching the box, therefore, is the primary aim of most teams. Now, while there is no area on the rink that perfectly maps onto football’s penalty box, hockey’s slot (home plate, scoring chance area, whatever you want to call it) shares some similarities. 5v5 Shots taken from the slot have, in the 23/24 HA season, about a 10.4% chance of resulting in a goal while shots in the rest of the offensive zone have about a 1.8% chance of lighting the goal lamp. HA 23/24 average xG value by shot location Thus, being able to get to the slot, to high danger areas is quite important if one wants to generate high danger shots. Why Djurgården though? 5v5 Team Metrics If we look at the team shot metrics I included in last week’s post, we can clearly see that one team that is likely to struggle in this area (be it offensively, defensively or both) is Djurgården. Their share of shots is elite while their share of expected goals is hovering around average. If we look a bit more closely and split the shot and xG shares into their components, we can see that Djurgården is averaging about 0.3 xG less per 60 minutes than you’d expect based on the amount of shots taken alone and are allowing about 0.28 xG more than you’d expect based on the amount of shots allowed. Both of these value are easily bottom of the league. Let’s take a look at the individual players who have played at least 100 5v5 minutes for DIF this season. Looks like that gap is present across pretty much the entire lineup.2 Most players perform worse, sometimes significantly worse, if you include shot quality in the on-ice metrics compared to when you look at just the shots for and against. So, offensively the average DIF shot is less dangerous and the average shot allowed by DIF is more dangerous than league average. What could lead to that? Does Djurgården avoid the slot like Pontus Netterberg avoids the lineup sheet or is there something else going on? You mentioned something about slot entries? Using slot shots isn’t a new concept in hockey analytics. The thing that might translate well from football however is using entries into the high danger area as a (hopefully) leading indicator of generating shots from high danger areas. So a Slot Entry is any time a player ends up with possession of the puck in the slot area. This can be accomplished in a few ways: somebody passes the puck into the slot somebody skates the puck into the slot somebody recovers a puck in the slot (loose puck, off a turnover, etc.) Some examples from Djurgården’s latest game vs Karlskoga (at 5v5): and in video form: Two things to point out: Firstly, in the video above, I think you can clearly separate the slot entries into two distinct categories, plays made during offensive zone possessions and plays made off the rush. We would expect the skills necessary to facilitate slot entries to differ slightly between the two scenarios. One being mainly playing against 5 defenders in-zone, the other playing against backtracking defenders attacking the offensive zone with speed. Secondly, teams (or players) who play more get more of a chance to create in the slot, that’s why we usually adjust our metrics for time on ice. But a simple time on ice adjustment is not the kind of adjustment we should be looking to here. The skill we’re trying to isolate is being able to turn possession in the offensive zone into, for lack of a better term, dangerous possession by moving the puck into a more dangerous area of the ice. We don’t want to include a team’s skill at forechecking or at bridging the NZ2, so adjusting for opportunity here means adjusting for the number of possessions the team has in the offensive zone (both of the rush and in-zone) rather than adjusting for time played. Using both of these leads to some interesting results. We want to now split the slot entries into Rush and Offensive Zone Possession and adjust for the number of possessions the team had in the OZ. Offense 5v5 Team Slot Entries The two areas are surprisingly correlated. So when you’re better (or in Djurgården’s case, worse) at generating slot entries off the rush, you are usually also better at generating slot entries vs a set defence. Let’s take a look at indi