How to read Skater Context Diagrams

I make skater context diagrams to see the variations in quality of teammates and opponents for skaters at 5v5. Here, for instance, is the skater context for the Stars forwards in 2016-2017.

The diagram shows every Dallas forward who played at least 50 minutes at 5v5 in the early part of the 2016-2017 season. The aim of the chart is to see who plays with and against relatively strong or weak players. As a proxy for player quality, we use ice-time in all situations, relying on head coaches to know their better players from their weaker ones. Defender minutes are shown on the horizontal axis, with heavy minute players to the right, and Forward minutes are shown on the vertical axis, with top-line players appearing higher. Every forward is listed twice, once for their teammates (in my facsimile of Dallas' victory green) and once for their opponents, uncoloured.

The top line of Benn-Eaves-Seguin, for instance, get the most minutes of any forwards, and they largely play with one another, so their green names are up at the top of the graph; they have the highest quality-of-teammate forwards. They also play the toughest quality-of-forward opponents, shown by their uncoloured names. Their competition, however, is not entirely first liners like themselves, they play weaker lines also and on average their competition is playing around sixteen minutes per night instead of almost eighteen like them. However, Benn-Eaves-Seguin typically play with defenders playing around 20 minutes a night; that is, a mix of first, second, and third pair defenders. Their opponents, however, skew a little stronger, to an average of 21 or 21.5 minutes a night. This suggests that their coach (or their opponents' coach) is deliberately engineering a matchup where they play tougher defenders. The magnitude is not enormous; only two or three shifts difference in "quality", but then matchups are not easy to get.

The "middle six" of Eakin/Sharp/Spezza/Faksa/Roussel/Hudler appear lower in the chart, they play fewer minutes, mostly alongside one another and hence have a lower quality-of-forward for teammates. They mostly play with weaker defenders against slightly tougher defenders, like the first liners. Eakin spent some time on the top line in some games so his usage is similar to both groups.

The "bottom six" of Shore/Cracknell/Dowling/Smith/McKenzie mostly played limited minutes, again with one anoter in various permutations, so their green names appear low. Inevitably they played considerable minutes against various lines from Stars' opponents, so their quality-of-forward competition was much tougher than their quality-of-forward teammates, despite still being the "easiest" competition on the team. Korpikoski and Ritchie, who played up and down the lineup, appear between the two groups.

The values themselves are calculated using only 5v5 situations, even though all-situations icetime is used as a proxy for player quality. The graph doesn't tell us anything about which players were used on the penalty kill (inevitably against top competition, trying to score) or on the power-play (usually against weaker competition, not trying especially to score).