Two nights ago, when no one was looking, I tweeted out a telling statistic to understand how teams have reacted to the salary cap post-lockout.
In 2005-06, the lowest even-strength TOI/G (minimum 20 GP) was 2:30 (Jesse Boulerice). This year, it was 5:50 (Anthony Peluso).
— Benjamin Wendorf (@BenjaminWendorf) April 20, 2015
Boulerice wasn’t the only one scraping the bottom of the barrel in 2005-06; Colton Orr was nearby with his 2:49 per game, and you didn’t have to look much further to see Andrew Peters (3:15) and Eric Godard (3:27). In fact, 19 skaters played over 20 games that season and recorded even-strength TOI/G lower than Peluso’s from this year. Teams have realized that, in a salary-capped league, even league-minimum dollars can’t justify players who cannot be trusted with regular minutes.
This was a fairly stark evolution of player usage, but it led me to wonder if there were any other things we could see by looking at finer-grained data from 2005-06 to the present. The salary cap was a game-changer because it pushed teams at the top and bottom closer together, and that compelled teams to stop employing players they couldn’t trust at evens; what are some other areas we see the pressure of parity?
From the get-go, we already know that parity has pulled down scoring levels equivalent to pre-1967 expansion, as well as pulled team shooting talent tighter together than we’ve ever seen it. But thanks to War-on-Ice.com, it’s become far easier to run tests on game-to-game data in measures like Corsi-For percentage and offensive zone starts. These measures can tell us a bit more about change in the course of play and player management. In this case, I took every individual player game from 2005-06 to the present and, focusing on even-strength (5v5) minutes and removing the bottom 2.5% time on-ice games from each season, I was able to look across around 42,000 player performances per year. The first thing that jumped out, again, was the change in player ice-time:
With the standard deviation holding steady, there is little doubt the pressure is pushing out players from the bottom — but it’s also pushing players at the top. Comparing the top and bottom 25% in 5v5 game TOI:
As teams refine their decisions about their bottom 6/bottom pair players, they’re increasing their usage, and using the same decision-making to exhibit more confidence in their top players. But…these are decisions, not necessarily improvements. Where do they appear to be picking the better players?
This chart is a lot to digest, but here’s the gist: I maintained the player groups I used above (top 1/4 in 5v5 game TOI, bottom 1/4 in 5v5 game TOI) and compared their usage and Corsi-For% results from each of the years 2005-06 through 2014-15. In addition to their average result, I included the standard deviations (indicated by column grouping with “SD” in label).
Start with the first grouping, of the first four sets of columns: league-wide, the top 1/4 players continue to participate in a war of attrition (more often than not, against opposing top players). The bottom 1/4 dipped, then declined slightly over the last decade — but when comparing the progression of either top or bottom’s zone starts, you can see that they’re moving in very different directions. The bottom 1/4, once granted generous offensive zone time, now seem to be given more defensive roles, while the top 1/4 have had slightly better offensive zone starts. More importantly, considering the zone starts, the bottom 1/4 seem to be getting deployed with greater success and utility than they were in the past.
The latter six sets of columns are focused in the spread of Corsi-For% and offensive zone starts among these groups. Across the board, we’re seeing competition getting closer (in the three SD CF% sets) and deployment becoming more consistent (in the three SD ZS% sets) as standard deviations decrease. This is parity and the influence of the analytics movement, without a doubt. Teams are making these usage decisions, and returning the results above. The fact that the zone starts are making a drastic shift, yet the CF% returns are not as stark, re-raises important points: zone starts and their impact are muted by the limits of the measure and characteristics of the game like the 60-40 rule for possession in hockey. Nevertheless, we are comparing apples to apples with the zone starts over time; that shift is real, muted impact or not.
It’s very unlikely we’ll ever reach the point in the NHL where every team is optimizing its lineup according to possession or any other measure controlling the puck, but we will know we’re moving towards it if the indicators of parity (closer results, predictable player usage, increased influence of luck on results) begin to show themselves. We gain a glimpse of that pressure in parity by observing the past decade of a salary-capped NHL — a league of the same size for 15 seasons — and how player usage and performance has evolved.