December 13 2012 02:55PM
Part of what I think contributes to the success of offensive starts like Jonathan Toews and the Sedin twins is the fact that they've had very dependable third lines behind them in recent years. I recently went to Behind the Net, parsed through ice time data for offensive zone start rate and quality of competition to see if there was a league-wide trend towards adopting a system where the third liners were given a higher level of importance.
Turns out, yes, actually, even over the last three years, it's very noticeable. In the 2009-2010 season, third and fourth liners were pretty interchangeable, playing against middling competition, usually in offensive zone situations. What's happened since then is that replacement-level players and fourth liners have seen less ice time, while first liners and third liners have picked up the bulk of the difference.
I went to Behind the Net and sorted forwards by total even strength ice time (you can find it on these pages). I also downloaded their Corsi Rel QoC and zone start numbers, as well as time on ice per game into a spreadsheet, and totalled up players by hypothetical "line" they're on. First liners, for instance, were 1-through-90 in ice-time, second liners were 91-through-180, third liners were 181-through-270 and fourth liners 271-through-360. Every player below that was considered "replacement".
Of course, injuries throw a wrench into the situation, but injuries do factor into a player's usability by a coach over the course of the season. By totalling up offensive zone start and quality of competition rates, as well as averaging per-game ice-time, I looked to see what the major difference in player roles was over the last three seasons.
Here's the first chart. This is simply ice-time per game. You'll note a greater disparity between first liners and fourth liners between 2010 and 2012:
I'll assume the overall ice times don't add up to 60 due to the differing number of "replacement" players through the years. The important thing to consider is that players on lines 1, 2 and 3 have seen their ice time grow, while the 4th liners and replacements have seen it drop. This is consistent with a specific case, the Toronto Maple Leafs, who saw about two minutes of less ice-time between 2010 and 2012.
As for usage, this next graph is the average Relative Corsi of those line's opponents. Again, the first three lines have seen more difficult minutes at the expense of fourth liners and replacement players:
Offensive zone start rate is something else however:
NHL first liners have seen a lot more offensive zone minutes in recent years. Offensive zone rate is the percentage of overall starts in the offensive zone to the total number of starts in either the offensive or defensive zone (contributor Rob Pettapiece showed this does impact goal totals). First line players started 51.4% of those shifts in the offensive zone in 2010, up to 53.3% in 2012. Who has suffered the most? Third line players, who have dipped from 48.9% to 47.6%.
The fourth line was sheltered more in 2011 than 2010 or 2012, and the reverse effect was found for the second line, but the clear trend is the first liners getting offensive zone starts that third line players have been giving up.
My major takeaway is that the decrease of ice-time and overall roles of importance to fourth line and replacement players shows that NHL coaches are becoming better at dividing up the labour. Two-way talent is of greater importance than it was.
If there is a competitive advantage to exploit, I may look at that fourth line. I see those players as having little importance, but if a team has the second line personnel to drive play, (I think of Vancouver's Chris Higgins-Ryan Kesler-David Booth as an example) perhaps it's worthwhile to piece together a sheltered fourth line of young scoring talent rather than goons who will knock heads.
Fourth liners are playing less, in less critical situations, presumably because of the systematic reduction of the fighting specialist's minutes. NHL teams have yet to find a suitable replacement for that role, although perhaps the real lesson here is that the NHL is 90 players too large, and should cut itself down by four teams to ensure every player is useful...