Mitch Marner (who had a very average to below average game) TOI: 27:23 including nearly a 4 minute shift to end the game in a LOSS.
Marner had 29 shift & 27:23 TOI/g = Average shift time: approx 57 secs
Then I compared that to other star players on winning teams yesterday.
Sidney Crosby - 24 shifts & 17:10 TOI/g = Average shift time : approx. 43 seconds
Alex Barkov - 22 shift & 15:30 TOI/g = Average shift time : approx 42 seconds (and this game went 5 minutes in OT)
You don't need to be a rocket scientist to know and understand what a lot of NHL coaches know and that is shorter 40-45 second highly productive shifts are more effective than long drawn out shifts that last a minute or more.
Keefe main coaching strategy is to play Marner and Matthews together and then play them as much as possible to address all situations .. He did that against Columbus last year to break a trap and shutdown system and try and score, and he is doing it this year to try a fix a flawed dysfunctional PP etc etc . This is against low scoring opposition when he has a deep team at his disposal and a high scoring roster.
Keefe doesn't use his 2nd, 3rd and 4th lines very effectively via line matching to attempt to create mismatches, and generate more offense, even with home ice advantage and last change possible. He just taps Mitch and Auston on the shoulder and tells them to get out there to try and mask is coaching weaknesses, with their skill and talent level.
The whole value of having a deep roster is to use all 4 lines advantageously and strategically to overpower your underdog opposition, not play a couple players 25-27 minutes and the rest of the line-up 10 -12 minutes. Allowing your opposition to just focus on stopping a couple players to win the game.