APBRmetrics

For entire playoffs, both the Warriors and Cavs used 5 lineups for about 44% of the minutes. For the Warriors the most used 5 lineups accounted for about 70% of the net pt total. For the Cavs about 86%. The net pt / min. of the most used 5 lineups were 320-370% of the average performance for the rest of lineups. But both choose to spend 56% of the time with those lower efficiency on average lineups. A total of 118 other lineups for GSW, 112 for CLE. Room for improvement? There are player minute limitations and it would need further study to take full and fair account of context; but I'd think so. I keep seeing that the most used 5 lineups do the most good and all the rest of the lineup jockeying combined is a junior partner at best; and on worse than the very best teams it is usually a negative, sometimes a huge one.

One should look at splits for the most used 5 lineups and try to identify the best and worst matchups and usage situations to try to improve the performance. Same with the most used among the rest too.

Warriors lineups used over 5 minutes total against Cavs, 13. 8 good, 5 bad. Qualifying Cavs lineups, 10. 4 good, 6 bad. So the winning coach picked right a bit over 60%. Loser 40%.

Warriors 13 lineups used over 5 minutes, plus 43. The other 52 lineups used for less than 5 minutes each? I believe 0, or 1 different. Adding nothing or virtually nothing. Cavs "bigger minutes" set -58. Blatt won the super dink lineups battle (those used less than 1 minute per game) by 15 overall (perhaps in part by using just 35 instead of 52) but lost the main war using lineups over 5 minutes by a lot more. Kerr's mostly micro tactics over Blatt's micro micro and micro micro micro tactics that played seconds or even less than one second.