Whereas the lineup of T. Duncan | D. Green | K. Leonard | T. Parker | T. Splitter is just not stellar compared to the best in the playoffs as previously noted,
T. Duncan | M. Ginobili | D. Green | K. Leonard | T. Parker
is performing outright horrendous in the playoffs (-31 pts per 100 possessions) http://bkref.com/tiny/6ahdd
Very modest minutes but I question how much they should be counting on that lineup to lead them to a title. It was very good in 22 minutes of regular season "testing". That is of course not really adequate testing for your #2 playoff lineup. Not by a mile or 10. Pop instead used 568 lineups in the regular season. One over 4 minutes a game. 4 others over 1 minute a game. No way one can really analyze and choose / semi-rationally guess on any lineups based on data when a coach does that with an entire season of opportunity to test lineups. Some coaches adequately test 1-2 lineups. A few test a handful of others to the point were one can make some speculation on good bad or in the middle. Any team could do a better job of strategically testing key lineups if they valued that information. They don't appear to value it (in addition to in the game coaching for the win) given the way they handle the available minutes of a season so scattershot.
eWins per36 rates tm mpg Eff% Sco Reb Ast Stl TO Blk e480
.43 Lebron James Mia 41 .486 16.9 12.4 9.4 1.3 1.8 1.3 2.50
.23 Chris Bosh Mia 33 .463 13.8 8.8 3.4 3.3 .5 1.1 1.66
.22 Tim Duncan SAS 34 .403 13.7 14.7 2.7 .0 1.6 1.6 1.56
.20 Danny Green SAS 32 .967 23.4 3.7 .5 .6 .6 1.1 1.47
.15 Ray Allen Mia 25 .897 26.0 3.0 .0 .0 .7 .0 1.44
.14 Kawhi Leonard SAS 35 .404 8.1 13.7 1.6 1.1 .5 .0 1.00
.13 Tony Parker SAS 37 .472 16.9 1.6 5.4 1.0 2.5 .5 .83
.12 Mario Chalmers Mia 32 .551 16.9 3.1 2.7 .6 .6 .0 .88
.11 Dwyane Wade Mia 33 .435 14.8 2.4 5.5 .5 1.6 .5 .77
.10 Mike Miller Mia 18 1.17 22.6 4.0 1.1 .9 .9 .0 1.43
eWins per36 rates tm mpg Eff% Sco Reb Ast Stl TO Blk e480
.10 Chris Andersen Mia 14 .727 26.1 8.6 .0 1.1 2.3 1.1 1.79
.02 Norris Cole Mia 15 .389 6.8 4.6 6.6 .0 .0 .0 .36
.02 Manu Ginobili SAS 24 .474 13.1 1.6 2.8 2.2 2.9 .0 .17
.01 Matt Bonner SAS 8 .500 9.1 8.8 .0 .0 .0 .0 .43
.01 Udonis Haslem Mia 16 .636 8.6 7.0 .0 .0 .0 .0 .14
-.00 Gary Neal SAS 22 .500 13.1 3.4 .8 .0 2.3 .0 -.02
-.01 Boris Diaw SAS 10 1.00 5.1 1.7 4.4 .0 .0 .0 -.13
-.03 Tiago Splitter SAS 24 .458 7.4 2.4 .7 .0 1.5 1.5 -.27
Shooting: Mia .541 - .497 SA
Rebounds: Mia 41.0 - 40.5 SA
Turnovers: Mia 7.0 - 10.0 SA
In Game 2, Miami scored 55% of the points.
They gained 82% of the game's eWins and 113% of the Pyth-Wins. With the score tied and 15 minutes to go, both teams had about 34% of a win.
Thru game 2 the Spurs starting lineup is -2 in about 20 minutes, at least according to the gameflows. Will wait n see how it looks later in the series.
Substituting Ginobili for Splitter was terrible in game 1 and appears to have not been used in game 2. Reacting to "small sample size"- appropriate or not? The alternatives to that by and large didn't work in game 2.
If coaches don't test lineups to a much higher degree than they usually do then they are left to use these even smaller sample sizes or their "gut" / "experience". Larger sample sizes (hundreds of minutes in the regular season) may still be relatively small and relatively unreliable compared to what is desired but they may be better (when used with all other available information & judgment) than the alternatives. Hence my calls to use available minutes for more disciplined and targeted lineup research.
Coaches also react to what the opposing coach does. So yeah, if 2 opposing coaches set out to test their favorite lineups, come hell or high water, then we'll see larger sample sizes.
Since coach A considers himself a genius, and coach B calls himself an innovator, we'll continue to see small lineup samples -- particularly when opposing lineups are part of the considered combination.
When computer programs are allowed to do the coaching, then we'll see longer stretches of lineup Q vs lineup 21.
Miami still has 2nd best lineup used over 50 minutes in the playoffs but has not used it in the finals. So I dunno whether to "count it". Spurs' best such lineup has moved up to 9th best from 12th a few games ago. Heat starting lineup has fallen to 11th from top 3. The team whose main lineup finishes highest will get an edge from that. Perhaps decisive. Perhaps it will end up in the top 6 as usually occurs for title winners. Or maybe not.
Lineup matchups can be simulated to huge sample sizes, for what that might be worth. I imagine Bob C. is doing that for Miami. Dunno if the Spurs try anything like that or not.