APBRmetrics

Walker - Smart is mildly positive but 2nd worst of 20 biggest minute pairs. Not going to be great with every guard pair, though might be ways to improve it.

Something that I have rarely seen: a detailed video review of a LINEUP. How about you? I can think of Warriors death lineup and... little else. Maybe a Sixers lineup. Maybe Celtics. Hardly anything else on one lineup. May be out there and I am unaware. Not just one or a few plays. The whole mix. Offense and defense too would be nice / appropriate.

Should be a common cut. I could probably make more recommendations about how and whether to use a lineup if I had the selection tool and took the time to study a lineup's video comprehensively for season.

Pistons are doing poorly. Have a great performing biggest minute lineup but used less than 4 minutes per game (6 minutes when Griffin is healthy). Have a great all bench unit but used less than 2 minutes / gm. Losing because of all the effectively random coach guessing and probably assembled talent / mix. The good stuff isn't all guaranteed good. Testing further / harder is the only way to find out.

16 of 20 most uses pair players are negative so few things are likely good. But if there are a few, the importance of emphasizing them is great.

Not enough 3s. Too many turnovers. Weak defense.

Popovich's offensive and defensive design / results are hurting the Spurs. One of the very least Moreyball shot distributions and possibly the least. 6th best offensive efficiency last season but slipped 8 spots to barely above median. Did defenses adjust to it? Without the results of last season, it is easier to criticize / harder to try to justify. Defense of a really poor team, probably design and effort.

So is his lineup management. 9 of 11 most used lineups negative. A great appearing bench lineup used only a bit over 1 minute per game.

Trade players? Yeah I probably would. But Popovich is holding Popovich and the Spurs back.

16 of most used 20 player pairs negative. The top 5 pairs look like they belong to a bottom 5 team.

They don't turn it over and defensive rebound but the rest of the factors are average to bad.

Their bench is great when uses properly. How would they do starting? I dunno, but they should play a lot more at minimum. I'd try a total lineup flip at least briefly to check what happens. But they probably won't. Too much drama and work for Pop.

"No meaningful test of Young with another PG (Turner or Goodwin)."

... 200 seconds of recent testing with Goodwin. Woohoo.

Crow wrote: ↑Thu Dec 26, 2019 7:03 am Blazers with 9 of 11 most used lineups positive but they still have a negative net points average. Cuz they are not using what works enough (nothing over 5 minutes per game) and using too many meh to bad lineups. 184 lineups used. More than 100 are negative for season. You don't need that many lineups. You can't reasonably test more than a few handfuls. And yet they try it. As the norm. Use a few dozen if you insist on stretching past a reasonable definition of reasonable. Hundreds of lineups to a great extent shows you don't know what is best.

Blazers using starting lineup far far more and well in last 10 games. 18 minutes per game. Didn't dig deep last time. But the advice was on target. Behavior changed after adding Anthony. So might have been motivated by wanting to make him comfortable/ happy rather than some analytic / reality -based conversion. At least they did the right thing.

Crow wrote: ↑Thu Dec 26, 2019 6:30 am What is working for Magic is 4-5 starters or 0, 1 and 2 at most starters. I'd try to eliminate a lot of the in between dink lineups and test some of the best starter heavy and bench heavy lineups far more. The biggest minute lineup is working. Could increase its minutes by 50-100%.

Magic 4-3 in last 7 games. 1 lineup used 6 minutes per game. 1 other over 3 minutes gm. Only 3 lineups used in majority of games and only 1 positive. Injury to Issac (very recently) but the best bigger minutes lineups were without Issac.

The best performing bigger minute lineups recently? Almost entirely with 4 or 1 starter. Should do this more.

Crow wrote: ↑Thu Dec 26, 2019 6:38 am Current starting lineup for Hornets has played 34 minutes so far for season. Getting to it late. Mildly negative. The version with Monk over Bridges is vastly better in similarly small sample. I'd try both a lot more. Neither player is good on individual stats but Monk way better on raw plus minus on court and on-off. Monk has never started an NBA game. If you are not willing to try that... why? What Borrego has been doing generally hasn't worked well, so try more of the untried but in a vigorous way. Test and test right. Use what works more til it doesn't.

I've seen some 2 (and 3) PG lineups around league do well. May need to study that more.

Version with Bridges still performing poorly recently in big minutes. Version with Monk got 54 seconds of total testing in last 9 games. Testing...

Crow wrote: ↑Thu Dec 26, 2019 7:16 am Of the handful of lineups tested over 1- just 1- minute per game, the only strong ones have Paul, Schroder and SGA together. They have been fabulous together. In short testing. 169 minutes of test. An average result of plus 14.5 points per 100 possessions. If I wanted to win, I'd take that 5.5 minutes per game testing up to 15-20 minutes per game.

Did they increase use of this guard trio? Nope. 6 minutes per game in first 30 games and exactly 6 min / gm in last 5 games.

Crow wrote: ↑Thu Dec 26, 2019 7:52 am LaVine, Satoransky, Dunn together plus 9pts / 100p in 313 minutes of testing. Test it more.

Went from less than 10 minutes per game usage in first 32 games to over 20 minutes / gm in last 4. Yay. Working even better / great.

Crow wrote: ↑Thu Dec 26, 2019 8:23 am Brunson - Wright, good. Brunson - Doncic, neutral. Could be other things but can Doncic share? He can with Wright in modest minutes. Why the big difference? Find out.

Brunson - Donic used 4 minutes per game in first 29 games before my comment. Over 6 minutes / gm since. Did well in that time. What did they change or find out, if anything? Test and study more.

Crow wrote: ↑Thu Dec 26, 2019 7:55 am Augustin - Fultz. Plus 8 in 125 minutes. Give it far far more testing.

Tested about 4 minutes per game before comment. 9 minutes per game since. Doing great.

Crow wrote: ↑Thu Dec 26, 2019 8:14 am Lowry - VanVleet modestly negative. Too small (shows some on rebounding) and not enough 3 point shooting around them? Those were my fast takes.

Lowry - VanVleet suddenly doing great since comment. Shift away from drawing net ftas vs.opponents to net greater 3 point activity. Coincidence or significant? Could study more.

"Do the job" and "do it well" and "it will get noticed".

Crow wrote: ↑Thu Dec 26, 2019 6:19 am Timberwolves lineups and player pairs with Covington generally weak. Either play some of his 5 best bigger minute lineups that look positive more (and cut random dink lineups) or give up and trade him. No Covington lineup used even 3 minutes per game for season.

Haven't succeeded with him, in part because haven't tried any disciplined selectivity with him.

Since comment , Covington has been used 9 minutes per game in 1 lineup. Did really great. No Towns, no Wiggins.