Timberwolves coach Chris Finch almost certainly has not been hacking into my web browser, but he did seem to be describing my behavior on Tuesday.
I had asked him after practice, as the 6-8 Wolves prepared to play Wednesday in Orlando, what the relative benefits are of looking at five-player lineup data this early in the season compared to clusters of how two or three players fare together.
"You can drive yourself crazy looking at a lot of five-man lineups, because circumstantially, you can't always get back to those anyway, and they're a small sample size at this point in the season, anyway," Finch said. "I think it's more about two- and three-man lineups that you want to focus on."
The NBA's web site makes all this information readily available, and some of us are more guilty than others of scouring the data while concocting new starting lineups out of players who have barely played together – something I talked about on Wednesday's Daily Delivery podcast.
The Wolves' preferred starting five of Rudy Gobert, Karl-Anthony Towns, Jaden McDaniels, Anthony Edwards and D'Angelo Russell has played 177 minutes together this season — a growing and relatively meaningful sample. No other group has played more than 36 minutes together, and the next-highest total is 16 minutes.
Break things into two-man or three-man chunks, though, and you get larger minute totals and more usable data. You see that no two-player group comprised of those five starters has a positive net rating, owing in large part to their minus-6.2 rating as a five-man group (meaning that per 100 possessions, they are being outscored by 6.2 points).
You see that Edwards with backup point guard Jordan McLaughlin (plus-25.0 net rating in 106 minutes) and Towns with McLaughlin (plus-20.3 in 99 minutes) have been clicking, albeit in samples less than one-third the size of Edwards-Russell (minus-6.7 in 336 minutes) or Towns-Russell (minus-5.9 in 341 minutes).
Smaller groups at this point might still tell us more about what the Wolves need than what they have.