Photo by Elsa/Getty Images In his day job, ProfessorB is an award-winning social scientist. His writing has appeared in the New York Times, Washington Post, and other major media outlets. But he also dabbles in NetsWorld. In their first 24 games (before facing Golden State), the Nets utilized 222 different five-man lineups. Most of those lineups have been on the court for just a minute or two all season, and only nine have played even 20 minutes together. While it is fun to debate their relative merits, we have far too little evidence to seriously evaluate them as units. But we know that lineups matter. Successful teams thrive on synergy, not just the sum of their separate parts. Hard as it is to evaluate five-man lineups, much more information has already begun to accumulate on the performance of two-man combinations. Mikal Bridges and Spencer Dinwiddie have played 579 minutes together, hardly a small sample. Most of the 66 pairings of the Nets’ 12 rotation players have played at least 130 minutes together. Evaluation of these player pairings provides a fruitful middle ground between focusing entirely on individual performance (ignoring synergies between players) and attempting to find meaning in the performance of hundreds of distinct five-man lineups, most of which have played only fleetingly. Which pairs of Nets have played well together so far this season, and which have struggled? Non-metric multidimensional scaling is a handy tool for visualizing relationships among objects—in this case, the relative success of pairs of players as measured by their net rating (points scored minus points allowed per 100 possessions on the court together). The aim of the scaling exercise is to reduce the 66 two-man ratings to just two dimensions, placing players close to each other if they’ve played well together and further apart if they’ve been less successful. Each pair’s rating is weighted by (the square root of) the number of minutes they’ve played together; so the mapping is most accurate for the pairs of players whose ratings are most reliable. 
The Nets’ strongest pairing so far this season has been between Spencer Dinwiddie and Day’Ron Sharpe. Their net rating has been stellar, +19.0 in 221 minutes on the court together. (While some other pairs have even better net ratings, they have had far less playing time and hence are less reliable.) Moreover, their location near the middle of the map reflects the fact that both have played well with most of their other teammates. Dinwiddie has recorded positive net ratings with 9 of the 11 other rotation players (all except Watford and Claxton), while Sharpe has recorded positive net ratings with 8 of 11 (all except Claxton and Simmons, both in very limited minutes, and Finney-Smith). The largest distances in the figure generally indicate less successful pairings. Some of these seem unsurprising. For example, analysts and fans were speculating even before the season began about whether Nic Claxton and Ben Simmons could succeed together, given their offensive spacing issues. While Simmons and Claxton have managed just 16 minutes together due to injuries, their net rating of -17.6 (and offensive rating of 100.0) has done nothing to quell concerns about their effectiveness together. Their positions on the margins of the map also indicate that both have been ineffective in most of their other pairings as well. Claxton has just three positive net ratings (with Johnson, O’Neale, and—barely—Bridges), as does Simmons (with Dinwiddie, Walker, and—in just 6 minutes—Watford). Perhaps more surprisingly, the Nets’ two veteran two-way wings, Dorian Finney-Smith and Royce O’Neale, likewise appear on opposite ends of the map. That distance reflects their dismal -12.9 net rating in 345 minutes together—the most substantial record of subpar performance by any pair so far this season. Perhaps their respective strengths and weaknesses are too similar to make for an effective pairing. However, the distance between them also reflects different patterns of success with other teammates. Finney-Smith’s best minutes have been with Johnson, Dinwiddie, Claxton, and Bridges. Indeed, that five-man lineup is the Nets’ most frequent so far this season, with a net rating of +8.7 in 107 minutes together. It was the starting lineup at the end of last season and again while Cam Thomas was out for 9 games last month. O’Neale’s success, on the other hand, has come mostly with bench players—Watford, Sharpe, and Walker—along with Johnson. Any four of those five, with Dinwiddie or Smith as point guard, would provide a nicely balanced lineup. But, please, less O’Neale with Finney-Smith—and, for that matter, less O’Neale with Bridges, too. (Their -5.4 net rating in 389 minutes together is another one of the Nets’ most frequent subpar pairings.) This sort of mapping can also suggest promising lineup possibilities that haven’t already been tried. The five players clustered in the center of the figure—Dinwiddie, Bridges, Johnson, Watford, and Sharpe—have not played a single minute together; but why not? Watford and Sharpe have played very well as a pair (+10.4 in 111 minutes), and would be a viable alternative to Finney-Smith and Claxton (-0.1 in 215 minutes) supplementing the solid 1-2-3 lineup of Dinwiddie, Bridges, and Johnson (+5.6 in 331 minutes). Various lineups close to this one, but including Lonnie Walker or Cam Thomas, have performed well in limited minutes. It’s the coaching staff’s job to make adjustments over the course of the season, building on what has worked well, tweaking or shelving what has worked less well, experimenting on the basis of imagination and evidence, however tentative. That’s true for individual players’ minutes, but perhaps even more true for combinations of players. Given the vagaries of a long NBA season, no team has the luxury of settling completely on lineups or rotations. But keeping an eye on who plays well with whom should help the Nets maximize the effectiveness of their gritty, hard-working roster over the next four months and into the playoffs. Source: https://www.netsdaily.com/2023/12/17/24 ... ith-others