Blake Murphy Sports Writing

Title: How Does Pace of Play Impact FanDuel NBA Contests?
Date: October 30, 2015
Original Source: FanDuel
Synopsis: I debuted at FanDuel, where I’ll have a bi-weekly analytics column, trying to determine what – or who – impacts the pace of individual NBA games.

Call it the Michael Carter-Williams effect.

As a rookie on the moribund Philadelphia 76ers, Carter-Williams put up ludicrous counting stats by freshman standards. Given the opportunity to start and average 34.5 minutes for a team that had little means of scoring efficiently – the 2013-14 Sixers would ultimately rank 30th in offensive rating – the No. 11 pick was free to loot in the riot, so to speak. He averaged 16.7 points, 6.2 rebounds, 6.3 assists and 1.9 steals, and while he would have hurt a season-long fantasy player’s percentages, he wasn’t inefficient enough to negate the robustness of his raw numbers.

That anaemic Philly offense also happened to play at the league’s fastest pace (the estimate of a team’s number of possessions per-48 minutes), averaging 99.2 possessions per game. Those extra possessions did Carter-Williams a great service in his eventual Rookie of the Year victory, and had the tempo-boost been predictable, it may have helped FanDuel players find value when Philadelphia was in an environment to really push the pace.

Philadelphia played at a fast pace as much as they could that season, but would Carter-Williams still have provided value if the Sixers were hosting the Memphis Grizzlies? What about in Memphis? The Grizzlies played at a glacial pace, averaging 89.9 possessions per game – would that slow Carter-Williams and the Sixers down? Would last year’s slowest team, the Utah Jazz, be able to force a grind-it-out pace against the league’s fastest team, the Golden State Warriors?

Coaches talk all the time about adjustments to the opposing team’s style of play. When styles really clash, particularly in terms of size, the game can become a chess match of constantly changing lineups, in part aimed to dictate the tempo of the game.

It would serve daily fantasy players well to know, going into a game, if one team’s preferred pace can be expected to “win out,” so to speak, and force the tempo to one pole rather than the teams meeting in the middle. Over-unders are a useful tool in this regard, but determining which team dictates pace should help further understand game totals and, more importantly for FanDuel players, the total possessions in a game.

To determine which team has a greater influence on pace, we pulled game logs from the 2014-15 season and compared them to the season-long pace ratings for each team. While this hides the potential for in-season changes in style of play, a team’s 82-game pace is the best approximation available, and we used that number when determining who influenced pace.

We examined five different potential impacts on pace: the home team’s pace, the visiting team’s pace, the faster team’s pace, the slower team’s pace, and the average of the two teams’ paces. The assumption going in was pace would approximate the average of the two teams’ paces, but maybe it’s the faster teams that force an opponent to speed up, or maybe the home team has an environmental edge in setting the tempo (shout out to Game Ops).

As it turns out, pace appears to be fairly volatile game-to-game. Taking the average of the two team’s paces can explain about 23 percent of the variance in a game’s pace, but that’s a fairly small impact despite being the largest of any pace factors. The slower-paced team seems to have a slightly more pronounced impact than the faster-paced one, but it’s not a major difference.

To drill down further, we isolated for only those teams in the top- or bottom-five in pace, asking the question of whether teams playing at the extremes exhibit more control than those in the middle. When a top- or bottom-five team is involved in the game, the average of the teams’ paces can explain about three percent of the variance in a game’s pace. When the 20 teams in the middle played against each other, pace was essentially unpredictable.

There also wasn’t much of a relationship between a team’s overall pace and the variance in their game-to-game pace – while the Warriors and Jazz had narrow ranges, the overall R-squared for pace and range of paces was just 0.013. Whether teams varied in pace or stayed to a tight range seemed to be largely team dependent (it’s possible there’s a confounding variable at play, and I’d imagine a team’s free-throw rate is worth looking at for a further study).

This is a bit of a disappointing result, though not a bad reminder that as much as styles dictate fights, so can the flow of the game. Fast teams can play slow when called upon to do so, and slow teams can play fast when required – how that effects team-level output is another article, as the idea here was to project possessions a player may be on the court for in a particular game. The pace a player’s team plays at may not be the most accurate gauge for a single game, but the combined pace of the two teams playing gets us at least part of the way there.

Now the trouble comes in determining what pace teams will play at this season. The league got its first reminder not to take the preseason too seriously on Wednesday, when the Toronto Raptors and Indiana Pacers – the preseason’s slowest and ninth-slowest teams, respectively – played at a frenetic, 106.5-possession pace.

Ahead of last season, a team’s preseason pace explained roughly 49 percent of the variance in regular season pace, per a study I conducted for Nylon Calculus. The predictive rate shrunk when examining changes in pace rather than just pace, so it may be best to use preseason pace rates as a vague proxy of the tempos to come.

And, as with most things, a case-by-case qualitative analysis should accompany any preseason extrapolation. The 76ers and Warriors were among the preseason’s fastest-paced teams, which is hardly surprising. The presence of the Los Angeles Clippers at number two and of the Washington Wizards, complete with a new offensive philosophy, at number four, are ranks that warrant further examination once real games are underway. The Grizzlies and Jazz, meanwhile, remained sloth-like, likely portending a continued preference for the leisurely.