The hidden side of politics

Pelton mail: Is this the best postseason rookie class?

Reported by ESPN:

This week’s mailbag features your questions on the Utah Jazz‘s second-half turnaround, playing with foul trouble and more.

You can tweet your questions using the hashtag #peltonmailbag or email them to [email protected].

“Between Ben Simmons, Donovan Mitchell, Jayson Tatum, et al, playoff production among rookies seems unprecedented thus far. After the first week of games, how does this rookie class stack up historically?”

— Matt Owens

Once upon a time, it was pretty common for more experienced rookies to contribute right away, so I don’t think this year’s class will stand out historically. After all, through his first two games Tatum was playing only slightly better than replacement level and Mitchell rated at league average because of his below-average efficiency.

What does stand out, particularly relative to recent seasons, is how much these three players are being asked to do. Through their first two games and Simmons’ third, all three were averaging at least 37 minutes per game. You have to go back to 2004 to find a postseason when three rookies from the NCAA averaged even 30 minutes per game (Carmelo Anthony, Marquis Daniels and Dwyane Wade) and that’s as many as have reached 30 minutes per game as in the past seven playoffs combined (Harrison Barnes, Malcolm Brogdon and Taurean Prince are the three to do it in that span).

Given the trend toward less prepared rookies and fewer minutes for star players, the last rookie to average as many minutes per game in the playoffs as Mitchell logged through his first two games (39.0) was Derrick Rose in 2009. So this year’s top rookies are definitely out of the ordinary.

“I was looking at the stats for this year’s Jazz team, and realized that they had a plus-8.8 differential in the second half of the season. Are there past teams that have turned their season around so drastically over half a season — they had a negative differential before then — and if so, how did they do in the playoffs?”

– Josh Larson

Going back to 1984, the first year the playoffs expanded to 16 teams, the Jazz had the third-largest improvement in point differential between the first and second halves of the season of any team — though only the second-best of the past two years, because the Miami Heat had the biggest turnaround last season.

This year’s Utah team has rightfully been compared to the 2003-04 Heat, whose in-season turnaround was also driven in part by a high-scoring rookie guard (Dwyane Wade, in Miami’s case). The Heat parlayed that second-half success into a first-round win over the New Orleans Hornets. Is that typical?

Let’s take a look at the other teams in the top 10 who reached the playoffs, comparing their success points (awarding a point for a win, subtracting one for a loss and giving four points for each series played so the numbers can’t be negative) to what we’d expect from a team with their playoff seed.

There doesn’t appear to be much of a trend here. The Heat winning a series and taking two games off the top-seeded Indiana Pacers in the second round beat expectations, but four of the other five teams fell slightly short of what we’d expect from a team with their seed. The big exception here are the 2015 Cavaliers, who utilized a roster remade with two midseason trades and LeBron James‘ tendency to step it up in the playoffs to reach the NBA Finals.

Overall, point differential in the first and second halves of the season have basically the same correlation with playoff point differential adjusted for opposition. So in general there’s no particular reason to believe strong finishers will perform better in the playoffs than teams that start well, aside from roster changes like Cleveland’s trades and Rudy Gobert‘s return to health.

I don’t think it’s a satisfying answer for either side, but the single dominant factor in this series remains shot making. After Game 1, I noted that Second Spectrum’s shot-quality metrics suggested the Trail Blazers had actually gotten pretty good shots. Based both on quantified shot quality (qSQ, which measures the performance of an average shooter given the shot location and type and location of nearby defenders) and quantified shot probability (qSP, which also attempts to take into account the shooter’s ability), Portland should have shot much better in that game.

Lo and behold, the Blazers have shot much closer to their expected mark the past two games. There has been only one problem: New Orleans has shot much better than expected in those two games. Here’s a full chart of qSQ and qSP throughout the series.

Now, there’s somewhat more to the story than this. Game 3 was the first time in this series New Orleans got better shots based on quantified shot probability, and the Pelicans also had seven more shot attempts by virtue of Portland’s 24 season-high turnovers. So even given average shot making by both teams, we’d have expected New Orleans to win that game. It’s unclear how much of that is the Pelicans playing at home and how much has to do with growing frustration among the Blazers.

I also suspect there are some benefits to New Orleans’ length on the perimeter and the way Damian Lillard in particular has rushed shots for Portland that fall outside this analysis. Still, I think it’s clear that this series would look very different if more shots had fallen for the Blazers in Game 1 and fewer for the Pelicans in Games 2 and 3.

“Is there a correlation between players getting in early foul trouble (i.e. Victor Oladipo getting two quick fouls Wednesday night and getting pulled out early) and then fouling out later in the game? Looking for more insight into why coaches pull players out so early when there’s a high chance they don’t end up fouling out anyway. Another variation is if there are any stats to back up a common notion that players who get in early foul trouble tend to play more passively — i.e. fewer rebounds/steals/blocks?”

— Abhinav Dantuluri

Naturally there’s a correlation, though it’s a little tough to answer the question that way because coaches do typically pull players in foul trouble. So it’s not necessarily telling that Oladipo did not foul out once this season. To really understand the alternative scenario, we’d need a world where coaches never adjusted to foul trouble until the player fouled out; that way, we’d know how many minutes were being foregone by pulling the player early.

There’s a better answer for your second question. As I’ve noted in a past mailbag, in 2011 the researchers Allan Maymin, Philip Maymin and Eugene Shen looked at the performance of teams with players in foul trouble and found that when a starter was on the court with at least one more foul than the quarter being played (i.e. two fouls in the first, three in the second, as caused Nate McMillan to pull Oladipo), his team played much worse than expected. Their conclusion was that “foul trouble makes a star player perform worse than an average player.”

Based on that research, I agree with Zach Lowe’s take on this week’s Lowe Post that the decision to take Oladipo out wasn’t as obviously wrong as most of the reaction on Twitter suggested. Still, Oladipo is a low-foul star player, and his backup (Lance Stephenson) a below-average one. In this case, I think keeping Oladipo in the game would have been advisable.