Pelton mail: Should rookies take a year off to develop?

Part of the Process for the Philadelphia 76ers was not having Ben Simmons or Joel Embiid available for their true rookie seasons. 

This week’s mailbag features your questions on redshirt rookies, starting lineup patterns, and more.

You can tweet your questions using the hashtag #peltonmailbag or email them to peltonmailbag@gmail.com.

“Can NBA rookies benefit from taking a redshirt year? In a very small sample size, due to injury, players like Blake Griffin, Ben Simmons, Joel Embiid, Julius Randle and Nerlens Noel have benefited from redshirt years. These players all understood the pace of NBA life, strengthened their bodies, and worked on weaknesses in their game without dealing with the NBA regular season. They also avoided the loss of confidence due to suffering setbacks getting used to playing higher levels of competition during a grueling schedule. What do you think?”

— Joseph Brown

I’m hesitant to say these players have necessarily benefited from their year off because, naturally, we don’t have the alternative case for comparison. Would Embiid, for example, have been just as good in season three had he spent the previous two seasons developing on the court? We’ll never know.

What we can do is compare how players performed during their first NBA action (as measured by their player win percentage, the per-minute version of my wins above replacement player metric akin to player efficiency rating) with what my SCHOENE projection system forecast for them based on their translated college stats.

Led by the Philadelphia 76ers duo of Joel Embiid and Ben Simmons, whose season is obviously not complete, these players have generally outperformed their projections. That said, before we start recommending a mandatory redshirt year for NBA rookies, it’s worth noting that this sample is heavily biased toward top picks. Three of the six players were taken No. 1 overall; Embiid and Noel might have joined them if not for their injuries. So it’s possible that I’m simply underrating these elite prospects based on their college performance, or how much stars improve at this age relative to non-stars.

The results weren’t nearly as good with the two non-lottery picks who fit into this category. Daniel Orton played just two games in the G League as a rookie, while Elliot Williams missed his entire first season but isn’t included because he played just 149 minutes in his subsequent rookie campaign before suffering yet another severe injury. So it’s worth continuing to watch as the sample grows larger over time.


Quite the opposite, in fact. Offense tends to improve over the course of the season, and therefore naturally so do shooting percentages. Here they are broken down by month from 2007-08 through 2016-17:

One question I’ve always had is what explains this improvement over the course of the season: better shooting or easier shots? The fact that free throw shooting is relatively stable outside of the small October sample suggests the latter explanation, but data provided by Second Spectrum allows us to actually take into account shot difficulty using the quantified shot quality (qSQ) metric. The qSQ of a shot is the expected effective field-goal percentage (eFG) based on the location and type of shot and location of nearby defenders.

Using last season’s qSQ data, and comparing it to actual eFG, it becomes clear that teams get better shots over the course of the season, but that their ability to convert them also improves rapidly early in the schedule.

So there you go: The value of fresh legs at the start of the season is apparently outweighed by the lack of rhythm in game situations.


“I have a question about the diminishing importance of the centre position (if that is what is happening) and whether the statistics on minutes played by more traditional centres tell us anything about this. Anecdotally, it seems like many traditional centres are starters, but their total minutes is often less than the sixth and seventh men on their teams. So whilst they are starters, their minutes seem to be rarely in the top five of their team. If this is a trend, is it a recent phenomenon or has it always been like this? If overall their usage is relatively low, why is the probability of them starting so high every game?”

— Matthew Carbines

To answer whether this is a recent phenomenon, I decided to come up with a statistical definition for the kind of player Matthew is discussing: someone who is in the team’s top five for games started (as a percentage of games played, minimum 20 games) but not in the top five in minutes per game (again using the same 20-game minimum). Having found such players back through 1981-82, the first year for which Basketball-Reference.com has data on starts, I then sorted them by position. Here’s how that looks graphically:

It has always been more common for players listed as centers to start games but not be among their team’s top five in minutes per game; they account for more than 40 percent of all such seasons since 1981-82, including one season — 2005-06 — when a full half of teams had a qualifying center. In that context, the rate is actually relatively down the past few seasons, with an average of 10 low-minutes centers per season.

Why would centers in particular start but play relatively few minutes per game? Foul trouble is surely part of the equation; sometimes, centers would be in the top five in minutes per game if they could stay on the floor without fouling. It may also be logical to start a traditional center because rim protection (and, at one point, post defense against star offensive centers) are at a premium when starting lineups are on the court. But more than anything, I think this speaks to a belief for coaches in starting with the base lineup and adjusting from there, something that also plays out in playoff series with teams playing smaller as the series goes on.


I think there are two key factors at play here. First off, in its infancy the league surely couldn’t have afforded to hire the crew of stat teams that now help input stats using a computer program. Odds are we’re talking about one scorekeeper who would have found it impossible to keep track of everything we now include in the box score, particularly given the faster pace of play at which the NBA played after adopting the shot clock.

Second, there presumably wasn’t the same kind of interest in or demand for statistics at that point. The advanced statistics we now calculate weren’t even a gleam in anyone’s eye in the 1950s. Steals, blocks and turnovers might have been useful to writers describing the action, but I doubt the demand really existed at the team level during an era when player-coaches were commonplace.

While I’ve offered some defenses for the NBA, it’s tough to understand why the upstart ABA was able to track turnovers for players and teams from its inception in 1967-68 while the NBA didn’t add team turnovers until 1973-74 and first recorded individual turnovers in 1977-78, a season after the merger with the ABA.

The lacking early statistics are a shame because unlike baseball’s robust statistical history, we can’t come close to capturing the complete value of legends like Wilt Chamberlain and Bill Russell without blocked shots, or Jerry West without steals, or so on and so forth.


“Why couldn’t all back-to-backs be on the home team’s schedule?”

— Michael Doliner

This is an interesting possibility given that there’s evidence that back-to-backs with travel are particularly difficult, which makes sense because teams must fly out after the game and typically don’t arrive in the next city until late. Home back-to-backs, by contrast, shouldn’t affect players’ sleep schedules much.

That said, a relatively small proportion of back-to-backs are played at home, and I think the explanation is financial. First, a big reason back-to-backs evolved on the road is because the minimize the time teams spend in hotels, reducing cost. (From a less bottom-line perspective, they also get players, coaches and other team employees back home to their families faster.)

Second, I suspect teams believe that playing back-to-back games at home tends to depress single-game attendance and makes it more difficult for season-ticket holders to attend both games. So I think we’re more likely to see a further decrease in back-to-backs than we are the league move more of those games to home.

Leave a Response