Like NBA father, like NBA son?
There’s a theme to this week’s mailbag, which answers questions about the league’s legacy players. First, we take a look at how similar fathers who played in the NBA are to their sons who have reached the league, using sibling duos as a point of comparison.
Which father and son duos are most alike? Which are most different? And which traits tend to carry over the most from one generation to the next?
Then we take a look at how much more likely these kinds of relationships are in basketball, where having a parent who played professionally dramatically increases the child’s chances of playing in the NBA.
Finally, I answer a question about whether point differential holds its predictive power when it comes to team records at the extremes given the outlier performances we’ve seen from two teams over the past two seasons.
Throughout the NBA season, I answer your questions about the latest, most interesting topics in basketball. You can tweet me directly at @kpelton, tweet your questions using the hashtag #peltonmailbag or email them to [email protected]
“Do NBA fathers and sons often have similar strengths? thinking about Seth, Steph and Dell Curry all being great shooters or Gary Payton 1 and 2 being excellent defenders”
Thanks to Basketball-Reference.com generously supplying the data it tracks on NBA relatives, I was able to answer this question by calculating the similarity scores between stats for all sets of fathers and sons who both played at least 500 NBA minutes since individual turnovers were first tracked in 1977-78.
On average, the similarity score between fathers and sons was 80.7 out of 100, which indicates some general similarity but not particularly close. For example, I generally use a minimum cutoff of 90 when using similar players to forecast development as part of my SCHOENE projections. And by contrast, I also got a list of siblings from Basketball-Reference.com and found that they were somewhat more similar on average (84.1 similarity score).
There are some sons who scored quite similar to their fathers. The closest match in my sample at a 97.1 out of 100 was between forwards Harvey Grant and Jerami Grant, one of Harvey’s two sons who have played in the NBA. (The other, guard Jerian Grant, was fairly average in terms of similarity.)
The Grant similarity might reflect a challenge in this method as the game evolves: Even though my similarity scores are standardized to league averages at the time, the same skill set that made the 6-foot-8 Harvey Grant a perimeter player now has his 6-foot-8 son playing often on the interior, including center at times in his career.
Despite The Mitten having similar defensive prowess to his legendary father, aka The Glove, the two Gary Paytons actually have one of the lowest similarity scores in the group at 50.3. A little shorter than his 6-foot-4 father, GPII (6-3) plays more of an interior-oriented game that has worked with the Golden State Warriors and doesn’t use plays nearly as frequently. The younger Payton’s steal rate is also off-the-charts on a per-minute basis (2.6 per 36), even as compared to the last guard to win NBA Defensive Player of the Year (1.9 per 36).
The Curry family is interesting in that Stephen and Seth went in slightly different directions from their father. For someone remembered primarily as a spot-up shooter, father Dell Curry was a high-volume scorer in his prime, maintaining a career 25% usage rate. Steph (29%, with more playmaking) has pushed that in one direction, whereas Seth has (until this season) gone the other way with an 18% career usage rate. They both end up pretty average in terms of similarity.
#NBAAllStar is a family affair.
— Golden State Warriors (@warriors) March 7, 2021
I expected Golden State’s Klay Thompson to be more of an outlier since his father, Mychal, was an interior-oriented player during the 1980s. They’re on the low side, but the sons of Hall of Famers Tim Hardaway (Tim Jr.) and Bill Walton (Luke) score as more different from their famous dads. Also near the bottom: the late Kobe Bryant, who easily outscored his father, Joe.
As for elements that carry over most from one generation to the next, height is naturally high on the list. Rebounding is the most stable trait between fathers and sons, with my measure of how much a player favors playing inside (free throw rate minus 3-point rate) coming in second. Usage, steal rate, block rate and turnover rate tend not to show much carryover at all.
Lastly, since I put together the data, let’s talk about similar siblings. It’s no surprise that sets of twins (Joey and Stephen Graham, Caleb and Cody Martin, Marcus and Markieff Morris and Jarron and Jason Collins) often have similarity scores to each other of better than 95. Brook and Robin Lopez will surely be thrilled to know they’re lower in similarity (93.6).
However, the two most similar siblings in my sample are also second-generation NBA players who continued their father’s legacy of NBA sharpshooting: Brent and Jon Barry, who were much more similar to each other (98.7) than either was to their dad, Rick (80.9 and 79.3, respectively).
“There are so many professional basketball players who have a parent or other close relatives who were professionals when they were children. From Luka Doncic to Devin Booker, from Carlos Delfino to Gary Payton II, and of course, Steph and Seth Curry, they all had parents who went professional ballers. Is there any way for you to assess whether having being born or grown in a family with ballers give you an edge, a know-how, a feeling for the game that put you already in a better position to succeed? Can you quantify that?”
Consider this: Of the 503 players currently under contract in the NBA, 27 had fathers who played in the league according to the Basketball-Reference.com data, or more than 5% of the league — a group that doesn’t include the international players you mentioned whose fathers played professionally but not in the NBA. It doesn’t take fancy math to know NBA players don’t make up that much of the world’s population. Your odds of making it to the NBA are probably somewhere north of 3,000 times more likely if your father played in the league.
As to why that’s the case, well, height is the first place to start. The NBA population is much taller than the world at large and height is hereditary. Despite regression to the mean between generations, if your dad was a 7-footer, there’s a much better chance of you having the height necessary to be successful as a basketball player. (Although, oddly, the group of fathers in the question above were slightly shorter on average than the typical NBA player.)
Wealth is a factor here as well. As much as we like to romanticize the notion of sports as a level playing field and the examples of great players such as the Los Angeles Lakers’ LeBron James coming from humble upbringings, that’s not the reality. A study by Joshua Kjerulf Dubrow of the Polish Academy of Sciences and Jimi Adams of Arizona State University in the International Review for the Sociology of Sport found that higher class background was associated with better chances of making the NBA over the timeframe they considered (1994 to 2004).
All of that makes it a little difficult to tease out the specific influence of the effect you mentioned: that sons of professionals are around the game from a young age. They’ve presumably got access to facilities and great coaching throughout their lives. They also know what is required to be a professional player and what to expect — something as simple as having been in locker rooms. There’s surely value to that, as well.
“With the Utah Jazz underperforming their point differential and the Oklahoma City Thunder overperforming their point differential, is the expected wins formula poorly calibrated with edge cases? Or are those two data points simply outliers?”
This question is actually from the end of last season, when the Jazz had the league’s best point differential and the Thunder the worst, but I decided to bring it up now because … well, it’s happening again.
Through Wednesday’s games, Utah’s plus-9.1 differential was second only to Golden State, while Oklahoma City is getting outscored by 7.1 points per game. Yet the Jazz were third in the West standings and tied for fifth overall, while the Thunder’s 6-8 record ranked 10th in the West ahead of the Sacramento Kings, who had outscored their opponents on the season.
Looking at full seasons over the past decade, it’s clear these performances are outliers. (Oklahoma City in 2020-21 was the one team with a point differential worse than minus-10 to win more than 25% of its games; Utah doesn’t quite stand out as much.)
So why are the Thunder and Jazz outliers? Typically, a team’s wins and losses will both come by similar margins on average, meaning their point differential is a product of how often they win or lose. Neither team conformed to that last season.
Utah’s average win in 2020-21 was by a league-high 16.2 points, the fifth-highest mark in NBA history. Meanwhile, Oklahoma City’s average loss came by 18.3 points per game, the second-worst ahead of the 1992-93 Dallas Mavericks. Meanwhile, the Thunder’s wins came by an average of 6.7 points.
Those same trends have continued. Oklahoma City’s wins this season have been by an average of 6.3 points and their losses by an average of 17.1 points, while Utah has the reverse splits (wins by 17.3 points, losses by 7.0 points). And that’s how your point differential looks very different from your record.