Data on the Rising Stars Challenge
Jan 24, 2015; Milwaukee, WI, USA; Milwaukee Bucks forward Giannis Antetokounmpo (34) drives past Detroit Pistons forward Caron Butler (31) in the fourth quarter at BMO Harris Bradley Center. Mandatory Credit: Benny Sieu-USA TODAY Sports
[Ed. Note: Like yesterday’s variance conversation, we’ve also been discussing the Rising Stars Challenge in our Nylon Calculus email thread. Michael Murray and Layne Vashro each put together a quick write-up of their analysis.]
Michael Murray: The Rising Star Challenge during this year’s all-star weekend will pit a team of U.S.-born first and second-year players against first and second-year players from around the world. The game has previously held a number of formats, always being a game of young players. Some previous MVPs of the game include John Wall, Kyrie Irving, Kevin Durant, Carmelo Anthony, Gilbert Arenas and Allen Iverson. Each team has 10 players, with three rookies on the U.S. side and four for the World. More stats like these after the jump. Here are this year’s rosters:
Many have pointed to the World Team as being particularly strong, so I looked up some of my favorite “one-number” metrics for each player to see if the U.S. Team is really that big of an underdog.
Yikes, looks like the U.S. sqaud is outmatched in every one of these statistical categories. These numbers don’t give the best read on where the fault is with the U.S. Team. Win Shares makes the U.S. offense look bad compared to the World while Real Plus/Minus makes the U.S. defense look much worse than the World’s. I decided to break the teams out by best line-ups by each stat type to see if the U.S. can hide some of its statistical weakness on the bench. The top two PERs belong to the U.S. Team, but they only have three of the top ten. The starting five by Win Shares aren’t even close. BUT, the U.S. team’s top five by RPM barely gets edged out 4.6 to the World’s 5.85.
Of course the heaping spoonful of salt to be taken with this is that season stats have little bearing on an exhibition showcase. But the numbers back up those expecting a big showing the World Team.
Layne Vashro: Everyone has their own opinion of which young players should/should not have been selected for the Rising Stars challenge. Personally, I felt that two players in particular, Jusuf Nurkic and Marcus Smart, were obvious candidates who should have made it over some of the players chosen.
However, this kind of debate is always a bit silly, since the actual criterion squishy. My two favorites above can both be reasonably rejected based on criteria like “minutes played” where I would reject some of the chosen players based on their dismal scores on various statistical metrics. Instead of forcing my own biases onto investigating which players should/should not have made it into the Rising Stars challenge, I tried using the criterion revealed by the actual selections.
To do this, I took all of the eligible players and used a simple automated model-comparison program to find the statistical values that best explained the selections. It turns out that most of the discrimination boils down to three basic statistics: Total points, total assists, and total rebounds. Nothing else was very important after accounting for these measures.
I then used the resulting model to calculate the predicted probability that each eligible player was going to be selected for the Rising Stars challenge and compared that to the actual selections. Here are the results:
As you can see, most of the selections are consistent with the revealed criteria. However, there are a few outliers. Most notably, it looks like Solomon Hill was snubbed. Despite putting up numbers like the other players selected for the competition, Hill was left off the list. On the other end of the spectrum, we have Dante Exum and Bojan Bogdanovic who seem like less likely participants than a number of players who were left off the list.