The 2020 NBA Draft is coming up Wednesday, November 18th. Countless hours have been spent scouting and evaluating players, determining team needs, and making mock drafts. No other team sport is as dependent on its star players, and the draft is supposedly the great equalizer. The Cleveland Cavaliers were the worst team in the NBA in the ’02-’03 season, drafted LeBron James that spring in the ’03 draft, and were playing in the NBA Finals four years later. A great draft pick can radically change the tides of an NBA franchise.
The Lottery
Each year, the first 14 picks in the draft are assigned via a lottery system, with the teams with the worst records in the prior season receiving the highest odds to select first. Historically, the importance of these “lottery picks” has been sometimes undervalued by league executives, with teams trading top picks away for mediocre players. Recently, you could argue that they might be overvalued: in the Philadelphia 76ers “process” era from 2013 to 2016, the team was intentionally built to be as bad as possible in order to increase the team’s chances of winning early draft picks (in these three seasons the team had 47 wins and 199 losses, including the 10 win, 72 loss ’15-’16 season, which ranks as the third lowest winning percentage in NBA history).
The consensus for this year’s draft is that it is weak at the top of the class, but has a lot of potentially strong options in the later part of the lottery and the rest of first round. In this post, I’m not going to predict who will be the best player in this year’s draft, which player should go to which team, or anything like that. I am only looking at how much of a crapshoot the draft lottery is in general. As we’ll see, there are great players in each class – the drafts that we’ll be looking at have featured at least a multi-time All-Star and usually a Hall of Famer – the tough part is finding this needle in the haystack.
The Redraftables
One of my favorite things to listen to this spring/summer has been Bill Simmons’ Book of Basketball 2.0 “Redraftables” podcast series. The name pretty much sums it up: with the blessing of hindsight, Bill and a guest repick a given year’s draft class based on the careers that the players have had. These redrafts aren’t always perfect – sometimes the picks turn into a contest to name the most obscure player and make a case for him, sometimes the picker’s personal biases get in the way of picking the consensus better player, and sometimes they are just mean – but, for the most part, I think that they do a better job of ranking players based on the full scope of their career than any one statistic can capture. (For a great podcast by the players, I recommend Knuckleheads, especially the Shaq and Dikembe Mutombo episodes).
For example, in the 2002 Redraft, Udonis Haslem is taken with the 8th pick, ahead of Drew Gooden and Luis Scola, both of whom have better marks in advanced metrics, like value over replacement player (VORP), win shares (WS), and box plus/minus (BPM), and basic stats, like average points or rebounds per game. Maybe it’s because a lot of my philosophy on basketball was/is informed by Simmons himself, through his Book of Basketball or the columns written on his website(s), but I think that Haslem is the best player of the three and had the best career. I value Haslem’s playing meaningful minutes for and being an important player on three Championship teams higher than the additional VORP that I supposedly get from Gooden or Scola. I really like advanced metrics and am certainly not one of the people denigrating basketball analytics (in fact, I’ll later rely on these stats to help make comparisons), but I think where a player is picked in these Redraftables provides one of the best means we have of grading the players in each draft.
I listened to all of the Redraftables and recorded the picks, then pulled the actual draft order from Basketball Reference. This comprises every draft from 1996 (Kobe Bryant, Steve Nash, Allen Iverson) to 2010 (Paul George, Gordon Hayward, John Wall). The Redraftables only re-pick the top 14 or 15 players in each draft, which is usually more than enough, most times they are scraping the bottom of the barrel in the late picks.
The Results
How Many Lottery Picks Stay Lottery Picks?
The first thing I looked at was whether a player who was selected in the actual draft was taken again in the redraft (FYI – a mispelling of “redraft” can yield “redfart”). Of the 215 total redrafted picks, 115 of these players were reselected with hindsight. That sums up the NBA draft lottery – just shy of half of the players drafted did not live up to expectations and would not be selected in the lottery again (or maybe at all).
Funny Seeing You Again!
Another simple thing I looked at was how many lottery picks each team had in this 15 draft span. As stated before, the teams with the worst records get the earliest draft picks. That means, barring trades, that the teams with the most lottery picks in this timeframe were also the teams with the worst records. The Los Angeles Clippers led the pack with 15 lottery picks in this timeframe, followed by the Golden State Warriors with 13, the Toronto Raptors with 11, and the Chicago Bulls and Minnesota Timberwolves with 10 (I’m not combining the lottery picks for franchises that moved cities or changed names). Conversely, only two teams had just a single lottery pick in this timeframe, the Los Angeles Lakers and the San Antonio Spurs. From 1997 to 2010, one of these two teams represented the Western Conference in the NBA Finals 11 times.
To refine this search further, I also took a look at how many teams had multiple top 3 picks in this draft timeframe. The Bulls had 5 and the Clippers and Vancouver Grizzles had 4. Looking at the list of these three teams’ picks, it’s clear why they kept finding themselves back in the top 3 picks; none of these draft selections really panned out until Chicago’s 2008 pick of Derrick Rose and LA’s 2009 pick of Blake Griffin.
Where Does A Top 3 Player Come From?
We can’t be too hard on these teams for failing repeatedly to get a top 3 player with a top 3 pick. Of the 45 top 3 picks in our draft range, only 18 actually turned out to be one of the best three players in their draft class. This means that the average likelihood of getting a top 3 player with a top 3 pick is about 40%. Another way to think about it is that each year, roughly one of the three players taken with a top 3 pick will actually go on to become one of the top 3 players in his draft class.
The likelihood continues to drop as we move to the later picks in the lottery. Interestingly, the 4th-6th pick and 7th-9th pick present about the same probability of yielding a top 3 player, at 13% and 18%, respectively. It is worth acknowledging that statistically it seems that the 7th-9th are preferable to the 4th-6th, but ultimately I think this is more randomness than real trend. What is significant is that by the time we get to the late lottery (10th-12th and 13th-14th picks), there doesn’t seem to any advantage in landing a top 3 player than if you were picking at any other point in the draft.
In a nutshell, history says that in this year’s draft one of the top 3 players will be taken 1st-3rd, one will be selected 4th-9th, and one will go 10th or later.
Why The Obsession With Top 3 Players?
We keep talking about top 3 players, but why? How much better are the best 3 players in a draft than the rest of their class?
In order to judge this, I’m going to use the advanced metric box plus/minus that I mentioned earlier. You can read much more here, but basically BPM is “the player’s contribution in points above league average per 100 possessions played.”
My dataset has the player’s BPM calculated over their entire career, so it tends to reward players who developed quicker or retired closer to their peak, because they don’t have as many weaker seasons from early or late in their career dragging their score down. For example, Dirk Nowitzki’s career spanned 20 seasons; he had a tough transition to the league his rookie year and he played ~5 “twilight” seasons where he was past his prime. His career BPM is 4.5, but during his three-year-peak from 2004-2007 his BPM was about 8.0. So, just keep in mind that these BPM’s are usually a bit low from a player’s career peak.
Now that I’ve explained BPM a bit, I can get back to answering the original question – “How much better are the best 3 players in a draft than the rest of their class?”
The answer can be seen in the histogram below. This counts the number of players who have BPM’s that fall between -4 and 9. For example, at the far right of the chart, we can see that we have one player in our dataset that has a BPM between 8-9 (LeBron James). The different colors represent when the player was picked in the Redraftables (i.e. the ranking of the player in that year’s class).
We can see that, typically, the best players in a draft class have the highest BPM. This shows that either: a) BPM is a pretty good means of assessing the quality of a basketball player, b) the draft order in the Redraftables was based on highest BPM, or c) a mix of both. I think that the answer is probably c.
We can use the average BPM of each pick to assess the average difference in quality between these picks. The average BPM for the 1st Redraftables pick (AKA the best player in the draft) is 4.4, which, according to the BPM explanation corresponds to something between an All-Star and All-NBA caliber season. Similarly, the average second-best player in a draft class has a BPM of 3.3, somewhere between “good starter” and All-Star. The third-best player’s average BPM is 2.1, putting it solidly in the “good starter” range. The BPM from the 4th to the 14th best player drops steadily from 1.4 to -0.7, meaning “decent starters or solid sixth man.”
The 2008 draft class is perfect illustration of this: Russell Westbrook is the number 1 player and has a BPM of 4.8; Brook Lopez is the number 4 player and has a BPM of 1.4. Regardless of your feelings about Russ (he may be traded as I’m typing this), this helps explain why top 3 players are so important in each draft class – from the best in the class to the fourth-best in the class, the talent drops off from All-NBA player to decent starter.
How Often Do Teams Make The Right Pick?
Now that we’ve shown the importance of a top 3 player and explained BPM, we can see how good NBA general managers and scouts are at picking the top talent with the top spots in the draft.
In the histogram below, I show the BPM for all players drafted between 1996 and 2010 who have played at least 2000 minutes.
We can see that top 3 picks tend to be better than other lottery picks, and lottery picks tend to be better than all the other picks. This could mean two things:
1) Teams have enough eye for talent that they can see the difference between a future bench player (the average outcome of all picks) and a good starter (the average outcome of a top 3 pick).
2) Because a player was drafted so high, teams feel that they have more invested in a player and work harder at developing him, giving him more playing time, coaching, and attention.
Again, I think that in reality, it is probably a combination of the two.
What’s a “Weak” Draft Anyway?
As mentioned earlier, if you’ve heard anything about this year’s draft class, you’ve heard that it is “weak.” So the final thing I wanted to look at is what constitutes a weak draft?
The bar plot above shows the cumulative BPM for the lottery picks in the Redraftables and in the actual draft. Obviously, the blessing of hindsight allows the Redraftables lottery to glean the maximum BPM out of each class. Without a crystal ball in every front office in the league, we could never expect this to happen in reality.
Some years, even with hindsight, can’t help but have negative BPM’s. For example, the 2000 and 2002 draft classes just didn’t pan out like some of the others did. I think everyone would agree to call these drafts weak.
Some classes are strong. The 1996 and 2003 draft each featured 3-4 Hall of Famers, including Kobe Bryant and LeBron James.
More interestingly still, some draft lotteries missed bigger on the available talent than others. The 1999 draft’s second-best player was Manu Ginóbili, who was taken with the 57th pick. The 2006 draft saw multi-time All-Stars Rajon Rondo go 21st, Kyle Lowry go 24th, and Paul Millsap go 47th. It will be interesting to see where this year’s class ends up in 20 years.