Are Players Who Fall on Draft Day Really Steals?

Every year, several players fall a lot farther in the draft than projected. And as they do, NBA fans wonder why they are dropping and hope that one of them slips to their favorite team. Why? Well, because that player is a "steal" at such and such draft spot. In this past 2016 draft, the two notable "fallers" on draft day were Skal Labissiere and Deyonta Davis. At the Clips Nation Draft Night Meet Up (which was a ton of fun by the way), everyone present was increasingly excited that one of them would reach the Clippers at 25-- and hopeful that Doc would pick accordingly. When the pick came, many of us were unhappy that Doc picked Brice Johnson, even though it would have been considered a good pick just a couple hours earlier. But is there really anything to the idea that these "steals" actually outperform their draft position?

My idea of a "fall" was a drop from projection to actual draft pick of nine spots or more. Nine seems like a rather arbitrary number, but 10 seemed like a bit much, and anything less was too close. As for the projection, the only measure I used was Draft Express's final mock draft—which was shockingly accurate for a couple of drafts, especially the first dozen picks or so. While not the be all and end all, Draft Express (DX) is the most respected, accurate, and generally best regarded draft prediction on the internet. A compilation of several draft sites may have created a somewhat different result, but I felt that Draft Express's mock draft would accurately convey how people felt regarding draft day "fallers". Finally, I didn't go below the projected 40^th pick: below that I think severe drops are much more common and less surprising.

The drafts I used were from 2008-2013. Pre-2008 is getting into somewhat older times before "fan scouting" and draft analysis was common, and I felt mock drafts might not be as accurate. Teams/GMs probably had significantly different methods and tools back then as well, so any analysis would have been less relevant to today . 2013 was the other cut-off because third year players are already established, and while some of the players who are considered "good picks" now might not be such in a few years (and the reverse), I think the bulk of the basic analysis will hold. The following results are based primarily on consensus view of the player, as well as a general understanding of how much value the average pick at that draft position provides to a team.

The first observation that must be made is that the number of players significantly dropping has declined over time: 10 in 2008, then 9, 6, 7, 4, and 5 in subsequent years. There are several possibilities for this. The first is that DX got better at predicting picks. This doesn't ring true to me, because they were already nailing most of the early picks in the late 2000's. The second is that general managers started scouting more, updated their draft models, and generally became better pickers, and that corresponded better with the analytics heavy DX. This would make sense if mock drafts were of what teams should do, but instead they are a prediction of what they will do, and that means drafting "better" doesn't really effect the mock draft forecasts. The third possibility is that general managers now draft more safely, and don't want to let a supposed "steal" fall through their grasp even if it's not someone they even scouted or considered (Bulls in 2012 with Marquis Teague, I believe).  It could be any or all of these factors, or none, but it definitely appears that the number of players slipping considerably in the draft has decreased over time.

When examining the players themselves, one thing quickly becomes clear: a relatively high number of them provided good value at their pick position. There are 40 players on this list, and 19 of them were good picks at their respective draft slots. These range all the way from All-NBA stars (DeAndre Jordan and Draymond Green) to solid role players such as Nick Calathes and Kyle Singler. Moreover, the bulk of these players were drafted in the bottom of the 1^st round or in the 2^nd round—and a success rate of nearly 50% is incredibly high for picking that late. So in other words, the next time a player drops 9+ spots to your team, chances are good that he will outperform his pick selection.

There are some qualifiers, however. Of players originally projected to be in the 2^nd round—who then just fell much farther in it— only 4/13 were good picks. This is still better than the average success rate on 2^nd round draft picks, but much lower than that of the whole sample. On one hand, the simple fact is that since they were already projected lower, there was even less reason to expect them to be good, so the drop really isn't very meaningful. Of course 2nd round picks are less likely to stick in the NBA than 1st round players. On the other hand, the adjustment for pick number means that expecations are lower; these players didn't live up to even more moderate hopes.

The most interesting nugget I found, however, had to do with the size of the drops in question. I was curious if bigger drops might mean an even higher success rate: falling to a much lower position should have made it easier for these players to outperform their draft position as well as given them an even larger chip on their shoulder. I found the opposite. Of players who fell more than 15 picks, only 3 panned out. Three!! What does this suggest? Well, teams really do know what they are doing sometimes. Whether because of attitude issues found in interviews, or medical red flags in examinations, or getting severely outplayed in workouts, these players were passed up by team after team. And for the most part, they didn't "punish" the teams who rejected them, suggesting they really did have some kind of issue preventing them from success. Flipping that number reveals that 16/26 players who dropped in projections by 9-14 spots were good picks. That comes out to 61.5%, which is a simply ridiculous rate of success at any part of the draft outside the top five overall picks.

On the whole, there is no question in my mind that players who fall in the draft (by more than a few spots anyway) generally provide good value. Not every player who tumbles down the board will become the "steal of the draft", but there is a reasonable chance that they will outperform expectations for their actual draft position. And in the NBA draft, that's about as much as one can ask for.