NBA draft stock started as a personal passion project. As NBA fans, we wanted to try to dig in to what makes successful NBA prospects. Historically, mock drafts have had a really hard time moving off of pre-season biased prospects. When a player is considered a top 5-10 selection, they have to do really, really egregious things to be cycled downward in a meaningful way. And if that player has interesting physical measurables? There’s virtually no chance. They also tend to have more subjective criteria for evaluating players.
So why does that matter? NBA teams tends to make the most mistakes at the middle and the end of the lottery and our inclination was that far too often, a team is unable to put that pre-season bias aside. It appears to be much tougher to make a draft selection based on more concrete things (like how prospects performed vs. their historical counterparts, on a possession basis), for players that specifically underperform their pre-season expectations. There are strong indicators that prospects just really cannot (realistically) improve from certain possession baselines and if they do, it has less to do with a team being right and more to do with downright luck (in our opinion). We try to explore what those possession baselines are, compare them to their position on a historical basis and make a determination for how likely that player is to succeed.
Our assessment considers things like age, height / length, athleticism, strength of schedule and then assess how those characteristics could help or hurt a prospect at the next level. I’m not going to get in to details (today) on exactly how we do that but a quick explainer is our baseline prospect is: average age on January 1st of their rookie year (21.00), average height / length (for position), average athleticism (for an NBA player), based on historical combine data. A players possession based production is then either increased (above those baselines) or decreased (below those baselines) depending on the severity in either direction. For example, a PG under 6’0” needs to (historically) do impactful things at an extraordinary high level to stick in the NBA. So from what we see, that player needs to really counteract their size disadvantage with skills that are not just good but great for their position. The shorter you are, the more and more production you need to counteract it. You can counteract height with athleticism / length and to a lesser degree (youth)…..so if a player is shorter (like a Donovan Mitchell) but possesses great length and athleticism, he functionally plays at much taller than his listed height, so that is considered. Strength of schedule also decreases a prospects statistical output..less for consensus lottery players like Damian Lillard but more and more as you hit 1st round, 2nd round, UDFA thresholds. Why? Similar to baselines on height, length, athleticism, a lesser prospect (on the draft scale), at a smaller school, needs to REALLY outperform larger school prospects projected to be picked in the same area and they need to do so with a statistical handcuff. The reason? They, historically speaking, need to counteract the degree to which they can accumulate their possession based stats, against lesser competition, in an easier manner…..to show they have actual NBA prospects as a player. If all of this doesn’t make sense, we’ll publish how we come up with these numbers some time in the future.
This stands for “Adjusted Game Score”. The calculation is based off of John Hollinger's game score methodology and considers (mainly) two things…..a players’ possession based contribution, less points scored and a rough approximation of their shooting potential….generally speaking, a weighted mix of 3PTA / FT% and age. This GS/ADJ stat can be either a positive or negative, depending on where a players falls in comparison to their GS/NET. For example:
If......a player has both a high GS/ADJ and a high GS/NET, they tend to be the best players from a given draft. Think Anthony Davis. Someone who can impact a player on both ends.
If......a player has a high GS/ADJ and a low GS/NET, they are a much larger candidate to wash out, because the typical profile of that player is one who contributes so little scoring value in college, that it is unrealistic to expect them to hang on in the modern NBA, as a really sub par offensive player. BUT THE ASSITS? Our historical data says NBA prospects need to score to a baseline level in college, regardless of their creation skills. Think Kendall Marshall. A distributor / defender type that just cannot score, at all. Defensive players are more apt to stick in this metric, as their specialized skill is harder to find.
If......a player has a lower GS/ADJ and a higher GS/NET, their offense needs to project to the NBA to stick because it intimates they are unlikely to impact on the defensive end. Think Kemba Walker. A highly explosive offensive player that can be more of a liability against upper echelon NBA offensive talent.
If......a player has a high GS/ADJ and an average GS/NET, they tend to profile as great role player candidates because they produce high impact stats (less scoring) but still historically score at a rate that makes them passable offensive players. Think Steven Adams. A quality starter that contributes across the statistical spectrum but can still chip in enough on offense to provide some defensive gravity.
If......a player has a low GS/ADJ and a low GS/NET, it is typically a death sentence for a prospect. Will a few of them survive? Sure. Will a GM keep his job if he continually drafts these types? Probably not.
This stands for “Net Game Score”. The calculation is based off of John Hollinger's game score methodology and is the sum of a players scoring AND non scoring contributions. Are you saying, doesn’t game score already do that? My answer is kind of…….at the college level, offense can really mask one dimensional players. This is specifically where this calculation tends to shine…..determining who should be taken, at what position, if your team has a positional need in the same tier. This calculation tends to almost always make the right selection when deciding between a closely grouped tier of players, at a given position. In tie situations, GS/ADJ breaks them, further strengthening the likelihood a team chooses the best player. We’ll eventually start posting historical comparisons for everyone to see. Like anything, it isn’t full proof but it does a pretty good job of better assessing what appear to be like players from a game score perspective…..and deciding which is a better player to bet on.
Relative to their NBA position, how has a prospect performed vs. their positional peers. 0.80 is an average prospect profile at a players’s given position. Above 0.80 is better than average…..the higher the number, the more likely we think that player is to turn in to an impact player at the next level. The lower the number, the longer the odds for that prospect to stay in the NBA. 83% of All-Star and top 10 VORP (relative to their draft year) players are average or better in POSS/DIFF.....so it proves as a good baseline when assessing a player. There are certainly instances where you bet on more raw players below this threshold but you should have compelling arguments to do so.