My Perfect Prospect

A summary of NileHoops prospect evaluation processes.

My Perfect Prospect

My general basketball theory can be indirectly extracted from a combination of my writing portfolio, Twitter profile, YouTube videos on my 2025 draft board, and podcast appearances with my dear friend Finn.

Once a level of prestige is procured in the draft space, there is an expectation for one to communicate their evaluation beliefs in a manifesto. Ever the contrarian, I’ve found myself less than enthused for my turn. My approach diverges from the consensus at both amateur and likely professional levels, and communicating my methodology could lead to a minor shift in discourse and practice. A pillar of my process is to avoid pontificating about the concept of analysis; instead, analyzing as many players as possible, at extraordinary depths, is my method. Turning the process of player evaluation into a theoretical evaluation strays too far for my liking, especially when the task at hand can be undertaken on a case-by-case basis, where each individual and team unit stands in isolation under the umbrella of organized basketball.

For example, in my Portland State follow-up piece, I sat with my adopted franchise signing a player one could deem the worst statistical D1 transfer prospect in the nation: Mozae Downing-Rivers from Missouri State. As kind as their staff had been to me, their neglecting his statistical profile and rostering him made me feel like the hope and strategy I had mustered for the Vikings was for naught. In a significant shift of perspective, pondering Downing-Rivers’ strengths instead of dreading his flaws, I ultimately made a breakthrough in my approach, as seen below, regarding perimeter defenders' utility of fouls and the predictive value of a physical, bullyball style of defensive playmaking compared to a defender who is more foul-averse.

Being able to respect any evaluation assignment as a sandbox and not an examination with designated, correct answers allows for growth.

I refuse to undertake a philosophical drain akin to those before me. My following has come mostly from prospect evaluation, and in carefully defining the values that would lead to my designating a prospect as ‘perfect’, I will have completed my obligatory philosophical placation.

NCAA is King

My ideal prospect would play in the NCAA, almost without exception. It is by far the most frequent middle point from teenage amateur play to professional basketball, and thus allows for the most standardization in next-level projection. My ideal prospect in the German BBL, Chinese CBA, or even the G-League would come with a substantially lower confidence interval. I do not care in the slightest about prospects “playing against grown men”, and I have never used this as a reason to be higher on an international prospect than a stateside alternative. Outside of inconsistent talent distributions, many of which are weaker than a competitive NCAA schedule, league-wide stylistic distinctions between American and international strains of hoops are a reason for uncertainty.

An instructional continuation to prioritizing NCAA prospects is knowing which NCAA prospects are to be analyzed. Narrowing the field of prospects to only those with the most NBA viability is a mandatory aspect of evaluation. Based on work from some of the most gifted, thoughtful analysts in the field, I have created an intuitive three-step vetting process that removes something like 98 percent of the field from draft contention.

As stated, the best way to master the practice is to explore as many in-depth prospect and conceptual evaluations as possible. This is a time-restricted practice, and only seriously analyzing the most relevant fraction of college basketball players should reduce waste. I don’t believe it can be fast-tracked by draft models, specifically because I have yet to come across a model that accounts for as many factors and relationships as are necessary for accurate projection. An in-depth model, especially on the public front, will not outperform a thoughtful and skilled analyst for this reason. Mastering modeling versus assembling a comprehensive research-driven process is also a misallocation of skill by CompSci scouts; this intensive utility for summarizing basketball into all-in-one metrics is not necessary, replicable by combinations of handily available metrics, and surpassed by sharp analysis using them in tandem.

Athletic Indicators

Simply requesting my ideal prospect to be 7-foot-5, 300 pounds with a +10 WS would be on-brand for the artist who's given the world hits like Zach Edey and Khaman Maluach as #1 overall prospects in their respective classes, while draining the rigor from the exhibition. As we further explore, it will become clear that size raises the ceiling for a player, as opposed to creating the floor.

My expectations of a player’s athletic indicators materializing on-court are more concerned with meeting functional thresholds than identifying the greatest athlete ever. A standard of sufficient dunk totals (with dunk rate and dunks per rim attempt as underlying continuations), stock rates, and rebound accumulation being met is more important than any player being 100th percentile in a single category. This combination of vertical and lateral ability is the difference between otherwise similar players’ success in the NBA.

OFFENSIVE PROCESS

Shooting (or lack thereof)

I am a proponent of ‘Moreyball frequency’ in prospect scoring. A 100th percentile frequency is extreme, and points to a lack of counters at a degree hazardous to functionality. At the same time, a player whose shot attempts come exclusively at the rim and from behind the three-point line is more likely to be maximizing effective scoring attempts, regardless of team environment. My perfect prospect does not concern themself with midrange shots. In fact, not only does my dream prospect not concern themselves with attempting middies (note that touch-based counters like floaters and hooks are of some importance, as possessions die without them), they mustn’t be concerned with making ANY shots from the field! Outside of shooting percentages reminiscent of playing with a blindfold on, 3PT and non-rim 2 efficiency are not mandatory.

What is mandatory is free-throw shooting.

Making free throws is the ultimate separator between otherwise equal players, rivaled only by the aforementioned athletic indicators. In both cases, the additional ancillary trait has an immediate impact and is indicative of further upside.

Possession Economy

I’ve made it a consistent trait of my process to drill into low-minute players at every level. It’s the most ambitious way to find under-the-radar talent and reap the buy-early benefits. With this, it’s mandatory to identify the most concrete, low-variance ways players affect the game on a possession basis. Shooting and making a single three-pointer naturally comes with a 100 FG%/150 TS% and could be the difference in a team winning and losing a single game, but it does not speak to any sustainability across a career, season, game, or even the next shot attempt.

In contrast, generating stocks, maintaining security of the ball relative to playmaking obligation (AST:TO), and gaining even more possessions with offensive rebounds is the skill intersection that a player can show off in low minutes, and translation can be expected.

I intend to depict a prospect that would ensure winning on a per-minute basis; these are the pillars. Along with an above-average sample in free-throw efficiency, my perfect prospect is beginning to take shape.

Stress Testing

As comfortable as I am using smaller samples in analysis when necessary, I also heavily subscribe to the importance of a player’s impact being stress tested versus starters and over extended periods. As I cannot force a coaching staff to play rotations optimized for my analytical pleasures, this is a dream.

I am less age-pilled than the draft analyst field; a great player being younger is influential, a young player being weaker is much less of an excuse for their shortcomings, and a great player being older is not a deathblow to their outlook. If it takes a multi-year sample to gauge how effective a prospect will be, then I would implore them to go back to school, not only for my confidence intervals in their projection to become more accurate, but also to avoid the austere hammer of the NBA Clock. Once a player signs an NBA deal, there is no way to reverse the poor play that comes with developing on the job. On Finn’s 2017 Draft podcast, I note that De’Aaron Fox was one of the best BartTorvik-era prospects who 100% should have gone back to school for his sophomore year. Joining the NBA as a 6'3, 170-pound guard with no semblance of a jumpshot and only mediocre length (6'7 WS) was bound to struggle from day one, and he did just that.

Outside of running hot from three during the 2019 and 2024 seasons, Fox has been statistically disapointing; a stelath case for Worst Defender in the League in a few seasons across his career, and offensive performance not substantial enough to counteract, outside of Mike Brown saving his life for three seasons (where his 3-year ORAPM still capped outside of t30, with a total RAPM just inside the league’s top 100). I believe the same fate looms for Ace Bailey for not returning to the NCAA level, a prospect whose few positive skills will easily be countered by his underdeveloped NBA frame and flaws that will crush his impact into ashes.

I say this to emphasize that defeating the NCAA before advancing to the league is a trait that I admire and prefer, no matter what age, though the younger the better. The varying degree of talent across college conferences adds a smaller layer of uncertainty to the process, but a thoughtful understanding of league talent and a filtering of games versus talent irrelevant to pro translation helps. I filter for games versus the top 220 teams in the nation, with logic being as simple as removing the bottom third of competition. If one day, it is presented that there is a better cutoff point for NBA predictive ability while keeping a solid sample for every relevant player, I will transfer my efforts abruptly.

‘Flashes’ (Star Run)

While sustained, replicable success throughout a college season would be the ideal scenario for projecting a prospect, this rarely happens. 40 games of invariable performance from a person with a still-developing prefrontal cortex is impossible to expect, more so at a high level of play.

Furthermore, projecting a teenager’s peak performance based on the totality of one or two NCAA seasons and unstandardized HS/AAU metrics is a risky endeavor, as lulls in performance or uninspiring team context mask outlier traits only sometimes apparent. Insert the Star Run.

I’ll do a live walkthrough of the practice with Jarrett Allen’s profile.

A freshly-turned 19-year-old whose best basketball looks like this without abusing lower talent levels or having elite talent raising his team’s floor seems indicative of a higher outcome than his raw stats or unremarkable advanced stats may indicate. Allen would have only further validated this process with alien-like measurements/testing at the Combine and ascending into the status of one of the most impactful bigs in the world at 27 years old.

False positives can materialize when shrinking an already small college sample; most often, from a player going on a three-point heater for a few weeks, a Box Plus-Minus boon. My favorite example was Kentucky’s Justin Edwards being just as good a shooter as Reed Sheppard for ten games, with little change in the rest of his profile.

Even accounting for his RSCI rank, not a single NBA program was duped, and he went undrafted. He has a true case as the worst player in the NBA with the 76ers today.

Regarding my perfect prospect, it would take a full season to christen them in this way, but an intense sample of games versus meaningful competition where all of my preferred traits are amplified would lead me closer to that designation.

Does my Perfect Prospect already exist?

With my thesis settled on what my perfect prospect would look like, the most logical choice was to query for an approximation. In order of importance, these are the traits we are looking for:

1. Forcing turnovers
1. An optimized shot diet
1. Free-throw excellence
1. Above-average offensive rebounding
1. Dunks and dunk-adjacent athletic performance
1. Offensive turnover aversion
1. Height/Weight/Age/Wingspan/Measurements (in that order)
1. A Star Run
1. Strength of schedule and minutes played
1. Jumpshots going in

The initial six factors are the easiest to query in BartTorvik, and the final four would be best considered with a more subjective view once prospects are identified.

The most intense isolation of my most principal prospect traits leaves two above thousands before and since. Paul George’s sophomore campaign and Keon Ellis’ senior siege stand alone as the nearest true approximations to My Perfect Prospects.

Aside from being realized on-court likenesses of my theory, Ellis and George stand for a handful of other conceptual beliefs I’ve promoted in this writing. Neither were NBA prospects upon their arrival to college basketball or after their freshman seasons, playing multiple seasons before eventually dominating.

A 22-year-old qualifying for this lofty query is more impressive than an 18-year-old only able to complete one segment. Ellis’ pro outlook would not have been taxed much by his age; rather, his first-percentile 167-pound frame and unspectacular combine testing would have deflated his case. George’s opting out of the combine leaves only his impressive height, weight, and wingspan as anthro highlights.

Distinctions in their Perfect Prospect seasons do not end there, naturally. Their star runs provide the most intense imprints of their profiles.

George generally clears Ellis, though the latter’s ridiculous three-point volume and turnover aversion point to the more intense possession-maxxer. One was completed against the 2010 Western Athletic Conference, and the other against the 2022 SEC. See how I could abbreviate one of those, and had to spell out the other? Box Plus-Minus accounts for opponent strength to an extent, but this is a drastic difference in rigor.

As an anti-model proponent, I questioned the ethics of devising a relatively basic spreadsheet rendition of My Perfect Prospect. Completing this task, limiting the field to players with at least twenty free throws made and six dunks as seen here, was less reflective than expected.

Perhaps, adding every required datapoint would have stimulated my efforts more, but the intensity of identifying weights, wingspans, and ages for three thousand seasons of college basketball prospects would have engulfed the objective. The chase for ‘perfection’ in prospects became futile by the end, as I observed just how much I loved analyzing prospect environments without a fixed goal. I still hope that my attempt at a simple weighted metric to identify players I would especially prefer helps one understand my process.

My perfect prospect is not inherently a high-volume, load-bearing offensive engine. Procuring basketball’s most preferred on-ball agents is such a fulfilling endeavor, as the goal becomes crafting the ideal scenario for their weaknesses to be nullified and their strengths amplified. Whatever skill/anthro intersections lead to a player being the lead chef of an elite unit are fluid and should be critically analyzed before being identified. Identifying my perfect initiator requires a mental limitation of athletic alchemy; until players like James Harden, Stephen Curry, or Shai Gilgeous-Alexander were actualized, there was simply no way to conjure their abilities. I believe I’ve shown here that one could manifest the two-way Moreyball possessionmaxer who leaves no points at the free-throw line.

Nile!