Lessons from the 2025 NBA Draft Cycle
When you evaluate a basketball player, what do you see? Do you take in the highlights, note the schematic or technical execution or simply look for the skillsets you value? There are infinite ways to watch and evaluate, something I believe is underappreciated in the draft space.
That’s what this annual column is for (see 2022, 2023 and 2024 versions): How did I evaluate this NBA draft cycle, how does it compare to years’ past and how will I adjust for 2025-26? Where does my process fit into the greater universe of basketball understanding, and how can it get better? These are questions I believe more should ask.
I published the below graphic at the end of my last piece with The Stepien before it shut down. It is easy to get bogged down with the evaluatory framework I outline on the far right. You read consensus views from major outlets, then either take them at face value or come up with a shortcut reason why this is not the case. Many use the middle approach, relative scouting, ordering by category / archetype (i.e. primaries first) and then comparing within that group. I have aimed for something I like to call absolute scouting, that is, looking at a player’s current ability while assessing with an eye to what they may become. This process can be amplified by the other two, but is ultimately the “truest” evaluatory framework, giving the player himself center stage.
But even within absolute scouting there are endless approaches. This column tracks the evolution of my evaluatory framework to better project NBA careers at the time of the draft.
Forbidden Knowledge
My thesis last year went something like this: if a player is productive on the court, making things happen almost by accident, exhibits high feel, and also demonstrates a high level of athleticism, that is the type of prospect I would want to invest in. The draft is about chasing outliers, and outliers tend to show themselves in those three arenas.
This strategy led to some major out-of-consensus calls. The highest profile call was placing Zach Edey #1 atop my 2025 NBA draft board. Edey was productive in a way I expected to translate at the next level, particularly his rim volume, offensive rebounding and screen-setting. He has made significant improvements to both his feel and athleticism over his college career, items you could notice even evolving over the course of his rookie season, and despite a nagging foot injury. Finding a way to be useful for an above-average Grizzlies team while drastically shifting from his college role, I still have high hopes for the big.
I also had Jonathan Mogbo as a clear-cut lottery talent, finishing as my #5 prospect, then drafted by the Toronto Raptors to kick off the second round. Mogbo finished #18 in the class in minutes played, able to get rotation and occasional starter playing time. While only 22nd in the class in points scored, Mogbo is #8 in the class in rebounds, #7 in assists, #5 in steals and #12 in blocks after his rookie season. While still a bad player overall – he was one of the worst finishers in the league, exchanging his lob finishes in college for off-the-dribble lays too far from the hoop – Mogbo has quickly proven he can do as many non-scoring things on the court as anyone in the class. In some ways, he’s adapted from mid major to NBA competition better than I expected. I’d still bet on him, particularly given his immediately above-average defense and the weakness of the 2025 class otherwise. Should the passing continue to click and his teammates become more comfortable finding him on lobs, the path to offensive value is there. He came out of the gate as one of the NBA’s most bothersome defenders.
My other two big swings near the top were Oso Ighodaro and Terrence Shannon Jr. as late lottery bets, consistent rotation players. While minutes for both were up and down, both showed enough for me to remain encouraged.
The Ten Dimensions
This year, however, I wanted to become more literal in assessing player value. This led me to inspecting the game by “dimensions of impact,” where I categorize each type of contribution into ten groupings. I based these on how one interacts with the ball and court in literal manners, inspecting each realm in close detail, creating clear as possible delineations among categories.
The ten categories within three skill groupings:
- On-Ball Interactions: Three point shooting, midrange shooting, rim finishing, handle, passing
- Off-Ball Interactions: Grabbing/deflecting the ball, ground coverage, positioning
- Physical Force: Pace-force, strength-force
However, no two traits have the same impact of the game; I would have to weight each area of impact. I constructed these weights and inputted values for each player on my board with endless tinkering, informed by tape watch, statistical assessment, philosophical inquiries into how the game is won. The heaviest weights went to three point shooting and positioning, two areas of impact I only added more and more weight to as I back-tested to current and historical NBA players.
But this has limits, too. Namely, skills on the court interact with each other in varied and unusual ways. Even though both combinations provide additional value, a player’s ability to finish at the rim is more advantageous when mixed with a strong handle than if it were mixed with great rebounding. While rim touch + rebounding equals putback potential, handle + rim touch means an extremely deadly drive threat. A team is able to scale that up and gain secondary benefits off of that more than the other combo. There are synergies and frictions across skillsets that make performance better or worse. You can’t just add up skills.
How Good Are You?
My solution? To throw away the ladder, yet again, to construct a new one. Having advanced my ability to inspect skillsets by interaction type, how each player impacts the game became much clearer. But basketball evaluation is even more mystical than that, especially in the absence of a well-constructed statistical model.
So, my answer was simple: ask myself, “how good is this player, on offense and defense?”
I decided to use Estimated Plus-Minus projection as my peg, not taken too literally but a useful impact curve where one can ascertain, as long as with context on role, a rough approximation of how good a player is. Not perfect, but more dynamic than skillset grading.
I’ve said it before, but draft projection is primarily an exercise in imagination. Even if I graded each player’s current skillset perfectly, there are more complex interactions between qualities shown today as it relates to future skill development. For example, a high feel, coordinated player may be more likely to develop a shot than one who doesn’t have those underlying foundations.
It is also extremely difficult to anticipate where development may arise, to the point of it being easier to rather say, here are the way this player might improve, and here are the odds of each happening. I back into this assessment by projecting into multiple scenarios: the future has not been set.
A New Dynamic
My philosophy has generally been extremely pro-risk, for two simple salary arbitrages, in addition to the fact that I can’t get fired: 1.) a team gets its most value out of paying only a max contract amount to a player worth far more, and 2.) if a player doesn’t work out, his minutes go to zero, limiting the downside impact. These are two HUGE incentives, as it is very difficult to compete if you aren’t getting plus-max value out of one player, making even multiple busts less damaging.
However, my strategy did evolve somewhat towards the end of this cycle. I changed my board to become dynamic, first ranking the top of my board by 80th percentile outcomes only – still keeping it high risk for the players whose talent makes it worthwhile. Then, towards the end of the lotto, the assessment becomes 50/50 between a player’s 80th percentile outcome and 50th percentile outcome. By pick 30, my calculation will only be considering median outcome, omitting the ceiling factor at all.
I made this change for an obvious reason I had been ignoring: it is simply impossible to develop an entire roster of projects at the same time. If a player isn’t deemed as high ceiling, he simply will not get the developmental reps to push through to those higher percentile outcomes. In this way, it is more worthwhile to take the bird-in-hand once you get past the obvious star bets. My changed formula accounts for this.
In addition, I should not be ranking my board, as an outsider not working for a team, based on salary arbitrage opportunities, rather than by how I expect the long-term results to shake out.
The Winners and Losers
My outlier calls this year included four bigs or big wings with shooting questions: South Carolina’s Collin Murray-Boyles, Georgetown’s Thomas Sorber, Creighton’s Ryan Kalkbrenner and Arkansas’ Adou Thiero. They went #9 to Toronto, #15 to Oklahoma City, #34 to Charlotte and #36 to Los Angeles Lakers, respectively, but I would have taken all four much higher.
Collin Murray-Boyles is perhaps my boldest take, finishing #2 on my Big Board. “CMB” is a tank at Draymond Green dimensions, and has shown a non-shooting skillset, defensive acumen and physicality that indeed do remind one of the Hall of Famer. Draymond is one-of-one as a processor, but Murray-Boyles has lightning quick reaction time and excellent understanding of the floor, too. He does not have the vertical pop of Charles Barkley, but CMB does mimic him in carving out space around the basket, constantly. The most important commonality is the physicality and processing speed, and CMB is far ahead of his age for both.
For a glimpse into the degree of impact CMB had on South Carolina’s woeful squad, I calculated the number of points at the rim SC would score or allow when Murray-Boyles was on or off. South Carolina scored THIRTEEN more points at the rim when he was on than off, and allowed SEVEN fewer points at the rim in the same scenarios. That offensive figure is more than double the second most among his 2025 comps, and defensive figure third to stalwarts Thomas Sorber and Amari Williams.
CMB has perhaps the best hands in the class, and they synergize nicely with not just his defensive but also his offensive game. Murray-Boyles learned how to better manipulate the ball when driving to the basket over the season, using his intelligence for when to attack to find seams just large enough to let his stellar touch take over. CMB was in the 85th percentile for layup efficiency, and top ten in the country in rim finishing among anyone with 150+ makes. Only Derik Queen was close among underclassmen, and CMB is six months younger despite being the higher grade.
Murray-Boyles is able to conduct traffic, palming the ball in the high post, one spin away from the hoop. He will operate more out of the short roll in the NBA, and thankfully with better shooters (even with the Raptors’ subpar personnel, they exceed his 31.6% three-point shooting college team). He is better than a connective passer, able to hit small windows and create advantages with his sense of timing, leading his teammate into space.
CMB provides rim protection, elite rebounding, on-ball disruption (he is particularly strong blitzing and recovering) and leads the defense when guarding away from the ball. He is the best defensive prospect in a class full of very strong bets in Cooper Flagg, Thomas Sorber, Noah Penda, or perhaps second to Joan Beringer. He does that while being one of the best driving big men in the country, putting up a very strong 0.92ppp on over 100 drives. He thrived out of isos as the season went on, scoring nearly five points per game out of the play type over South Carolina’s final six games. Check out the versatility in the clips below.
It is rare to find obvious defensive disruptors of this level who also have this kind of offensive potential. He almost certainly won’t be a very useful three point shooter, but he has nearly everything else (I’m even hopeful about the midrange).
On the downside, I thought players like Ace Bailey, Egor Demin, Nolan Traore, Hansen Yang and Will Riley went over-drafted. A common theme for these players is being young and high-risk while needing a good amount of touches to approach their ceiling. With my new system, their upside outcomes do not quite drag them up the list high enough to use the 80th percentile calculation, rather being graded on their less thrilling median outcomes.
Four of the five are skinny for NBA players at their heights, with the exception of the slow-footed Hansen Yang. Returning again to our synergies, a weaker frame mixes extremely poorly with on-ball potential, unless you’re a Haliburton-esque conductor, or Shai-esque scorer, both nearly perfect at capitalizing on space creation specificity. This does not mean the path is closed – I’m especially still high on Ace Bailey as a late lottery option – but not the bets I would make with the group compared to where they were drafted unless you can spot the magic. All five have magical moments, no doubt, but lack consistency and are likely to face struggles as they adjust to NBA physicality.
CMB is, on the other hand, extremely difficult to tilt off his spot, making those on-ball reps more consistent and allowing him to explore the studio space in a safer manner. I was also high on Javon Small, Max Shulga and Joan Beringer, all with BMIs higher than all but Yang from the group of players I was lower on. Small’s physicality allows him to drive and dunk through traffic, set up offense without being knocked off his spot. Shulga is broad-shouldered, allowing him to wall off drives and switch up. Joan Beringer, despite being one of the youngest in the class, has been able to bulk up some, on his way to becoming one of the NBA’s best rim protectors. If I have one regret so early, it would be not ranking Beringer in my top ten. The defensive instincts and physical tools give him an extremely lofty ceiling, even with mediocre offense, and he already seems good enough to say his defensive floor is safe, too.
An Eye to 2026
2025 was a fantastic class to evaluate, extremely deep in starter bets. I ended up ranking Jase Richardson around 20, and even so would not be surprised at all if he carved out a starter spot, overcoming his 178-pound frame by being so effective and technical playing off the ball. He was a painful player to rank even that low, given how high his feel for the game is and proven technique, though I remain concerned about his lack of a right and limited defensive ceiling.
2026 promises to be thrilling at the top, as Cameron Boozer, Darryn Peterson and AJ Dybantsa all vie for the top spot. My early leanings rank them in that order, with Boozer vs. Darryn vs. Flagg being very tough to discern.
My biggest adjustment will be getting more accustomed to projecting peak impact, but I want to tweak my ratings system to become even more risk-averse as you go down. I will do so by implementing a 20th percentile outcome which becomes the ranking priority starting at pick 30. With each round of new tape or statistical analysis, making those projections gets a little easier.
As usual, I expect that evaluation criteria to evolve over time.
Tags:
Related Articles
