The Analytics Revolution in Oklahoma City: How Data Drives Drafting Future Stars

The Oklahoma City Thunder have quietly assembled one of the most exciting young rosters in the NBA, built largely through a disciplined, forward-thinking drafting strategy. While the team’s stunning collection of future draft assets often grabs headlines, the engine behind their success is a deep commitment to player analytics. By marrying traditional scouting with advanced data, the Thunder consistently identify and develop players who outperform their draft position. This approach has transformed them from a small-market rebuild into a model of modern team building.

The Evolution of Basketball Analytics

Basketball analytics didn’t start with the Thunder. The modern era began in the early 2000s when teams like the Houston Rockets, under Daryl Morey, pioneered the use of advanced metrics to challenge conventional wisdom. The "Moreyball" philosophy emphasized three-pointers, layups, and efficiency—a disruption that forced the entire league to adapt. Over the past two decades, the data revolution has expanded far beyond basic scoring and rebounding. Today, teams utilize player-tracking cameras in every arena, capturing movement, spacing, and decision-making in real time. The sport has moved from simple box-score stats to complex models like Real Plus-Minus (RPM), Player Impact Plus-Minus (PIPM), and defensive rating adjusted for pace. For a deeper look at how analytics have reshaped the NBA, check out ESPN’s retrospective on the analytics decade.

The Oklahoma City Thunder were early adopters of this data-driven mindset. General Manager Sam Presti, who came from the San Antonio Spurs organization—a franchise known for its analytical rigor—understood that small-market teams cannot afford high lottery misses. To compete with deep-pocketed markets like Los Angeles and New York, Presti had to gain an edge through superior information and evaluation. That meant building a front office capable of turning raw data into actionable player projections.

Oklahoma City's Data-Driven Draft Philosophy

The Thunder’s draft strategy is not a single algorithm but a layered system. They start with a comprehensive database of college and international player statistics, then overlay scouting reports, medical histories, and psychological evaluations. The analytics team, led by a dedicated director of basketball analytics, works alongside scouts and coaches to build predictive models for player development. The goal is not simply to find the most talented player available but to identify players whose skills translate to winning NBA basketball and whose trajectories can be accelerated by the Thunder’s developmental infrastructure.

Key Advanced Metrics the Thunder Prioritize

While the Thunder’s proprietary models remain confidential, publicly known metrics that correlate strongly with NBA success include:

  • Player Efficiency Rating (PER): A per-minute measure of overall statistical production, adjusted for pace. While flawed for defensive evaluation, a high PER in college often translates to NBA impact.
  • True Shooting Percentage (TS%): A more accurate measure of scoring efficiency than field-goal percentage, as it accounts for three-pointers and free throws. The Thunder place a premium on players who score efficiently, even from non-traditional positions.
  • Defensive Win Shares (DWS) and Defensive Box Plus/Minus (DBPM): Metrics that estimate a player’s defensive contribution. The Thunder have consistently drafted players with strong defensive college profiles, such as Luguentz Dort (undrafted) and Cason Wallace (10th pick in 2023).
  • Assist-to-Turnover Ratio (AST/TO): A proxy for decision-making and ball security. The Thunder value playmakers who take care of the ball, a trait they saw in Josh Giddey and Shai Gilgeous-Alexander (acquired via trade but scouted using similar models).
  • Usage Rate and Rebounding Rate: These help contextualize a player’s production within their team’s system. The Thunder look for players who produce at high usage without sacrificing efficiency, as well as rebounders who dominate their position—key for modern switchable lineups.

These metrics are not used in isolation. Instead, the Thunder’s analytics team runs regression models to see which college stats best predict NBA success. For example, they have found that a high steal rate in college strongly correlates with NBA defensive value, which is why they often target undersized guards with active hands. They also weight performance against strong competition more heavily than stats amassed against weak schedules.

International Scouting and Analytics

The Thunder have also applied analytics to the international draft market, where scouting is sparser and data can be noisy. Josh Giddey, selected 6th overall in 2021 from Australia’s NBL, is a prime example. His raw numbers—10.9 points, 7.4 rebounds, 7.5 assists per game—were unusual for a 6'8" guard. But the underlying analytics showed elite passing vision, high rebounding rates for a guard, and a solid AST/TO ratio. The Thunder’s models projected that his playmaking would translate, even if his shooting was a concern. They were willing to bet on the data. For more on how teams evaluate international prospects using analytics, see The Athletic’s deep dive on international scouting.

Case Studies of Thunder Draft Picks

Examining recent Thunder draft selections reveals how analytics have guided their decisions, often bucking conventional wisdom.

Josh Giddey (2021, 6th overall)

Coming into the 2021 draft, Giddey was viewed as a high-risk, high-reward prospect. He had a low usage rate in the NBL but an astronomical assist rate. His TS% was below average due to poor three-point shooting, but the Thunder saw his ability to draw defenders and kick out as a translatable skill. In his first two seasons, Giddey became a triple-double threat and a key initiator. His rookie efficiency stats (10.1 PER, 47.2 TS%) were poor, but his 6.4 assists per game placed him in the 94th percentile among rookies. The analytics predicted that as his shooting improved, his overall value would rise—and the Thunder invested heavily in his development.

Chet Holmgren (2022, 2nd overall)

Holmgren was the consensus top prospect in his class, but his slight frame (7'1", 195 lbs) raised durability concerns. The Thunder’s analytics team focused on his defensive metrics: in his lone season at Gonzaga, he posted a 4.5 block percentage and a defensive rating of 90.1. His DBPM was +7.2, one of the highest ever for a college big man. The data strongly suggested that his shot-blocking and floor-spacing would be elite. Even after missing his entire rookie season with a foot injury, the Thunder never wavered. In his debut season (2023-24), Holmgren averaged 2.3 blocks per game and shot 37% from three, validating the pre-draft analytical models.

Cason Wallace (2023, 10th overall)

Wallace was not a flashy prospect, but his defensive analytics were staggering. At Kentucky, he averaged 2.0 steals per game and had a 3.7% steal rate, ranking in the 97th percentile among guards. His DBPM was +3.9, and he demonstrated excellent off-ball navigation. The Thunder, already deep in guards, saw Wallace as a perfect fit for their switch-heavy defense. His offensive numbers were modest (11.7 ppg, 4.3 apg), but his AST/TO ratio (2.5) and TS% (57.8%) indicated efficiency. In his rookie season, Wallace became a rotation staple immediately, often guarding the opponent’s best perimeter player. This is a classic Thunder pick: a player whose defensive analytics were elite but whose traditional stats didn’t jump off the page.

Other Notable Picks

The Thunder’s success extends to second-round and undrafted players. Luguentz Dort (undrafted 2019) was originally a scouting find, but analytics quickly highlighted his defensive versatility—he posted a 2.0% steal rate and a 1.3% block rate for a guard in college. Those metrics indicated he could become a perimeter stopper. Similarly, Isaiah Joe was claimed off waivers in 2022; his college three-point percentage and catch-and-shoot efficiency (42.3% from three in college) made him a target for data-driven teams. Joe has since become one of the league’s best spot-up shooters.

The Role of Player Development and Analytics Integration

Drafting is only half the equation. The Thunder invest heavily in their G League affiliate, the Oklahoma City Blue, where they apply the same analytics to track player development. They use Synergy Sports data to break down each player’s shot selection, defensive matchups, and decision-making in game situations. This data is then used to create individualized training plans. For example, if a rookie guard struggles in pick-and-roll situations, the coaching staff can program specific drills and film sessions. The Thunder also utilize wearable technology and biometric data to monitor workload and injury risk—critical for a young core expected to play heavy minutes.

This development pipeline is a key reason why the Thunder can take chances on raw prospects. They know that if a player has a high floor based on analytics, the team’s infrastructure can lift his ceiling. Sam Presti has often spoken about the importance of "organizational alignment" between analytics, scouting, and coaching. The data doesn’t make decisions; it informs the decision-makers. For a detailed look at how the Thunder integrate analytics into their daily operations, read this official Thunder feature on their analytics department.

The Future: Machine Learning and Predictive Modeling

As data collection improves, the Thunder are positioned to exploit the next frontier: machine learning. Using historical data from thousands of players, teams can build models that predict career arcs. For example, a model might take a college player’s height, wingspan, shooting percentages, steal rates, and age to generate a probability of becoming an All-Star. The Thunder have already invested in quantitative analysts and data scientists who work alongside basketball personnel. They are likely using clustering algorithms to identify player archetypes and Bayesian statistics to handle uncertainty in small samples (e.g., a player who only played one season after a COVID-shortened year).

Another emerging area is the use of player-tracking data in the draft process. The NBA’s Second Spectrum cameras capture every movement, generating metrics such as speed, distance, and spatial awareness. For international players who don’t have such tracking data, teams may use video analytics to approximate these measurements. The Thunder have been at the forefront of adopting these tools. For insight into the role of AI in NBA scouting, see Sports Illustrated’s article on AI and the draft.

Challenges and Limitations of an Analytics-Driven Approach

While powerful, analytics are not infallible. One major challenge is sample size: college seasons are short, and international leagues vary wildly in competition level. A dominant big man at a small conference may struggle against NBA athletes. The Thunder’s models adjust for strength of schedule, but outliers still occur. Additionally, injuries and off-court factors are hard to predict with data. The Thunder have been burned on this front—for example, Aleksej Pokuševski, drafted 17th in 2020, had intriguing analytics (ball-handling at 7'0") but never developed due to a combination of injury and lack of physical maturity.

Another limitation is the human element. Chemistry, coachability, and mental toughness are difficult to quantify. The Thunder therefore still rely heavily on interviews, background checks, and psychological testing. They’ve been known to draft players with high "basketball IQ" scores from tests administered at the NBA Combine. The balance between data and intuition is delicate. The most successful analytics departments, like the Thunder’s, understand that numbers are a guide, not a gospel.

Conclusion: The Competitive Advantage of Small-Market Analytics

The Oklahoma City Thunder have turned analytics into a competitive weapon. In a league where big-market teams can lure free agents with money and glamour, small-market franchises must win through the draft. The Thunder’s data-driven approach has allowed them to consistently find value—whether it’s a top-five pick or a second-round sleeper. Their model isn’t just about accumulating picks; it’s about making each pick count by understanding the underlying probabilities. As the NBA becomes more analytically saturated, the Thunder’s edge may narrow, but their commitment to constant innovation suggests they will remain ahead of the curve. For fans and analysts alike, watching this team’s rise is a case study in how data can build a dynasty from the ground up.