Availability Is the Real Variable: Re-Reading the NBA Pricing Board Through ADP Data

96 games across three seasons. Joel Embiid, one of the highest-paid centers in NBA history, has taken the floor for exactly 39% of possible games. That number does not live in the points column, the shooting-percentage column, or any highlight reel. It lives in a blank line on the schedule, and where the market reads that blank line as injury, I read it as price. Every number I touch carries a scar. This time the scar is not on Embiid's knee. It is in the way we price him. Before watching the game, watch how the data breathes. Embiid's data is breathing very slowly. Each preseason, millions of fantasy players are forced to do what NBA general managers do year-round: put a price on a player. Their tool is ADP, Average Draft Position, the average spot at which a player is selected across thousands of mock drafts. ADP is the crowd's price, updated daily, and in a Yahoo league that requires two starting center slots, it creates a market with its own rulebook. That rulebook is specific. The starting lineup requires PG, SG, G, SF, PF, F, C, C and two UTIL slots. Two center slots, only one flex guard slot and one flex forward slot. That structure pushes big-man value above ADP, because center supply is scarce at twice the demand. A manager who understands the rules drafts bigs early. But when I look at this season's ADP board, I see something more notable than roster strategy: a crisis in the pricing of availability. That summer was empty, but the data never rests. In 2026, when stadiums closed, I tracked the Bundesliga restart and learned one thing: every variable outside the pitch leaves a trace in the data. The crowd vanished, home-win rate fell from 46% to 32%, average goals dropped from 3.1 to 2.4. Now the missing variable is no longer the crowd. It is availability. Here are the three most important data columns in the entire pricing cycle. Lauri Markkanen. The last three seasons: 55, 47, and 42 games. That totals 144 of 246 possible, or 58.5% availability. The more frightening number lies deeper in the history: across nine seasons, Markkanen has cleared 60 games only three times. His career high in games played remains 68, set in his rookie season. Statistical noise does not produce a nine-season pattern. This is a structural signal. Markkanen's ADP is 39.2. In the mock draft I tracked, one manager selected him at pick 27, 31% above ADP and the largest rise on the board. It is the signature of a market in transition: the crowd has cast him aside after years of disappointment, while a smaller group is buying back in. I do not object to the buy-back. I object to the price. At this percentile, you are paying the price of a 75%-available center for a 58%-available one. Joel Embiid. 96 games over three seasons, 39% availability, roughly 32 games per season. For a max-contract player, this is the most extreme availability deficit among stars. His ADP is 51, but in the mock draft he fell to pick 65, a 27.5% slide and the largest on the board. The market is pricing in rest and decline. In my view, it still has not priced in enough. The logic is clear when you look at roster structure. If Embiid genuinely shifts into a mode of managed rest, fewer minutes, and lower usage, that is three compounding negative adjustments, not one. And three compounding negatives are rarely fully reflected in a 14-pick fall. The question is not whether Embiid is talented. The question is how much of the payroll is paying for replacement-level production. In the real salary world, a star at 39% availability means nearly three-fifths of the schedule is filled by mid-tier production while the invoice stays at star pricing. Stephon Castle. The opposite case, and the most underrated bright spot in the whole picture. 18 double-doubles and 5 triple-doubles in one season. Five triple-doubles from a guard is a strong leading indicator of durable multi-category volume, and that is exactly what volume-based scoring formats pay for. Castle was taken at pick 39, while his ADP was 55.3 and his fantasy points per game ranked tied-51st. The market is paying for trajectory, not current production. That is a defensible price. But here is what most analysis skips: Castle does not carry health risk, he carries role risk. In San Antonio, a backcourt of Fox, Castle, Vassell, and Harper is competing for the ball and minutes. Castle's ceiling is capped by his own team. A four-man backcourt competition is a season-long structural question, and it will be resolved, either through minute allocation or through a trade. Place the three players in a single frame: Embiid at 39%, Markkanen at 58.5%, Castle carrying role risk. The shared axis is not talent. The shared axis is availability. And that is something per-game stats never say out loud. This is where I have to state plainly what the original analysis never said: ADP is not value. ADP is a contaminated price, and it is contaminated systematically. ADP is produced by blending different league formats, category and points leagues. In a standard Yahoo points league, missed shots typically carry no penalty, so efficiency matters less than volume. But that is an assumption, and the original analysis never disclosed its league's scoring settings. If turnovers are penalized, as many points leagues still do, the entire efficiency-does-not-matter thesis partly collapses, and Castle's high-turnover games stop being a point in his favor. The correlation-versus-causation trap lives here. A player with high fantasy points is not necessarily a good player; a player with good efficiency is not necessarily a high-fantasy-points player. The same data, two entirely different risk instruments. In the fantasy world, a rested star is merely a draft-position bargain. In the real contract world, that same player is an uninsurable liability that cannot be traded at par. The same number, two readings. I found the Russian curse in the past, and it was only a calculation. This time the curse carries no country's name. It carries a time window's name: the final three to four weeks of the regular season. Chaos on the floor always has a hidden order, and the hidden order here is the schedule. When a contender rests its cornerstone and a lottery team shuts down its players, both are doubly penalized right in the decisive stretch. Embiid is minute-managed because his team no longer needs him to carry the regular season. Markkanen's usage is reduced because Utah is rebuilding and has an incentive to lose. The same mechanism, two forms, one consequence: a player loses value exactly when value is being tested. The signals I will track in the next cycle: Embiid's minutes per game over the first 20 games, and Markkanen's usage rate. If Embiid stays below 28 minutes, the rest thesis is confirmed by data rather than speculation. If Markkanen's usage falls while his efficiency rises, we are witnessing a scoring archetype shifting roles, good for real basketball, bad for anyone pricing him by volume. The game will answer. But the data will answer first.