New York Liberty 2026: 14 Exploited Pick-and-Roll Possessions per Game and the Hole Inside the Drop Coverage Rule

Third quarter, 6:12 on the clock. Han Xu drops back behind the three-point line, keeping her distance from the ball handler instead of stepping up. The opposing guard takes a screen at mid-court, turns left, and has a full two metres of clear space to shoot. The ball goes in. The official box score records two points for the opponent, nothing more. No line records that Han Xu did exactly what her team's defensive rule asked of her, and that the rule itself created those two metres.

I rewound that clip four times in one evening. The first viewing to confirm the gap. The second to check whether Han Xu lost her balance. The third to see whether anyone rotated to help. The fourth to count how many times in the same quarter the Liberty repeated that exact structure. Four viewings, one conclusion: the player was not wrong, the system was. I once recounted a clip four times, and the error belonged to the source, not to me. This time the source error was not a rebound figure but an entire defensive rule packaged as "individual effort".

I learned to read the game that way in February 2026, as a freelance reporter on the NCAA beat. I mis-recorded Zion Williamson's rebound total in the Duke–Virginia Tech game, recounted the tape four times, and found the error sat with the host organisation's data feed. The correction went up on a personal blog with 240 readers, and it led to a statistical research assistant role the following season. Since then, every number I use has to clear two independent sources before it goes on air.

The New York Liberty entered the 2026 WNBA season with a roster built to win a title: Breanna Stewart on the wing, Jonquel Jones in the paint, Sabrina Ionescu and Courtney Vandersloot sharing ball-handling duties, Betnijah Laney taking the perimeter assignments. Sandy Brondello had two of the league's best rim protectors at centre. In theory, that is a roster that lets a coach impose an ambitious defensive system.

That roster was not cheap. Those four stars absorbed most of the salary cap, forcing the front office to fill the rest with minimum contracts and young players. When the money sits with four names, depth becomes the only elastic variable, and every error inside the defensive system gets amplified because there is no adequate backup plan. A roster this expensive does not lack effort; it lacked a correct rule for the personnel it actually had.

Then nine straight losses arrived. The staff talked about execution and effort. The media talked about roster chemistry and shots not falling. Both were true at the surface, and neither explained why the same roster leaked in the same pattern, repeatedly, against different opponents.

I rewatched all 40 regular-season games and cross-checked every pick-and-roll possession against Second Spectrum tracking data. The system logs the position of all ten players and the screener at every hundredth of a second. I used it to answer one narrow question: when the Liberty chose drop coverage, who was defending the second ball handler, and what happened next.

Drop coverage is a simple rule. The big retreats to protect the rim and does not step up to the level of the screen; the on-ball guard chases over the top to stay attached. In exchange, the defence concedes mid-range shots. That trade is sound when the dropping big is Jones or Stewart, because both are quick enough to hold the rim and still step out to contest.

With Han Xu, the trade flips. She stands 2.06 metres with long arms, but her reaction time when forced to change direction in open space is slower than the league's centre baseline. When she drops deep, the gap between her and the chasing guard becomes a corridor exactly the width of a high-quality mid-range shot.

The data gave me a far clearer picture than the coaching box described. Han Xu was the drop defender on 14 pick-and-roll possessions per game. Opponents scored 1.17 points per possession on those plays. The league average for the same situation is roughly 0.95. Multiply it out: those 14 possessions produced about 16.4 points per game from a single channel, in a WNBA game that lasts only 40 minutes.

The "low man" mechanism deserves detail, because it is what the Liberty left empty for most of the nine-game slide. When the ball goes to the roller in the middle of the floor, the weak-side defender must sink one step and sit between the roller and the rim, ready to cut off the pass. If that defender stays glued to the corner shooter, the roller has an open lane. In my tape review, in at least six of the nine losses, the Liberty's weak-side defender stood still. Not out of laziness. Because the team rule prioritised staying attached to the corner shooter over tagging the roller, and that rule was designed for a centre who could handle a one-on-two situation by herself.

Opponents did not need to hide anything. I counted three teams repeating the same action chain on more than 40 percent of their first-half pick-and-rolls: screen at mid-court, guard turning toward the strong hand, roller diving inside. When an opponent repeats a structure that often, it is not luck. They had watched the tape and knew exactly where nobody was covering.

The Liberty tried switching on those same actions. Opponent efficiency fell to about 0.89 points per possession. But switching has a price: Jones and Stewart get dragged out to the perimeter, foul more, and defensive rebounding drops. Either choice costs something, and across those nine games they paid by not choosing decisively.

The official box score has no column for "player exploited in drop coverage". Han Xu still recorded blocks, still produced highlight recoveries, and her aggregate defensive numbers looked acceptable. A rebound the league records wrong still counts — if you are willing to rewind. Here, what was recorded wrong was not a rebound but the cause of the points conceded.

Three weeks after the podcast series aired, the Liberty changed the rule. Han Xu stayed closer to the rim, the chasing guard accepted going under screens instead of over them, and the weak-side player was assigned to tag the roller. Sandy Brondello declined to comment on my data. The team announced no adjustment; it simply implemented one. That series drew 80,000 listens, five times a normal episode.

The easiest conclusion for the public was that Han Xu is not quick enough to defend in space and should be cut from the rotation. My data says otherwise. When I isolated the possessions where the chasing guard stayed connected to the ball handler, opponent efficiency fell to 0.98 points per possession, even with Han Xu dropping just as deep as before. Most of the damage came from the defensive layer above her, not from the 2.06-metre centre underneath.

I audited myself before publishing. Fourteen possessions per game is a small sample, and scoring efficiency depends heavily on whether opponents make or miss. I re-ran the calculation across all 40 games, split by each opponent's three-point shooting rate, and the gap between 1.17 and 0.95 held. A small sample is not the same as being wrong; a rushed conclusion is.

Transfer-window noise makes this story hard to see clearly. When a team loses nine straight, the market immediately talks about cutting players, swapping centres, adding a contract. Agents have an incentive to keep the story centred on their clients, because a move generates fees and generates headlines. But look at the cap sheet: the Liberty had no room to replace an expensive centre mid-season. The only open path was fixing the rule. That is the path they took, three weeks later.

Croatia at the 2026 World Cup taught me the same lesson. I rewatched all seven of their matches and logged that Ivan Perišić ran 12.3 km per game but sent only 31 percent of that running toward the opponent's goal. I wrote 19 pages of internal notes; my editor spiked it as too dry, and then Croatia reached the final. Croatia were not the team that ran the most — they were the team that ran in the right direction. The Liberty were the same: not the least effortful defence in the league, but the most misdirected one.

In 2026, coding 612 NBA games from the empty-arena period, I found free-throw accuracy among players under 25 dropped 2.8 percent on average, while EuroLeague showed almost no change. Change the environment and the numbers change. A defensive rule is a kind of environment too, and it distorts individual numbers in ways a box score never displays. People see a mistake and laugh; I see a mistake and go looking for the source.

The 2026 season will answer the bigger question: will the Liberty change the people or change the rule? If they keep drop coverage and only swap centres, that nine-game slide will return under a different name. If they keep Han Xu and repair the layer above her, we will have proof that a correct system can cover a slow-footed big — and that the only thing that needed fixing was the way we read the box score.