Kentucky 2-1 Louisville: The .079 to .420 Swing and a Lesson on Sample Size in NCAA Women's Volleyball

Three sets of a match that has not yet finished, and one pair of numbers that made me stop rewinding. Kentucky's attacking efficiency in Set 1 was .079. In Set 2, it was .420. Same roster, same opponent, separated by one intermission. A swing of 5.3 times — the largest volatility in the entire data set I collected from this match.

I sat with the numbers, trying to find a cheap explanation. Personnel change? No record. Scheme change? No evidence. Changed serve pressure from the other side? Entirely possible, but nobody logged that metric.

That was the moment I recognised something I repeat in every analysis session: most of what we call a "tactical adjustment" in NCAA women's volleyball is really just a team finding its rhythm again. And to prove that — or refute it — I need more than three sets.

Data never lies; only hasty readers do.

Context: an in-state derby without precedent

This is an NCAA Division I women's volleyball match, early in the regular season, non-conference. But calling it "an early-season friendly" misses the point entirely.

Kentucky is ranked No. 4 nationally. Louisville is No. 3. Both are elite programs in the state of Kentucky — a state where women's volleyball is not a minor sport. According to the head-to-head history I compiled, the two teams have met 67 times since 2026, with Kentucky leading 33-29. That is near-perfect parity across nearly half a century.

What makes this match different: for the first time in the history of the series, both teams entered a match while ranked in the national top five. This is not decorative detail. In American college volleyball, early-season rankings are built on extremely small samples — often after just two or three weeks of play. Louisville is listed at No. 3 in the Power 10 poll for Week 3. In other words, the "top five" label both teams carry is a label with high uncertainty.

The match was broadcast live on ABC. This is a more important industry signal than many realise. A regular-season match, with no tournament berth at stake, was placed on national television. The arena was sold out. To me, that is a sign that American college women's volleyball is entering a new commercialisation cycle, and derbies like this are products designed to be sold.

One technical note is mandatory before we get into the numbers, because I know many readers will make the wrong comparison: the NCAA calculates hitting percentage as (kills − errors) / attempts, and does not deduct blocked shots. The FIVB standard we are used to at world championships does deduct blocks. That means the .079 or .420 figures here read higher than the concept of "attack efficiency" Vietnamese audiences usually see. This is a different statistical convention, not an error.

Finally, a scope limitation: the document I am analysing is a live report, frozen at the point where three sets had been completed. The match is unfinished. Kentucky leads 2-1. Every conclusion below is bounded by that fact.

The core: a chain of data evidence

Set 1 went to Louisville, 25-19. But the scoreline does not tell the whole story. Louisville opened with a 4-0 run and led 14-5. That is not a good start — that is a rout. Louisville hit .324 in the set. Kentucky hit .079.

To put the gap in perspective: .079 means roughly one net point per 12.6 swings. In college women's volleyball, .150 is considered serviceable for a set. .079 is the level of a set in which almost every rally drifts to the other side.

The person spreading discomfort through Kentucky in Set 1 was Chicoine, an outside hitter. In a single set she recorded 4 kills, 3 blocks and 2 digs. Three blocks in one set from an outside hitter is unusually high. The outside pin is rarely a major blocking contributor — that is usually the middle blocker's domain. When an outside hitter blocks three times in the opening set, it usually means the opponent is attacking into a very predictable spot, and she is reading it early.

In Set 2, the picture flipped entirely. Kentucky won 29-27. Tracking the report, the set featured 16 ties and 7 lead changes. Louisville saved two set points. Kentucky converted on its third.

A set with 16 ties and 7 lead changes is not a normal set. It means both reception systems held up under pressure, and neither side could build a long enough break to pull away. In such sets, the outcome is usually decided by one moment of serve pressure or one high-ball handling at the end.

And Kentucky's attacking efficiency in Set 2 was .420. From .079 to .420. I checked several times because the figure looked like a typo. It is not a typo.

In Set 3, Kentucky kept control. Middle blocker Washington recorded 5 kills in the set — a significant attacking load for a middle, showing the quick middle attack had been brought into regular operation. Kentucky built a 21-14 lead. Louisville answered with a 5-1 run to cut the gap to 22-15.

This is where I must stop and flag source data quality: if Kentucky led 21-14 and Louisville scored 5 unanswered while Kentucky scored 1, the score should be 22-19, not 22-15. The 22-15 figure is internally inconsistent with the sequence described in the same sentence. As an analyst, I am obliged to mark this detail as "data pending verification."

Through three sets, DeLeye — Kentucky's outside hitter — had 18 kills, or 6 kills per set. That is a very high output level.

But this is where methodology matters more than the raw box score. 18 kills does not tell me how many times DeLeye swung. An attacker with 18 kills on 30 swings is a star. 18 kills on 55 swings is a serious distribution problem. The report does not provide the denominator, meaning I cannot conclude anything about DeLeye's true efficiency — only about volume.

That is exactly why I keep repeating one principle: I do not argue with emotion, I argue with sample size.

And what is the sample size here? Three sets. No team block totals. No dig metrics. No perfect-pass rate. No ace-to-error ratio. No individual attacking efficiency for any player. This data set is missing nearly every dimension needed to evaluate a volleyball match at a professional level.

The only thing I can defensibly claim statistically is the magnitude of Kentucky's attacking swing. Everything else sits in the "insufficient data" zone.

The contrarian angle: correlation is not causation

I want to say plainly something sports media often avoids: the .079 to .420 swing is attractive for storytelling, but it may mean nothing at all.

Hitting percentage at set level is an extremely volatile metric. With sample sizes of only a few dozen swings in a set, a handful of kills or errors can push the figure from .080 to .400 without any tactical change at all. This is regression to the mean, not evidence of a revolution on the coaching bench.

In other words: if you believe Kentucky "solved" Louisville in Set 2, this data set does not support you. It only shows Kentucky hit much better. Those are different things in causal terms.

The 2026 World Cup taught me a lesson: a model does not need to be big, it needs to be right. There, I learned that assigning causation to a statistical phenomenon is the most common mistake of the hasty analyst. Three volleyball sets is far too small a sample to describe a team's nature.

One more point on source data quality. The report calls Bultema "a Kentucky transfer and middle blocker," yet lists her on Louisville's roster. The two pieces of information directly contradict each other. There are only two possibilities: either she transferred from Kentucky to Louisville (meaning the description is wrong), or the roster listing is wrong. In American college volleyball, the transfer portal is a talent-redistribution mechanism operating powerfully at the elite tier, so the first possibility is entirely plausible. But I do not have enough data to confirm the direction of this transfer.

There is one more detail to flag: the match date is given as Sunday, September 20. September 20 is a Sunday — that detail is internally consistent. But the year attached to it needs independent verification before use. I mark this as data pending verification.

Error is not the enemy; it is the silent teacher of every model.

There is another contrarian layer here, and it concerns the team that won Set 1. Louisville led 14-5, hit .324, then lost Set 2 in a nail-biter. A team that loses a set after leading 14-5 is usually assigned a psychological flaw. But viewed through data, the more plausible cause is that they kept the same attacking structure while the opponent adjusted, and when Kentucky's serve pressure rose late in the set, their distribution became more predictable.

I emphasise: this is inference, not conclusion. There is no positional distribution data to verify it.

Interestingly, the data limitations cut both ways. I also cannot confirm that Kentucky "changed its defensive system." There are no team dig or block figures for Kentucky. What is described as a defensive comeback is, on paper, an offensive-line correction.

And if I had to pick one explanatory structure, I lean toward the simplest: Kentucky began the match tight, in a packed and hostile arena, against an in-state rival ranked one spot above it. Set 1 was the consequence of that state. Set 2 was a return to normal. The adjustment that was truly needed was not on the tactics board, but in the head.

Fortunately, this is a testable data sample. If Kentucky sustains efficiency around .250 or better in Sets 4 and 5, the rhythm hypothesis holds. If they collapse back toward .100, we know we are dealing with a team of extremely high variance.

Signals to track from the next round

Kentucky leads 2-1, and in the NCAA's best-of-five format, that is not a safe margin. A Louisville win in Set 4 would force a 15-point decider. That scenario is entirely within reach for Louisville, a team that just showed it can generate counter-runs.

I will track four signals, not by feel, but by the numbers.

First, DeLeye's true hitting efficiency, not her kill count. If her 18 kills came on excessive volume at low efficiency, Kentucky is depending on a single attacking point, and that is a vulnerability exploitable over the second half of the season.

Second, Louisville's set-closing ability. They lost Set 2 after saving two set points. If this recurs in Set 4, we are looking at a behavioural pattern, not an accident.

Third, the balance of Kentucky's attacking distribution. If Washington and the remaining pin hitters sustain the volume they showed in Set 3, that signals depth. If DeLeye continues to account for more than half the kills, I will add it to the single-point-dependency risk list.

Fourth, and perhaps most important to me because it concerns the industry: broadcast metrics. A regular-season match on ABC, in a sold-out arena, between two top-five in-state programs. If this match's viewership exceeds the season average, that is evidence that sponsorship and television investment money is flowing more strongly into American college women's volleyball, and derbies like this will increasingly be pushed into prime slots.

Every figure on the transfer board is a story untold.

As for this match, the biggest question is not who wins. Kentucky leads 2-1 and controls its own fate. The question I really want answered is this: is the .420 in Set 2 the truth of this team, or just a lucky data point in too small a sample? The regular season is still long. And I will wait for enough sample to answer.