How WAR Works in NCAA Volleyball

The Problem

Volleyball has no WAR. Baseball has it. Softball has it. Even hockey prospect models can lean on goals-above-replacement frameworks. But volleyball — a sport where every point involves a passer, a setter, and a hitter — has no established framework for measuring individual value.

The challenge is structural. In baseball, a batter stands alone in the box. In volleyball, a kill has three contributors: someone passed the ball, someone set it, and someone attacked it. How do you assign credit? How much of a kill belongs to the hitter who swung, vs. the setter who placed it, vs. the passer who kept the ball in system?

We built a model that answers this empirically — no arbitrary knobs, no “feels right” weightings. The credit split, replacement levels, and conversion rates all come from the data.

Points Are Our Runs

In baseball, the intermediate currency is runs. A home run is worth ~1.4 runs, a walk ~0.3 runs, and you can convert runs into wins via Pythagorean expectation. The entire WAR framework rests on this currency.

In volleyball, the equivalent currency is the rally point. Every rally ends with a point scored. Every point has contributors. The model's job is to figure out how much each role — hitter, setter, passer — contributed to winning those points, and how many wins that contribution is worth above a replacement-level player.

The Credit Split

Before building individual WAR, we need to know: how much does each role affect winning? We measured this by sorting all 78 teams into quartiles by each role's primary stat, then comparing win percentages.

Teams with top-quartile hitters win 73.1% of their matches; bottom-quartile hitters win 41.3%. That 31.8-point gap is the largest of the three roles, giving hitters 48% of the credit. Setters account for 29.5%, passers 22.5%. The shares are simply each role's gap divided by the total gap across all three (0.662).

This isn't surprising if you've watched volleyball. Hitting is the terminal skill — the swing that ends the rally. But the model confirms that setting and passing are not just supporting acts. A team with elite passers and setters but average hitters still wins significantly more than the inverse.

Hitter WAR

Hitting percentage is the wOBA of volleyball— a single number that captures offensive efficiency. It's (kills − errors) / attempts, and it correlates with team winning more strongly than any other individual stat.

The formula is straightforward:

Where 0.1414 is the replacement-level hitting percentage (bottom quartile of qualified starters across 495 hitters), and 0.02453 is the conversion factor that translates extra net kills into wins. That conversion factor is itself the product of two empirically-derived numbers: 0.7009 (hitter credit per raw unit) × 0.03500 (WAR per point).

Worked example — Kennedy Martin (Penn State OH):

Blocking bonus: Any player averaging 0.50+ blocks/set gets additional credit above the 0.3613 replacement level. This primarily benefits MBs but also rewards front-row setters in 6-2 systems who contribute meaningful blocking.

Top 10 Hitters

N=713 qualified hitters. Avg WAR: +1.06. Note that 7 of the top 10 are outside hitters — the position that sees the most attempts and has the most opportunity to accumulate value.

Setter WAR

Setter WAR uses assists per set above the replacement level of 2.37 A/S (bottom quartile of 124 qualified setters). The conversion rate is lower than hitting — reflecting the credit split that gives setters 29.5% of the pie.

Multi-axis credit:The model doesn't lock players into a single role. Any setter who also hits (100+ attack attempts) or blocks (0.50+ B/S) receives credit for those contributions independently. This matters most for 6-2 offenses, where two setters alternate — each sets from the back row and hits/blocks from the front row. A 6-2 setter running 9+ A/S while hitting .250+ on 300 attempts earns WAR from both axes.

The assist problem: Assists partly reflect hitter quality. A setter running an offense with Kennedy Martin and Malaya Jones will rack up more assists than one feeding average hitters. We checked this by regressing team hitting percentage on individual hitter quality, then looking at whether the setter moves the residual. The correlation is r = 0.22— modest but real. Setters do affect team hitting percentage beyond what the hitters themselves explain, but the effect is small. This is why pure setter WAR is compressed — multi-axis credit is what separates the best setters from each other.

Top 10 Setters (Total WAR)

N=186 qualified setters. Every setter in the top 10 earns multi-axis credit — the days of a pure-assist setter ranking first are over. Claire Ammeraal's hitting contribution (1.74 WAR) actually exceeds her setting contribution (1.73) — she's essentially a two-way player who happens to be listed as a setter. The spread is now 1.07 WAR from #1 to #10, much wider than the 0.67 WAR spread in the old single-axis model.

Passer WAR

Passer WAR measures reception errors per set below the replacement level of 0.351 RE/S (bottom quartile of 91 qualified passers). Fewer reception errors means cleaner balls in system, which means better swings for hitters.

We validated this connection directly: after controlling for individual hitter quality, team reception error rate still correlates with team hitting percentage at r = −0.29. Clean passing measurably improves the hitting around it.

The defensive sub filter:Not every libero or DS listed on a roster actually receives serve. Many are defensive-only subs — they come in for back-row digs but never stand in the serve-receive pattern. These players show 0.000 RE/S not because they're perfect passers, but because they never receive. The model requires 2.0+ digs/set to qualify as a passer. Below that threshold, a player is classified as a defensive sub and receives no passer WAR. This filter removed 160 defensive subs and left 111 real passers.

Top 10 Passers

N=111 qualified passers (liberos + defensive specialists with 2.0+ digs/set). Like setters, the passer WAR range is compressed — the best passer in the country is worth +1.95 wins, roughly a third of what the best hitter is worth. This is consistent with the 22.5% credit share.

Validation: The 2025 MLV Draft

The best test of a player-value model is whether it predicts how real decision-makers behave. The 2025 MLV (Major League Volleyball) draft gives us that test.

MLV's first round was 8 picks: 6 outside hitters and 2 middle blockers. Our top 10 is 7 outside hitters, 1 opposite, and 1 middle blocker. The positional value distribution matches.

Six of the eight first-round picks appear in our top tier. The model correctly identifies the OH-heavy value distribution that MLV's front offices independently converged on. Carlson's multi-axis total (+2.49, including front-row hitting credit) makes her the highest-WAR setter drafted — consistent with being the first setter off the board.

Where the model disagrees:Hayden Kubik (MLV #3, Nebraska OH) doesn't appear in our top 50. Tennessee went .333 in 2024–25, and the model penalizes production on losing teams. MLV is projecting future value; the model measures what happened. That's a feature, not a bug — the model is production-based, and the gap between our ranking and MLV's reveals where scouts are adding projection on top of production.

Key Numbers

What the Model Misses

• Setter assists are noisy— Assists reflect hitter quality as much as setter quality. The r=0.22 residual effect is real but modest. Multi-axis credit mitigates this (hitting and blocking are individual skills), but the pure-setting axis still inherits noise from the hitters around it.
• No defensive credit beyond blocking and reception errors— Digs, serve receive quality (beyond errors), and floor defense are not captured. A libero who turns a hard-driven ball into a perfect pass gets no more credit than one who barely keeps it alive. This is a data limitation, not a philosophical choice.
• Production, not projection— WAR measures what a player didthis season. It doesn't account for age, physical tools, or development trajectory. A senior with +5.0 WAR and a freshman with +5.0 WAR are very different draft prospects.
• Passer WAR is still compressed— The digs/set filter removes defensive subs who never receive, but the remaining passer pool is still narrow (111 players vs 713 hitters). A truly elite libero is worth roughly 2 wins — important, but hard to differentiate from a merely good one statistically.
• Short-sample effects— NCAA volleyball seasons are ~30 matches. A hitter who goes on a two-week tear can swing a full win of WAR. The model is best used as one input alongside scouting, award history, and team context.