The Eggbeater Power Test: A Water Polo Protocol for Rise Power and Hold Time

A center forward can win every sprint on the roster and still get scored on from three feet away, because she can't hold her chest above the surface long enough to get a hand up before the ball is past her. That gap doesn't show up on a 20m swim time or the standard vertical-height-out-of-water measurement most clubs run as a single best effort. It shows up in the fourth quarter, on the third straight possession where she has to rise, get an arm up, sink back to reload, and do it again twelve seconds later without her elevation falling apart.

Two qualities are buried inside that one movement, and most testing treats them as a single number: how fast and high a player can drive up on one all-out eggbeater effort — rise power — and how long she can hold that position once she gets there, kicking against her own body weight and the water pulling her back down — hold time. A player can post an excellent rise and a mediocre hold, or the reverse, and the weaker of the two is usually what shows up as a missed block or a late pass. The protocol below separates them, tests both in the pool, and scores each on its own.

Why One Rise Height Doesn't Tell You What You Need to Know

Sanders (1999), modeling the water polo boost — the eggbeater-driven rise this test measures — from film analysis of skilled players, found peak hip and knee angular velocities during the insweep of each leg's cycle were the primary drivers of vertical velocity. Just as important, the timing offset between the two legs' cycles mattered nearly as much as raw leg speed: poorly timed alternation let force from one leg cancel the other. That model explains what a good boost needs mechanically. It says nothing about how long an athlete can keep producing it, since it came from single maximal efforts on film.

That's the blind spot a one-off height test carries into the pool. An athlete with strong hip drive but weak local endurance in the sculling musculature can post a height nearly identical to a well-rounded player on the first trial, then be the one sinking on the second block attempt of the same possession, arms still up, torso dropping. Coaches usually read that as a fitness problem. Just as often it's a narrow gap in how long the pattern holds near maximal output — one a single rise trial was never built to catch.

Equipment and Pool Setup

No lab equipment needed. The test needs deep water, a fixed vertical reference, and a way to mark the same spot on the athlete's torso every time.

The ball overhead does two jobs: it keeps the arms out of sculling range, and it gives testers an unambiguous cue — if it dips before the torso tape drops below the target line, the trial is void, no debate needed.

Step-by-Step Protocol: Rise and Hold

1. Warm-up (10-12 min): Easy swim, dynamic hip/shoulder mobility, 3 submaximal rises building to ~80% effort.
2. Mark the resting waterline: Ball overhead, athlete treads with no deliberate sculling. Mark where the torso tape sits against the pole at rest.
3. Familiarization: One ~80% effort rise confirming the athlete can drive up with the ball fixed overhead, no arm-assisted sculling.
4. Phase 1 — Rise trials: On the go signal, one maximal eggbeater-only drive, holding the peak half a second to mark it. Record distance from the resting mark to peak. Run 3 trials, 90 seconds rest between; keep the best.
5. Time-to-peak (optional): With video or an IMU, note time from first upward drive to peak — typically 0.4-0.7s in trained players, feeding the power estimate below.
6. Full rest before Phase 2: 3-5 minutes so residual fatigue doesn't drag down the hold score.
7. Set the hold target: 85% of the athlete's own best Phase 1 height. One fixed height for the whole team rewards already-powerful athletes twice; scoring against each athlete's own peak isolates the quality you're after.
8. Phase 2 — Hold trials: Rise to the 85% mark and sustain it, continuous kicking, ball overhead. Clock starts on crossing the mark, stops when it drops below for over half a second or the ball touches water. Run 2 trials, full rest between; record the longer.
9. Void and rerun any trial with visible hand sculling, wall/rope contact, or an early ball drop.

Total time per athlete runs about 15-18 minutes, so a squad of 12-14 fits in one session with two testers on parallel lanes.

Turning Height and Time Into a Power Estimate

Height alone rewards a slow, grinding rise almost as much as a fast one, which misses the explosive quality coaches care about. Capture time-to-peak alongside height, and you get average velocity and a rough power output.

Average rise velocity: v = h / t, h in meters, t in seconds from first upward drive to peak.

Average power against gravity: P = (m × g × h) / t, m in kilograms, g = 9.81 m/s². This treats the rise as work lifting the center of mass against gravity, and it's a floor on true output, not a ceiling — a meaningful share of eggbeater leg drive goes into accelerating water rather than lifting the body directly, the same assumption behind any field-based jump power formula. Treat it as a comparable index across trials, not a force-plate-grade figure.

Worked example: an 82kg athlete rises 0.24m in 0.55s — velocity ≈0.44 m/s, power ≈351W. A teammate rises higher, 0.30m, but takes 0.9s, a slower effort. Velocity drops to 0.33 m/s and power to about 263W despite the taller rise. Height alone would have scored the second athlete higher; power correctly flags the first as more explosive.

What the Research on the Boost Actually Shows

Sanders (1999), publishing in the Journal of Applied Biomechanics, is the closest thing water polo has to a mechanical blueprint of the boost. Working from film-derived joint kinematics of skilled players, the model found vertical velocity at peak height tracked most closely with peak hip and knee angular velocities during each leg's insweep, and with how tightly the two legs' cycles were offset from each other. The limitation matters: the model came from a small sample of already-skilled performers using film-derived angles rather than force-plate data, so it explains what a good boost looks like mechanically without saying how that capacity holds up across sustained efforts — the gap the hold-time half of this protocol fills.

Falk, Lidor, Lander, and Lang (2004), following elite and sub-elite junior water polo players over two years for the Journal of Sports Sciences, included a vertical leg-power measure in their fitness battery and found it one of the qualities separating players later selected for national squads from those who weren't, the gap widening rather than narrowing across the follow-up. The authors were careful about what that could support: squad selection wasn't random, so the study can't fully separate physical talent from training alongside a higher-level group, and a measure discriminating selected from non-selected players in one talent pool doesn't hand you universal cutoffs for every level. That's why the bands below are coaching reference points, not published thresholds.

Reading the Two Scores Together

The ranges below come from the general magnitude of leg-power output in the boost and eggbeater literature plus field-testing experience running this rise-and-hold split. Treat them as a starting point, not a cutoff that decides playing time.

The band a player lands in matters less than the relationship between the two scores. Rank the squad on each metric and look for players whose two ranks are far apart. Near the top on rise power but near the bottom on hold time signals a specific, trainable local-endurance gap, not a general power problem. The reverse — modest rise, strong hold — usually points to an athlete compensating for a lower ceiling with pacing and technique, with real room to add hip and knee power without losing what already works.

Mistakes That Quietly Wreck the Numbers

What to Train Depending on Which Score Is Low

A low rise-power score responds to explosive strength and rate-of-force-development work: loaded jump squats, hip thrust jumps, resisted single-leg eggbeater sculling against a drag parachute or band. These build the peak hip and knee angular velocities Sanders' model points to as the main driver of the boost.

A low hold-time score needs a different stimulus — local muscular endurance and lactate tolerance in that same musculature, trained under sustained rather than ballistic load. Extended holds at roughly 70% of the target line, sets of 3, progressing target time 10-15% every two weeks, close this gap faster than more explosive dry-land work ever will.

Retest every 4-6 weeks. Falk et al. (2004) tracked athletes across a two-year window precisely because gaps like these close over a season, not a week. And resist prescribing the same block for every low score — a rise problem and a hold problem are different athletes needing different work.