Lacrosse Shot Speed Radar Measurement: Correcting Release-Angle Error

Why One Radar Spot Reads Two Different Speeds From the Same Drill

A Division II assistant coach ran into this two falls ago during a crease-finishing session. One radar unit sat dead-center behind the cage for the entire practice. Shots from X, behind the goal line, working a quick roll, logged in the mid-50s mph. Bounce shots from the top of the box on the next rep, same shooters, same rotation, logged in the high 70s and low 80s. The staff's first instinct was to flag the inside finishers for a strength deficiency and add more medicine-ball work. That diagnosis was wrong before the new program's first rep.

Nothing about the shooters' mechanics changed between reps. What changed was the angle between each shot's true flight path and the radar's fixed line of sight. Doppler radar reports only the velocity component traveling directly along its own beam, so a shot crossing the goal mouth at a sharp bearing loses a chunk of its reading to simple trigonometry, not effort.

Akiyama and Yamamoto ran into this constraint directly when they studied physical predictors of shot velocity in 20 lacrosse athletes. Using a microwave Doppler unit to log ball speed, they found moderate-to-strong correlations between shot velocity and medicine-ball throw distance (r = 0.58-0.66), countermovement vertical jump height (r = 0.44-0.46), grip strength (r = 0.33-0.46), and isokinetic knee extension strength (r = 0.20-0.45) (Akiyama & Yamamoto, 2019). The detail that matters for testing protocol: their radar was realigned to each athlete's fixed shooting position for every trial, not left in one spot to cover every release point on the field. The study's own limitation, a single-institution sample of 20, means those correlations need replication, but the alignment discipline behind the data collection is the part any coach can copy immediately.

The Cosine Effect Doesn't Ask Whether You're Shooting From X or Up Top

Doppler radar measures speed by comparing the frequency shift of the signal it gets back to the frequency it sent out. That shift only equals true ball speed when the ball is moving directly along the radar's beam. Any angle between the beam and the actual flight path, call it θ, knocks the reading down by a factor of cos(θ), regardless of how fast the ball is really moving.

The scale of the loss surprises most people the first time they see it laid out. A 15-degree offset, easy to end up with when a radar is aimed at the goal in general rather than one release point, already discards about 3.4% of the true reading. Widen that to 30 degrees, roughly what separates a shot from X off one pipe versus the other, and the loss passes 13%. A shooter capable of a genuine 90 mph release can watch the display read anywhere from 90 down to 78 mph depending purely on where they stood.

Every value in that table is pure trigonometry. No radar model fixes it, and no shooter's technique changes it. It only ever tracks where the unit was pointed relative to where the ball actually went.

Crease Position Changes the True Angle Faster Than Most Setups Adjust For

Lacrosse compounds the angle problem in a way tennis or baseball radar setups rarely face, because shots legitimately originate from all around the box in one session: X and GLE behind the cage, both wings, up top near the restraining line, a crease roll from point-blank range, and a bounce shot that changes direction mid-flight. A radar bolted to one tripod spot might sit within 5 degrees of true for a top-of-box shot and 35 degrees off for a shot from X on the far pipe, in the same ten-minute block, tripod never touched.

Bounce shots add a second wrinkle. The ball's velocity vector changes at the bounce point, so a reading taken after the bounce measures a different trajectory than the one that left the stick. If the goal is release velocity, the unit needs to be gated to catch the ball in the short window before the bounce, not averaged across the whole flight.

Distance compounds it at the close end of the box. Crease finishes from inside 3 meters put the radar close enough that stick-check and follow-through motion can register before or alongside the ball itself, contaminating the peak reading the same way racket-head speed contaminates a badly placed tennis radar. Indoor box lacrosse adds multipath reflection off boards and low ceilings, which some consumer units average into the display rather than reject.

A Setup Standardized by Crease Zone, Not Bolted to One Spot

The fix isn't a better radar. It's treating each crease zone as its own measurement problem with its own bearing correction, built around one fixed base position:

Pair the zone table with a fixed base protocol: mount height at 1.0-1.3 m to match stick-head release height, a locked tripod rather than a handheld unit, peak-speed hold mode, one unit of measurement logged on the sheet. Mark each zone's bearing with tape or a cone on the turf so the angle reproduces session to session instead of getting eyeballed fresh every practice.

One detail catches most testing programs off guard: dominant- and non-dominant-side shots from the identical spot release at a different point relative to the body. Vincent and colleagues documented this when they compared shooting mechanics across high school, collegiate, and professional men's players; trunk lean, shoulder rotation, and crosse angular velocity all differed measurably between dominant and non-dominant releases in the same athletes (Vincent et al., 2016). A bearing tuned for a right-handed release from the wing needs re-checking, not just re-aiming, for a left-handed shooter working the same spot.

A Two-Minute Angle Check Before Any Zone Goes Live

Run this before trusting a single number out of a new crease rotation, new field, or new radar position:

• Have one shooter hit two shots at matched, moderate effort: one from the zone closest to dead-center, one from the widest-angle zone in the rotation. A wide-angle reading more than 8-10% below the centered reading, with no obvious effort difference, means the bearing needs correcting before real data gets logged.
• Sight down the radar barrel yourself toward the release point for that zone. If you can't trace a reasonably straight line from the radar head through the ball's actual release point, the bearing is off before the first real shot is taken.
• Run a stationary drop check between zone changes. A ball dropped from shoulder height should read near zero at rest with one clean spike on the drop; a unit returning noise on this basic check needs a battery or firmware check before it goes near a live drill.

This adds barely any time to a warmup but catches most bearing errors before they get logged as real performance differences between zones.

What Realistic Shot-Speed Numbers Look Like Once Angle Is Controlled

Once bearing error is controlled, shot-speed numbers become genuinely comparable across zones, sessions, and, cautiously, players. Field benchmarks for men's outdoor lacrosse cluster roughly like this by position:

The gap between collegiate and elite levels isn't purely a strength gap. Vincent and colleagues found professional players producing crosse angular velocities roughly 21% faster than high school or collegiate players in the same shooting motion analysis (Vincent et al., 2016), a mechanical efficiency difference layered on top of raw output. Treat a 5-8% swing between sessions under matched conditions as normal noise, and watch the gap between an athlete's average and peak shot across ten reps more closely than any single best number.

The Mistakes That Show Up Most in Crease-Zone Testing

1. One radar position covering every zone in a rotation drill without re-bearing. Fix: use the zone table above and mark bearings on the turf so recentering takes seconds, not a guess.
2. Logging a post-bounce reading as release speed on bounce-shot reps. Fix: gate the reading to the release-to-bounce window, or note on the data sheet which reps were bounce shots and treat them as a separate category.
3. Testing crease rolls from inside 3 meters without accounting for stick-check contamination. Fix: pull the unit back to 5-6 meters for any inside finish, even if that means repositioning between reps.
4. Mixing dominant- and non-dominant-side shots into one average. Fix: log side separately; the mechanical difference is real and documented, not measurement noise.
5. Comparing an indoor box lacrosse session to an outdoor field session as if the radar setup were identical. Fix: log the surface and enclosure type; wall and ceiling reflections indoors add variability that has nothing to do with the shooter.

Building a Protocol That Moves With the Drill, Not Just the Radar

A shot-speed number only earns its place in an athlete's file if the next number can be trusted enough to compare against it. That means writing the zone-by-zone setup down like a lab sheet: radar model, bearing per zone, mount height, distance from the shooter, unit of measurement, and which reps were bounce shots or non-dominant side.

Retest on the same cadence used for other power qualities, every 3 to 4 weeks inside a training block works for most programs, and retest with the exact zone sheet from before, not just the same radar in roughly the same spot. A program that rotates fields, tripods, or testers without documenting the setup generates drift that reads exactly like real performance change, with no way to separate the two without the paper trail.