How to Test Bench Throw Power: The Standard Protocol for Upper Body Explosiveness

Elite rugby athletes average bar velocity above 2.4 m/s at 30% 1RM bench throw, roughly 22% faster than amateur peers (Baker, 2017, n=247). Unlike conventional bench press, the bench throw eliminates the deceleration phase by allowing the bar to leave the hands, exposing true peak power output. PoinT GO's 800Hz IMU sensor, attached to the bar center, captures bar velocity, peak power, and acceleration profiles at 0.00125-second resolution. This guide covers 1) the biomechanical definition of bench throw, 2) safe equipment setup, 3) the standardized measurement protocol, 4) sport- and weight-class-specific benchmarks, and 5) programming applications using [velocity cutoff method guide](/en/guides/velocity-cutoff-method-guide) principles. We also show how to use bench throw outputs together with [1RM calculation methods](/en/guides/1rm-calculation-methods) to back-calculate maximum strength estimates without max testing.

Because the bar is intentionally released, bench throw requires hardware that prevents uncontrolled bar drop. Attempting it on a standard free-weight bench is associated with wrist, shoulder, and facial injury risk. Recommended setups:<br/><br/>1) Smith machine or PlyoPower system - bar travels in a vertical guide, eliminating lateral deviation.<br/>2) Magnetic bar catch - automatically captures the bar at apex or guides it back safely.<br/>3) Floor mat for impact absorption in the unlikely case of a drop.<br/>4) PoinT GO IMU mounted at bar center or one sleeve, with wireless range below 5 meters.<br/><br/>Free-weight bench throws are not recommended. All data must be collected on consistent equipment to allow valid comparison across sessions.<br/><br/><table><thead><tr><th>Equipment</th><th>Safety</th><th>Data Quality</th><th>Cost</th></tr></thead><tbody><tr><td>Smith machine</td><td>High</td><td>High</td><td>Mid</td></tr><tr><td>PlyoPower</td><td>Very High</td><td>Very High</td><td>High</td></tr><tr><td>Free bar (not advised)</td><td>Very Low</td><td>Variable</td><td>Low</td></tr></tbody></table><br/>Warm up with scapular stabilization from [power clean technique](/en/exercises/power-clean-technique) progressions, 2-3 sets of push-up variants, and light medicine ball chest passes to prime the nervous system.

Procedure: 1) Lie flat with feet firmly planted; 2) grip width 1.5x shoulder width; 3) load to 30% 1RM, the loading shown to maximize peak power (Baker & Newton, 2018); 4) lower the bar to the chest in a controlled eccentric phase (~0.5-0.6 m/s); 5) without pause at the chest, throw the bar upward at maximum velocity; 6) accelerate hard enough that the bar separates from the hands.<br/><br/>Each set consists of 3 reps with 3 minutes rest between sets. Perform 4 sets, recording both peak and average values.<br/><br/>The PoinT GO IMU streams these metrics in real time:<br/>- Mean bar velocity (m/s): average during the concentric phase<br/>- Peak bar velocity (m/s): max within the concentric phase<br/>- Mean power (W): bar mass × mean velocity<br/>- Peak power (W): bar mass × peak velocity<br/>- Time to peak velocity (ms)<br/><br/>Frost et al. (2019) reported IMU agreement with optical motion capture at ICC 0.94 and mean error within 0.03 m/s. For autoregulated work, use these outputs with [autoregulated velocity training](/en/guides/autoregulated-training-velocity) principles to adjust loads weekly.

Numbers must be interpreted relative to sport, weight class, and training phase. Rugby forwards prioritize absolute power (W); pitchers prioritize bar velocity (m/s); throwers care about both.<br/><br/><table><thead><tr><th>Level</th><th>Bar Velocity (m/s)</th><th>Peak Power (W)</th><th>Relative Power (W/kg)</th></tr></thead><tbody><tr><td>Elite (rugby)</td><td>&gt;2.4</td><td>&gt;850</td><td>&gt;9.5</td></tr><tr><td>Advanced</td><td>2.1-2.4</td><td>700-850</td><td>7.5-9.5</td></tr><tr><td>Intermediate</td><td>1.8-2.1</td><td>550-700</td><td>5.5-7.5</td></tr><tr><td>Novice</td><td>&lt;1.8</td><td>&lt;550</td><td>&lt;5.5</td></tr></tbody></table><br/>A key interpretation point: bar velocity and power respond differently to load. At 30% 1RM, velocity dominates; above 50% 1RM, power dominates. To diagnose [force-velocity imbalance](/en/guides/force-velocity-imbalance-explained), test across multiple loads and plot the slope.<br/><br/>Upper body explosiveness should be cross-referenced with lower body and rotational outputs. Pair bench throw with [medicine ball throw test](/en/exercises/medicine-ball-throw-test) and [rotational power measurement](/en/exercises/rotational-power-measurement) to build a complete profile.

Bench throw data is not just a measurement; it is a load-prescription tool. González-Badillo et al. (2019) demonstrated that when bar velocity drops more than 15% below the day's baseline, the set should be terminated to avoid neuromuscular fatigue accumulation - the core logic behind the [velocity cutoff method guide](/en/guides/velocity-cutoff-method-guide).<br/><br/><strong>Power phase (weeks 1-4):</strong> 30% 1RM, 5 sets x 3 reps, target velocity above 2.2 m/s. Drop load 5% when velocity falls.<br/><strong>Speed-strength phase (weeks 5-8):</strong> 50% 1RM, 4 sets x 4 reps, target above 1.5 m/s.<br/><strong>Strength phase (weeks 9-12):</strong> 70% 1RM, 4 sets x 4 reps, target above 1.0 m/s. Consult the [how to program strength block 12 weeks](/en/guides/how-to-program-strength-block-12-weeks) framework for sequencing.<br/><br/>Bar velocity gains accelerate when paired with deliberate eccentric loading. The principles in [why eccentric training builds more muscle](/en/guides/why-eccentric-training-builds-more-muscle) apply: 3-second eccentric phases potentiate the subsequent concentric throw via stretch-shortening cycle enhancement.<br/><br/>For long-term monitoring, follow the [athlete testing battery guide](/en/guides/athlete-testing-battery-guide). Re-test 30% 1RM bench throw every 4 weeks; once weekly variability stays below 5%, advance to the next phase. This rhythm keeps adaptation predictable and overtraining off the table.