Redefining the Bench Press: From Standardized Lift to Testable System

A documented evolution

The bench press became a controlled system in layers

Strength history is especially vulnerable to a neat “first” that the documents cannot support. Joe Roark’s archival review instead shows overlapping practices: early lifters pressed while supine, contests sometimes required a pullover or used a handoff, and photographs document both plain benches and benches with uprights before the modern commercial form stabilized [1].

Jan Todd’s powerlifting history places a “press on floor” or bench press among 1940s odd lifts and identifies the first AAU Senior National Powerlifting Championships in 1965 [2]. The International Powerlifting Federation was founded in 1972 and held its first official World Championships in 1973 [3]. Those milestones did not invent supine pressing; they helped turn a varied practice into a governed sport.

Reference condition

Standardization made the lift comparable

The current IPF rules define a flat, level bench, including a 29–32 cm width and a 42–45 cm height, alongside commands, legal contacts, grip limits and depth requirements [4]. The point is measurement fairness: a valid lift performed on one sanctioned platform should mean substantially the same thing on another.

The competition condition is therefore a reference, not a universal control for every research question. Transfer back to that reference must still be measured when a variation is used to build competition performance.

Programmable variation

Once the reference existed, changing the task became informative

Madsen and McLaughlin used high-speed two-dimensional cinematography with 36 lifters—19 experts and 17 novices—in 1984. The experts lifted 79% more, lowered the bar more slowly, used a bar path closer to the shoulders and organized the press differently. Yet the groups did not differ significantly in shoulder torque. That null finding matters: superior performance did not require every proposed mechanical risk marker to move in the expected direction [5].

In 12 nationally and internationally competitive athletes, Saeterbakken and colleagues compared six-repetition maximum performance across flat, incline and decline benches and grip variations. Incline pressing reduced the load that could be lifted, while grip width did not significantly change prime-mover activation in that trained sample [6]. In a separate one-repetition-maximum experiment with 14 experienced men, Larsen and colleagues showed that grip width altered strength, displacement, horizontal force direction and calculated joint moments [7]. Together, the studies show why “grip” or “angle” cannot stand in for a full protocol.

Support-surface evidence is equally protocol-bound. Koshida and colleagues found lower peak force, power and velocity during dynamic pressing on an intentionally unstable support [10]. Goodman and colleagues found no difference in one-repetition maximum or selected muscle activation between a stable bench and an exercise ball [11]. Those results are not contradictions to be averaged into a slogan. They used different tasks and outcomes, and neither tested conventional pad geometry.

Instrumented era

The bench is now an experimental condition, not background furniture

Mausehund and colleagues showed that omitting lateral barbell forces can substantially distort net joint-moment calculations [12]. Noteboom and colleagues then modeled shoulder loading across scapular pose, grip width and shoulder-abduction targets in 10 experienced athletes; several modeled glenohumeral and acromioclavicular force components changed across conditions [13]. These are important mechanics findings, but modeled force is not pain, injury or diagnosis.

Direct interface research

What the Launch Pad Research Adds to the Experimental Record

AMM’s published program directly tested the support interface rather than treating the bench as background equipment. The work followed the design question posed by AMM’s co-founder and Launch Pad inventor: if load, grip, range, rack and backrest angle can be changed deliberately, why should the torso-support surface remain unexamined? Across one acute crossover and two randomized training trials, the support condition was experimentally defined while outcomes were measured at acute and longitudinal levels.

Together, the three studies establish acute interface-dependent differences and two controlled longitudinal performance differences under their tested conditions. The evidence progression makes the support surface a measured experimental variable rather than an assumed constant. Feature isolation, mediation of longitudinal adaptation and clinical outcomes such as pain or injury require separate experiments.

The next experimental question

Joint Ops turns configuration into a research problem

Joint Ops is not the next citation in a Launch Pad proof chain. It is a distinct, selectable system in active prototype research with no published outcomes. The relevant question is whether an indexed configuration changes a prespecified variable relative to another indexed configuration—and whether that change survives familiarization and replication.

The bench press has endured because it can be standardized without being frozen. Its next useful designs will not win by declaring the flat bench obsolete. They will win by making the changed condition measurable, the comparison fair and the limits visible.