Skill & distance progression: the evidence

CA-P-01Standard push-up load

PrescriptionA standard push-up loads roughly 64–66% of bodyweight at the start position, rising toward 75% as the elbows flex. Used as the reference point every other push-up rung is placed against.

Evidence gradeC

EffectGround reaction force at the normal position measured 66.4% of bodyweight [1]. An independent force-plate study of six variations found the regular push-up around 64% of body mass, and reported that these data can be used to progress push-up intensity as a percentage of body mass [2].

PopulationEight healthy men [1]; twenty-three recreationally fit young adults, mean age 22.5 [2].

Sources[1][2]

CaveatsTwo studies in small samples, measuring acute kinetics rather than training outcome. A push-up is not one load: it climbs as you descend, so a single percentage is a simplification of a curve.

CA-P-02Knee push-up load

PrescriptionAbout 53% of bodyweight. Placed below the standard push-up in the ladder on this measurement.

Evidence gradeC

Effect52.9% of bodyweight at the start position, against 66.4% for the normal variant. Muscle activity in both triceps brachii and pectoralis major was lower during the on-knees variant [1].

PopulationEight healthy men.

Sources[1][2]

CaveatsThe lower EMG alongside the lower force is what makes this a genuine reduction in demand rather than a change in leverage the muscles do not feel. Still a single small study of acute kinetics.

CA-P-03Hands-elevated push-up load

PrescriptionFalls continuously from the standard push-up down to about 41% of bodyweight as the hands rise. Roughly 55% at a 30 cm surface, 41% at 61 cm.

Evidence gradeC

EffectHands elevated on a 30.48 cm box and a 60.96 cm box produced lower ground reaction forces than every other push-up variation tested [2]. A separate study modeling incline push-ups found peak force decreasing as the incline steepened, with mean maximum force at 10° exceeding 30° by 71.12 N [3].

PopulationTwenty-three recreationally fit young adults [2]; twenty-seven male university students [3].

Sources[2][3]

CaveatsThis continuity is why MVIII uses an incline push-up rather than a knee push-up as its main scaling rung: an incline is a dial that fits whatever surface somebody has, where a knee push-up is one fixed load. Neither study measured a wall push-up, so its load is inferred from the far end of the same scale rather than measured.

CA-P-04Feet-elevated push-up load

PrescriptionAbout 70% of bodyweight at a 30 cm surface, up to 74% at 61 cm.

Evidence gradeC

EffectPush-ups with the feet elevated produced higher ground reaction forces than all other push-up variations tested [2].

PopulationTwenty-three recreationally fit young adults.

Sources[2]

CaveatsThe gap from a standard push-up to the hardest feet-elevated version is about ten points of bodyweight, which is a much smaller step than any of the steps below the standard push-up. Progress past a full push-up is therefore mostly a matter of repetitions rather than leverage, which is why MVIII switches to a rep scheme there.

CA-P-05Band-assisted pull-up, load removed

PrescriptionA long band takes off enough load to roughly double or triple the repetitions somebody can perform. Placed below the negative pull-up in the ladder.

Evidence gradeC

EffectParticipants performed 11.35 repetitions unassisted, 17.61 with a band arranged vertically (feet inside the band) and 28.03 with the band arranged horizontally (feet on top) [4].

PopulationThirty-three university kinesiology students, 26 male and 7 female.

Sources[4]

CaveatsRepetitions are a proxy for the load removed, not a measurement of it. The two band arrangements differ enough that the same band is effectively two different rungs depending on how it is tied, and the app cannot tell which way anybody tied theirs.

CA-P-06Eccentric-only training and concentric strength

PrescriptionNegatives are programmed as loaded practice at the bar, late in the ladder. They are not programmed as a faster route to the first repetition.

Evidence gradeA — contested application

EffectAcross 27 studies, eccentric-only training produced greater gains in eccentric strength than concentric-only training (Hedges' g 1.51; 27% versus 10%; p < 0.001). No difference was evident between the training modalities for changes in concentric contraction strength [5].

PopulationHealthy adults, 27 studies, 162 study results.

Sources[5]

CaveatsThis is the strongest evidence in the document and it cuts against the usual argument for doing negatives. Pulling yourself over a bar is a concentric action, and eccentric-only training did not improve concentric strength more than concentric training did. The isokinetic protocols in these studies are also not identical to a slow lowering from a bar. A negative remains a defensible way to accumulate time under load at the movement; the claim that it is the fast route is not supported.

CA-P-07Training frequency

PrescriptionLadder work is placed on the member's existing training days, two to three exposures per week. No additional sessions are asked for.

Evidence gradeB

EffectA meta-analysis of weekly training frequency found the existing data does not provide a strong correlation between increased weekly frequency and maximal strength gain in upper- and lower-body exercises for a mixed population [6].

PopulationMixed trained and untrained.

Sources[6]

CaveatsFrequency is how weekly volume gets distributed rather than an independent driver, which is precisely what supports inheriting the member's existing schedule instead of prescribing more days.

CA-P-08Submaximal set effort for a rep-count goal

PrescriptionFive sets at approximately 50% of the best single set, when the goal is a repetition count and the movement itself is already achievable.

Evidence grade— (no controlled trial located)

EffectNo trial isolating this scheme was located.

Population—

Sources—

CaveatsThis practice is widespread and has a well-known name. It has no controlled trial behind it that could be found. The number is a convention MVIII follows and states as a convention. It is recorded here rather than omitted, because a number the app prescribes with nothing behind it should be visible as such rather than absent from the table.

CA-R-01Interval training versus continuous running

PrescriptionA pace goal is built from repeats at the goal pace rather than from continuous running below it.

Evidence gradeB

EffectMatched for total work and frequency, four-minute high-intensity intervals produced approximately a 7% increase in VO2max over eight weeks and outperformed continuous moderate running, which produced no significant change [7].

PopulationForty trained men.

Sources[7]

CaveatsOne trial in trained men. The size of the effect in untrained runners, and the transfer specifically to holding a target pace rather than to VO2max, are not established by it.

CA-R-02Weekly distance increase

PrescriptionA distance goal is approached in steps of roughly 10–15% of the target. This is a way to approach a distance, not an injury-prevention measure.

Evidence gradeC — contested

Effect532 novice runners preparing for a four-mile event were randomized to a graded thirteen-week program built on the ten percent rule or to a standard eight-week program. Running-related injury occurred in 21% of the graded group and 20% of the standard group. No difference [8].

PopulationNovice runners, n = 532.

Sources[8]

CaveatsThe most repeated rule in running has been tested and does not do what it is famous for. MVIII uses steps of this size so a target is approached rather than jumped at, and never presents them as protective.

CA-R-03Session volume for a pace goal

PrescriptionRepeats totaling about 1.5 times the goal distance in one session.

Evidence grade— (no controlled trial located)

EffectNo trial isolating session volume for a goal-pace session was located.

Population—

Sources[9]

CaveatsWidespread coaching practice with nothing published isolating the total. The polarized training-intensity meta-analysis cited speaks to how hard versus easy work should be distributed across a week — a 75–80% low-intensity and 15–20% high-intensity split, with polarized distribution favored for VO2peak only in interventions under twelve weeks and in highly trained athletes — and does not speak to how much work belongs in one session. It is cited as the nearest adjacent evidence, not as support. # PART B — Findings

CA-F-01A first pull-up is a long project

FindingAfter twelve weeks of combined strength and aerobic conditioning, four of seventeen college-age women could perform a pull-up. Upper-body strength rose approximately 36% and body fat fell. Successful participants had significantly greater 1RM strength and strength-to-mass ratios [10].

Evidence gradeC

PopulationCollege-age women, n = 17 completing.

Sources[10]

CaveatsA single study in one population. It is the most honest number in this document: general training moves everything in the right direction and frequently does not reach the skill within a season. This is why MVIII shows the rung a member is on rather than a predicted date, and why a target date on a goal is never rendered as a deadline or a countdown.

CA-F-02Pulldown work transfers to pull-up performance

FindingEight weeks of lat pull-down resistance training improved pull-up endurance performance and reduced antagonist coactivation [11].

Evidence gradeC

PopulationRecreationally active male college students.

Sources[11]

CaveatsA single trial in one population. It supports the early rungs of the pull-up ladder being loaded pulling of any kind rather than bar work exclusively, which is what makes the ladder usable by a member who owns no bar.

CA-F-03Straight legs are the harder lever

FindingA hanging leg raise presents a longer resistance arm at the hip than a hanging knee raise and therefore demands more hip flexor and abdominal torque at the same bodyweight.

Evidence grade— (mechanics; no comparative trial located)

Population—

Sources—

CaveatsThis is mechanical reasoning rather than a measured comparison, and it is the entire basis for that rung order. It is listed so the reasoning is visible rather than implied.