Long Jump Run-Up Velocity and Board Accuracy Testing

A regional qualifier I worked with jumped her season best on her second attempt, called back a foul by about four centimeters. Her longest legal jump that day was 41cm shorter and missed the automatic qualifying mark by a margin that stung. Nobody had a bad day mechanically. Her final-six-stride velocity had been climbing all block, exactly the trend her training log said to expect. What the log didn't have was any record of how often that faster approach was actually landing inside the board rather than past it.

That's the number this article is built around: not top-end approach speed by itself, but the rate at which a given speed produces a jump that's both legal and well-placed. Call it a speed ceiling — the point past which more velocity starts trading away more legal jumps than it's worth.

The Gap Between Fast and Legal

Most programs I've seen test approach speed on its own testing day, get a clean number, and treat improvement on that number as unqualified good news. It usually is, right up until a jumper starts approaching the edge of what their stride-adjustment system can handle at that velocity. Past that point, the same training block that raised the speed number is also quietly raising the foul rate, and the two trends don't show up on the same chart unless someone deliberately puts them there.

The fix isn't testing speed less. It's testing it alongside a second number — the percentage of reps at a given velocity that come back both legal and inside a reasonable target zone from the board — so a coach can see the exact point where one curve starts working against the other instead of finding out at a meet.

Why the Final Six Strides Decide It

Full-approach velocity is mostly set well before the board — strides 11 through 6 out are where the bulk of raw speed gets built, and that zone is the one most tightly linked to jump distance. The final six strides do a different job. That's the window where the jumper visually locates the board and adjusts stride length by a few centimeters to arrive with the takeoff foot in the right spot. It's a closed-loop correction made under time pressure, and the faster an athlete moves through it, the less time there is to make that correction before the foot lands.

This is why two athletes with identical full-approach speed can post completely different foul rates. One has a stride-adjustment system that scales cleanly with velocity; the other's regulation window compresses past its limit at that same speed and starts missing. A single approach-speed number averaged over the whole run-up can't tell those two athletes apart. Isolating the final six strides — and pairing that reading with what actually happens at the board — can.

The Research Behind the Trade-Off

Makaruk, Starzak, and Sadowski (2015), publishing in Human Movement, studied 70 national-level long and triple jumpers (39 men, 31 women) and split each group by how variable their footfall pattern was across the final approach strides. Among the men, the low-variability group committed significantly fewer fouls than the high-variability group; among the women, the low- and medium-variability groups posted significantly better takeoff-board accuracy than the high-variability group (p < 0.05 for both comparisons). The same dataset found approach velocity measured across strides 11-6 correlated strongly with distance in elite male long jumpers (r = 0.72, p < 0.001). Read together, those two results are the whole argument for a speed ceiling: velocity buys distance, but it's stride-to-stride consistency in the closing strides — not raw speed — that decides whether an individual athlete converts that distance into a legal mark. The authors are explicit that accuracy research in horizontal jumps is still a thin literature, so treat the effect sizes as a strong directional signal rather than a fixed number to design around.

Moura, Moura, Moura, Moura, and Brandão (2024), in Frontiers in Psychology, tracked 10 elite long and triple jumpers (five men, five women, mean age 27.14 ± 4.25) across both training and competition. The counterintuitive finding: these athletes ran a faster approach in competition than in training and fouled less doing it — a 43.79% ± 16.35 failure rate in competition versus 66.53% ± 16.86 in training (p = 0.02, Cohen's d = 0.92, a large effect). That matters for how a speed ceiling gets built, not just whether one exists: a ceiling measured only in low-pressure practice reps can undersell what an athlete handles once arousal shifts things on meet day, but it can just as easily get pushed past its limit if adrenaline sends the athlete faster than the practice data predicted. The authors flag their own sample of 10 as a real constraint on generalizing the exact numbers, and note their variability measure covered only the final six strides, potentially missing earlier-phase adjustments some elite jumpers rely on.

A Protocol for Testing Both at Once

The goal is a set of reps where velocity and board outcome are recorded on the same jump, not on separate testing days that never get cross-referenced.

Run 8-10 full-approach jumps per session, spaced with full recovery (3-5 minutes) so fatigue isn't quietly narrowing the velocity range you're trying to test across. A single session won't cover enough velocity variation to build a useful curve on its own — the real value comes from pooling several sessions' worth of reps, sorted by velocity, once there's a large enough sample.

1. Warm up and mark the approach normally — same checkmark, same starting stance as competition.
2. Record final-6-stride velocity on every rep, not an average for the whole run-up.
3. Film or chalk-mark board contact on every rep, reading toe-to-line distance to within roughly 1cm.
4. Tag each rep clean or not using the legal-plus-target-window rule above.
5. Log the context — practice or competition — since Moura et al. (2024) found that context alone shifts the numbers.
6. Repeat across several sessions until there are at least 20-25 tagged reps spanning a real range of velocities.

Finding an Athlete's Speed Ceiling

Once reps are tagged clean or not, group them into velocity bins (roughly 0.2 m/s wide works for most athletes) and calculate the clean-jump rate inside each bin. Here's what that looked like for the jumper from the opening story, pooled across three practice sessions before her adjustment block:

Set a threshold — 75% clean is a reasonable default for most competitive athletes — and the ceiling is the highest velocity bin that still clears it. For this jumper, that put the ceiling right around 8.9 m/s, not the 9.3+ m/s she'd been chasing in her fastest reps. That's a specific, testable number to train around, instead of a vague sense that speed and fouling both feel like they've gotten worse this block.

Benchmarks: What a Reasonable Ceiling Looks Like

These are starting reference points, not targets to chase for their own sake — Makaruk et al. (2015) found the accuracy relationship held within an athlete's own variability group rather than at one fixed number across a whole population, and that logic applies here too.

The pattern that matters more than any row in this table: an athlete's ceiling should track upward over a training career, but it rarely moves in the same training block as the raw speed gain that's pushing against it. Expect the ceiling to lag the speed number by weeks, not days.

Mistakes That Blur the Ceiling

Coaching the Ceiling Up, Not Just the Speed

Once a ceiling is identified, the answer usually isn't to abandon the faster velocity bin — it's to spend a block raising the clean-jump rate inside it before running through it in competition. For the jumper from the opening story, that meant four weeks of final-4-stride precision work at her 9.0-9.2 m/s bin specifically: short approaches from a fixed mark, scored only on board accuracy, velocity held at the exact range that had been producing fouls.

Her clean-jump rate in that bin went from 50% to 81% over the block without her full-approach speed changing at all. At her next meet, her best legal jump came in at 9.1 m/s on the final-6-stride reading — a velocity that would have fouled roughly half the time four weeks earlier. The speed had never been the problem. It had simply outrun what her board control was ready to handle, and the ceiling told her exactly which pace to train rather than avoid.