Fast Isn’t Enough: Rethinking Pole Vault
What three studies reveal about the gap between running quickly and clearing the bar.
Everyone in the vault knows a faster run gives a higher bar. The number attached to that belief is real and it's larger than most people expect: 0.54 metres of peak height for every extra metre per second. What almost nobody says out loud is where that height gets collected. Two thirds of it arrives above the grip, on the pole, well after the runway has stopped mattering. Speed sets what's available. It doesn't hand any of it over.
What speed actually buys
The cleanest measurement of this comes from Linthorne and Weetman, who took one experienced male vaulter through seventeen jumps and varied his run from two steps to sixteen. That produced approach velocities from 4.5 to 8.5 m/s, a deliberately wide spread, because you cannot see the shape of a relationship from the narrow band an athlete normally trains in.
It came out linear, and the slope carried a confidence interval of plus or minus three centimetres. For a single-athlete study in a technical event, that is a remarkably clean number.
Figure 01 / Interactive
Peak height against approach velocity, and where the height comes from
Approach velocity / tap to change
At 4.5 m/s the athlete is still running, planting, jumping and inverting. The basic actions never disappeared anywhere across the range. What changed was the scale of everything.
A metre and a half faster, and the relationship has not bent. The line stays straight right through the range, which is unusual and is what makes the slope so easy to use.
Still linear. But the energy lost at take-off has been climbing the whole way up, because a faster arrival means a harder collision to absorb.
The athlete still gained energy on every single vault. The size of the gain shrank slightly as the run got faster. Speed has a price attached, and it gets paid at the plant.
Here's the split that changes what you do with the information. At competition speed, only about a third of the extra height came from gripping higher. The other two thirds came from push height, the distance travelled above the grip.
The runway creates the opportunity. Most of the height is collected during inversion and push, which is the phase furthest from where the speed was generated.
From Linthorne & Weetman (2012), seventeen jumps by one experienced male vaulter, filmed in the sagittal plane, run-up length varied from 2 to 16 steps. Height gains in the panels are calculated from the reported slope of 0.54 m per 1 m/s. This is a single-athlete study. The authors compared their vaulter against six others of varied ability and found a similar pattern, which is reassuring but not the same as a group finding.
The two thirds nobody schedules
That split quietly reorganises a training week. If most of the return on a faster approach is collected above the grip, the athlete who spends every session chasing runway speed is investing heavily in the smaller share.
Push height is not one skill. It's the swing off take-off, the compression of the athlete against the loaded pole, the turn, the inversion, and the extension off the top hand as the pole recoils. Each of those has its own timing and each can be trained. What they have in common is that they need a pole, a pit and a coach's eye, where run-throughs need a runway and a stopwatch. The cheaper thing to schedule is not the thing carrying two thirds of the benefit.
None of this argues against developing speed. Speed sets the ceiling, the relationship is real, and it is measurable to within three centimetres per metre per second. The argument is about what happens to that speed afterwards. An athlete who arrives at the box faster but inverts no better has bought the opportunity and left most of it on the mat.
Same speed, different result
Which brings in the study that makes the whole picture awkward. Theodorou and colleagues filmed twelve elite vaulters, seven men and five women, at an indoor international championship, using a panning camera at 300 frames per second.
The design is the strong part. Rather than comparing good vaulters with less good ones, which mostly tells you that good vaulters are good, they compared each athlete's own successful and failed attempts at the same bar height. Same athlete, same bar, same session. Only the outcome differs.
Figure 02 / Interactive
One athlete, one bar height, two outcomes
Attempt / tap to change
Step length baseline
Step velocity baseline
Step frequency, pole-carrying leg baseline
The successful attempt. Nothing here is remarkable, which is the point. The three variables sit in their usual relationship to each other.
Step length no significant change
Step velocity no significant change
Step frequency, pole-carrying leg significantly higher
Two bars have not moved. The athletes were not slower on the attempts they missed. They arrived at the box at equivalent velocity.
What changed was how that velocity was assembled. Step frequency on the pole-carrying side went up, and the ordinarily stable trade between frequency and length came apart in the last strides. The authors call it a perturbation in the interaction of the two variables.
Schematic. Bar lengths show direction and the pattern of change, not measured values. From Theodorou et al. (2023), 12 elite vaulters at an indoor international championship, panning camera at 300 fps. The study found no inter-limb difference in absolute step length or step velocity between outcomes, and significantly higher step frequency on the pole-carrying leg at failed attempts. Asymmetry values themselves did not differ significantly between successful and failed attempts.
High velocity in the final phase of the approach is essential, but it is not the sole determining factor for a successful attempt. Theodorou et al., Journal of Human Kinetics, 2023
Why the carrying side is the fragile one
There is a mechanical reason the pole-carrying leg is where this surfaces.
As the pole comes down through the last few metres, gravitational torque on the system rises and the combined centre of gravity of vaulter and pole shifts forward, ahead of the ground support. That leaves less time for the swing leg to recover into a normal sprinting position and drive the foot backwards before it lands. Carrying the pole on one side also restricts the arm swing, which feeds into leg coordination and hip rotation. Elite vaulters absorb most of this. They trade a little step length for a little step frequency and hold velocity steady, which is exactly what the data shows on the attempts they clear.
The accommodation has a limit. When it's exceeded, the authors suggest the consequences land at take-off: compromised positional requirements, more variable take-off foot placement, or interference with the vertical impulse that transfers energy into the pole. All of which means an athlete who is running well and missing heights may not need more speed or more aggression. They may need a pole drop that leaves the last two steps alone.
The same variability turns up in injury histories
The three source studies never cite each other, so this connection goes unmade. Edouard and colleagues collected competition biomechanics from national-level French vaulters, including U17 and U20, and matched it against twelve months of retrospective injury history.
The parameters associated with a higher proportion of injury history cluster in the same place. Speed between ten and five metres out. Last stride adjustment. Stride length variation. Ground contact time. Lower grip-hand height at take-off. Plus weekly training volume, which surprises nobody.
Read that alongside Theodorou and one practical theme emerges. The final steps of the approach are where clearances are lost and, on this evidence, they are also where injury history concentrates. Stride adjustment and stride length variation are not neutral technical curiosities. They describe an athlete arriving at the box having improvised, and improvisation at that speed carrying a five-metre pole is expensive in both currencies. This is an association in a retrospective design rather than a causal finding, so hold it loosely. It does point the same way as everything else here.
The pole is a variable, not a fixture
Both of the studies above treat the pole as a constant. Warburton's doctoral work at the University of Western Australia does not. Working with eight elite male and female vaulters, he built an instrumented plant box with a three-dimensional load cell to measure forces at the base of the pole, plus a purpose-built rig to characterise how individual poles bend, store and return energy.
Figure 03 / Interactive
What an instrumented plant box found, and what it couldn't
Finding / tap to change
A pole behaves differently depending on how you load it. Testing showed more energy both stored and lost under slow dynamic bending than under quasi-static bending. The same stick, two loading conditions, two answers.
A flex rating is real information and it is not a description of behaviour under an actual vault. The number on the pole is measured in a rig nobody jumps in.
The method used to estimate body segment inertial parameters significantly changed the energy calculations. Which means the answer you get about an athlete's energy through the vault depends partly on which population model you borrowed to describe their body.
Warburton concluded that subject-specific measurement is preferable for elite athletes, whose proportions diverge meaningfully from population norms. Among the variables examined, total vaulter energy at maximum pole bend emerged as affecting performance.
Pole ground reaction force could not reliably separate successful vaults from unsuccessful ones. Force profiles looked similar across genders and across differing outcomes. Several variables expected to correlate with peak height didn't.
It remains a genuinely useful analytical and feedback tool. It is not a verdict. And a theme runs through the whole thesis worth the attention of anyone who reads group means: aggregating athletes obscured meaningful individual findings.
From Warburton (2015), doctoral thesis, University of Western Australia, eight elite male and female vaulters using a custom instrumented plant box with a 3D load cell. Eight athletes. The thesis is explicit that its own value lies in the method and in the individual analyses rather than in generalisable group conclusions.
One system, three instruments
Strip the three studies back and they are describing the same object from different angles. The vault is an energy transfer problem. Kinetic energy from the approach becomes elastic strain energy in the bending pole, which returns as gravitational potential energy that lifts the athlete over the bar. Simple to state, difficult to do.
Linthorne and Weetman quantify what velocity supplies to the system. Theodorou shows that rhythm preserves it, and that velocity delivered in a disrupted pattern cannot be fully used. Warburton shows that pole and athlete are one system, and that the instruments you choose to measure either one will decide what you are able to know about both.
What follows for the next training session
Nothing here requires a load cell or a 300 fps camera. Most of it is a question of where attention goes.
- Keep developing approach velocity. It sets the ceiling and the relationship is real, quantified and linear across a wide range.
- Treat the pole drop as its own technical skill with its own session time, not as the lead-in to the plant.
- Watch the last two steps specifically, and watch the pole-carrying side. A rising step frequency there is a signal, and it will not show up in a run-up timing gate.
- When an athlete is running well and missing heights, check rhythm before adding speed or aggression. More of both is the intuitive answer and on this evidence it is often the wrong one.
- Give inversion and push proportionate time, because that is where two thirds of the benefit of a faster run gets collected.
- Treat pole selection as a performance variable rather than a paperwork one, and expect flex ratings to describe less than they appear to.
- Log take-off foot placement variability across a session. Stride adjustment and stride length variation showed up in the injury association data, so this is worth tracking for two reasons at once.
- Analyse athletes individually. Warburton reported that aggregating athletes obscured meaningful individual findings, and nothing in this literature is large enough to hand you a group answer worth applying blind.
Which is the honest place to finish. One experienced male vaulter across seventeen jumps. Twelve elite vaulters at a single indoor championship. Eight athletes on an instrumented plant box. That is the evidence base for an event contested worldwide for more than a century, and it is small enough that these findings are better treated as good hypotheses about your athlete than as facts about them. The one number here that arrives with real confidence is the slope, plus or minus three centimetres, from a sample of one.
References
- Linthorne NP, Weetman AHG (2012). Effects of run-up velocity on performance, kinematics, and energy exchanges in the pole vault. Journal of Sports Science and Medicine 11(2):245-54.
- Theodorou AS, Panoutsakopoulos V, Exell TA, Cassirame J, Sanchez H, Kotzamanidou MC (2023). Success to clear the bar in elite pole vaulters is affected by step frequency perturbation. Journal of Human Kinetics 88:41-49.
- Warburton TK (2015). Energy and pole ground reaction force contributions to pole vault performance. Doctoral thesis, University of Western Australia, Perth.
- Edouard P, Sanchez H, Bourrilhon C, Homo S, Frère J, Cassirame J (2019). Biomechanical pole vault patterns were associated with a higher proportion of injuries. Frontiers in Sports and Active Living 1:20.
- Hanley B, Gravestock HJ, Hopkinson M, Paradisis GP, Merlino S, Bissas A (2022). Kinematics of the final approach and take-off phases in world-class men and women pole vaulters. Frontiers in Sports and Active Living 4:835659.
- Frère J, L'Hermette M, Slawinski J, Tourny-Chollet C (2010). Mechanics of pole vaulting: a review. Sports Biomechanics 9(2):123-38.
Note on citations: the step frequency study appears in Journal of Human Kinetics volume 88, not 87 as it is sometimes listed.