Which take-off features a video really yields, which camera position and frame rate you need — and why clean optics prove nothing about aerodynamics.
Video is ski jumping's most important feedback tool — and at the same time its most overrated one. It reliably shows you geometry: joint angles, timing relative to the table edge, symmetry. It does not show you what happens aerodynamically. And that co-decides the distance.
The evidence is uncomfortable: lift and drag values of the initial take-off position vary strongly even between elite jumpers and, according to simulation, decisively influence distance [1]. Two take-offs can look practically identical in a freeze frame and still carry differently.
What video does very well by contrast: checking the measured markers. From take-off into the early flight position, jumpers extend the hip to 86 to 109 degrees and the knee to 136 to 171 degrees, incline the trunk from 49 to 41 degrees and angle the skis from 20 to 51 degrees [2]. Those are visible quantities — and therefore your camera's actual job.
A methodological section on analysis — not execution or correction instruction for the hill.
Only assess quantities that are biomechanically defined. The take-off sets initial flight speed, early flight angle, angular momentum and aerodynamic body position, and the force must sit at the optimal moment at the optimal point of the table [5]. Everything else — "looked explosive", "seemed stable" — is impression, not analysis.
First, timing against the table edge: where does extension begin? Second, joint coupling: do hip and knee open together with trunk inclination and ski angle, or in two portions [2]? Third, leg axis head-on: do the knees stay over the feet [3]?
Do the arithmetic once and you will never argue about frame rate again: 90 km/h is 25 metres per second [4]. Every frame per second less enlarges the gap in which precisely your trigger point disappears. And the table already points around 10.5 degrees downwards [4] — a camera not looking perpendicular to the plane of movement measures systematically wrong angles.
Here is where video ends. Lift shortens take-off time by around 11.3 to 14.4 percent, and the correct use of aerodynamic forces before and during take-off is a precondition for a good flight position [1]. None of that is visible in the frame. Anyone inferring good aerodynamics from clean optics has not understood the scatter in the elite data [1].
What the image leaves open, you get through measurement: the imitation jump delivers vertical take-off velocity (R = 0.718 to competition performance) and makes knee valgus quantifiable (R = 0.729 imitation jump, R = 0.685 squat) [3]. Video plus force measurement beats video alone — clearly.
Never judge one jump. Because the aerodynamic properties of the take-off position scatter even among the world's best [1], a single jump is worthless as proof. Only across a series does pattern separate from chance.
Video analysis is a tool for the training dialogue — not a licence to implement technique changes yourself. This guide explains and motivates, it never replaces supervised training in a club or at the hill, and it deliberately contains no execution instruction for take-off, flight or landing.
Separate understanding from execution. A cleanly analysed video does not make you hill-ready and justifies no hill size. Changes to take-off technique belong to the coaching team, to controlled environments and to the hill size you are cleared for — never to a self-experiment.
And the warning that follows from the data: because aerodynamic effect is invisible in the frame [1], a video can certify a change as "good" that is aerodynamically more expensive. Without accompanying measurement and coach judgement, video feedback is half a tool.
Build yourself a standard capture that you do not change for a whole year: the same camera position perpendicular to the plane of movement, the same height, the same frame rate, the same marker at the table edge. The value of video analysis does not arise in the single clip but in comparability across months — and you destroy that with every "let me stand somewhere else today". here means: freeze the measurement system first, then talk technique.
Above all geometry and timing: the start of extension relative to the table edge, the joint opening of hip, knee, trunk and skis [2] and, from the front, the leg axis [3]. Not assessable are aerodynamic quantities — lift and drag of the take-off position scatter strongly even in the elite [1].
As much as possible, and the reason is arithmetic: at take-off big hills see around 90 km/h [4], that is 25 metres per second — at 30 frames per second, around 0.8 metres between two frames. Since the force window at the table lasts only fractions of a second [5], the trigger point disappears between frames at low rates.
Because the relevant fault patterns lie in different planes. Timing and joint opening are visible from the side [2], leg axis and symmetry only head-on — and knee valgus was measurably associated with worse performance in elite jumpers [3]. One perspective always delivers half a diagnosis.
The imitation jump with force measurement: there, vertical take-off velocity correlated with season and competition performance (R = 0.718), and valgus tendency with worse performance (R = 0.729 imitation jump, R = 0.685 squat) [3]. That supplements exactly the quantities missing from the image.
The First Tenths of a Second: From the Table into Flight