Daryl Fitzgerald's Roadman interview contains a memorable saddle-height case: a rider lowered the saddle by about 7mm and went one minute faster on a regular 20km loop with roughly three watts' difference between efforts.
That is what Fitzgerald reported. It is also one uncontrolled case, not proof that a 7mm drop makes cyclists faster or that every amateur's saddle is high.
The useful lesson is the process surrounding the story: Fitzgerald watched the rider, formed a specific hypothesis, the rider made a documented change, then repeated a familiar effort. The unsafe lesson would be to copy the 7mm number without the rider, video, starting coordinate or complete position.
For the wider position assessment, read the Roadman bike-fit guide. For the setup sequence, use how to set saddle height.
What Fitzgerald actually said
In Roadman's interview with Daryl Fitzgerald, Anthony asked what jumped out first when Fitzgerald looked at an amateur. Fitzgerald answered “always saddle height” and said the usual mistake in his experience was too high.
He described riders whose chosen height exceeded what they could control given their hamstring and lower-back range. In his account, they could slide toward the saddle nose and change pelvic movement to reach the bottom of the stroke.
Later, he described the Australian rider's regular time-trial loop. The transcript records Fitzgerald saying: “Your saddle's a touch too high.” He said the rider lowered it by about 7mm and, on the same loop, was one minute faster for approximately three watts' difference.
When Anthony asked about home changes, Fitzgerald advised against 5–10mm jumps and said: “Go one mm at a time.” Elsewhere in the same exchange he referred to 1–2mm changes and acknowledged that rider sensitivity differs.
The interview also contains a useful counterexample to universal rules. Fitzgerald reported some elite riders losing 20–30 watts in an immediate test after moving to shorter cranks, then recovering the watts when their original cranks were restored. Again, that is practitioner-reported experience—not a controlled crank-length trial—but its logic is sound: test the individual and retain a route back.
Evidence label: observation, anecdote or research?
These claims do not all carry the same weight.
| Claim | Evidence type | What it supports | What it cannot prove |
|---|---|---|---|
| High saddles were common in Fitzgerald's amateur fittings | Practitioner observation | The questions he prioritised in his practice | Prevalence among all amateur cyclists |
| One rider improved after a 7mm drop | Individual case reported in interview | A reason to assess and test that rider | Causation or a 7mm rule for others |
| Some riders strongly notice 1–2mm | Practitioner observation | Small changes can be a cautious test | Everyone can detect or benefits from 1mm |
| Shorter cranks cost some clients 20–30W in immediate tests | Practitioner cases | Individual responses can contradict a trend | Population effects of crank length |
The same loop does not automatically mean the same conditions. Weather, wind, pacing, fatigue, equipment, body position, traffic and measurement error can affect a time trial. A three-watt difference also does not establish equivalent physiological effort.
That does not make the story useless. It makes it a hypothesis-generating case rather than a treatment effect.
What saddle-height research says
A systematic review of 41 saddle-height studies found strong evidence for configuring height using dynamic knee-angle measurement and confirmed that height changes alter lower-limb kinematics (Bini and Priego-Quesada, 2022). It found limited evidence concerning injury risk, and height changes smaller than 4% of leg length produced trivial-to-small changes in lower-limb loads with no effect on oxygen uptake or efficiency in the reviewed evidence.
That does not identify one best height. It supports a dynamic, method-specific assessment instead of claiming one inseam percentage or pain pattern finds the answer.
Static measurements can also differ from angles captured while a rider pedals. A study comparing static and dynamic approaches found method-dependent differences, so an angle without its method is ambiguous (Swart and Holliday, 2020).
The broader cycling-overuse evidence is even less certain. A systematic review found conflicting evidence between seat height and symptoms and no strong evidence linking any single bicycle, body or load measure to overuse pain or injury (Visentini, McDowell and Pizzari, 2022).
The responsible conclusion is:
- saddle height changes movement;
- dynamic assessment is preferable to an unlabelled static rule;
- a rider-specific test can be informative;
- one coordinate does not guarantee power, comfort or injury prevention.
Is “too high” more common than “too low”?
Fitzgerald said too high was the usual direction in his practice. That is valuable first-person expertise from a high-volume fitter. The episode does not publish a denominator, sampling method, measurement definition or distribution of fit results, so it cannot establish a general prevalence.
Roadman should therefore write “Fitzgerald said he usually saw saddles set too high” rather than “most amateurs are 5–7mm too high.” The first accurately attributes an observation. The second invents a population rule and precise range.
There is also no contradiction between valuing practitioner pattern recognition and asking for evidence boundaries. The pattern tells us what to check. The assessment tells us whether it applies to the rider in front of us.
A safe home saddle-height test
1. Preserve the original coordinate
Mark the seatpost. Measure from a repeatable bottom-bracket reference to a named point on the saddle, following the seat tube line. Photograph the tape measure and the saddle/seatpost interface.
Also record:
- saddle model, setback and tilt;
- crank length;
- shoe, insole, pedal and cleat;
- any pedal washers or adapters;
- the date and bicycle;
- recent equipment or maintenance changes.
A new saddle or pedal system changes more than the seatpost number. Rebuild the whole foot-to-saddle relationship rather than copying one coordinate blindly.
2. Record normal pedalling dynamically
Place the bicycle securely on a trainer. Position the camera square to the bicycle at approximately joint height. Use the rider's normal shoes, cadence, power and hand position. Film long enough for the rider to settle.
Record the method if you measure knee angle:
- which anatomical landmarks;
- which crank position or event in the stroke;
- static or dynamic capture;
- camera side, height and distance;
- software or manual method.
Do not diagnose height from hip movement alone. Record pelvic control, ankle strategy, whether the rider slides, and whether the movement changes with fatigue or power.
3. Identify a reason to test
A height test is most interpretable when:
- the saddle or seatpost moved during travel, service or a crash;
- a new bicycle did not reproduce a previous comfortable position;
- crank, saddle, cleat, shoe or pedal stack changed;
- the rider cannot control the position under their normal load;
- a symptom or loss of control followed a documented height change;
- a fitter's dynamic assessment supports a small trial.
If the position is comfortable, controllable and stable, “Daryl's client was faster lower” is not enough reason to disturb it.
4. Make one small reversible change
Fitzgerald's 1–2mm approach is useful because it avoids a large blind jump. The exact increment still depends on measurement repeatability, the size of the suspected error and the fitting context. A ruler reading to 1mm does not guarantee that the complete rider position was reproduced to 1mm.
Write down the hypothesis and the move. Do not change setback, tilt or cleats at the same time. Re-tighten the seatpost to the bicycle manufacturer's instructions; use a mechanic for uncertain carbon assembly, torque, damage or slipping hardware.
5. Compare a controlled ride
Use a submaximal trainer session or familiar route before a time trial. Keep power, cadence, hand position and duration as similar as practical. Record:
- control and stability;
- pressure and comfort;
- breathing and ability to hold the intended position;
- symptom onset and progression;
- response over the following day.
An FTP test before and after is not a clean saddle-height experiment: learning, pacing and fatigue can overwhelm a millimetre-scale effect. A maximal test is also a poor first safety check.
6. Restore or escalate
Return to the recorded baseline if the change is clearly worse or creates a new issue. Stop making home adjustments when several variables interact, the result is inconsistent or pain persists.
Use a fitter for complete rider-position assessment, a mechanic for component safety and a clinician for diagnosis or treatment.
What about saddle setback, tilt and cleats?
Saddle height cannot be isolated conceptually from the rest of the fit. Moving the saddle fore or aft changes the effective distance to the pedal. Changing saddle model or tilt changes where the rider actually sits. Cleat and pedal stack change the foot-to-axle relationship.
This is why a good fit record includes every contact point. Do not move the saddle to solve reach without checking what that does to the pedal relationship. Do not change cleats and height together, then attribute the result to one of them.
For shoe and pedal measurement, use the Roadman cleat-position guide. For the “one change every amateur should make” claim, use Roadman's evidence-led decision page.
When pain changes the decision
Pain is not proof that the saddle is high or low. Reduce the aggravating load and record the pattern, training history and equipment timeline. Seek clinical assessment for trauma, persistent or worsening pain, marked swelling, weakness, numbness, locking, giving way, night pain or symptoms away from cycling.
The cycling knee-pain checklist gives the full boundary. A fitter can assess position; they should not use the Fitzgerald anecdote as a medical diagnosis.
The Roadman conclusion
Daryl Fitzgerald gave listeners a valuable practitioner lens:
- he looked at saddle height early;
- he usually saw too-high positions in his fitting context;
- he preferred small changes over large jumps;
- he tested the rider rather than defending a fashionable component choice;
- he used a repeatable loop to collect a response.
Roadman can preserve all of that without claiming every amateur is 5–7mm high or that one 7mm drop caused a one-minute improvement.
The best use of the episode is not to copy the result. It is to copy the discipline: preserve the baseline, make the hypothesis explicit, change one variable, repeat a comparable test and keep the evidence label attached to the claim.
Evidence and limitations
The interview is a primary source for what Fitzgerald said, not a clinical trial. The case details are self-reported within the episode and were not independently audited. Published saddle-height studies support dynamic measurement and show kinematic effects, while performance and injury conclusions remain limited.
This article was transcript-checked on 25 August 2026. Daryl Fitzgerald did not review or endorse Roadman's evidence analysis unless explicitly stated elsewhere.
Sources
- Roadman interview: The 1 Bike Fit Change That Costs Cyclists Watts
- Saddle-height methods and outcomes systematic review (PMID 34706617)
- Static versus dynamic knee-angle assessment (PMID 32022807)
- Cycling overuse-factor systematic review (PMID 35151569)
- Cycling position optimisation systematic review (PMID 39285616)