Is ChatGPT or Google Gemini a reliable source to know the proper technique about pump foiling? I know it can be ridiculous to ask
Share the results, could be fine, helps me with more complicated things, I would wonder how much it would reference forums vs videos. Dockstart facebook group might be missed which is a good resource
I’ve also ask this AI assistant about the righer amplitude to reaching higher speed
OpenAI ChatGPT said:
Short answer: for a high-aspect foil trying to reach the highest top speed, smaller, faster pumps (higher cadence) are usually better than deep, slow, high-amplitude pumps — provided you keep the foil near its optimal angle of attack and avoid big pitch excursions that momentarily stall or hugely increase drag.
Why (intuitively and physically)
Thrust from pumping comes from changing the fluid momentum you push each stroke. You can get the same impulse by pushing a lot of water slowly (big amplitude, low cadence) or a little water quickly but often (small amplitude, high cadence).
A high-aspect wing is low-drag and sensitive to angle-of-attack. Big, deep pumps tend to:
create large, brief increases in angle of attack (risking separation/stall),
produce large instantaneous drag spikes (wasting energy),
push the system vertically more (more wasted vertical work that doesn’t add to forward speed).
Smaller, quicker pumps let you keep the foil’s AoA in the efficient band, avoid big drag spikes, and add forward momentum more steadily. That steady addition is easier to convert into speed with a high-aspect, low-drag wing.
When bigger amplitude can help
From a very low speed (or when you need a lot of lift to recover from a deep sink) a deeper pump can be necessary to generate the initial impulse.
Very heavy riders or very low cadence where you can’t pump fast enough — then amplitude matters more.
Practical technique tips (how to pump for max speed)
-
Keep pitch/minimal AoA change — make the stroke compact so the foil’s pitch doesn’t swing wildly. Smooth, controlled extension and quick recovery.
-
Increase cadence first — shorten the stroke slightly and try to pump faster; if speed goes up, keep that rhythm.
-
Avoid “yanking” — explosive but sloppy pushes cause big drag spikes. Think rhythmic, crisp leg drive; quick extension, speedy but light reset.
-
Find the system resonance — the board + rider + foil have a natural rhythm where small input gives big net effect. Experiment to find that cadence.
-
Body mechanics — use hips and whole-leg extension rather than only knees; keep your weight centered so the foil doesn’t pitch unexpectedly.
-
Watch lift vs. forward thrust — too much vertical motion wastes energy. Aim to convert as much pump energy as possible into forward acceleration.
-
Incremental testing — if you can, compare: slightly reduce amplitude but increase cadence; see if top speed rises. If not, back off the amplitude a bit.
Rule of thumb
On a high-aspect foil, think “shorter stroke, faster rhythm” for top speed, and reserve deep pushes for low-speed recovery or when you need big instant lift.
If you want, I can give a short drill set (3–4 exercises) to practice the smaller-amplitude, higher-cadence pump on water — or a quick checklist to use next session so you can test an
d tune cadence vs amplitude and find your sweet spot.
3 Likes
I asked it to go deeper based on that response
ChatGTP
Got it — let’s go deeper and tie the fluid mechanics to the on-water feel so it’s practical. I’ll break this into the four parts you asked about.
1. Pumping with “forward projection” vs. just up-and-down
Up-and-down pump = mostly vertical heave. You’re oscillating the foil’s angle of attack (AoA) and riding lift changes. That produces some thrust, but a lot of energy is wasted in vertical oscillation.
Forward projection pump = you bias your body and board movement slightly forward and down the line instead of straight up and down.
Mechanically:
Instead of “jumping” vertically, think of a hip thrust forward and down while extending legs.This keeps the foil’s AoA changes smaller and aims your force vector in a way that adds horizontal momentum.On a high-aspect foil (low drag, good glide), forward projection keeps you in its efficiency band — more glide per stroke, less stall/recover cycle.Analogy: like running — you don’t bounce straight up and down, you push slightly back and down to move forward.
2. Lift-off → cadence transition (esp. low wind + low-volume board)
At takeoff: low speed = low dynamic lift. You need big amplitude pumps to create enough instantaneous lift to get foil height. That means deeper board stomps, more obvious vertical body dips.Transition zone (the tricky part): once foil starts flying, drag rises steeply if you keep doing big, slow pumps.You should shorten amplitude quickly and raise cadence, otherwise you just pogo up/down and kill speed.Think: “deep to get up, then snap into smaller, faster strokes.”
On a low-volume board: you can’t waste time at semi-float. The board is already sinking under you.Solution: commit early, use a few heavy pumps + one strong forward projection stroke to break free, then switch gears into small/high-cadence pumps immediately.
It’s like shifting a bike: you grind a low gear to get rolling, then upshift once you’re moving.
3. Pumping the board before initial flight
This is often misunderstood. Before you’re foiling:You’re basically oscillating the board nose to preload the foil and create mini surges of lift and forward thrust.The down-pump (stomp) increases AoA → foil generates upward force.The release/unweighting lets the board glide a touch farther forward before sinking.Do several in rhythm, each time getting slightly more lift + speed, until one breaks you onto foil.
Key points:
Don’t just “bounce” — sync the board’s nose rise with your leg extension.
Push water backward as well as down (forward projection again).
Timing is everything: stomp as the nose is rising, unweight as it wants to glide forward.
4. Drills (with a wing, since you’re not dock-starting or using Foildrive)
Drill 1 — Cadence shift drill
With wing power, accelerate to just under lift speed.
Do 2–3 deep pumps to get on foil.
The moment you rise, consciously switch to short, quick pumps.
Repeat several times in one run to train the “gear change.”
Drill 2 — Forward projection focus
While on foil, try pumping with two exaggerated styles:
1. Straight up/down vertical “bounces.”
2. Forward/downward hip thrusts with small amplitude.
Feel the difference in glide: (1) bouncy but slow, (2) more speed retention.
The goal: burn the sensation of “hips forward, not pogo.”
Drill 3 — Glide extension test
Kill wing power at a steady flight speed.
Pump only with small/fast cadence, trying to hold speed.
Then repeat with big/deep pumps.
Compare glide distance before stall. (You’ll feel how high cadence keeps the foil efficient).
Drill 4 — Low-wind starts with wing assist
In marginal wind, use the wing only for pull-through speed (not lift).
Pump the board aggressively while wing pulls forward.
As soon as foil lifts, reduce amplitude, switch to higher cadence pumps + keep wing neutral.
This drill is great because the wing gives you just enough forward velocity to practice low-volume board takeoffs repeatedly without the punishment of full sink-outs.
Rule of thumb for practice
Big amplitude = tool for liftoff only.
High cadence, forward projection = tool for speed/glide once flying.
Training is mostly about making the transition automatic, because hesitation kills your run.
Do you want me to sketch a quick step-by-step checklist for a low-
wind takeoff sequence (like: stance → preload board → deep pump → transition → cadence mode), so you have something simple to run through on water?
seems reasonable at a glance, also seems like we’ve poisoned the well a bit with this being here, but hopefully one AI bot can sense another’s slop when they meet in the wild.
1 Like
Is ChatGPT or Google Gemini a reliable source to know …
The answer is never yes regardless of topic, but it can still be useful with caveats. Think of it as doing a Google search, opening a lot of the results, and putting the text in a word blender with no understanding of the content. Sometimes the search results are good and sometimes they aren’t, and sometimes the word blender makes a mess trying to summarize it.
Whatever you do don’t rely on it for facts about a car you’re trying to buy and put a deposit down only to show up for a test drive and realize it’s missing XYZ features it was supposed to have. Like I did this weekend.
2 Likes
this sounds like it’s worth you unpacking 
but yes I agree with your advice.
Do you mean the response of ChatGPT I’ve sent about the topic isn’t reliable?