Estimating ride time from power and FTP

Why "20mph average" is the wrong tool for a hilly route, and what a segment-by-segment simulation does instead.

Last updated: September 22, 2026

The problem with a single average speed

The easiest way to guess ride time is to pick one average speed for the whole route and divide distance by it. It's also usually wrong the moment a route has real terrain. A rider who holds 20mph on flat ground might hold 8mph on a 7% grade and 35mph on the descent back down -- and none of those numbers are close to 20. Average them incorrectly (by distance instead of by time) and the error compounds further, because you spend far more time per mile climbing than descending.

A route with 500ft of climbing per 10 miles and one with 3,000ft of climbing per 10 miles can have nearly the same total distance and wildly different finish times. A flat average speed has no way to tell them apart.

Simulating the ride instead of averaging it

Ride Planner instead walks the route segment by segment -- typically every few hundred feet, using the actual elevation profile from RideWithGPS or your GPX file -- and, for each segment, works out how fast you'd travel at your target power given that segment's grade. The physics behind each segment accounts for:

• Gravity -- the power cost of climbing scales with grade and your total weight (rider + bike + kit), which is why the same 200 watts feels completely different on a 1% grade versus an 8% grade.
• Rolling resistance and air drag -- drag grows roughly with the square of speed, which is a big part of why descending fast takes little power (gravity is doing the work) while holding a high speed on the flats takes a lot.
• Headwind or tailwind -- computed along the route's actual heading at each point, not just guessed from the start line (more on this below).
• Coasting on descents -- on a steep enough downhill you're not pedaling at your target power at all, you're limited by how fast you're comfortable descending, which the simulation accounts for separately from the calorie estimate.

Summing the time across every segment gives a total ride time that reflects this specific route's actual shape, not a generic guess.

Where FTP and intensity fit in

FTP (functional threshold power) is roughly the highest power you can sustain for about an hour. Riders rarely target 100% of FTP for a multi-hour ride -- that would be a very hard effort -- so Ride Planner asks for FTP plus an intensity fraction (for example, 0.7 for a moderate all-day pace) and multiplies them to get your target power in watts. If you already know the wattage you plan to hold and don't want to think in terms of FTP at all, you can skip straight to entering a fixed target power instead.

Neither number needs to be exact. The simulation is most useful for comparing "how much longer will this route take than my usual ride," or "what pace do I need to average to finish before dark" -- both of which are far more sensitive to the route's terrain than to being off by 10 watts on your target power.

What it doesn't model

The simulation doesn't know about traffic lights, unplanned stops, drafting in a group ride, or how you're feeling that day. A rotating paceline option adjusts for the aerodynamic benefit of sharing pulls in a group, but anything outside the physics of you, your bike, the road, and the wind is left to your own judgment. Treat the result as a well-informed estimate for planning purposes -- start time, daylight, fueling stops -- not a guaranteed finish time.