saddlebag
Every number in saddlebag is an estimate worked out from what you enter. This page says how each one is worked out and what it rests on: published research, a published model, or saddlebag's own estimate where no study gives the number.
Published model
The watts each slope asks for at your speed: lifting you and the bike against gravity, rolling resistance from the tyres, and air drag. The same sum gives your speed in each gear and the gear-by-grade chart.
Drivetrain losses are left out, so the watts you need on the road run a little higher. Rolling resistance and drag start from typical values for your bike and tyres, not measured ones.
Published model
The pacing plan rides climbs harder and flats easier, because a watt spent climbing saves more time than the same watt on the flat. Descents are coasted.
Flats at 72% of the climbing effort is saddlebag's own split.
Published model
Above your FTP you spend from a limited reserve of work (W′); below it the reserve refills. saddlebag tracks it climb by climb to find where you would run dry and whether you can make it at all.
A simplified version: FTP stands in for critical power, the reserve's size scales with FTP, and it refills at a steady rate rather than the published curve.
saddlebag's estimate
On long rides the watts you can hold slowly drop. saddlebag lowers your FTP by about 4.5% for every 1,000 kJ you have ridden after the first 1,000, and never by more than 20%.
That fade is studied as durability; the rate is saddlebag's own estimate.
saddlebag's estimate
For the finish time, saddlebag assumes you can average about 83% of your FTP over an hour's ride, 74% over two hours and 65% over four.
A coaching rule of thumb, not from a study.
saddlebag's estimate
Easy is up to your FTP, Steady up to 110% of it, Hard above that. Too hard means you could not keep the bike moving up the climb, even in your easiest gear. A route's verdict is its hardest climb.
These cut-offs are saddlebag's own.
saddlebag's estimate
A climb starts where the road rises 3% or more and counts if it gains at least 20 m over 300 m. Elevation is smoothed, and rises under 3 m are not counted, the way Strava and Komoot count climbing.
These thresholds are saddlebag's own.
Data source
The forecast for the hour you will reach each part of the route. The wind along your direction of travel is averaged into one headwind for the watts. Heat changes how much you should drink, not the watts.
From research
8 to 10 g of carbs per kg of body weight a day before a long ride, or 5 to 6 g for a short one. Studies pooled in a 2025 review found muscle glycogen tops out once intake passes about 8 g per kg.
From research
One day of loading fills a trained rider's glycogen. Rides saddlebag expects to take 4 hours or more get a second day for margin.
From research
About 30 g of carbs an hour for rides of 1 to 2 hours, 60 g up to 2.5 hours and 90 g beyond. The gut absorbs about 90 g an hour at most, and only with a glucose and fructose mix; your pantry's mix sets your ceiling.
The fructose shares that move the ceiling to 75 and 90 g are saddlebag's own cut-offs.
From research
Harder riding burns more carbs, but you cannot absorb more per hour, so the on-bike rate follows how long you ride, with a small step down for easy rides. The extra burn is covered by loading and recovery.
Published model
Climbing costs extra energy: your weight and the bike's, times gravity, times the height gained, at about 24% efficiency, about 60% of it from carbs. It shows as extra energy for loading, not extra food on the bike.
24% efficiency and the 60% carb share are typical values, not measured for you.
From research
1.0 to 1.2 g of carbs per kg each hour for about 4 hours after riding, with 0.3 to 0.4 g of protein per kg, but only when your next hard ride is within about 8 hours. Further away than that, normal meals refill you just as well.
From research
Drink to thirst, about 0.4 to 0.8 L an hour, finishing no more than about 2% lighter and never heavier. On rides of 1.5 hours or more, about 300 to 600 mg of sodium an hour. Overdrinking, not underdrinking, is what causes dangerously low sodium.
saddlebag's estimate
saddlebag starts from about 1 L an hour and raises or lowers it with the feels-like temperature and how hard you ride, between 0.4 and 2.5 L an hour.
The steps are saddlebag's own. Sweat rates vary hugely between riders in the same weather, so a sweat test or a watch's sweat-loss reading beats this estimate. The full prediction equation was not used because it over-predicts fluid needs.
Data source
Carbs in cooked rice, pasta, potato, sweet potato and pandesal, and the sodium in Pocari Sweat (about 480 mg a litre).
Published model
Sessions are set in zones by share of your FTP, from recovery under 55% to VO₂max at 106 to 120%. The ladder is adapted from Coggan's power levels, with a sweet spot zone at 88 to 93%.
From research
FTP from a 20-minute test is 95% of your average power. The ramp test estimate, 75% of your best minute, is a common convention.
From research
Four or six weeks in foundation, build and peak blocks, mostly easy riding with a few hard interval sessions. How hard the plan gets follows how hard the target climb is.
The session blueprints and the link from climb difficulty to plan level are saddlebag's own.
Manufacturer guidance
Starting lives for parts (chain 3,000 km, brake pads 2,000 km and so on) and service intervals come from manufacturers' manuals and workshop guides. Your own riding, weather and roads change them, so measure wear where you can.
Last checked 2026-10-19. Forecasts, not guarantees.