About Base: Why a Hybrid Training App Had to Be Built

The problem, stated precisely

An athlete doing both halves is running three programs against two shared budgets.

Recovery is one budget. A hard interval session and a heavy squat day draw on the same account. Nothing tells you the balance, so you find out by feel — which works until the week it does not.

Energy is the other. The surplus that grows muscle is startlingly small: a pound of lean tissue costs roughly 2,500 kcal, and a trained lifter adding 0.5 lb a week needs about 179 kcal a day of usable surplus. A single 40-minute run costs 400 to 500. If maintenance was estimated in a week without running, three runs a week silently erase the surplus entirely.

The athlete reconciles this by hand, in a spreadsheet, on Sunday nights. It works for about six weeks — which is roughly when the reconciliation stops being interesting and starts being a chore, and the spreadsheet quietly stops getting updated.

What Base does about it

Four things, and they are all the same thing viewed from different angles: measure the outcome instead of predicting the input.

1. Maintenance is measured, not estimated. Trend weight change over fourteen days, converted to energy, subtracted from actual intake. That counts the running whether or not you logged it and whether or not your watch guessed well.
2. Endurance zones come from your own threshold, estimated from your own efforts, carrying their own confidence label — and every session is graded afterwards against what it was actually prescribed.
3. Recovery is scored against your own thirty-day baselines, never a population chart, and it produces a decision rather than a number: how many reps to back off today's target, and whether to deload.
4. Every derived number opens its own arithmetic. Not a marketing bullet — a hard constraint. A number you cannot interrogate is a number you will eventually stop trusting, and these formulas are opinionated enough that hiding them would be dishonest.

Principles, and what they cost

Never block on data entry. Every field that can be pre-filled from history is pre-filled. You should be confirming numbers, not typing them. The daily interaction budget is sixty seconds, and features that would break it live in Settings or were cut.

Show the math. Every derived number — maintenance, recovery score, sleep score, trend weight, e1RM, threshold, session grade — has a plain-English explanation of its formula, with the same wording as the comment next to the code that computes it.

No silent failures. If the health store returns nothing, the app says "no sleep data for last night". It does not render a zero and move on. A wrong number is worse than a missing one, and every fitness app that renders a zero for missing data is teaching its users to distrust every other number on the screen.

Never pool the two halves. Strength and endurance are different capacities in different units. They share budgets — showing them against those shared budgets is the whole product — but a blended "fitness score" could not answer either question and would hide the trade-off you are making. Why there is no single fitness score covers this at length.

Local-first. Every read and write hits a database on the device; the app works with the network off indefinitely. Sync is a background reconciliation, never in the path of a user action.

What Base is not

• Not a coach. It will not write your program. It will tell you what the numbers did and what the research says about them.
• Not a session runner. Apple Watch and Garmin already handle live pacing, zone alerts and structured intervals well. Base decides what the session should be and grades it afterwards — the half they do not do.
• Not a social network. No feed, no kudos, no leaderboard. Those are a different product with a different business model.
• Not a medical device. It gives no medical advice and makes no diagnostic claim.
• Not for single-discipline athletes. If you only lift, a lifting app is better. If you only run, a running platform is better. Base earns its place only when both halves are real.

Status

Base is in a TestFlight beta on iOS 17 and later, with Android alongside it. There is no price to announce, no rating to quote and no customer count to cite — and nothing on this site claims otherwise, which is a deliberate choice rather than an oversight. Every number on every page here is either arithmetic you can check or a cited finding you can look up.

What to do next

What hybrid training is is the long version of the problem this app was built for. How Base works is the workflow. Every screen is the app itself.