What you actually burn, once the endurance work is counted.

Why the usual activity multiplier fails a hybrid athlete

Almost every calorie calculator asks one question about activity — sedentary, lightly active, very active — and multiplies your resting rate by a fixed number. That works acceptably for someone whose week is roughly the same every week. It fails for a hybrid athlete for a specific reason: the thing that varies most in your week is exactly the thing the multiplier flattens.

Consider two weeks for the same 180 lb athlete. Week one: four lifting sessions, no running. Week two: four lifting sessions and 30 miles of running. That running is roughly 2,300 kcal, or about 330 kcal a day. Both weeks get "very active" and the same target. One of those targets is wrong by more than the entire surplus you were trying to run.

What this calculator does instead

It splits the day into four parts and adds them:

Running is estimated at roughly 1 kcal per kilogram per kilometre — the long-standing figure that makes running cost depend on how much of you there is to move and how far, and barely at all on how fast. Lifting is treated as about 5 METs averaged across the whole session including the rest between sets, which is lower than most people expect and is the honest number: a set is 30 seconds and the rest is three minutes.

Where this number is still wrong

It is a prediction, and predictions of energy expenditure carry error bars of several hundred calories between two people of identical size. The sources of that error are real and unmeasurable from a form:

1. Non-exercise movement is the most variable component in all of human energy expenditure, and it adapts. Eat less and you fidget less, often without noticing.
2. Mifflin-St Jeor was fit on a general population, not on athletes with above-average lean mass. It tends to under-predict for muscular people.
3. Training economy improves. The same ten-mile run costs a trained runner meaningfully less than an untrained one.

The number that is not a guess

There is a way to get maintenance calories without predicting anything: measure the outcome. Over a two-week window, take how much your smoothed bodyweight changed, convert it to energy at roughly 3,500 kcal per pound, and subtract that from what you actually ate. What is left is what you actually burned.

weightChangeLb  = trendWeight(today) - trendWeight(14 days ago)
energyBalance   = weightChangeLb * 3500 / 14        // kcal/day
maintenance     = avgDailyIntake - energyBalance

That figure needs no assumption about your activity level, no MET table and no equation fit on other people. It needs a fortnight of honest logging and eight weigh-ins, which is exactly what Base measures automatically and re-derives every day.

What to do next

Take the number above as a starting point, eat at it for two weeks while weighing yourself most mornings, then compare what the scale did to what you predicted. The gap is your correction, and it is worth more than any calculator.