One tap on the run. Or two afterwards. That's the whole ask.

Mark it at 170bpm without breaking stride

One tap mid-run. Eight things that actually go wrong, three groups, and the whole screen turns the colour of the one you picked for eight seconds, so you know it landed without reading anything.

It works with no signal, and the mark finds its own way onto the run when your phone next syncs. No timer, no counter, deliberately: a number on screen turns logging into a score.

Your history is in there before you log anything

Connect Strava and Absorb imports 24 weeks that day, then keeps reaching back across a full year: the terrain, the weather, the hour and the effort of every run.

Nine runs is noise. Thirty is a shape.

Climbs when fatigued. Hard efforts. Five-hour days. Absorb compares the issue rate in a condition against your own overall rate and tells you which is which.

Every number arrives with its denominator

"4 of 5 runs had issues", never "80%". Below two runs it says nothing at all rather than guess. No score, no grade, no ranking put on you.

Then run a real test on yourself

Absorb offers what the research knows, each option with a citation and an honest evidence grade, several graded down in the source with the reason attached. Pick one, and from then on it tags the runs that qualify.

It reports counts, never a verdict, and it compares you against your own runs without the change, not against an average of other people.

You can call a test off, too. It keeps watching afterwards, because your next runs are what decide whether you were right.

And you find out what you've never tested

Heat you've never trained in. Hours you've never strung together. Absorb marks them before race day does the asking, while there's still time to go and run them.