Thresholds and zones

Almost every number in TriCoo is measured from a single point — your threshold. Zones are cut from it, the intensity of every session is worked out against it, and the load score follows from that. Which is why the threshold gets so much attention here: get it wrong and you get everything wrong at once.

What a threshold is and why it matters most#

Your threshold is the intensity above which fatigue starts to snowball. Put simply: the pace, power or heart rate you can hold for about an hour.

The cost of an error here isn't hypothetical. If your threshold is 5% too high, the load score of every future session comes out about 9% below the true value in running and cycling, and almost 14% below it in swimming. Worse, the error runs one way: an inflated threshold makes every prescribed session harder than the plan says it is. That's exactly why the platform won't add load until it's sure of your threshold.

Each sport has its own threshold#

FTP is the threshold on the bike, in watts. CV, critical velocity, is the running equivalent: the speed you can hold for a long time. CSS is threshold pace in the pool. LTHR is the heart rate threshold, kept separately for running and for cycling.

These numbers aren't interchangeable: critical power is a model in its own right and doesn't get written into the FTP field. There's no heart rate threshold for swimming at all: heart rate in the water is treated as an unusable signal, and a number there would mean nothing. On the bike, threshold heart rate is usually 5–10 beats lower than for running. If you have neither a power meter nor a threshold heart rate of your own for the bike, the platform takes the running one minus 7 beats — purely in order to work out the load of the ride. That number never becomes a threshold of yours: it isn't saved, it's marked as an approximation, and it disappears as soon as power or a cycling threshold of your own turns up.

How a threshold comes out of ordinary training#

You don't need a test for it. The platform looks through your sessions for near-maximal efforts: it needs at least three usable efforts from three different ranges — 2–5, 8–12 and 15–20 minutes — within the last six weeks. An effort counts as near-maximal if it comes from a session you marked as a race or a test, or rated 8 or higher out of ten. An ordinary session with no such marking feeds into the estimate too, but it stays low-confidence.

In swimming there are two points rather than three: a session of intervals rarely holds three continuous efforts of the right length, which makes the swim threshold the shakiest of the automatic estimates. It has a second anchor, though: if you swim 30 minutes or longer without stopping, the threshold can't be slower than that pace. Such a swim raises the threshold up to that mark — under the usual rules, no more than once every two weeks and by at most 2 % per step — but never lowers it: a steady swim at a comfortable pace shows the pace you chose, not your limit.

One rule sits above all of this: a session unfit for calculating load can't be evidence for a threshold. If a single effort is all that survives the filtering, the threshold won't go up — though it can still come down.

Confidence: how fresh, how confirmed, and how to get out of low#

Confidence answers the question of how fresh and how solid the number is:

• high — a confirmed test or race, or two agreeing estimates no more than 5% apart, within the last six weeks;
• medium — one confirmed effort or an imported number over the same period;
• low — an unconfirmed effort, or an estimate between 43 and 84 days old;
• stale threshold — the estimate is more than 84 days old, or there's no usable data at all.

The second, independent question is whether the number has been checked against your own training. An imported value may turn out to be someone else's default or a guess from years ago, so until it's confirmed it runs your sessions by feel and serves as a one-way guide — “no faster than this” — while no hard zones are prescribed from it. Confirming is easy: the numbers are shown one at a time, with the sport and the units, and an “edit” beside each. A screen you scrolled past doesn't count as agreement.

There's a third state as well — “your threshold disagrees with your training”: threshold heart rate and threshold pace describe two different people. The number isn't corrected on its own then: the platform takes numeric control away from it and waits for a test or a new confirmed estimate.

There's one way out of low confidence: a field test, a race, or a fresh all-out effort. The tests are short — 20 minutes on the bike, 30 minutes running (threshold heart rate is taken from the last 20 minutes), and 400 m and 200 m swims in the pool with a rest between them. Follow the test step by step: each step is its own lap by the button on the watch (warm-up, rest, 400 m, rest, 200 m, cool-down); the threshold is computed only from such an execution, and a single all-out split gives no threshold — you will see its result in the coach thread. Doing tests is optional: the plan works without them.

You don't have to find the moment for a test yourself — the platform suggests it. When a threshold isn't leading the plan by number, hasn't been checked for a while, or a race is close, the coach puts a "Threshold field test" card into the thread with the steps and a button. The test takes the place of a regular session of the same discipline next week — nothing is added on top, the rest of the week stays as is. The suggestion doesn't come during a recovery week, a taper or right after illness, and not before you've done at least two hard sessions of that discipline in the last four weeks: an all-out effort without a habit of intensity isn't needed. Decline it and the platform comes back to the question in a month. Once the test is done, the threshold is computed from its protocol and applied at once as confirmed; a spoiled test — ragged pace, a descent, segments of the wrong length — doesn't become a number, and the platform will suggest a repeat later.

Zones are cut from the threshold#

Internally the platform works with five zones and shows three: easy, moderate, hard. On the bike the boundaries fall at 55, 76, 91 and 106% of FTP. In running and swimming the percentages are different, and that isn't sloppiness: the cost of speed in water rises roughly as a cube, while in running it's nearly linear — identical percentages would mean different things.

Three levels instead of five is a conscious choice. For deciding how to run today you need nothing more, and extra boundaries create false precision: every one of them rests on an estimated threshold, not on a laboratory measurement.

Until there's a threshold, we work by feel#

This is a full working mode, not a placeholder. Each zone has a matching effort rating out of ten: easy work is 2–3 on the bike and 3–4 in running and swimming, moderate is 5 on the bike and 6 in running and swimming. The numbers aren't invented: both scales are anchored to the same reference point — an hour of work at threshold.

Hard zones aren't prescribed at all without a threshold: there's no honest equivalent for them in feel.

Why an increase waits for confirmation and a decrease doesn't#

The asymmetry here is about direction, not size. A decrease is applied at once and quietly: holding on to an inflated number means making every session harder than it claims to be. An improvement is applied on its own as well, but by no more than 2% per 14 days, with a notification, an explanation of what it's based on, and an undo button. Undoing sets a 42-day quarantine: until that expires, this estimate won't come back by itself.

Load that has already been calculated isn't recalculated — a new threshold applies going forward. The rest of the current week, though, is recalculated straight away rather than waiting for Monday: otherwise the plan would stand on one number while what you did was counted against another.

Indoors and out on the road#

The platform always remembers where a session took place. But a separate indoor threshold only appears after six or more comparable confirmations of a consistent difference, not after a couple of hard evenings in a stuffy room. Until then there's one threshold, and overheating shows up in your effort rating and your readiness rather than in a hidden multiplier in the maths.