Cardio base and cardio form
Two numbers that answer the questions a weekly load total cannot: has the training actually built anything, and is the athlete fresh enough to use it today.
This page assumes you are familiar with cardio load and with acute and chronic load.
The gap these fill
The app already tells you what an athlete did — session by session, week by week, and as Fatigue over 7 days and over 21 days. What it did not tell you is where that leaves them.
Two questions came up again and again, and neither had a number attached:
- "Is my athlete in better shape than they were in March?" A cardio load chart shows the work. It does not show the base that the work built, and it certainly does not say whether 118 is a good number for this athlete.
- "Can they handle a hard session today?" Recent load tells you what they have paid. It does not tell you what they have in the bank to pay it from.
Cardio base answers the first. Cardio form answers the second.
| Metric | Where it lives | What it answers |
|---|---|---|
| Cardio base | Reports chart, and the athlete card | How much aerobic base the athlete is carrying |
| Cardio form | Reports chart, and the athlete card | How fresh they are on top of that base |
Both are expressed as cardio load per day, in the same TFF unit as the two fatigue charts, so all four sit on one scale and can be read against each other directly.
Cardio base and cardio form are built from cardio load alone, across every sport. They say nothing about what the training did to muscles, tendons or bones — that is muscular and structural load. This is why the metrics are named cardio form rather than simply "form", and it matters more than it sounds: see the mountain week below.
Cardio base
Cardio base is the training the athlete has accumulated and adapted to — a recency-weighted average of cardio load per day, over roughly the last six weeks.
Recent training counts for most, and older training fades away gradually rather than dropping out on a fixed day. There is no window that "ends"; six weeks ago still counts a little, three months ago counts very little, last year counts for nothing.
Read it as a level, not as a state. That distinction is the whole point of the pair: cardio base is what the athlete has, cardio form is where they stand on top of it.
- it rises through a block of consistent training
- it keeps rising for a while into a taper, because the base is still there after the volume comes off
- it falls when training stops, over weeks rather than days
- it is the number that answers "did the block work"
Week 1 (start of the block) : 62
Week 6 (mid-block) : 88
Week 12 (peak volume) : 118
Week 15 (mid-taper) : 121
Race day : 114
Week 18 (two weeks off) : 92
Note that the highest value is not on race day but a fortnight before it, and that two full weeks off cost less than the last six weeks of the build earned. That is what a slow-moving base looks like — it takes a long time to build and a long time to lose.
It is not the same as "Fatigue over 21 days"
Both are a recency-weighted mean daily cardio load in TFF, and it is reasonable to expect them to be close. They are not, and the gap can be large.
The 21-day chart is a short window with most of its weight on the last few days. Cardio base has about six weeks of memory plus a long tail. On an athlete whose training has been steady the two land within a few percent of each other. On an athlete whose training has changed a lot recently they diverge sharply — on one real athlete, on the same day, the 21-day figure read 25 and base cardio read 260.
That is not an error in either number. The 21-day chart says "almost nothing in the last three weeks"; cardio base says "but there is a great deal of base underneath it". Both are true, and the fact that they can disagree by ten times is exactly why the athlete in that example is worth looking at.
Use Fatigue over 7 and 21 days to see what an athlete is doing now. Use cardio base to see what they have built. When they disagree, the disagreement is the finding.
Why six weeks
Forty-two days is the established convention for peaking, and it is the length at which a taper reads correctly. It is a convention, not a physiological constant, and we do not claim otherwise — different athletes adapt and detrain at genuinely different rates, and we do not fit the number per athlete.
The practical reason for choosing it over a shorter one is separation. On a reference marathon build, a six-week base puts race-day form at +36 against −13 and −8 in the two down weeks of the block — a clear gap. A three-week base puts the same three days at +24, −3 and +0.3, where the down weeks hover around zero and read as "about the same as race day". They were not.
Cardio form
cardio form = cardio base − the cardio load of the last few days
Expressed as cardio load per day, and it is often negative.
- Negative means the athlete is absorbing more training than their base has settled at. This is where an athlete should spend most of a build; it is not a warning.
- Positive means they are carrying more accumulated base than recent cost — freshness. It is the state you want on race morning.
- It goes deeply negative in a peak week and climbs through a taper.
Week 1 : −4 starting out, roughly in balance
Week 6 : −18 absorbing a solid block
Week 12 (peak volume) : −41 the hardest fortnight of the year
Week 15 (mid-taper) : +9 volume off, base still there
Race day : +36 the taper worked
Week 18 (two weeks off): +52 nothing left to absorb — and nothing being built
Race day and a fortnight of doing nothing both produce a positive number. They mean opposite things. How to tell them apart is the most important thing on this page.
It is what the athlete brings into today
Cardio form is the state the athlete starts the day in, before today's training.
This is deliberate. Form is the number you use to decide what today's session should be, so it must not already contain that session — a number that drops the moment the athlete does the workout you are trying to plan is no use for planning it.
Two consequences worth knowing:
- The number does not move during the day. It reads the same at 08:00 and at 20:00. An athlete who logs a long run and then checks their form will see no change, and that is correct.
- Today's training appears in tomorrow's number. The date the value belongs to is shown beside it.
Read the shape, not the number
Form is shown with a short window beside it rather than as a bare figure, because it swings naturally across the days of a week. A Sunday long run moves it by more than most coaching decisions would. A single day's value is easy to over-read; the shape over three weeks is not.
The window is the part that tells you whether form is moving or parked — and that turns out to be the difference between a taper and a problem.
A high positive form is not automatically good
This is the one genuine trap in the metric, and it is common rather than exotic.
A high positive form means one of two opposite things:
- the athlete tapered off a strong base — peaked, ready to race; or
- the athlete stopped training and the base eroded — detrained.
Nothing in the number itself distinguishes them. Measured across thirty real athletes, more than a quarter were reading as strongly fresh at any given moment, and when the list was sorted by form the genuinely peaked athlete came fifth — everyone above them had simply stopped training.
Athlete A : cardio form +28.6 — 93rd percentile of their own base → peaked
Athlete B : cardio form +68.4 — 19th percentile of their own base → detrained
B's number is more than twice A's. B is in considerably worse shape.
Three things on the card resolve it, and none of them is the form value.
1. The comparison against the athlete's own history
Beside cardio base, the app reports where today's value sits against that athlete's own past — a percentile, and how long since they were last this high. This is a personal yardstick, not a comparison with anyone else, which is the only kind that means anything for a number in arbitrary units.
| cardio base percentile high | cardio base percentile low | |
|---|---|---|
| form positive | peaked — the taper worked | detrained — the base has eroded |
| form negative | building on a strong base | early build, or overreaching |
The comparison always states the period it covers — "92nd percentile of the last two years", "highest in 4 months". Read the window with the claim: highest in four months is a real achievement for an athlete in their first season and a modest one for an athlete with five years on the app.
2. The direction of cardio base
The app also reports which way cardio base itself has been moving over the last few weeks — rising, steady, falling, or sharply either way. This is expressed as a percentage per week, so it means the same thing regardless of how much history it had to work with.
| cardio base steady or rising | cardio base falling | |
|---|---|---|
| form positive | taper — freshness is arriving | detrained — the base is draining away |
Note that steady means "not falling", not "peaked". Where the percentile is available it is the better signal; the direction is what you have when it is not.
3. The shape of the three-week window
The window separates the two cases on its own, with no history required beyond the last three weeks:
| Detrained athlete | Peaked athlete | |
|---|---|---|
| Cardio base across the window | −27.9 % | +4.4 % |
| Days with positive form | 21 of 21 | 3 of 21 |
| Form across the window | +123 → +134, flat | −27 → +28.6, crossing zero |
A taper is form moving. Detraining is form parked at a high value while the base drains underneath it. Once you have seen the two shapes side by side they are hard to confuse.
Read cardio form next to the structural ramp
Cardio form describes the cardiovascular side only. It is blind to impact loading by construction, and there is one situation where that blindness is actively dangerous.
An athlete spends a week in the Alps. Long days, low heart rates, thousands of metres of descent.
- cardio load falls, because the heart rate stayed low
- so cardio form climbs, and reads "fresh"
- meanwhile structural load is the highest it has been all year, and the quads are wrecked
Cardio form is not wrong here — the cardiovascular system genuinely is rested. It is simply answering a different question from the one you are asking.
This is why the structural ramp rate is shown beside cardio form rather than somewhere else on the screen. High form together with a high structural ramp is the combination to stop at. High ramp on its own, during a build with form deeply negative, is ordinary and means nothing is wrong.
When the numbers are not there yet
Neither metric appears immediately, and the comparison beside them takes considerably longer than the metrics themselves. This is deliberate: a six-week average computed over a handful of days is not a six-week average, and a number a coach can read is a number a coach will act on.
| Days of history | What the athlete has |
|---|---|
| 1–41 | Nothing. No cardio base, no form, no chart line. |
| 42 | Cardio base and cardio form appear. The three-week window has a single point. |
| ~48 | The direction of cardio base can first be reported. |
| ~62 | The three-week window is fully populated. |
| ~162 | The comparison against the athlete's own history appears. |
So there is a real stretch — roughly the athlete's first four months on the app — in which the card shows a large number and cannot yet say what it means. During that period, read direction and shape, not the value.
For the first few months, cardio base is computed over a shorter memory than its name implies, simply because there is not yet six weeks of history behind every day of it. The number can sit noticeably above or below where it will eventually settle — most of all for an athlete who joined during a taper, came back from a long break, or ramped up hard straight away.
Form's direction is trustworthy from the first day it appears. Its exact value, and the percentile beside it, sharpen as history accumulates.
Blank values and gaps
A blank cardio base means fewer than 42 days of history behind that date. On the Reports chart the line simply starts where the data starts — it is not drawn as zero, because zero would be a claim that the athlete had no base rather than that we do not know.
The chart is never drawn into the future. A monthly chart of the current month is plotted at the last complete day, and periods that have not started yet are left empty. There is no forward projection in this feature.
A gap in the data reads as freshness. If an athlete's watch stops syncing, or they train without recording, the app sees no load — and no load is exactly what a taper looks like. This is rare in practice (of thirty real athletes, one had gone quiet for a week or more) but it is real, and it is another reason not to treat a high number as an achievement without checking the athlete's recent activity.
Where to find them
Cardio base is a chart type in Reports, available daily, weekly or monthly, over ranges up to six months. On a weekly or monthly chart the value shown is the level at the end of the period, never an average across it — averaging a level over a month describes no day the athlete actually had.
Cardio form is a chart type there as well, and the chart is where its shape becomes visible: a taper is form rising and crossing zero, while detraining is form parked high while cardio base drains beneath it. Two numbers cannot tell those apart — two lines can, so plot it against cardio base and read the pair rather than the value. The warning above still applies with a chart in front of you: a line going up is not automatically a better athlete, and the trap is easier to fall into when a rising line makes it look like progress. On the athlete card it appears as a number plus its three-week window.
Common questions
Q: My athlete's cardio form is −40. Should I back off?
A: Not on its own. Deeply negative form is what a genuine training block looks like, and −40 during peak volume is the metric working. What matters is whether it is planned, whether base cardio is still rising, and whether the athlete is recovering between sessions. Form only becomes a concern when it is deeply negative and cardio base has stopped rising — that combination is cost without return.
Q: Cardio base says 260 but "Fatigue over 21 days" says 25. Which is right?
A: Both. The 21-day figure is dominated by the last few days; cardio base has six weeks of memory and a long tail. A gap that size means the athlete has trained very little recently on top of a substantial base — a deep taper, a break, or an interruption. The disagreement is the information.
Q: How is cardio form different from "Relative Fatigue (rTFF)"?
A: They answer neighbouring questions in different ways, and they will not agree. rTFF is a ratio of the last 7 days to the last 21, neutral around 9–11, and available only in daily periods; it catches a sudden change in load. Cardio form is a difference, in the same unit as cardio load, neutral at zero, measured against a six-week base, and available daily, weekly or monthly; it is built to read a build and a taper. Use rTFF to notice that this week was unusually heavy, and cardio form to see where that leaves the athlete overall. Neither is a check on the other, so a disagreement between them is not a fault.
Q: My athlete's form did not change after their long run. Is it broken?
A: No. Form is the state they brought into today, so today's session appears in tomorrow's number. That is what makes it usable for deciding today's session in the first place.
Q: Can I compare one athlete's form with another's?
A: No. These are arbitrary units on a personal scale — an athlete with a large base has larger form swings in both directions. Compare each athlete against their own history, which is what the percentile beside cardio base is for. The one figure that is comparable between athletes is the direction of cardio base, because it is a percentage rather than a level.
Q: Why is there no muscular or structural version of these?
A: For the same reason there is no muscular acute-to-chronic ratio — the interpretation is cardiovascular and does not transfer. Structural tissue recovers over days to weeks rather than in 24–48 hours, so a six-week base against a one-week cost is calibrated to the wrong timescale entirely. Structural load has its own tool, the ramp rate.
Q: The percentile has not appeared and the athlete has been with me for months.
A: It needs about 162 days of history. Until then the app deliberately shows no comparison rather than a comparison drawn from too little data. Use the direction of cardio base and the shape of the three-week window in the meantime — both work from day 42.
Q: Is this the same as TrainingPeaks' CTL, ATL and TSB?
A: It is the same family. Cardio base corresponds to a chronic (fitness) figure and cardio form to a form or balance figure, computed with the same style of exponentially weighted average and the same six-week constant. The input is our cardio load rather than TSS, so the numbers are not interchangeable with theirs, and we report form as what the athlete brings into today rather than including today's session.
Best practices
DO:
- Read cardio form with the percentile beside cardio base — the pair, never the number alone
- Read the three-week shape, not a single day's value
- Expect form to be negative through a build, and treat that as normal
- Check the structural ramp before calling a high form "ready"
- Check that the athlete's data is actually arriving before trusting a sudden jump in freshness
- Use cardio base to answer "did the block work" and the fatigue charts to answer "what are they doing right now"
DON'T:
- Treat a high positive form as an achievement — it is a state, and one of its two causes is bad
- Compare form or cardio base between athletes
- Read cardio base against Fatigue over 21 days as if they should match
- Expect the number to move after today's session
- Act on a single day, or on the first few weeks after the metric appears
Summary: the simple version
- Cardio base is what the athlete has built — a six-week average of cardio load per day.
- Cardio form is cardio base minus the last few days' cost: what they bring into today.
- Negative form is normal; it is what absorbing training looks like.
- Positive form means either peaked or detrained, and the percentile, the direction and the three-week shape are what tell you which.
- Both are cardio only. Check the structural ramp before you conclude an athlete is ready.
Technical notes
For coaches and data enthusiasts:
- Both are exponentially weighted moving averages of daily cardio load, in the Banister / TrainingPeaks form: each day's value carries forward and the day's load is blended in. This is a different smoothing from the 7- and 21-day fatigue charts, which use a fixed window with a tapering weight.
- Cardio base uses a 42-day time constant; the fast side of cardio form uses 7 days. Cardio form is the difference between the two, in cardio load per day.
- The fast side of cardio form is therefore not the number the Fatigue over 7 days chart shows. They measure the same idea with different smoothing and do not agree, which is why it is not published as a separate figure on the card.
- Sport filters do not apply. Like the fatigue charts, these are deliberately whole-athlete — cardio load is the only additive cross-sport currency in the system.
- The value is read at the end of a period, not averaged across it, on every chart resolution.
- A rest day inside an athlete's history is a genuine 0 AU and is averaged in as one. Days before the athlete's first recorded session are not data and are excluded entirely — the calculation starts at their first completion.
- The 42-day gate is on history, not on the calculation. Values below the gate are withheld rather than estimated.
- The percentile ranks today's cardio base against up to two years of that athlete's own daily values, and appears once about 120 rankable days exist behind the gate. The period it covers is always reported with it.
- The direction verdict is a geometric rate — percent per week over up to the last six weeks — with bands at 5 % and 10 % per week. Because it is a rate, it does not drift with how much history was available.
- There is no per-athlete parameter fitting. The 42- and 7-day constants are the same for everyone. Individual adaptation and detraining rates genuinely differ, and fitting them per athlete needs far more data than most athletes have.