October 17, 2026 · 13 min read

Building fitness or wearing myself out? Reading my season in the Form tab

The Progress page in WattLog.pro has eighteen chart cards. I built every one of them — and when I opened the page on my own account a few days ago, I caught myself unable to say what most of them were telling me. So I sat down with my own 2026 season and went through the Form tab card by card, checking in the code how every number is produced. This is the first article in a series about the Progress page.

The Form tab in 30 seconds

The Progress page has two tabs. Performance tells you what you can put out. Form answers a different question: is the training sinking in, and are you overdoing it? Each card on this tab answers one specific question:

If you only have time for one card, make it the PMC. The rest deepen the answer it gives.

PMC: three lines, one story of the season

The PMC (Performance Management Chart) computes everything from one number per day — the sum of TSS from all of that day's rides. Three values come out of it:

• CTL — “Fitness”. An exponential average of daily TSS over about 42 days. It rises and falls slowly. It measures how much training your body has absorbed.
• ATL — “Fatigue”. The same over about 7 days. It jumps after every hard ride and drops after two or three days off.
• TSB — “Form”. The difference: yesterday's CTL minus yesterday's ATL. It tells you how fresh you are this morning.

CTL_today = CTL_yesterday × (1 − 1/42) + TSS_today / 42

The idea comes from Banister's model, in which performance is fitness minus fatigue [1]; Allen and Coggan carried it over to cycling with a power meter [2]. I write more about the chart itself in the article on reading the PMC. This is what my season looked like:

The chart shows the whole season in four acts.

April: from zero to very tired in two weeks

I started the season on 27 March with a CTL of 1. In the first two weeks of April I rode almost every day, so ATL shot up much faster than CTL. On the morning of 14 April my TSB was −37 at a fitness of 13. A red line far above the green one is the typical picture of the start of a season: fatigue arrives in a few days, fitness in a few weeks.

June: the first peak

On 18 June CTL reached 48. That was the best point of the first half of the season — regular ERG rides during the week, longer rides at the weekend.

July: the most important lesson on this chart

From 1 to 26 July I did not get on the bike once. The green line on the chart above falls gently but relentlessly: on 7 August CTL was 19, which means I lost 60% of my fitness in seven weeks. Fitness builds slowly and fades slowly — but “slowly”, with a 42-day average, still means about 2% a day without training.

August and autumn: rebuilding

I came back in August and, after eleven days of rebuilding, rode to St. Anne's Mountain — 144 km on which I fell 1.3 km short of the summit. That morning I had a CTL of 28 and a TSB of −16, and the ride was worth 400 TSS. The next day TSB fell to −58, the lowest of the whole season. Since September I have ridden regularly, 270 to 450 TSS a week, and on 9 October CTL reached 49.4 — higher than in June.

TSB, the “Form” tile

Above the PMC chart there are three tiles with today's values. The TSB tile changes colour: above +10 it is green and says “Peak form”, below −10 it is red and says “Fatigued”, in between it says “Neutral”. This is my season against those thresholds:

Over the 200-plus days of the season the tile was red on 94, neutral on 73 and green on 38. Two things are easy to misread here.

A red tile is not an alarm. A TSB between −10 and −30 is the normal state while building fitness: you are training more than your body has absorbed yet, and that is exactly why fitness rises. The real warning sign is a TSB below −30 for several days in a row — and that is the threshold the alert described below reacts to. I write more about using TSB before an event in the article on TSB and tapering.

A green tile is not always good news. From 26 June to 28 July the tile was green on 32 of 33 days. For more than a month it said “Peak form”, while I was losing 60% of my CTL. TSB rises when fatigue disappears — including when it disappears together with the training. A positive TSB makes sense for a few days before an important ride, not for a month. So always read TSB together with the direction of CTL.

Recovery alerts

Alerts appear above the PMC, but only when one of the rules fires. They are computed from the same CTL, ATL and TSB lines:

“High fatigue load” is the only critical alert, and for me it appeared exactly where it should: in April, when I jumped into the season too fast, and after St. Anne's Mountain, when one ride added more fatigue than the whole week before it.

“Sharp training load increase”, on the other hand, is over-eager after a break. When ATL is close to zero, a 40% increase is one ordinary ride. After a holiday this alert says “you're coming back, be careful” more than “you're overdoing it”. On 15 October, when I collected the data for this article, I had no alerts at all — and that is the most common, healthy state of this card. How to spot overtraining before the chart shows it, I describe in the article on avoiding overtraining.

Training calendar

Below the PMC is the calendar: every day of the season, shaded by training load. An empty square is a day off, not an easy ride.

Below the calendar the app counts the streak of active weeks — weeks with at least one ride. A day off does not break the streak; a whole empty week does. I had 28 active weeks out of 30 in the season, and the two empty ones are mid-July. The longest streak is 15 weeks, the current one 13. This calendar is the simplest card on the page and at the same time answers the most important question for an amateur: do I ride consistently at all? The July gap is visible straight away, without understanding any formula.

The Plan switch above the calendar shows the same against your active training plan: a day completed, missed, a different ride than planned, or a ride outside the plan.

ACWR: how fast your load changes

ACWR (acute:chronic workload ratio) is the average daily TSS of the last 7 days divided by the average of the last 28. A value of 1 means this week was exactly like the last month. Above 1 you are training harder than your recent normal, below it — easier. Days off count as zero, and the card shows nothing until 28 days of history have built up.

The chart says the same thing as the PMC, only more bluntly. It went highest after the July break: when the 28-day average is close to zero, every ride gives a ratio of 2, 3 or 4. On 19 August, the day of St. Anne's Mountain, ACWR was 2.57 — my week was two and a half times harder than my monthly normal. From this chart's point of view, the ride was a bad idea before I even left the house.

On 15 October I was at 0.82 — just above the bottom of the usual range, because the last week was lighter than the monthly average. Had it dropped below 0.8, the app would have said “Below your normal”, and that would not have been an alarm either: during a taper or a recovery week ACWR is supposed to drop — that is the point.

One honest caveat. The ACWR bands come from injury research in team sports [3, 4], and the method itself is heavily criticised today — among other things because a ratio of two averages behaves strangely with small numbers [5]. My August is a good example. Read ACWR as a description of how fast your load changes, not as an injury-risk verdict. The app says so under the chart, too.

The other three cards

The Form tab has three more cards that answer narrower questions. I describe them in more detail in the article on endurance; here only briefly:

• Load by energy system splits CTL and ATL into the aerobic, glycolytic and neuromuscular systems, according to each ride's time in power zones. It is a model, not a measurement — read the proportions between the lines, not their height. I cover it in more depth in the article on training proportions.
• Cardiac drift compares the ratio of heart rate to power in the first and second half of a steady ride lasting at least 30 minutes. Below 5% means a good aerobic base. More in the article on cardiac drift.
• Heart rate trends show the average and maximum heart rate of every ride against your own normal from the last 10 readings. A run of rides clearly above normal at a similar effort points to fatigue, illness or heat.

What I take away from this

Next in this series, the Performance tab: the power curve, critical power and the rider profile — in other words, what kind of rider I am according to my own data.

Sources

1. Calvert TW, Banister EW, Savage MV, Bach T. A systems model of the effects of training on physical performance. IEEE Trans Syst Man Cybern. 1976;SMC-6(2):94–102.
2. Allen H, Coggan AR, McGregor S. Training and Racing with a Power Meter. 3rd ed. Boulder, CO: VeloPress; 2019.
3. Hulin BT, Gabbett TJ, Blanch P, Chapman P, Bailey D, Orchard JW. Spikes in acute workload are associated with increased injury risk in elite cricket fast bowlers. Br J Sports Med. 2014;48(8):708–712. doi:10.1136/bjsports-2013-092524
4. Gabbett TJ. The training—injury prevention paradox: should athletes be training smarter and harder? Br J Sports Med. 2016;50(5):273–280. doi:10.1136/bjsports-2015-095788
5. Impellizzeri FM, Tenan MS, Kempton T, Novak A, Coutts AJ. Acute:chronic workload ratio: conceptual issues and fundamental pitfalls. Int J Sports Physiol Perform. 2020;15(6):907–913. doi:10.1123/ijspp.2019-0864

← All blog posts