Start dates are all it needs
Write the first day of bleeding, one per line, as 08/01/2026. Add the days of bleeding and a note if you want them kept: 08/01/2026 | 5 | cramping first day. Both extras are optional, order does not matter, and a date entered twice is merged with its notes joined. A cycle is measured from one start date to the next, so two dates give one cycle and the first date in the list never has one of its own. Three cycles is the least worth reading and six gives an average that stops jumping every time you add an entry. A gap of more than ninety days is treated as a missing entry rather than a cycle and left out of the average, which is flagged so you can check.
Why the next date is given as a window
Cycle length is not a fixed property. It moves with sleep, illness, stress, weight change, travel and medication, and it moves inside the same person from month to month. A single predicted date built from an average therefore misses routinely, which is why three figures are shown instead of one: the last start plus the average, the span between the shortest and longest cycle you have recorded, and the span within one standard deviation of the average. The last of those is usually the practical one — it is where cycles have tended to land if the variability you have already seen carries on. Treat any of them as a window to pack for or schedule around, not a date to depend on.
This is not a fertility or contraception tool
It deliberately does not estimate ovulation and does not display a fertile window, and that is a design decision rather than an omission. Ovulation shifts by several days between cycles in the same person, so working backwards from an averaged cycle length locates it only loosely. Sperm can remain viable in the body for several days, which widens the days that matter well beyond any single estimated one. The combination makes calendar arithmetic a poor guide in both directions: relying on it to avoid pregnancy fails at a high rate compared with other methods, and relying on it to time conception wastes cycles. If you are avoiding pregnancy, a clinician can go through methods that work. If you are trying to conceive, ovulation testing and a clinician will tell you far more than any date arithmetic on a page like this.
What the counts against ranges are
Four specific things are counted: cycles under 21 days, cycles over 35, bleeding of 8 days or more, and a spread of 10 or more days between the shortest and longest cycle. Those lines are commonly printed but they are not the only ones. The FIGO 2018 classification of abnormal uterine bleeding puts normal cycle frequency at 24 to 38 days, narrower at the bottom and wider at the top than the older 21 to 35 range used here, so a 37-day cycle is flagged on this page and would not be under that classification. FIGO also sets the regularity boundary by age — within 9 days at 18 to 25 and at 42 to 45, within 7 days at 26 to 41 — and this page does not ask your age, so it prints the spread and lets you apply the right line. A count is not a finding, one cycle outside a range is unremarkable, and no absence of flags means anything is fine. What the counts are genuinely good for is walking into an appointment able to say when something changed, with the dates in hand.
Questions people ask
Does this predict when my next period will be?
It does arithmetic: the last start date plus an average of the cycles you recorded, shown alongside the range your own cycles have actually covered. That is not a prediction and it will miss. The range is the useful part, and it is honest about being wide when your cycles have been.
Can I use it to work out my fertile days?
No, and it deliberately does not show them. Ovulation shifts by days between cycles and sperm survive for several days in the body, so a fertile window derived from average cycle length is not reliable enough to act on — not for avoiding pregnancy and not for timing conception. Speak to a clinician about contraception, and use ovulation testing with clinical advice if you are trying to conceive.
My cycles are all over the place. Is an average any use?
Less than for someone regular, which is exactly why the spread and the standard deviation are shown next to it. A wide window is not a failure of the tool; it is an accurate description of what your dates have done. If the spread stays at 10 days or more, or cycles keep landing outside the usual ranges, that is itself worth taking to a clinician, and this page counts it rather than interpreting it.
Is anything stored?
No. Everything is computed in your browser, nothing is transmitted, and a refresh clears it. Given what this record is, that is the whole point. To continue the log, copy the text output into a note on your own device and paste it back next time, adding new start dates at the bottom; the format reads straight back in. For an appointment, the printed sheet is easier to hand over.