Period predictor

A date without a margin is not a prediction

Every period tracker on earth will print you a date. Almost none of them will tell you how wrong that date is likely to be, which is the only part that actually helps when you are booking a holiday, choosing a wedding date or deciding whether to worry.

The arithmetic behind the date is trivial: add your cycle length to the first day of your last period, then keep adding. The honest part is the spread. Bull and colleagues analysed 612,613 ovulatory cycles from 124,648 women, tracked with daily basal body temperature and urinary LH rather than recall, and measured how much each woman's own cycle length moved. The answer was a standard deviation of 2.6 days — among women who would describe themselves as regular. Nobody's cycle is a metronome.

Those errors compound. Because each cycle's length is roughly independent of the last, uncertainty grows with the square root of the number of cycles ahead. Next month is worth about three days either way. Six months out, six days. Twelve months out, nine days — and eighteen if you want to be 95% sure. That is not pessimism about the method; it is what calendar prediction is worth, and a tracker that prints a bare date for next August is quietly overselling itself.

What normal actually looks like

The NHS says the average cycle is around 28 days and that 'regular cycles that are longer or shorter than this, from 21 to 35 days, are normal'. Periods are classed as irregular if the gap between them is under 21 days or over 35.

The measured data agree and sharpen it. In Bull's sample the mean cycle length was 29.3 days, and only 13% of cycles were exactly 28. Sixty-five per cent fell between 25 and 30 days, 19% ran 31 to 35 days, 8% ran 21 to 24, and fewer than 1% were longer than 50 days. A 32-day cycle is not a fault to be corrected; it is the second most common bracket there is. The average period lasted 4.0 days, which sits inside the NHS's 2-to-7 range.

Cycles get shorter as you get older

Between the ages of 25 and 45, mean cycle length falls by about 0.18 days per year — roughly three days shorter at 45 than at 25 — and almost all of that comes from the follicular phase shortening. The luteal phase barely changes. Month-to-month variability also shrinks through the thirties, from 2.9 days in the late teens and early twenties to 2.3 days at 35 to 39, before climbing again after 40 and reaching its highest recorded level of 3.1 days at age 45.

This tool predicts from a single average and does not model that drift, so a twelve-month run of dates will creep slightly if you are near either end of the range. That is one more reason the margin on the far rows is wide.

Why a late period is usually nothing

Stress, weight change, heavy exercise, illness, travel, disrupted sleep and hormonal contraception all move periods, and they do it by shifting ovulation rather than by stretching the second half of the cycle. That is why a period arriving four days late usually means you ovulated about four days late, and why the whole thing resets the following month.

If pregnancy is possible, the NHS says you can test from the first day of a missed period, or at least 21 days after unprotected sex, and adds that a positive result is almost certainly correct while a negative one is less reliable. If a test is negative and your period still has not arrived, test again a few days later.

When to take it to a GP

The NHS suggests seeing a GP if your periods are irregular, if they last longer than seven days, if you bleed between periods, if the change comes with weight gain or unusual hair growth, or if you are having difficulty getting pregnant. Three missed periods in a row, with pregnancy ruled out, is also worth an appointment. Irregular cycles can be a sign that ovulation is not happening every month — something no calculator can detect from dates, and something that is often straightforwardly treatable.

This is not contraception

Predicting bleeding is not the same as predicting fertility, and neither is reliable enough to prevent a pregnancy. Wilcox, Dunson and Baird found that even women who described their cycles as regular had a 1 to 6% probability of being in their fertile window on the day their next period was due, and that on every cycle day between 6 and 21 at least one woman in ten was fertile. If you are avoiding pregnancy, use a method designed for it.