Irregular cycles: what tracking can (and can't) tell you
Cycle tracking apps are built on a simple statistical idea: look at your recent cycles, find the average length, and predict the next one from that pattern. It works reasonably well when your cycles are consistent. It works a lot less well when they’re not. If you’ve ever had an app confidently tell you your period is due, and then be wrong by a week, this is usually why.
What actually counts as irregular
There’s a wide range of what’s medically considered typical. A cycle length anywhere from about 21 to 35 days is generally within normal range, and some month-to-month variation within that range is completely ordinary. Very few people have cycles that land on the exact same day count every single time.
“Irregular” usually refers to something more specific: cycle lengths that vary widely from one cycle to the next (not just a day or two, but a week or more), periods that stop and start unpredictably, or gaps of many weeks between cycles. Stress, significant weight changes, some medications, thyroid conditions, and conditions like PCOS can all contribute to this kind of variability. So can big life changes (travel, illness, a new medication), even without an underlying condition.
If your cycles have been consistently unpredictable rather than occasionally off by a few days, that’s worth a conversation with a doctor, not just a tracking problem to solve on your own. Tracking can give you useful data to bring to that conversation, but it isn’t a diagnostic tool.
Why prediction gets harder, not just less accurate
Here’s the part most apps don’t explain well: cycle prediction isn’t a single algorithm getting “less accurate” for irregular cycles. It’s a fundamentally harder statistical problem.
A count-back method (averaging your last several cycles and counting forward from your last period) assumes your recent history is a reasonable guide to what’s coming next. For a regular cycle, that assumption mostly holds. For an irregular one, the average of wildly different cycle lengths isn’t actually representative of any of them. If your last four cycles were 24, 31, 26, and 38 days, the average is a number that doesn’t resemble a single one of your actual cycles. It’s a mathematical midpoint of a highly variable pattern, not a meaningful prediction.
This is why an app can feel like it’s “getting worse” the more irregular your cycles are, even though the math hasn’t changed. The tool is doing the same thing it always does. It’s just that the underlying pattern it’s trying to summarize doesn’t compress into a single useful number the way a regular cycle does.
What’s still worth tracking
Prediction being harder doesn’t mean tracking is pointless. It changes what you should actually be using it for.
Pattern-spotting over prediction. Even when you can’t predict the next cycle precisely, a logged history lets you see things a memory can’t: whether your cycles cluster into a couple of different lengths, whether irregularity correlates with something in your life (a stressful season, a medication change), or whether a pattern is actually new versus something that’s always been true for you.
A real record for medical conversations. “My periods are irregular” is a vague thing to tell a doctor. “Over the last six months my cycles have ranged from 24 to 41 days, with two gaps longer than 45 days” is something they can actually work with. This is one of the more genuinely useful things tracking does: not predicting your next period, but building the record that makes a doctor’s visit more productive.
Symptom correlation, independent of cycle length. Mood, energy, cramping, and other symptoms can be worth tracking on their own timeline, separate from how well your cycle length prediction is performing that month.
Honest uncertainty instead of false confidence. The most useful thing an app can do for an irregular cycle isn’t a more confident-looking prediction. It’s a wider, more honest window, and a plain acknowledgment that this cycle is harder to call than a regular one would be. A narrow, precise-looking date range on an irregular cycle isn’t more accurate. It’s just more precisely wrong.
How Lua handles it
Lua’s predictions are built from your own recent cycle history (typically the last several cycles) and presented as plain-language answers rather than a chart you have to interpret yourself: your current phase, your likely next period, your fertile window. When your cycles are more variable, that shows up honestly in the estimate rather than being smoothed over into false precision. More on how that framing works in trying to conceive: what a fertile window estimate is actually telling you.
The tracking underneath doesn’t require your cycles to be regular to be useful. Bleeding days, symptoms, mood, and energy all get logged into one chronological history, which is exactly the kind of record worth having on hand if irregular cycles are something you end up discussing with a doctor.
Lua also imports up to 12 months of period history from Apple Health on first launch, so if you’ve been tracking irregularity for a while elsewhere, you’re not starting the pattern-spotting over from zero. Details are on our period tracker page.
Your logs stay private on your device. No account, no advertising. If you buy Lua+, Apple processes the payment; that receipt is not your cycle history.
Lua is on the App Store for iOS. Tracking is free, with no ads. Lua+ is optional and billed by Apple: it unlocks reminders for every method and full cycle stats.
This article is general information, not medical advice. If your cycles are persistently irregular, or you’re concerned about an underlying cause, talk to a doctor. Tracking can support that conversation, but it isn’t a diagnosis.
More from the blog
One app for your cycle, contraception, fertility and pregnancy.
Private, flexible and designed to adapt with you.