The Math of Check-In Cadence: Why a 30-Day Switch Has Different False-Trigger Risk Than You Think

$image.jpg$

The Question Every Deadman Switch User Eventually Asks

If you set up a deadman switch, the first decision you make is how often you have to check in. Two weeks. Thirty days. Sixty. Ninety. The slider is right there in the settings.

Most people pick a number that feels right and don't think about it again. The number that feels right is usually "thirty days, because that's a month." This is a perfectly fine answer for most people, but the reasoning that produced it isn't reasoning. It's a vibe.

If you actually sit down and ask what the right check-in cadence is, you'll discover it depends on a half-dozen variables you probably hadn't quantified. This post is the cold-blooded version of that calculation — not because we expect you to do the math, but because seeing the math lays bare what you're actually choosing between.

The Trade-Off, Stated Cleanly

The whole problem is a trade-off between two failure modes.

False positive: the switch fires when you're alive. You went on a long trip, missed your check-in, and the system started doing things you didn't want it to do. The cost: notifications go out, beneficiaries get nervous, you have to log in and undo whatever happened. Inconvenient but mostly recoverable.

False negative: the switch doesn't fire when it should. You're incapacitated, missing, or dead, and the system is still waiting because the check-in window hasn't elapsed. The cost: your family is locked out, possibly for months, possibly forever if your circumstances aren't the kind that produce a death certificate quickly. Much harder to recover from.

A short check-in window pushes you toward false positives. A long check-in window pushes you toward false negatives. Where you set the dial depends on how you weigh those costs.

Most people, asked abstractly, would say false negatives are obviously worse. Most people, in practice, choose check-in windows long enough to make false negatives the probable failure mode for them. The vibe wins over the math.

The Variables That Matter

Four variables drive the math. Let's name them.

1. Your baseline lifestyle. How often do you naturally go silent? Are you a daily-poster, weekly-check-iner kind of person, or do you regularly take month-long off-grid trips? The cadence has to be longer than your typical longest gap.

2. Your check-in friction. How annoying is it to check in? If checking in is one tap on your phone, you can have a short cadence. If checking in requires opening a laptop, logging in, navigating menus, it has to be longer to be feasible.

3. The escalation cascade. Most well-designed deadman switches don't fire abruptly at the cadence boundary. They send you reminders. They notify emergency contacts. They give you grace periods. The effective time-to-fire is your cadence plus the escalation tail.

4. Your recovery cost. If the switch does fire and you're alive, what's the actual cleanup? For some people it's a few embarrassed phone calls. For others it's untangling business emails or recovering accounts. The more painful the recovery, the more conservative your cadence should be.

A Rough Probabilistic Model

If you wanted to model this seriously — and you don't, but humor me — the math goes roughly like this.

Let p_silent(t) be the probability that, over the next t days, you'll be silent (no check-in) for reasons unrelated to incapacity. For most adults with normal jobs and lives, p_silent(7) is small — maybe 1-2%. p_silent(30) might be 5-10%, mostly accounting for vacations, hospitalizations, and life events. p_silent(90) for many people is in the 15-25% range.

Now let p_dead(t) be the probability that you become incapacitated or die in any given t-day window. Annual mortality for a 40-year-old is around 0.2%; for a 60-year-old it's around 1%; for an 80-year-old it's 5%. Convert to your window: p_dead(30) for a 40-year-old is about 0.017%, for a 60-year-old is about 0.08%.

The ratio you actually care about is p_silent(t) / p_dead(t). This is, loosely, the number of false positives you'd expect for every true positive.

For a 40-year-old with a 30-day window, the ratio is roughly 5% / 0.017% = ~300. Three hundred false alarms for every real one.

This sounds catastrophic. It isn't, because the escalation cascade catches almost all the false positives before they cause damage — a reminder email, a buddy-check call, a grace period. The system isn't designed to avoid false positives. It's designed to recover from them.

But the math is worth knowing because it tells you the system you're using is, statistically, almost always going to be reminding you to check in. The rare actual fire is just the long tail of a distribution dominated by everyday life.

The 30-Day Default Is Defensible But Not Optimal For Everyone

Thirty days is the default in most deadman switch products, and it's defensible for the median user. It's a window short enough to catch death within a survivable period for most beneficiaries. It's long enough that an active person's normal life doesn't trigger it constantly.

For a 25-year-old in good health who travels a lot, a 60-90 day window might be better. The probability of incapacitation is low, the probability of being out of touch is real, and the costs of false positives are higher than the cost of an extra few weeks of delay on a near-zero-probability event.

For a 75-year-old with a heart condition, a 14-day window is probably better. Mortality is higher, the escalation cascade is more likely to be needed, and the cost of a few extra reminder emails is small.

The right number for you depends on your specific position in the trade-off, and it's worth thinking about for ten minutes, even if the answer ends up being thirty days.

The Escalation Cascade Is The Real Innovation

The single most important thing in deadman-switch design — more important than the cadence itself — is what happens between your missed check-in and the actual fire.

A bare-bones system that fires immediately at the cadence boundary is terrifying. It's the equivalent of a smoke detector that calls the fire department on the first beep. Most fires don't need the fire department. Most missed check-ins don't need the switch to fire.

A well-designed escalation looks like this:

Day 0: Cadence elapses. System sends you a reminder email. "Hey, check in."
Day 2: Second reminder. SMS. "You missed your check-in. One tap to clear."
Day 5: Third reminder, with an explanation of what will happen and when.
Day 7: Notification to your designated emergency contact. "Hey, your friend hasn't checked in, can you reach them?"
Day 14: Final notice. "This is going to fire in 7 days unless you check in."
Day 21: Fire.

This means the effective time from your last check-in to the system actually acting is ~50 days for a 30-day cadence. The cadence is just the trigger; the escalation is the cooldown.

This design is dominated by the cost asymmetry between false positives and false negatives. False positives are cheap to recover from if you're given time and warning. The escalation builds in time and warning. The cadence itself is much less load-bearing than people think.

The Practical Takeaway

Don't sweat the cadence number too much. The escalation cascade is doing most of the work. What matters more, in descending order:

Make sure your escalation contacts are real, reachable, and have your phone number. A buddy-check at day 7 catches almost all false positives. If your emergency contact is your spouse and your spouse is also on the missing trip with you, the buddy-check isn't a buddy-check.
Make checking in low-friction. A check-in that takes 30 seconds and works from your phone gets done. A check-in that requires logging into a web UI gets skipped, and the system fires false-positively because of poor UX, not because of mortality.
Pick a cadence that fits your real life, not an imagined disciplined life. If you legitimately take 60-day trips off-grid, set a 90-day cadence. If you never go more than 10 days without phone access, you can run a 14-day cadence without false positives.
Re-tune annually. Your life changes. Your check-in cadence should change with it.

The math says the system is, statistically, almost always going to be in the middle of a false positive escalation. That's fine. That's the system working as designed. The rare moment it has to actually fire is the moment all the design effort was for. Don't optimize for the common case; optimize for the rare one.

Killswitch implements the escalation cascade this post describes. Tunable cadence, reminder emails, SMS, emergency-contact notifications, grace periods, and a single tap to clear a false positive. The math is on your side; the UX is built so you actually use it. Get started today