Need help now? Support now: crisis lines by country.

Mathematical Institute for Research of Pain · Problem No. 1

Flare or calm first?

A puzzle from the Mathematical Institute for Research of Pain: on a middling day, what is the chance a flare comes before a calm spell? Five states, invented daily chances, one exact answer.

It is impossible for a Dye, with such a determin’d force and direction, not to fall on such a determin’d side, only I don’t know the force and direction which makes it fall on such a determin’d side, and therefore I call that Chance, which is nothing but want of Art …

[John Arbuthnot], Of the Laws of Chance, London, 1692, Preface · Source

Why this question

On a middling day, someone in pain often wonders which way it will go: down towards a calm spell, or up into a flare. The chance that the flare comes first has a name, the committor, and it is one of the quantities the Institute studies. Here the rules are a small invented chain, so the chance can be worked out exactly, by hand.

The rules

Each day is in one of five states, in a row: Calm, Low, Moderate, High, Flare. Overnight the state moves one step down, stays where it is, or moves one step up, with these chances. They are invented for the puzzle, not measured in anyone.

The five-state chainFive circles in a row: Calm, Low, Moderate, High, Flare. Arrows above the circles go one step up towards Flare: Low to Moderate one quarter, Moderate to High one half, High to Flare one half. Arrows below go one step down towards Calm: Low to Calm one half, Moderate to Low one quarter, High to Moderate one quarter. Inside Low, Moderate and High: stays, one quarter. Calm and Flare are marked stop: the story ends there. Arrows above: one step up, towards Flare Arrows below: one step down, towards Calm Calm stop Low stays 1/4 Moderate stays 1/4 High stays 1/4 Flare stop 14 12 12 14 12 14 The five-state chainFive circles in a row: Calm, Low, Moderate, High, Flare. Arrows above the circles go one step up towards Flare: Low to Moderate one quarter, Moderate to High one half, High to Flare one half. Arrows below go one step down towards Calm: Low to Calm one half, Moderate to Low one quarter, High to Moderate one quarter. Inside Low, Moderate and High: stays, one quarter. Calm and Flare are marked stop: the story ends there. Above: one step up, towards Flare Below: one step down, towards Calm Calm stop Low stays 1/4 Moderate stays 1/4 High stays 1/4 Flare stop 14 12 12 14 12 14
The rules as a picture. Each number is the chance of that move overnight; the three numbers for each of Low, Moderate and High add up to one. Calm and Flare are where the story stops.

From Low: down to Calm 12\tfrac12, stays Low 14\tfrac14, up to Moderate 14\tfrac14.
From Moderate: down to Low 14\tfrac14, stays Moderate 14\tfrac14, up to High 12\tfrac12.
From High: down to Moderate 14\tfrac14, stays High 14\tfrac14, up to Flare 12\tfrac12.
Calm and Flare end the story: once either is reached, we stop.

The question

Today is Moderate. What is the exact chance that Flare is reached before Calm?

h(Moderate)h(\text{Moderate})chance asked for=  Pr⁡(=\;\Pr\big(τFlare<τCalm\tau_{\text{Flare}} < \tau_{\text{Calm}}Flare comes first  ∣  \;\big|\;X0=ModerateX_0 = \text{Moderate}today is Moderate)\big)

Here X0X_0 is today’s state, and days are counted from today, day 0. τFlare\tau_{\text{Flare}} is the number of the first day the chain is at Flare, and τCalm\tau_{\text{Calm}} the same for Calm; an end that is never reached counts as later than any day. Pr⁡(A∣B)\Pr(A \mid B) is the chance of AA given BB.

Give the answer as a fraction. Working matters more than the number.

For the keen

  1. Starting from Moderate, how many days, on average, until Calm or Flare is reached?
  2. Starting from Moderate, what is the chance that Flare comes first and is reached on or before day 3, counting today as day 0 (at most three overnight moves)?
  3. If every “stays” chance were set to zero, and the other two chances in its row scaled up in proportion, so each row still adds up to one, which of your answers would change, and why?
One hint

Write h(x)h(x) for the chance of reaching Flare before Calm when today is xx. Look one day ahead: today’s chance is the average of tomorrow’s chances, weighted by how likely each tomorrow is. That gives one equation for each of Low, Moderate and High, with h(Calm)=0h(\text{Calm}) = 0 and h(Flare)=1h(\text{Flare}) = 1 at the ends.

Watch some futures

Pick a starting state and run one possible future at a time. Each line is one invented run of days in this chain, until it reaches Calm or Flare. Watching is not solving: a handful of runs can mislead, which is why the question asks for the exact chance.

What this toy leaves out

In this puzzle the daily chances never change. In real pain they do change: with sleep, with treatment, and over the years. The Institute’s room on this question, Who Gets There First?, adds a context that keeps changing, such as a run of lost sleep, which tilts the odds; the line where a flare and a recovery are equally likely to come first leans with it. A fixed chain is only where the mathematics starts.

Etching: four small children on the ground beside a brick wall. One crouches forward, reaching towards dice lying on the ground, while the other three watch. Three lines of engraved verse run beneath the picture.
Giuseppe Maria Crespi, Four children playing dice, etching, 1705–15. The Metropolitan Museum of Art, New York, gift of Mr. Hans Calman, 1965 (65.568.2), via the Met’s Open Access collection; public domain. Half a century earlier, in 1657, Christiaan Huygens ended his short book on games of chance with five problems for the reader. In the last, two players pass counters back and forth on throws of three dice, and the question is which of them will first hold all the counters: the same question as on this page, with throws of dice in place of days.

Answers and credit

The answer, with the full working, will be posted on this page on Tuesday 24 November 2026.

If you would like credit, post your answer and working publicly and mention the Institute: @painmath@mathstodon.xyz on Mathstodon, @painmath.org on Bluesky, or @PainMathOrg on X. With the reveal we will name the correct public solutions we find, by the name or handle they were posted under. This page has no form and collects nothing: credit comes only from what you choose to post.

Teaching models, not forecasts

The chain on this page is invented. Its numbers describe no one, and nothing here is medical advice or a forecast about any person.

Need help now? Support now: crisis lines by country.

Except where a page says otherwise, the Institute’s own text and figures are shared under CC BY-NC 4.0. Pictures and quotations from others keep their own credits and licences.

Pain math · The invite page · Sign in

Paper