Pomodoro technique

In reference to my own practice:

  • Switched back to 25m/5m pomos in Jan 2022 as part of Program - finish my workday in the morning; focusing on longer contiguous blocks (6hr instead of 4hr, no long breaks) and higher-quality deep attention on harder tasks rather than trying to optimize efficiency
  • Switched to 40m/5m pomos on 2021-07-12
    • So goal for the morning block is 6 45m pomos.
    • 8h of working time is 12 pomos
      • 55m of break time spread within those 12
  • Switched to 35m/5m pomos on 2020-01-22
    • Goal for the morning with this block size is 7 40m pomos
    • 8h of working time is ~14 pomos
      • 65m of break time spread within those 14

The standard pomo break routine isn’t as (relatively) inefficient as it seems

On its face, the 25/5 split seems quite inefficient. You’re sacrificing 17% of your total working time to tiny breaks, too small to do anything serious! That feels like a terrible idea until you realize that if you don’t take intermittent breaks, you’ll often waste much more time than that in distraction, rabbit holes, poor focus, etc.

Still, it’s far from clear that 25/5 is the right ratio. In a 4 hour working block…

  • 25/5 has 40m of breaks (83% work)
  • 35/5 has 30m of breaks (87% work)
  • 55/5 has 20m of breaks (92% work)

It’s surprising how small these differences are. In my experience, 55/5 is much, much harder than 25/5 (and substantially less reliable), at least when my next steps aren’t crystal clear. And yet working in that much more difficult regime buys me only an extra 20m of functional working time. It’d be much easier (and more importantly, much more reliable) to achieve that same result by just tacking an extra pomo onto the end of a 25/5 series. In other words, if I want ~3h40m of focused working time, I’d rather sign up for a more reliable, less demanding 4h30m working block than a less reliable, more demanding 4h working block.

Now, in practice, the situation is muddier than this because breaks themselves have overhead. Sometimes it’ll take me a few minutes to get back into the groove after a break. This is mostly mitigated if I forbid myself from doing any unrelated cognitive activity during the break (no replying to anything but the most trivial messages; no reading news / social media; no solving unrelated problems; etc). Sometimes the breaks are very inconveniently timed, and I’m forced to break off a train of thought right in the middle of a sentence. Not sure what the cost of this is, really. Sometimes if there’s only a few minutes left in a block and what remains is difficult, I won’t even bother starting on the difficult thing, because I know I’ll get interrupted. This is explicitly suggested as part of the technique—you’re meant to just plan for what to do next instead. And sometimes that’s helpful advice. But you don’t want it happening too often. Longer blocks make that rarer.

Zooming out: what’s the “ideal”? If I know exactly what to do next for the whole four hour period, as is often the case on programming projects, the best structure is probably one five minute break after two hours, and 30s seated stretch/eye breaks every 30m. I’ll feel physically awful if I don’t walk around at all within a four hour block. So the best case for me is actually ~97% efficiency (not 100%). In a perfect world, I’d dynamically adjust the break distribution according to the character of the task: when programming, use bigger blocks; when doing hard creative work, maybe use smaller ones. But it’s hard to do that reliably.

On the margin, then, I think gains in “efficiency” easier to acquire by doing the following, rather than fussing too much about pomo breaks:

  • more aggressively controlling your calendar to ensure unbroken work time
  • more carefully choosing the best next action
  • more reliably avoiding mental static before working sessions (see It’s hard to hear yourself think)
  • manufacturing intensity through social pressures (collaboration, mentorship, deliverables, etc)

Unless the task is quite heterogenous or quite straightforward, I usually find the difficulty of each marginal pomo increases steeply starting from 7 or so; it often feels quite tough to stay on task for 9 and 10, and the quality of my attention is often pretty poor. Context switching is expensive, and it’s best to work in large contiguous blocks, but if you have 5 or 6 hours on your hands, it may be best to divide that into two big tasks. Practically speaking, I’ve struggled to do this. There’s a lot of “attentional residue” after the first big block; it’s hard to move into the second without a substantial break. So a substantial fraction of the time the second block sort of doesn’t happen, or happens poorly.

It’s hard to do difficult creative work for more than a few hours a day