Why a longer block is not automatically better
Demanding work has a warm-up cost. The first five to ten minutes of a block go into loading the problem back into your head, which is why a 90-minute block does not give you three times the output of a 25-minute one — it gives you one warm-up instead of three.
But that only pays off if you can actually hold attention for 90 minutes. If you drift at minute 40, a long block buys you 40 minutes of work and 50 minutes of guilt. The right interval is the longest one you can finish while still present, not the longest one you admire.
Interruptions shorten the right answer
If you are on call, on a support rota, or working somewhere people walk up to your desk, long blocks will be broken from the outside and each break costs you the warm-up again. Shorter blocks are more robust: an interruption lands between blocks more often than inside one.
The same logic applies to a noisy home. Match the block to the longest stretch you can realistically defend, not to the stretch you wish you had.
Treat it as an experiment, not an identity
Pick one interval and run it for a week — all of it, breaks included, which is the part most people skip. Then note two things: how often you finished a block without drifting, and how you felt at the end of the day. Drifting often means the block is too long. Finishing every block and ending the day with energy left means you can go longer.
Most people land somewhere between the famous numbers. 40/10 and 60/15 are perfectly good intervals that nobody wrote a book about.