I keep a reading log, which surprises exactly nobody who knows what I do for work, and it is, obviously, a spreadsheet. Title, author, start date, finish date, rating, and one column most people find unnecessary: pages per night, calculated automatically from the other two.

I didn’t set out to build a productivity metric for a hobby that’s supposed to be the opposite of productive. It started because I wanted to know, honestly, whether I was actually reading less than I used to or just felt like I was, and the only way I trust an answer like that is with an actual number instead of a feeling. Turns out I was right to be suspicious of my own feeling. I read almost exactly the same amount now as I did four years ago, somewhere between twenty and forty pages most nights, it’s just distributed differently, longer stretches on weekends and less on weeknights than my mental image of myself assumed.

The pages-per-night column has taught me a few uncomfortable things. Dense hard SF, the Egan novels especially, tanks the number hard, ten or fifteen pages some nights on Schild’s Ladder, and that used to bother me until I actually looked at the data and realized the slow books are consistently the ones I rate highest afterward. Slow reading isn’t a failure mode for me, it’s practically a leading indicator. Ficciones has some of the lowest pages-per-night numbers in the whole log and I’d reread it before half the books that flew by faster.

What the log can’t capture, and what I’ve stopped trying to force it to capture, is the nights where I read the same page four times because my brain was somewhere else entirely, work stress, an argument, whatever. Those nights show up as low page counts and normal ratings once I actually finish the book, and there’s no column for “the book was fine, I just wasn’t present for it,” so I’ve let that go. Some data you collect and some data you just accept you can’t clean.

Dana asked me once if tracking this stuff ruins the reading itself, turns it into another metric to hit. I don’t think it does, mostly because I never look at the log while I’m actually reading, only afterward, out of curiosity, the same way I’d check a forecast against actuals once the quarter’s closed. It’s not steering the experience. It’s just satisfying a question I apparently can’t leave alone: was that as much as I thought it was, or does it just feel that way. Usually the spreadsheet and the feeling agree. When they don’t, the spreadsheet wins, every time, and I’ve made my peace with that being a slightly strange way to read.