A month of 1:15 versus 1:17
I had been brewing pour-over at 1:16 for years without thinking about it, and in January I decided to spend a month alternating between 1:15 and 1:17 on the same beans to find out whether I could actually tell the difference and, if so, which I preferred. The answer turned out to depend on the roast more than on me.
Setup: 15 g of coffee every morning, so 225 g of water for 1:15 and 255 g for 1:17. Same dripper, same grinder setting, water at 93 degrees Celsius, same pour structure of a 45 g bloom followed by pours to 150 g and then to the final weight. I kept the grind fixed within each bag so that only the ratio changed, and I wrote a one-line tasting note in the log before I knew which cup it was, by having the two labeled jars mixed up the night before.
Week one and two, a medium roast from Colombia: 1:15 was richer, with more body and a caramel sweetness, but the finish was slightly heavy. 1:17 was cleaner, more tea-like, and I could taste the acidity more clearly, but it felt thin by the end of the cup. Preference over ten mornings: 1:15 six times, 1:17 four. Not a landslide.
Week three and four, a lighter roast from Ethiopia: here 1:17 won eight mornings out of ten. At 1:15 the cup was dense and the floral notes were buried; at 1:17 it opened up, with the jasmine and lemon that the bag promised actually showing up. 1:15 on this bean tasted like a different, lesser coffee.
So the conclusion is unglamorous: there is no right ratio, there is a right ratio per bean. Medium and darker roasts, which extract easily, want the stronger ratio before they become heavy. Light roasts want more water to let them open, and probably also a finer grind that I did not test this month. The useful outcome is that I now start every new bag at 1:16 and move one step toward 1:15 or 1:17 after the first two cups, depending on whether it tastes buried or thin. The log makes that decision in about two mornings rather than two weeks. Blind tasting, incidentally, was humbling. On three of the twenty mornings I guessed the ratio wrong.