Read more from Slate's Food issue.
To the next batch of dough I added vanilla and substituted palm sugar for white. Palm sugar: a misguided purchase that sat in my cupboard for months attracting ants. Not anymore! Palm sugar is the slightly funky-tasting granulated sap of the coconut palm, and it yielded swarthy, earthy, terrific shortbread. I was delighted. In the space of the next few manic hours I baked crispy Brazil nut shortbread (great), rye shortbread studded with candied ginger (not great), and brown sugar shortbread packed with dates (almost great). The defeats were as interesting as the failures, and my mind was whirring. Why weren't those ginger cookies tastier? (Too much ginger.) How could I have made them better? (Less ginger; try brown sugar.) I found myself lying in bed that night mulling new cookie flavors. It was like playing with paper dolls, creating crazy new outfits for my naked cookie ratio.
Unfortunately, it's hard to advance beyond shortbread with the 1-2-3 ratio, and I eventually grew restless. You can make only so many butter-rich, not-too-sweet cookies before you want to move on to something altogether different. But you can't easily extrapolate snickerdoodles, brownies, or tuiles, because once you start adding leaveners and eggs, you need a more detailed ratio. Or a recipe. These, Ruhlman obligingly supplies in the text of the book, but to look them up seemed like cheating.
Since there was dinner to think about, I moved on to pâte à choux (2 parts water: 1 part butter: 1 part flour: 2 parts egg) in order to make gougeres, airy cheese puffs that Ruhlman recommends you flavor with Gruyère or Parmesan. I went with aged Gouda and threw in some smoked paprika just to see what happened. (Work with ratios for even a day and you start resenting advice.) The tarted-up gougeres were a huge hit, as was the fettuccine (3 parts flour: 2 parts egg), which included both white and rye flours. I expected the experimental pasta to end up chewy and coarse, but it came out like pale gray silk. To finish the meal, I made crème anglaise (4 parts dairy: 1 part yolk: 1 part sugar) using goat's milk and honey and froze it in the ice cream maker. Goat-honey ice cream needs a new name, but it was otherwise practically perfect.
Does this seem like an insane amount of cooking for one weekend? It was, and it was a blast. Ordinarily I find cooking enjoyable and restful; this was exhilarating and slightly exhausting. With mad-scientist fervor, I baked a few cakes (1 part egg: 1 part sugar: 1 part flour: 1 part butter), including what I would consider my crowning achievement, a green tea sponge cake. Flavored with Japanese matcha powder, this was a confection of fluffy, buttery beauty, the color of honeydew, tasting distinctly of tea. Having never before in my life "invented" a cake, I was ridiculously pleased. Yes, I know someone else has probably already invented a green tea sponge cake, but don't rain on my parade. I'd had a breakthrough: After decades of following other people's recipes, the anti-recipe book helped me to invent a few of my own.
It's too soon to know how this thrilling fling with ratios will change the way I cook over the long term. I haven't gone back to them in the three weeks since, but I now find myself studying recipes to see if I can identify their underlying ratios. What is a recipe, after all, but an elaborate ratio someone thought delicious enough to write down? And I've begun to think that Ruhlman's narrow set of ratios might make a less useful starting point for improvisation than a traditional recipe. As I discovered with the cookies, getting fancy requires spending some quality time with the user's manual. If you're going to play around with variations on gingerbread, why not start with a gingerbread recipe?
All that said, green tea sponge cake? I didn't know I had it in me.