A baking story (introduction)

Posted on October 4, 2022

It’s a minor stereotype that developers turn into bakers, usually of the artisan bread variety. The mid-2020 “wfh sourdough” boom surely helped that along. It took me a bit longer to get into baking… I baked a few pan loaves in 2020 and 2021, sure, but I don’t think it really took off for me until I got into skillet-broiler pizzas in late 2021.

Anyway, of late I’m into mini baguettes and getting better. It’s been an interesting journey so far.

Misconceptions

I used to avoid baking because, among other things, I did a lot of my cooking by the seat of my pants and there’s no easy way to adjust halfway through baking (or proofing) a loaf of bread. To that end, I thought of baking as a very precise, lab-bench sort of scientific endeavour, where you follow a precise formula until it comes time to compensate for e.g. the hot and cool parts of your oven.

That’s not quite right, but it’s about as right as I expected it might be. The thing is, it’s a lot more fun than I thought.

Beginnings

I started off working from this 50:50 sandwich loaf, which I modified to fit my (smaller) loaf pan. It was… fine. Made some nice sandwiches, didn’t really inspire me to go further.

Later on, once I had a skillet-broiler pizza dough I liked, I started making single-serving focaccia with it. That made some really nice sandwiches! It also, incidentally, gave me some experience with a different kind of dough – different hydration, different cooking times, different ways of shaping it. I didn’t realize it at the time, but that was important context.

Finally, this summer, I started a “thirty days of home-cooked meals” challenge with a couple friends… right when I was developing a craving for banh mi. Fortunately, a friend had a banh mi roll recipe to offer, so I picked up a baguette pan and cooked a batch.

Context and creativity

So having gathered a bit of limited experience baking yeasted breads, and picked up a classic tome of the genre (Forkish’s Flour, Water, Salt, Yeast), the next step was obviously to ease into the french part of the genre with a nice dutch-oven boule, right?

Nah. I don’t entirely remember the circumstances, but several beers in one night around 10:30pm I scribbled down a recipe for a 20% whole wheat, 75% hydration dough which I figured would make good baguettes. I’m pretty sure this was inspired by making one, maybe two, batches of pizza dough with the same whole-wheat ratio and a poolish (a batch of dough pre-fermented at 100% hydration to act as a sort of instant sourdough), and part of my imagination just said “fuck it let’s do this next”. All sourced from the stuff I’d baked before (cooking times and temps from the banh mi rolls, basic ratios from the pizza dough but a bit more water, and so on), just… sprung fully-formed from my drunken imagination.

But did it work?

Ideation

Yeah, the first batch left me with some pretty tasty bread. The vents closed up pretty quick and the baguettes blew out the sides, and the crumb was pretty tight, and otherwise there were a lot of things I wanted to do better. But how?

Well, I bought some more books. I read some websites. I found a gem: the baguette community bake thread on The Fresh Loaf, which featured dozens of people baking some damn fine looking baguettes, way better than anything I’d managed. So I studied.

Their recipes and methods were way different from my own, and to my eye, way different from each other.

I read more. I baked some shokupan, and some fairly indifferent buns using whey from homemade labneh and way too much vital wheat gluten. And I built up some mental models for how things might work.

Iteration

I wanted a more open crumb. How to get it? Well, start with google.

First hits: “Practice. Work gently with the dough. Don’t flatten it, don’t knock the gas out.” I practiced gentle shaping. Didn’t get much more of an open crumb, but we’ll come back to it.

I baked some more. I thought some more. That gas has to expand quite a bit to create an open crumb, right? I need oven spring.

I bought a baking steel. Instead of transferring energy from a hot-ish oven through a sheet pan, I could heat up a chunk of metal that’d pump heat directly into my baguettes. I was really excited to try it.

I got similar results to before, but burnt the bottoms of the loaves. Back to the drawing board.

Imitation

I thought about “work gently with the dough”. I watched some youtube videos on dough mixing techniques, and was struck by how much the dough in the videos… stretched, compared to mine. I remembered the word “extensibility”, and googled some more.

Maybe my dough wasn’t extensible enough. Too elastic, too much protein, strong gluten networks that didn’t want to stretch in the hand or expand in the oven. I can’t make my AP flour lower-protein, but there are other tricks, like 1% nutritional yeast. Oh, and I can drop the vital wheat gluten I put in there to make the whole-wheat play nice from 4% to 2%.

Immediately the dough behaved better, like the stuff I’m watching on youtube. It’s more pleasant to mix and still windowpanes nicely, but it also shapes into baguettes more readily. (I bought a linen couche to help it keep its shape while it proofs.) I bake a batch, and… it’s better, a more open crumb and a bit more oven spring, but nothing like what I’ve seen other people bake.

Inflation

What else affects oven spring? Steam! My method of throwing a few ice cubes onto a quarter sheet on the bottom rack isn’t enough. I throw a couple towels into a loaf pan, cover them with boiling water, and use that to steam the oven for the first half of the bake. Success! …kind of, I end up with an incrementally more open crumb but not much else.

The baking steel covers most of the rack, maybe I need to steam from above. I take out the redundant quarter sheet, steam the next batch with the towel pan on top of the baking sheet. The bottoms scorch; that quarter sheet wasn’t redundant, it was also a heat shield. But the end result: …incrementally better.

Maybe I’m not going deep enough when I score the bread. I read up on the subject and my books and blog posts say the opposite: scoring too deep can create a flap too heavy for the gas inside the loaf to lift, and it’ll just fall back in on itself. Also, I’m scoring at too vertical an angle; I need to be at 20-30 degrees, like sharpening a knife.

I try all that… and get an incrementally better result.

Conclusion

Baking, it turns out, involves a lot of studying other people’s bread, trying out their techniques and recipes, not getting anything close to what you want, and backtracking. I’m building mental models of what’s happening with the bread, trying shit, seeing it kinda-sorta work, and updating my priors. And this all plays out on an interrupted time scale, not a continuous course-correction like stovetop cooking but one where planning and execution and results are separated by long enough that you lose your place a bit.

That explains the stereotype: Baking is like building software, only the results are a lot tastier and happen faster. It’s like therapy for devs in the same way Stellaris is like therapy for directors, only you can eat it.