The Starch Swap: What Potato Taught Me That Tapioca Couldn't

I've been slacking on my blog updates, but I've still been hard at work learning about and developing my own gluten free bread.

Not the kind built on xanthan gum and commercial yeast and a prayer. The kind that actually ferments, that smells like something, that has a crust you have to mean it to cut through. The kind that treats gluten-free grain as its own thing rather than an apology for not being wheat.

It has taken a lot of loaves to get here.

For most of that process, I was using tapioca starch as my primary starch component. Tapioca is popular in GF baking for good reason. It binds well, it gives you a chewy, cohesive crumb, it behaves predictably, and it's neutral in flavor. When tapioca starch gelatinizes, it forms a stretchy, elastic gel that mimics some of the textural qualities gluten would otherwise provide. It holds the loaf together. For a lot of GF bakers working without the structural safety net of gluten, that reliability is genuinely valuable.

And tapioca, it turns out, has a water retention problem.

What Tapioca Starch Actually Does

Tapioca starch is aggressive about holding onto water. It gelatinizes at a relatively high temperature and forms a tight, elastic gel. That's great if you want chew, and not so great if you want an open crumb and a crust that stays crisp after cooling. What I was getting was a loaf that looked promising coming out of the oven and then slowly, as it cooled, turned soft and gummy. The moisture that should have driven off during the bake was being held hostage inside the crumb.

This is partly a chemistry problem and partly a fermentation problem. Tapioca's water-binding capacity works against one of the most important things happening in a naturally leavened GF loaf: starch conversion by lactic acid bacteria. LAB need access to available starches to do their work: breaking complex carbohydrates down into simple sugars, producing lactic acid, transforming the flavor and digestibility of the bread in the process. When tapioca is locking moisture into a tight gel, it's also making the environment less hospitable to that conversion. You end up with a crumb that's dense and cohesive but hasn't fully fermented… And that shows up as gumminess, as a starchy aftertaste, as a loaf that feels like it's missing something even when it technically worked.

For a wheat sourdough baker, a soft crust is annoying. For a GF sourdough baker trying to prove that this bread is worth making, it's a problem. Students in a class don't just taste the bread. They look at it, they handle it, they form an opinion about what GF sourdough can be before they even take a bite.

I needed a crust that stayed a crust.

Fermentation Matters More Than People Think

Most GF bread recipes treat fermentation as optional at best and irrelevant at worst. Commercial yeast goes in for leavening, and that's considered sufficient. But fermentation in a naturally leavened loaf isn't just about rise — it's about transformation. Lactic acid bacteria break down starches into simple sugars, produce acids that develop flavor and extend shelf life, and make the finished bread easier to digest. That process takes time, the right temperature, and a microbial environment that's been allowed to mature.

In GF baking, this matters even more than it does with wheat. GF flours are largely composed of starches that are harder for the body to process when raw or under-fermented. LAB activity is doing genuine nutritional and textural work of softening the crumb structure, unlocking flavor compounds, converting what would otherwise be a dense starchy mass into something that tastes and feels alive. Skipping or shortchanging fermentation doesn't just affect flavor. It affects the entire character of the bread.

The problem is that starch breakdown is a slower, more metabolically demanding process than gluten fermentation. LAB need more time and a more stable environment to do it properly. Most GF recipes aren't built around that reality, they're built around speed and predictability, which is understandable, but it means the fermentation is doing a fraction of what it could.

The Switch to Potato Starch

Potato starch gelatinizes at a lower temperature than tapioca and doesn't hold onto water with the same grip. It produces a lighter, more open crumb (more crumbly, less elastic) and it releases moisture during the bake rather than trapping it. That means better oven spring, better crust development, and a loaf that actually looks like what it is when it comes out.

It's also a friendlier environment for fermentation. Because potato starch gelatinizes gently and releases moisture freely, the LAB have better access to available starches throughout the ferment. The microbial activity can actually reach what it needs to convert. In practice, this means a more complete fermentation, a less starchy flavor, and a crumb that reflects what the starter was doing rather than fighting against what the starch was doing.

The trade-off is that potato starch can make a crumb more delicate. Too much and you risk something that crumbles rather than slices. Which is why I'm not ruling out a 50/50 blend. I wanted to see what all potato could do first before I started splitting the difference.

The dough felt different immediately. Softer, lighter, more workable. Less like I was fighting it.

What the bake told me

The loaf came out with real dome height, a crust that held the banneton pattern, and color that actually looked like bread. The crumb was open and even with no gum line. No dense, wet band at the bottom that tells you the moisture never left. The crust was crunchy out of the oven.

It softened slightly on cooling, which tells me there's still moisture work to do. Nothing a longer uncovered bake at the end won’t solve. And the flavor read slightly starchy, which I'm reading as a mild underproof. The lactic acid bacteria didn't quite finish converting all the starches to simple sugars before the dough went in. That's a fermentation timing issue, not a starch issue, and it's fixable.

But the structure was there. The crumb was there. The crust, for the first time, was actually there.

Why this matters beyond the recipe

Here's what I keep coming back to: the reason tapioca became so dominant in GF baking is the same reason a lot of GF baking choices get made. Someone looked at wheat bread, identified what was missing, and reached for the ingredient that most efficiently replaced it. Tapioca gives you chew. So tapioca went in.

That's substitution logic, and it's not wrong. It produces real results. But it doesn't ask what these flours and starches are actually good at when you work with them on their own terms.

Potato starch doesn't try to be gluten. It just does what it does: releases moisture cleanly, gelatinizes gently, gets out of the way and lets the fermentation show up in the crumb. For a bread built on LAB activity and starch conversion and a sorghum starter that's been slowly developing its own ecosystem, that's exactly what I needed.

The swap wasn't just practical. It was a philosophy check.

This formula is still evolving. I'll keep documenting it here as it does. But if you've been baking GF sourdough and finding that your loaves look right and feel wrong: soft where they should be crisp, dense where they should be open, it might be worth asking what your starch is holding onto, and whether you actually want it to.