Underproofed vs. Overproofed Sourdough Crumb Chart: How to Read Your Loaf (Fool's Crumb Fix)

Why Your Crumb Is a Diagnostic Report

A loaf is the only honest audit a baker ever runs. The crumb does not flatter you: it records the exact hours, degrees and percentages your fermentation actually delivered. Two bakers can follow identical recipes and produce loaves that look like different species, because one of them read their dough at +50% volume and the other read it at +145%.

What follows is a four-mode forensic framework. Each mode has a distinct physical signature, a distinct mechanism, and a distinct numeric prescription. Match the signature first, then apply the number.

The Fermentation Ratio: One Number That Contextualises Everything

Every diagnosis begins by comparing what you did against what the kinetics required. The bulk fermentation engine predicts an ideal window from dough temperature and levain inoculation; the ratio of your actual hours to that window classifies the failure immediately:

• **Ratio below 0.78** — you fermented materially too little. Expect severe underproofing.
• **Ratio 0.78 to 1.22** — timing was within structural bounds. The fault lies in hydration, shaping or the cooling phase.
• **Ratio above 1.22** — you fermented materially too long. Expect proteolytic collapse.

At 75°F with 20% levain, the ideal window is 4.5 to 5.5 hours, so a bulk that ran 7 hours produces a ratio of roughly 1.30 and a verdict of overproofing even if the dough felt "perfectly risen".

Failure Mode 1: Severe Underproofing — The Fool's Crumb

The Fool's Crumb has a unmistakable geometry: a dense, wet, gummy paste band across the bottom third of the loaf, topped by a handful of enormous, chaotic tunnels near the upper crust. It is not a shaping artefact and it is not a hydration problem. It is a gas-distribution failure.

**The physics.** Yeast cells generate carbon dioxide at the gas-cell walls of the dough. Underproofed dough never develops enough cell density to lift the loaf, so the thermal expansion during the bake has nowhere to go. Heat drives the trapped moisture in the loaf's lower core upward as steam, punching violent, irregular caverns through the under-aerated upper mass while the unfermented starch at the base collapses into a wet paste. Meanwhile the unconsumed maltose pool never reaches the crust, so the crust bakes pale and matte instead of deep mahogany.

**Why the top tunnels and the bottom paste appear together.** The base of the loaf sits on a cool baking surface. Unfermented dough there is denser and colder than the crown, so it never warms enough to expand; the crown warms first and blows out. One loaf, two symptoms, one cause.

**Numeric prescription.** - Extend bulk fermentation at your actual dough temperature until the ideal window's midpoint, then confirm with a straight-sided 30g aliquot that expands to exactly the target volume rise — +50% at 75°F, +40% at 78°F, +25% at 82°F. - Boost starter vigour before mixing: feed the seed at a 1:2:2 ratio at 76–78°F and use it only once it has peaked within 6 hours. - Verify all four physical readiness cues before shaping: domed top surface, translucent sub-surface bubbles, a gelatinous marshmallow jiggle, and clean release from the bowl walls. - If your bulk ratio was below 0.78, do not compensate by shaping harder. Add the missing hours first, then re-evaluate the shape.

Failure Mode 2: Proteolytic Overproofing — When pH Breaks the Gluten

Overproofing is not "too much air". It is acid-catalysed structural demolition, and the acid is the operative variable, not the volume.

**The physics.** Wild yeast and lactic acid bacteria continuously deposit lactic and acetic acid as the dough ferments. In the early phase this is beneficial: acid strengthens the gluten network and improves gas retention. Past a threshold, however, the accumulated acid drives the dough pH below approximately 3.9, and at that acidity the flour's own endogenous protease enzymes become active. Those proteases cleave the disulfide cross-links that hold glutenin polymers in a single continuous network. The result is not a fragile dough — it is a dough whose gluten has been chemically cut into disconnected fragments, structurally equivalent to tearing a sheet of paper into confetti.

**The pH model.** The relationship between over-fermentation and dough acidity is well approximated as a first-order decline in pH against hours beyond the ideal window:

This is an operational model rather than a laboratory assay, but it captures the practical threshold well: at 75°F with an ideal bulk of 5 hours, a dough that ran 8 hours lands near pH 3.94 — right at the protease activation threshold — and a dough that ran 11 hours is down near pH 3.7, where glutenolysis is severe.

**Numeric prescription.** - Shorten bulk fermentation to the predicted window at your actual dough temperature and shape the instant the aliquot hits the target volume rise. - Eliminate the conical bowl optical illusion: place a 30g dough sample into a straight-sided 2oz aliquot jar immediately after mixing the levain, and read the rise there. In a flared bowl, waiting for a 50% height increase at 75°F inflates true volume by 67% — an overrun of roughly two full hours. - If your dough temperature exceeds 76°F, cool the mix with crushed ice via the DDT solver and keep the refrigerator at 37–38°F (3°C) to arrest thermal carryover during the cold retard. - After a severe overproof, discard the dough. It cannot be rescued by adding flour; the gluten network is already cleaved.

Failure Mode 3: Hydration-to-Protein Ceiling Collapse

Sometimes the crumb fails for a reason that has nothing to do with time at all: the dough was simply too wet for its flour to hold. This mode is distinguished by a soupy, shape-less loaf that puddles the instant it leaves the banneton.

**The physics.** Every flour has a protein-dependent absorption ceiling. Standard bread flour at 11.7–12.7% protein reaches full polymer saturation somewhere around 72–74% true hydration. Above that, the excess exists as unbound free water acting as a molecular lubricant between gliadin and glutenin chains, so the dough cannot develop elastic recoil during shaping. The starter is where most bakers accidentally breach that ceiling, because a 100% hydration levain smuggles in free water that never appears in the apparent hydration figure:

Working a 500g flour dough at 72% apparent hydration (360g main water) with 100g of 100% hydration levain gives 360g main water plus 50g hidden levain water (410g total liquid) over 500g main flour plus 50g hidden levain flour (550g total flour) — a true total hydration of 74.5% (+2.5 percentage points above the nominal recipe). Now add a 150g levain (30% inoculation: 75g flour + 75g water) instead and the same 72% apparent recipe lands at 435g / 575g = 75.7% true total hydration. The baker changed nothing visible in the main water ratio and pushed the dough 3.7 percentage points higher into the collapse zone.

**Numeric prescription.** - Recompute to 70–72% true total hydration and use the corrected apparent figure for your levain percentage. At 20% levain (100% hydration starter), hitting 71.0% true total hydration requires 68% apparent hydration (341g main water on 500g main flour); hitting 75.0% true total hydration requires 72.5% apparent hydration (363g main water). - Use the bassinage method: hold back 10% of mix water for the salt addition, which slows early hydration and gives gluten time to form before the free water arrives. - Upgrade the flour matrix to 85% bread flour at 12.7% protein plus 15% stoneground whole wheat, whose arabinoxylan pentosans absorb water far beyond what the gluten network can hold. - Dust with rice flour rather than wheat flour at shaping; wheat flour adds more gluten-free starch that behaves like extra hydration.

Failure Mode 4: Retrogradation and Early-Slicing Gumminess

This is the only mode where the fermentation was fine. The crumb fails in the two hours after the oven.

**The physics.** Wheat starch gelatinises between 185°F and 208°F during the bake, forming a viscoelastic gel that sets as the loaf cools. Amylose chains then recrystallise into a rigid, retrograded structure — the same process that makes day-old bread stale. Slice the loaf while its core is still above 110°F and you shear a molten starch gel into tacky, gummy streaks that cling to the knife. If the uncovered bake runs hotter than 480°F, the exterior scorches before the core has set, producing the classic burnt-crust-plus-gummy-centre combination.

**Numeric prescription.** - Bake covered at 450°F (232°C) for 20 minutes, then drop to 425°F (218°C) for 22–25 minutes uncovered until the core reaches 208–210°F (98°C). - Enforce a mandatory 2-hour cooling window on an elevated wire rack until the internal core temperature falls below 90°F (32°C) before cutting. - Slide an empty aluminium sheet beneath your Dutch oven to shield the base from radiant scorching while the core sets. - If your bulk ratio was above 1.22, treat the gumminess as a compounding factor: excess fermentation produces more dextrin, which amplifies the stickiness of a prematurely cut slice.

The Master Diagnostic Matrix

Reading the Chart, Not the Recipe

Two rules make this framework work in practice.

First, diagnose the signature before you diagnose the cause. The pattern of the crumb tells you which of the four mechanisms is operating; only then does the fermentation ratio or the hydration figure become meaningful. A dense base with tunnels has exactly one plausible cause regardless of what the recipe said.

Second, change one variable at a time. If the ratio says underproofed, extend bulk. If the ratio says overproofed, shorten bulk. If the ratio is fine and the crumb is soupy, change hydration. Chasing three variables at once guarantees you cannot attribute the outcome to any of them.

Load the Crumb Doctor, enter your actual dough temperature, your actual bulk hours and the observed pattern, and it returns the verdict code, the fermentation ratio, the corrected hydration figure and the exact prescription for that failure mode — with the matching bulk-rise target baked into the fix.