Brown Butter Banana Cookies bake log
What was actually made, versus what the card says. Companion to the cookie recipe.
- Bake001
- Date19 Sep 2026
- Oven350 °F
- Time15–17 min
- Count18 × 75 g
- Verdictfantastic
Bake 001 19 September 2026
Three things moved off the card: the banana over-reduced and was corrected with oat milk, the chocolate was short and made up with semi-sweet chips, and the portion came in at less than half the jumbo the card was written around. The smaller portion was the useful accident — 18 cookies is enough to run chill time as a controlled variable inside one batch.
The headline: they came out excellent, and the banana never caramelized. The formula is proven, and there's a whole flavor layer still unclaimed. That's the lead for Bake 002.
As built
| Card | Actual | |
|---|---|---|
| Bananas, raw | 2 | 2 |
| Banana roast | 15–20 min | 12 min — short |
| Reduction vessel | small skillet | saucepan, reused from the butter |
| Paste, finished | 130 g, amber, jammy | 105 g, pale yellow |
| Correction | — | +25 g oat milk, back to 130 g |
| Butter, raw → browned | 265 → 220 g | confirm browned weight |
| Butter at creaming | ~65 °F, plastic | held — scooped clean, no greasy spread |
| Sugars | 200 g + 60 g | 200 g + 60 g |
| Egg + yolks | 1 + 2 | 1 + 2 |
| Flours + malt | 305 + 40 + 25 g | 305 + 40 + 25 g |
| Soda / powder / salt | ½ / ½ tsp / 8 g | ½ / ½ tsp / 8 g |
| Dark chocolate | 227 g @ 70% | 120 g @ 67% |
| Chips | optional | ~107 g semi-sweet |
| Flaky salt | sprinkle | heavier — deliberate |
| Total dough | ~1,310 g | ~1,350 g |
| Per cookie / count | 163 g / 8 | ~75 g / 18 |
Mass reconciles within about 3% — scoop-average rounding plus a heavier hand on the chips.
Method deltas
- Banana under-roasted, then steamed rather than reduced. Twelve minutes against a 15–20 minute window, then cooked down in the deep saucepan left from the butter. The paste hit the right weight and held a furrow but stayed pale yellow — no caramelization, no Maillard. The banana's flavor ceiling was never reached.
- Over-reduced to 105 g, corrected with 25 g oat milk. The solids were never lost — all 25 g of shortfall was evaporated water — so the correction was complete.
- Oat milk rather than dairy. Hydration 1:1, and its maltose matches the malt. Arguably the better choice, not the fallback.
- Chocolate short — made up with semi-sweet chips, roughly 20 g more sugar and less bitterness.
- Flaky salt increased as the counterweight.
- Portion at ~75 g. 350 °F, 15–17 min, six per half-sheet, 5 min on the sheet.
- Staggered chill, turning the batch into a controlled test. Uncovered the first hour, covered after.
| Group | Chill | Baked | Result |
|---|---|---|---|
| A / B | 1 h, 12 h | 19 Sep | ✓ Excellent |
| C | 24 h | 19 Sep, eve | pending |
| D | frozen | 20 Sep | pending |
Banana : flour ratio
| Banana bread, Bake 001 | 1.58 : 1 |
| Banana bread, card | 1.39 : 1 |
| These cookies, 130 : 370 g | 0.35 : 1 |
A cookie carries roughly a quarter of what a loaf can. If a future bake reads cakey, this is the number to move.
Predictions and outcomes
| # | Prediction | Outcome |
|---|---|---|
| 1 | Banana-forward but not banana bread — malted banana caramel | hold until a caramelized paste |
| 2 | Chewier and denser than the original | ✓ chewy center, crisp edges |
| 3 | Less spread than the original | ✓✓ thick, domed, rolled edge — arguably overshot |
| 4 | Browns faster at 350 than the original at 375 | pending |
| 5 | Deeper, darker color overall | ✗ even medium golden — the pale paste added no color |
| 6 | Malt reads clearly; oat reads as background nuttiness | pending |
| 7 | Softer by day two | pending |
| 8 | Sweeter from the chips; heavier salt covers it | ✓ the salt covered it — keep the heavy hand |
| 9 | Two-textured chocolate | ✓ works — promote to technique |
| 10 | More set edge, less gooey middle at 75 g | ✓ an improvement |
| 11 | Group A spreads more and reads paler than Group C | pending |
| 12 | Risk: greasy spread from warm butter | ✓ avoided |
| 13 | Risk: oat milk restores water but not flavor | confounded — retest |
Verdict
Excellent cookie. Chewy center, crisp edges, sweet and salty in balance, two chocolate textures. The formula holds and 75 g is the better format — now the card's default alongside the jumbo.
What it isn't yet: a banana cookie at full strength. The paste was pale, and every banana-flavor question is confounded by that. Nothing in the formula needs changing to find out — just the roast and the pan.
Corrections made to the card
- Judge the roast by the peel, not the clock.
- Reduce in a wide skillet, never a saucepan — and brown the butter in that same skillet first.
- Two gauges on the paste: 130 g and the color of apple butter.
- Over-reduction is normal and cheap to fix — moved into the method.
- 75 g row added to the portion table as validated.
- Bar-plus-chips documented as a deliberate option.
Next-bake hypotheses
Bake 002 — one variable: the banana. Roast the full 15–20 minutes to a black, split peel, reduce in a wide skillet to 130 g of amber jammy paste, change nothing else. Then, in order:
- Still wants more banana → 25–30 g ground freeze-dried banana with the dry.
- Too sweet → run the full 227 g of 70% bar before touching sugar.
- Spread too tight → flatten the scoop to ¾″ before changing flour.
- Cakey → paste down to 115 g.
- Malt lost → malt to 40 g, flour down 15 g.
- Groups A/B/C read the same → the chill can come down.
- Group D matches C → scoop-and-freeze becomes the default workflow.
130 g amber paste, 220 g browned butter, 12–24 h chill, 350 °F.