The Method, In Full
This is the Methods page. Every Method Showdown post links back here instead of re-explaining how we test — so if you’ve landed here from another post wondering what a “3” on the texture scale actually means, you’re in the right place.
We don’t publish a “best” method because it seemed right in one kitchen, one time. We publish a winner because it beat 2-3 named contenders under the same conditions, on the same scorecard, every time. Here’s exactly how that scorecard works.
Controlled Conditions
Before any scoring happens, every contender in a showdown is made under matched conditions:
- Same frying temp and time — measured with a thermometer, not “medium-high heat,” because a 20°F swing changes browning and structure more than the substitution itself does.
- Same spice mix — so flavor differences come from the base, not the seasoning.
- Same oil type and same portion size — surface-area-to-volume ratio changes hold time on its own, so batch sizes get matched.
If a contender needs a genuinely different process — say, one dough needs a longer rest — that gets documented as part of its method, not treated as a controlled variable. The goal isn’t identical processes; it’s a fair fight where the only real difference is the substitution being tested.
Net Carbs
Net carbs are calculated the same way for every ingredient, every time: total carbohydrates minus fiber minus half of any sugar alcohols used (erythritol is treated as contributing negligible net carbs, per standard practice, since it’s not metabolized the way other sugar alcohols are). Figures come from ingredient nutrition data — brand labels where a specific product is called for, USDA data where it’s a generic ingredient like almond flour or besan. We link the specific source in-post when a number might raise an eyebrow, rather than asking you to trust a bare figure.
This is directional, kitchen-level math, not a lab assay. It’s precise enough to compare Method A against Method B fairly. It is not precise enough to be a medical or diabetic carb-counting reference — if you need exact numbers for a health condition, calculate your own from the specific brands in your pantry.
Texture Scale (1–5)
Texture is judged on structural integrity and mouthfeel relative to what the traditional version is supposed to do — a papdi is supposed to shatter, a samosa pastry is supposed to blister and crack, and the scale reflects distance from that job, not distance from “tastes like wheat.”
| Score | What it means |
|---|---|
| 1 | Fails its basic job. Crumbles, gums up, or won’t hold shape before it’s even plated. |
| 2 | Holds together but the mouthfeel is clearly wrong — dense, gummy, or leathery where it should be light or crisp. |
| 3 | Passable. Does its job on first bite but degrades fast, or has a texture flaw a taster would mention unprompted. |
| 4 | Close to the traditional version. Minor tell (slightly denser, slightly less shatter) but nothing that breaks the experience. |
| 5 | Indistinguishable from — or better than — the traditional version in a blind bite. |
Structural Hold-Time Test
This measures how long a food survives its actual job before failing — sogginess for anything meant to hold sauce or yogurt, structural collapse for anything meant to hold a filling.
Method: the finished item is plated with its real-world companion (chutney, yogurt, filling) at time zero, under open-air conditions at room temperature. It’s checked at fixed intervals — 5, 10, 20, 30, 60 minutes — until it fails. “Failure” is defined per snack before testing starts (for papdi: visible sogginess and loss of snap; for samosa pastry: filling breaching the shell or the shell losing its crack), so we’re not making the failure line up with whichever contender we want to win.
Hold time is reported as a range (e.g. “10–30 min”) corresponding to the last interval it passed and the first it failed, rather than a false-precision exact minute.
Taste / Flavor Authenticity Scale (1–5)
| Score | What it means |
|---|---|
| 1 | Doesn’t read as the target snack at all. |
| 2 | Recognizable as an attempt, but an off-flavor (bitterness, strong nuttiness, eggy notes) dominates. |
| 3 | Reads as the snack, with a noticeable but not dominant substitution flavor. |
| 4 | Close enough that a taster not looking for flaws wouldn’t flag it. |
| 5 | Blind-indistinguishable from the traditional version. |
Panel process: scored blind by a 2–3 person taste panel where possible — tasters know they’re comparing versions of the same snack, but not which method is which until after scoring. Contenders are labeled A/B/C/D and served in randomized order per taster. When a solo test is unavoidable, that’s disclosed in the post rather than presented as panel data.
What the Scorecard Doesn’t Claim
A method can win “best texture” and lose “best flavor match” in the same post — the verdict section reports both rather than collapsing everything into one number, because “best overall” often depends on what you’re optimizing for in your own kitchen. When methods tie or trade off, we say so.
Community Retests
When you run a method yourself and submit a retest, this is the same scorecard your data gets measured against — same net carb math, same 1–5 scales, same hold-time definition. That’s what makes your kitchen’s result comparable to ours instead of just another anecdote.
Next up: “Why Traditional Chaat Breaks Keto” — the actual carb science of maida, besan, and potato, before a single substitution enters the fryer.