Sensium blindLogic coaching cards 2026: 12,352 first / confidence / warning lines
Original Sensium research: 3,068 firstChecks, 3,068 confidenceSignals, 6,216 warningFlags — coaching-card families, not Study 3 word counts alone.
Sensium’s Band G stack already published grape confusion cuts (Study 1), catalog censuses including blindLogic word density (Study 3), structure cohorts (Study 6), and process/place/winemaking objects (Studies 9–11). This page is the next non-separator asset: blindLogic coaching cards — the first-check lines, confidence signals, and warning flags each dossier already encodes. It answers a different question than “how many blindLogic words?” It answers: what coaching moves does the catalog write before you name the grape?
Headline finding (snapshot 2026-07-09): Across 1,534 dossiers (0 empty on the three coaching fields): 3,068 firstChecks (avg 2; 28,485 words) · 3,068 confidenceSignals (avg 2; 14,071 words) · 6,216 warningFlags (avg ~4.05; 59,534 words). Plus engine-side 1,535 climateChecks and 1,535 oakChecks (avg 1 each). First-check family hits (multi-hit): structure 1,709 · aroma 1,528 · color 198 · climate 145 · oak/lees 17 · other 199. Warning classes: aroma_stop 1,774 · identity_lineage 1,592 · structure_stop 1,176 · other_stop 872 · oak_stop 578 · climate_ripeness_stop 166 · color_stop 58.
This is still catalog research, not population miss-rates. Study 2 (topic `137`) remains volume-gated. Do not cite these lines as “which checks candidates fail most.” Cite them as the coaching script behind each grape dossier.
Companions: Study 3 atlas · Study 6 structure · Study 1 field guide · tannin/acid/body · Study 8 aliases.
Methodology (read this before citing)
| Field | Value |
|---|---|
| Source | the grape catalog — every first-check coaching lines / confidence signals / warning flags (+ climate/oak counts) |
| First/confidence families | Keyword buckets: structure, aroma, color, oak/lees, climate, other (multi-hit) |
| Warning classes | identity_lineage · color_stop · structure_stop · aroma_stop · oak_stop · climate_ripeness_stop · other_stop |
| Exam spine | 12 planting-famous varietals with full coaching cards |
| Snapshot date | 2026-07-09 |
| Snapshot | Follow the method on this page. Counts move when the catalog is edited. |
| Not claimed | Live exam miss-rates, or that coaching lines replace Study 1 pair separators |
| Study 3 | Study 6 | Study 12 (this page) | |
|---|---|---|---|
| Object | blindLogic size / word density | Structure scale cohorts | Coaching-card content families |
| Question | How much prose? | Which structure traps? | What does the dossier tell you to do first? |
Why coaching cards after the census stack?
Google’s 2026 non-commodity bar still applies: dated snapshot, clear “not telemetry” label. Study 3 proved the atlas ships dense blindLogic. It did not publish the script: start with tannin before naming Cabernet; treat pale+light grip as a Pinot door; treat “Identity: Syrah is Shiraz” as a stop rule, not a trivia line. Study 6 groups dossiers by structure scales. Study 12 publishes the sentences that turn those scales into coaching moves.
Practically: open any dossier, read first-check lines before aroma poetry, then scan warning flags for the identity line and the structure stop.
Inventory snapshot
| Field | Items | Avg / dossier | Words |
|---|---|---|---|
| firstChecks | 3,068 | 2 | 28,485 |
| confidenceSignals | 3,068 | 2 | 14,071 |
| warningFlags | 6,216 | ~4.05 | 59,534 |
| climateChecks | 1,535 | 1 | 18,706 |
| oakChecks | 1,535 | 1 | 20,958 |
Warning-class mix
| Class | Count | Teaching use |
|---|---|---|
| aroma_stop | 1,774 | Wrong aromatic story for this grape |
| identity_lineage | 1,592 | VIVC / synonym / parent traps |
| structure_stop | 1,176 | Wrong tannin/acid/body/alcohol frame |
| other_stop | 872 | Residual editorial stops |
| oak_stop | 578 | Oak sweetness / dill-cedar traps |
| climate_ripeness_stop | 166 | Green / overripe / climate misreads |
| color_stop | 58 | Pale/deep/opaque color stops |
Family A — First checks (structure-first coaching)
First-check family hits lean structure then aroma — the catalog’s default blind order.
| Grape | First check (catalog) |
|---|---|
| Cabernet Sauvignon | Look for firm tannin and dark fruit concentration before naming the grape. |
| Merlot | Check whether tannin is polished rather than aggressive. |
| Pinot Noir | Start with color intensity and tannin: Pinot Noir is often pale with light grip. |
| Chardonnay | Judge texture first: Chardonnay often reads broader than aromatic whites. |
| Nebbiolo | Check whether tannin is high even when color seems moderate. |
| Sangiovese | Start with acidity: Sangiovese often shows clear acid drive. |
| Riesling | Start with acidity and alcohol balance: many examples are high acid and lighter-bodied. |
| Grenache | Check for elevated alcohol with relatively soft tannin. |
Drill: how to taste tannin/acid/body.
Blind cue habit: write one structure first-check before any grape name.
Family B — Confidence signals (short confirmers)
| Grape | Confidence signal (catalog) |
|---|---|
| Cabernet Sauvignon | High tannin with cassis core |
| Pinot Noir | Red cherry with mushroom/earth |
| Syrah | Pepper and blackberry together |
| Tempranillo | Dill and leather notes |
| Sauvignon Blanc | Gooseberry and cut-grass markers |
| Nebbiolo | Rose and tar combination |
| Chenin Blanc | Quince and wool markers |
These are confirmers, not separators. Use them after the first check lands — not as a substitute for Study 1 pair drills.
Family C — Warning flags (identity + structure + aroma stops)
Every exam-spine grape carries an identity_lineage line (VIVC / synonym / parent). Pair it with a structure or aroma stop.
| Grape | Identity stop (trimmed) | Structure / aroma stop |
|---|---|---|
| Cabernet Sauvignon | Cabernet Franc × Sauvignon Blanc — not Cabernet Franc / Merlot | Low tannin · Strong floral without dark fruit |
| Merlot | Magdeleine Noire × Cabernet Franc — not Carmenère mislabel | Sharp high-acid profile |
| Pinot Noir | Spätburgunder / Blauburgunder / Pinot Nero — same cultivar | Very high tannin · Dominant cassis and mint |
| Syrah | Syrah is Shiraz — naming/style fork, not genetics | Low alcohol · No spice/savory character |
| Tempranillo | Tinta Roriz / Aragonez / Tinto Fino… — Iberian naming | Very low tannin · Rose-tar dominance |
| Sauvignon Blanc | Fumé Blanc / Muskat-Silvaner — not Sauvignonasse / Savagnin | Heavy oak dominance |
| Nebbiolo | Spanna / Chiavennasca; parent of Freisa — not Freisa synonym | Soft tannin profile |
| Riesling | Rhine Riesling VIVC #10077 — not Graševina / Welschriesling | Pronounced tannin grip · Off-balance acid/fruit |
| Grenache | Garnacha = Cannonau — same cultivar | Very high tannin |
| Chenin Blanc | Steen / Pineau de la Loire — same cultivar | Low acidity · Strong tannic red-fruit profile |
Drill: Study 8 aliases · Study 1d identity traps.
Blind cue habit: read the Identity: warning before inventing a synonym story.
Engine-only note — climateChecks / oakChecks
Each dossier also ships ~1 climateCheck and ~1 oakCheck (engine-facing). Study 12 cites their counts for inventory completeness; teaching tables above prioritize the three user-facing coaching fields. The teaching tables on this page are the public snapshot; engine strings stay internal.
How to cite this study
- Name it: “Sensium Study 12 (blindLogic coaching cards), export 2026-07-09.”
- Link this URL.
- Name the field (firstCheck / confidence / warning class) and the grape.
- Keep the label: coaching-card content — not Study 3 word density alone, not miss-rates.
- Keep the snapshot date from the method table — do not invent a later count.
How this sits beside Studies 1–11 and Study 2
| Study 1 | Study 3 | Study 6 | Study 9 / Study 10 / Study 11 | Study 12 | Study 2 | |
|---|---|---|---|---|---|---|
| Object | Grape pairs | Atlas size | Structure cohorts | Process / place / method | Coaching cards | Wrong answers |
| Status | Complete | Complete | Complete | Complete | This snapshot | Volume-gated (`137`) |
After Study 11, inventing another thin aroma cut or a redundant size census would still be dishonest. Coaching cards are a new catalog object Study 3 only measured by length — the script the dossier already owned.
A two-week drill
| Week | Focus | Habit |
|---|---|---|
| 1 | First checks + confidence | Structure-first sheet for Cabernet / Pinot / Nebbiolo / Riesling / Grenache |
| 2 | Warning flags | Identity line + one structure stop + one aroma stop per exam grape |
Service: home flights · Blind · Grapes.
Frequently asked questions
Is this the same as Study 3?
No. Study 3 counts blindLogic size. Study 12 publishes coaching-card families and exam-spine scripts.
Is this the same as Study 1?
No. Study 1 edges are grape↔grape separators. Study 12 lines are per-dossier coaching moves (including identity warnings that support Study 1d traps).
Are climateChecks and oakChecks user-facing?
They are engine-facing companions counted here for inventory. Prefer firstChecks / confidenceSignals / warningFlags for teaching citations.
When does Study 2 ship?
When anonymized Train/Blind volume clears a documented threshold. Until then, prefer Studies 1 and 3–12 — and do not invent miss-rate tables.
Bookmark this page beside the Study 3 atlas and Study 1 field guide, open a dossier this week, and force one first check plus one warning flag before naming the grape. Structure first — confirm — then stop rules.