Sensium regional place separators 2026: 12,601 cues that finish the call
Original Sensium research: 12,601 regional place-narrowing cues across 2,975 expressions — commune twins and Old/New World forks after the grape is known.
Sensium’s Band G stack already published fifteen grape confusion cuts (Study 1), six catalog censuses (Studies 3–8 field guide), and process-glossary forks (Study 9). Study 3 counted how large the regional grid is. This page is the next non-separator asset: regional place separators — every place-separator cues the regional atlas already encodes. It answers a different question than “which grapes twin?” or “how many cells?” It answers: once the grape is known, which place↔place stop rules does the catalog write for the Region-Narrowing walk?
Headline finding (snapshot 2026-07-09): Across 2,975 regional expressions (425 varietals; flat 7 cells each), the atlas ships 12,601 separator edges (avg ~4.24 per expression). Collapsed undirected: 7,157 pairs (5,443 mutual / 1,714 one-way). Directions: favors_self 12,497 · favors_other 63 · either 41. A name-heuristic Old World ↔ New World slice finds 183 directed edges (teaching samples below — not a closed ontology).
This is still catalog research, not population miss-rates. Study 2 (topic `137`) remains volume-gated. Do not cite these cues as “which regions candidates miss most.” Cite them as the place stop rules behind Region Narrowing and the place blogs.
Companions: Study 3 atlas · Study 9 process forks · Left vs Right Bank · Barolo vs Barbaresco · Chablis vs Meursault · Sancerre vs Marlborough.
Methodology (read this before citing)
| Field | Value |
|---|---|
| Source | the regional expressions catalog — every place-separator cues cue |
| Collapse | Undirected pair = same `varietalId` + sorted paired region ids |
| Mutual | Both endpoints write at least one cue into the pair |
| Exam spine | 12 planting-famous varietals × 8 teaching pairs each (export sample) |
| OW/NW slice | Name heuristic on classic Old World + New World place strings — not a formal climate ontology |
| 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 place cues replace grape Study 1 pairs |
| Study 1 | Study 3 | Study 9 | Study 10 (this page) | |
|---|---|---|---|---|
| Object | Grape↔grape separators | Regional grid size (+ dossiers) | Process↔process forks | Place↔place cues |
| Question | Which grape twins? | How big is the atlas? | Which method mistakes? | Which place forks finish the call? |
Why place separators after Study 3 and Study 9?
Google’s 2026 non-commodity bar still applies: dated snapshot, clear “not telemetry” label. Candidates often get the grape and still fail the place — Pauillac vs Stags Leap, Barolo vs Barbaresco, Chablis vs Meursault, Sancerre vs Marlborough. Study 1’s grape pairs do not own those forks. Study 3’s census says the grid has N expressions at 7 cells — it does not publish the 12,601 stop-rule sentences between sibling places. Study 9 owns method mistakes (MLF vs lees), not commune twins.
Practically: after the grape call, open the dossier’s regional walk, read one place-separator cues, and force a place fork before inventing a village from vibes.
Regional totals keep growing via Expand waves. Counts move when the catalog is edited rather than treating any single blog table as eternal.
Inventory snapshot
| Metric | Value (export 2026-07-09) |
|---|---|
| Expressions | 2,975 |
| Varietals | 425 (min=max 7 cells) |
| Separator edges | 12,601 |
| Avg edges / expression | ~4.24 |
| Undirected pairs | 7,157 |
| Mutual / one-way | 5,443 / 1,714 |
| Top countries (by expression) | Italy 580 · Australia 421 · USA 346 · NZ 333 · France 274 · Spain 214 |
Climate-band expression mix (catalog labels): cool_moderate 916 · moderate_warm 898 · warm 573 · moderate 458 · cool 102 · hot 28.
Family A — Commune / sibling place twins (exam spine)
Same grape, neighboring or sibling appellations — the classic “you got Cabernet / Nebbiolo / Chardonnay, now finish the place” drill.
Cabernet Sauvignon · Merlot
| Place A | Place B | Catalog cue (trimmed) |
|---|---|---|
| Pauillac | Saint-Estèphe | Graphite/pencil-lead vs firmer, earthier tannin grain |
| Pauillac | Saint-Julien | Graphite austerity vs rounder, red-fruit balance |
| Margaux | Pauillac | Violet/floral finer tannin vs graphite/cedar depth |
| Pomerol | Saint-Émilion | Iron/truffle blue fruit vs limestone red-fruit floral |
Drill: Left vs Right Bank · Cabernet · Merlot.
Pinot Noir · Chardonnay
| Place A | Place B | Catalog cue (trimmed) |
|---|---|---|
| Vosne-Romanée | Gevrey-Chambertin | Exotic-spice perfume vs darker, sturdier earth structure |
| Chablis | Meursault | Steely oyster-shell, little oak vs richer nutty/buttery MLF+barrel |
| Chablis | Puligny-Montrachet | Kimmeridgian austere vs richer fruit, hazelnut, discreet oak |
| Puligny-Montrachet | Meursault | Tighter mineral white peach vs richer nuttier butter |
Drill: Chablis vs Meursault · Puligny vs Meursault.
Syrah · Nebbiolo · Sangiovese
| Place A | Place B | Catalog cue (trimmed) |
|---|---|---|
| Côte-Rôtie | Hermitage | Violet/floral (often Viognier) vs broader, darker power |
| Côte-Rôtie | Cornas | Floral aromatic vs darker, meatier, more tannic |
| Barolo | Barbaresco | More powerful/tannic/tar vs more perfumed, earlier-drinking |
| Chianti Classico | Brunello di Montalcino | Brighter savory tomato-leaf vs denser, longer-oaked |
Drill: Barolo vs Barbaresco · Chianti Classico vs Brunello · Syrah vs Shiraz.
Tempranillo · Sauvignon Blanc · Riesling · Grenache · Chenin
| Place A | Place B | Catalog cue (trimmed) |
|---|---|---|
| Rioja Alta | Ribera del Duero | Dill/coconut American-oak elegance vs denser, blacker, often French-oaked |
| Sancerre | Pouilly-Fumé | Brighter grass-grapefruit vs flintier smoky reductive |
| Mosel | Rheingau | Delicate slate, lower alcohol vs fuller, riper stone-fruit |
| Châteauneuf-du-Pape | Gigondas | Broader sun-baked galets vs firmer pepper-garrigue under Dentelles |
| Vouvray | Montlouis-sur-Loire | Shared apple-quince-honey; Montlouis often slightly leaner |
Drill: Sancerre vs Pouilly-Fumé · Gigondas vs Châteauneuf.
Family B — Old World ↔ New World forks (heuristic slice)
Same grape, classic European benchmark vs a New World twin the catalog already writes. Export name-heuristic count: 183 directed edges. Teaching samples (deduped):
| Grape | Place A | Place B | Catalog cue (trimmed) |
|---|---|---|---|
| Cabernet Sauvignon | Pauillac | Stags Leap District | Fresh cassis + cedar + graphite vs cooked fruit + chocolate + sweet vanilla |
| Cabernet Sauvignon | Coonawarra | Pauillac | Eucalyptus/mint on terra rossa vs graphite/cedar without eucalyptus |
| Merlot | Pomerol | Napa Valley | Iron/truffle savory vs cooked plum, mocha, sweeter oak |
| Pinot Noir | Vosne-Romanée | Willamette Valley | Exotic-spice concentration vs brighter cherry-pomegranate, finer tannin |
| Syrah | Côte-Rôtie | Barossa Valley | Pepper/savory/floral vs ripe blackberry, chocolate, sweet vanilla |
| Syrah | Côte-Rôtie | Hunter Valley | More violet-perfumed vs earthier leather-savory at similar weight |
| Sauvignon Blanc | Sancerre | Marlborough | Restrained citrus + wet stone + flint vs passionfruit/gooseberry/tomato leaf |
| Riesling | Mosel | Clare Valley | Delicate low-alcohol slate vs dry, higher-alcohol lime drive |
| Chardonnay | Chablis | Sonoma Coast | Restrained mineral vs ripe stone fruit, more visible oak |
| Chenin Blanc | Vouvray | Swartland | Quince/honey/wet-wool Loire spectrum vs denser, spicier Cape Chenin |
| Grenache | Châteauneuf-du-Pape | McLaren Vale | Garrigue/leather vs bramble, blueberry, sweeter American-oak vanillin |
| Tempranillo | Rioja Alta | California | Cedar-tobacco savory vs riper plum-jam, sweeter oak |
Drill: Marlborough vs Sancerre · cool vs warm climate.
Blind cue habit: after the grape, write one place fork (commune twin or OW/NW) before the final village poetry.
How to cite this study
- Name it: “Sensium Study 10 (regional place separators), export 2026-07-09.”
- Link this URL.
- Name the family (commune twin vs OW/NW fork) and the grape.
- Keep the label: regional place cues — not grape miss-rates, not Study 3 grid size alone.
- Keep the snapshot date from the method table — do not invent a later count. (Expand waves move the totals).
How this sits beside Studies 1–9 and Study 2
| Study 1 | Studies 3–8 | Study 9 | Study 10 | Study 2 | |
|---|---|---|---|---|---|
| Object | Grape confusion cuts | Catalog censuses | Process forks | Place-narrowing cues | Wrong answers |
| Status | Complete | Complete + field guide | Complete | This snapshot | Volume-gated (`137`) |
After Study 9, inventing another thin grape aroma cut or a redundant size census would still be dishonest. Place separators are a new catalog object the regional atlas already owned — the sentences Region Narrowing reads after the grape.
A two-week drill
| Week | Focus | Habit |
|---|---|---|
| 1 | Commune twins | Médoc communes; Barolo/Barbaresco; Chablis/Meursault/Puligny; Rioja Alta/Ribera |
| 2 | OW ↔ NW | Pauillac/Stags Leap; Côte-Rôtie/Barossa; Sancerre/Marlborough; Mosel/Clare |
Service: home flights · Blind · Grapes.
Frequently asked questions
Is this the same as Study 1 grape pairs?
No. Study 1 edges are grape↔grape. Study 10 edges are place↔place within a varietal’s regional grid.
Is this the same as Study 3?
No. Study 3 counts atlas/grid size. Study 10 publishes the separator cue text that finishes place calls.
Is the Old World / New World count exact?
No. It is a name heuristic for teaching samples. Cite the full 12,601 edge count for inventory; cite individual cues for teaching.
When does Study 2 ship?
When anonymized Train/Blind volume clears a documented threshold. Until then, prefer Studies 1 and 3–10 — and do not invent miss-rate tables.
Bookmark this page beside the Studies 3–8 field guide and Study 9 process forks, open a dossier’s regional walk this week, and force one place fork after every grape call. Grape first — method when needed — place last.