Sensium confusion magnets 2026: 12 grapes other dossiers point at most

Grape confusion pairs are usually studied as edges. Studies 1a–1e ranked mutual planting weight, neighborhoods, one-way hubs, identity traps, and planting-asymmetric mutuals. Study 1f flips the graph: confusion magnets — grapes with the highest inbound degree (how often other dossiers list them as a confusion target).

Headline finding: under default filters (planting rank ≤50, inbound degree ≥35), the catalog yields 22 magnets; the densest teaching cut is 12. Pinot Noir leads with 81 inbound edges (77 one-way). Chardonnay (70), Cinsault (62), Grenache (61), Syrah (60), and Airén (60) follow. Each magnet keeps only four outbound chips — so the product teaches “don’t call me X” from dozens of quieter pages that never appear on the magnet’s own Compare row.

This is still catalog-graph research, not population miss-rates (Study 2 / topic `137` when volume clears). Companions: 1a · 1b · 1c · 1d · 1e.

Methodology (read this before citing)

What this is not: a claim candidates guess Pinot or Chardonnay most often in live exams. It is a graph-structure finding: the editorial atlas concentrates “looks like …” warnings onto a small set of attractors whose own four-chip rows cannot reciprocate every inbound edge.

Why magnets deserve their own study

Study 1a elevates famous↔famous mutuals. Study 1c elevates mid-planted hubs that point out one-way. Study 1f elevates the targets of those arrows. Pinot Noir is the clearest case: Nebbiolo, Mencía, Pinotage, Blaufränkisch, and dozens more warn “don’t call Pinot” — yet Pinot’s four chips are already full of Sangiovese, Pinot Gris, Cinsault, and Barbera. Opening only Pinot’s dossier under-trains the inbound swarm.

The same pattern hits Chardonnay (quiet whites and aromatic whites both point in), Cinsault (pale soft reds), and Syrah (pepper/smoke look-alikes). Magnets are where Mistake Replay density should concentrate once Study 2 exists — until then, treat inbound degree as a study-priority signal, not a miss-rate.

Google’s 2026 non-commodity bar still applies: dated method, re-run command, clear “not telemetry” label.

The 12 teaching magnets

Inbound = dossiers listing the magnet. One-way in = magnet does not list that source back. Sample inbound shows the five most-planted sources.

Pale / soft / Pinot-family attractors

White attractors (oak-compatible + neutrals + coastal)

Dark / Mediterranean / Bordeaux attractors

(M) = mutual · (→) = one-way into the magnet

Three takeaways

1. Magnets are inbound problems — open the sources, not only the hub

If you only drill Pinot’s four Compare chips, you miss Nebbiolo→Pinot, Mencía→Pinot, and Pinotage→Pinot. Schedule source → magnet drills: open the quieter dossier first, write why it is not the magnet, then confirm on Compare.

2. Four outbound chips cannot absorb eighty inbound edges

That asymmetry is editorial, not a bug. Study 1c already explained one-way hubs; Study 1f names the attractors those hubs point at. Use inbound degree to pick which magnet weeks get extra Train focus.

3. Neutrals and pale soft reds are magnets too — not only nobles

Airén, Trebbiano, Cinsault, and Gamay sit beside Pinot/Chardonnay/Syrah in the inbound table. Candidates who only memorize “noble” attractors still fail quiet-white and pale-red flights. Cross-check 1a neutrals and 1e asymmetric mutuals.

How to drill in three weeks

1. Week 1 — Pinot family magnets (1–3). Pinot, Cinsault, Gamay. One source→magnet flashcard set per day (Nebbiolo/Mencía/Pinotage → Pinot; Pinotage/País → Cinsault).
2. Week 2 — white magnets (4–7). Chardonnay, Airén, Trebbiano, Vermentino. Pair with structure tasting so oak-compatibility does not become a default Chardonnay guess.
3. Week 3 — dark magnets (8–12). Grenache, Syrah, Sangiovese, Cabernet, Carignan. Revisit 1c one-way hubs that point into these magnets (Nebbiolo, Pinotage, Blaufränkisch).

A useful study habit: keep a “magnet week” calendar. When Train misses cluster on the same attractor, open that magnet’s inbound sample list from the export — not only its outbound chips.

Service note: magnets drill well as hub weeks with rotating sources (home flights; exam shopping). Place hubs help when the source is regional (Spanish, Italian, French).

How the six Study 1 cuts fit together

Use them as a stack. A month that only drills 1a still fails the inbound swarm into Pinot; a month that only drills 1f still fails Cabernet vs Merlot under time. When Study 2 ships, inbound magnets are the natural place to test whether graph attractors match population miss clusters.

What we will publish next (Study 2)

When anonymized Train/Blind wrong-answer volume clears a documented n threshold, Study 2 will publish miss-rates beside Studies 1a–1f. Until then, cite this page as catalog inbound degree, not live misread rates.

Frequently asked questions

How is this different from neighborhoods (Study 1b)?

Study 1b maps each planted hub’s outbound four neighbors. Study 1f ranks hubs by how often they appear as someone else’s neighbor. Pinot’s outbound row and Pinot’s inbound swarm are different study objects.

How is this different from one-way exam hubs (Study 1c)?

Study 1c selects mid-planted grapes that emit many one-way edges. Study 1f selects the grapes that receive them. Nebbiolo is a 1c hub; Pinot is a 1f magnet.

Why exclude Saperavi-scale Georgian magnets?

High inbound degree on very low-planted regional clusters can reflect local catalog density, not global exam priority. The default `max-pop 50` filter keeps the teaching list syllabus-relevant; widen `--max-pop` if you want the full long tail.

Can I cite this in a class?

Yes — attribute “Sensium confusion magnets, 2026-07-09” and link this URL. Keep the “not telemetry” label.

How do I reproduce the list?

`node scripts/data/export_confusion_magnets.mjs`

Open the Pinot Noir and Chardonnay dossiers this week, but drill from the sources — Nebbiolo, Riesling, Pinotage — in Blind before you open the magnet’s own four chips. Inbound first is the Study 1f habit.