“AFFORDABLE!?! Have you looked at the prices of these houses?! A $350,000 house is torn down and a $1,000,000 house goes up in its place!!”

Recorded 2026-09-02 by Stew. Purpose: methodology v1.13 says a story leaves the findings board when every claim on it is Not established because the public record cannot answer the question people ask. The test is a triage read of the published claims, so the call is not the editor’s alone. This record asks for that read.

Claim 1 (published as ip-teardown-price-gap, Not established, unanimous)

Circulating form, supplied by the founder from an Edmonton infill thread, source not captured:

“AFFORDABLE!?! Have you looked at the prices of these houses?! A $350,000 house is torn down and a $1,000,000 house goes up in its place!!”

Proposition as checked (methodology v1.10 magnitude re-run): on Edmonton lots where a house was demolished for infill, the housing built in its place is typically worth about three times the house that came down. The panel found Edmonton publishes no series joining a demolished house to its replacement; one reviewer hand-joined eight lots (ratios about 1.4 to 3.9) and called it an anecdote.

Claim 2 (published as ip-infill-affordable, Not established, split)

Proposition: new infill housing in Edmonton is affordable relative to the existing housing it replaces or to citywide prices. Published new-home prices do not identify infill by dwelling form.

What the read is for

For each claim: is it Not established because the record could have answered and did not, or because no public record can answer the question at the level people ask it? If the second, the story leaves the findings board and is listed on /considered as “Checked; the record cannot answer it”. Sources: src/content/claims/ip-teardown-price-gap.yaml, src/content/claims/ip-infill-affordable.yaml, src/content/stories/infill-prices.mdx, reviews/infill-prices/2026-09-02-magnitude/, reviews/infill-prices/2026-09-02-questions/triage-1.md.

GPT-5 Codex

Searches: No web or claim-truth searches. One local file inventory of the named review directories.

Sources opened: intake/candidates/infill-prices-published/intake.md, prompts/intake-triage.md, src/content/claims/ip-teardown-price-gap.yaml, src/content/claims/ip-infill-affordable.yaml, src/content/stories/infill-prices.mdx, reviews/infill-prices/2026-09-02-magnitude/brief.md, reviews/infill-prices/2026-09-02-magnitude/synthesis.json, reviews/infill-prices/2026-09-02-magnitude/disagreements.json, reviews/infill-prices/2026-09-02-questions/triage-1.md.

ip-teardown-price-gap

1. “When a $350,000 Edmonton house is demolished for infill, is what replaces it usually worth about $1 million?” A resident could ask this.
2. One founder-supplied quotation, possibly recalled or composite, came from an Edmonton infill thread. Its URL, platform, author, date and prevalence were not captured. It is weakly captured reader demand, not an editor-created hypothesis.
3. Permits, historical assessment rolls, infill reports and individual land-title records could support a full assessed-value reconstruction. They cannot supply an open, bulk transaction-price series. The narrower assessed-value ratio would still interest readers if labelled honestly.
4. Supported would surprise infill advocates. Contradicted would surprise critics. Partially supported could expose a real split between large single replacements and subdivided lots. Not established surprises little after the incomplete linkage.
5. The claim bears on zoning and affordability arguments affecting existing residents, buyers and displaced renters. It needs one decade-scale, lot-level linkage exercise, but no comparator cities. Several runs were completed, yet only eight lots produced illustrative ratios.

Not established reason: The record could have answered the declared assessed-value proposition, but the run did not build the conforming population series.

Triage: PARK because the public inputs exist, but a full linkage exercise must replace the eight-lot probe before readers can be given an answer.

ip-infill-affordable

1. “Can a median-income Edmonton household afford the new homes built on teardown lots?” A resident or reporter could ask this.
2. The captured quotation challenges “affordable” but does not mention median income or a 30 percent test. That formulation is the site’s hypothesis drawn from the founder-supplied quotation, not captured reader wording.
3. Statistics Canada, the Bank of Canada and City records supply the threshold, income, rates, taxes and assessed values. They do not identify a complete teardown-replacement cohort with title form, transaction prices, condominium fees, tenure or rents. Citywide benchmark affordability remains a narrower question readers would care about.
4. Supported would surprise infill advocates; Contradicted would surprise critics. Partially supported could reveal differences by housing form or tenure. Not established is unsurprising with the denominator and actual costs missing.
5. Buyers, renters and displaced households are affected, and affordability policy can shape zoning and housing spending. The claim requires lot linkage and unavailable property-level cost records, but no comparator cities. A prior model produced 88.4 percent from an inferred cohort, and the panel split over whether that cohort conformed.

Not established reason: No public record can answer the teardown-specific affordability question at the level people ask it.

Triage: PARK because existing sources answer citywide benchmark affordability, not the cost of the housing that replaced particular demolished homes.