The Housing Tribunal · November 2026 sitting

Buy now, or wait? Our AI argues it out in public.

Each month the states we cover go on trial (3 this month). Two AI advocates argue opposite sides from the same sourced numbers. A critic strikes what doesn’t hold up, and leaves the strikes showing. A judge rules and puts one forecast on the record. When the real data arrives, plain code grades it. You can make your own call on every question and try to beat the Tribunal.

  1. 1Pick a state below and read its verdict.
  2. 2Tap Higher or Lower on its graded forecast. No sign-up.
  3. 3Results arrive when Realtor.com publishes December 2026, usually early the next month.

Every verdict is written for buyers. If you’re selling, the same facts usually read the other way: a “wait” for buyers means sellers face more competition now.

November 2026 docket

3 states heard on data through October 2026. The critic struck 0 claims and rewrote 2. Forecasts are about December 2026 and are graded when Realtor.com publishes that month.

Ohio

For buyers: toss-up
Verdict confidence 55%

Both briefs describe the same loosened market: active listings are 26,471 in October 2026, up 18.1% from a year earlier, days on market edged up to 46 days, and list prices are nearly flat at +0.9%. But mortgage rates of 7.18% versus 6.22% a year ago offset that choice, and neither side shows a decisive trend.

Graded forecast

Will Ohio's active listings for December 2026 be higher or lower than December 2025's 19,811?

More active listings usually means more choice and more room to negotiate for buyers.

The Tribunal says higher, 88% confident in this forecast.

Your call

Watch the hearing →

Texas

For buyers: toss-up
Verdict confidence 55%

Both briefs rest on facts the Red Pen kept. Median list price of $350,000 is down 2.8% from a year earlier, a point for moving now. But active listings are up 0.5% and days on market are 68 days against 69 days, so conditions are close to last year's. The price softness is not yet matched by clearly faster sales or broader supply.

Graded forecast

Will Texas's median list price for December 2026 be higher or lower than December 2025's $350,000?

Lower asking prices favor buyers; higher ones favor sellers.

The Tribunal says lower, 65% confident in this forecast.

Your call

Watch the hearing →

California

For buyers: lean wait
Verdict confidence 58%

Both briefs rest on the same facts: the median list price of $719,000 is down 2.9% from a year earlier and down 0.8% on the month, while active listings are up 6.2%. Prices are still easing and supply is growing, which modestly favors waiting. The two-day rise in days on market fits the usual autumn pattern, so it carries little weight.

Graded forecast

Will California's active listings for December 2026 be higher or lower than December 2025's 56,016?

More active listings usually means more choice and more room to negotiate for buyers.

The Tribunal says higher, 80% confident in this forecast.

Your call

Watch the hearing →

The record so far

The Tribunal doesn’t grade itself. A script compares each forecast with the number Realtor.com later publishes on FRED, and the result stays on the scoreboard for good, misses included.

0forecasts right
0forecasts wrong
3awaiting data
1sitting held

How a sitting works

Five roles, and the two that decide facts and grades aren’t AI at all.

  1. The ClerkPulls each state’s numbers from FRED and numbers every fact. Nothing outside this packet can be cited.Code
  2. Two CounselOne argues conditions favor buyers moving now, the other that they favor waiting. Same facts, opposite briefs.AI specialists
  3. The Red PenThe critic. Checks every number, flags overblown wording and seasonal mirages, and enforces Fair Housing rules. Its strikes are published.AI specialist
  4. The JusticeWeighs what survived, rules, records the dissent, and commits to one forecast with a confidence level.AI specialist
  5. The ScorekeeperCompares each forecast with the published number and records hit, miss or void. An AI never grades a forecast.Code

Managing Agent  an AI editor that runs the whole sitting, decides what to do when something goes wrong, and logs each ruling below. The site editor reviews each sitting afterward.

The Managing Agent’s rulings this sitting

  • Why only three states? This is the launch sitting, a supervised dry run on Texas, Ohio and California. All 18 states sit from December.

Weekly game

The Rate Call

Every Friday a new round opens: will next week’s 30-year mortgage rate be higher or lower? Graded automatically the following Friday. Make your call →

How it’s built

This sitting: Managing Agent on Claude Opus 5.5, then Claude Fable 5.2 (launch sitting); specialists on Claude Sonnet. The evidence packet and the grading are plain code, published on GitHub.

The Tribunal describes market conditions from public statistics. It is not financial, legal or tax advice, it knows nothing about your situation, and its forecasts are often wrong, which is why we publish the record. Data: Realtor.com housing series and the Freddie Mac mortgage survey, via FRED. See our editorial guidelines and corrections log.