The Atlantic Stays at Five or Fewer Named Storms Through September
Published 27 August 2026
Through late August the 2026 Atlantic season had produced four named storms, Arthur, Bertha, Cristobal and Dolly, with no hurricanes, and NOAA maintains a below-normal outlook. A fifth, Edouard, formed on 31 August. This entry holds that the cumulative season total stands at five or fewer named storms on 30 September. September is climatologically the most active month of the season, with the 10 September peak inside the window, so the quiet-season thesis is being tested where it is most likely to break.
Stated probability 0.6 against a base rate of 0.5. The entry was published with a stated base of two storms, which was wrong. The true count at publication was four, and Edouard took it to five on 31 August. Failure therefore requires one further named storm before 30 September, not four. The stated probability of 0.6 was set against the wrong premise and is not re-cut after the fact; the call is graded at the odds as published.
Design and lineage
Checkpoint instrument inside the open ATL2026TOTAL thesis (11 or fewer for the full season). The August 4-storm checkpoint grades on 31 August; this entry re-arms the thesis through the hardest stretch. No chart-side claim at monthly resolution.
How it reconciles
The verdict publishes by 3 October 2026 on this page and on the public board, quoting the reference source named in the call. A failed entry is graded at full weight and stays on the board with its reasoning unedited.
Correction (11 September 2026)
This entry was published with the wrong storm count. It stated that the 2026 Atlantic season had produced two named storms through late August. The National Hurricane Center record shows four by that date: Arthur on 17 June, Bertha on 19 July, Cristobal on 12 August and Dolly on 27 August, which was the day this page went up. Edouard formed on 31 August, taking the season to five.
What it changes. The threshold and the falsifier are unchanged, because both were stated as an absolute season count and remain gradeable exactly as written. What changes is the difficulty. The page argued that failure would require four or more further storms in five weeks and put the odds near a climatological coin. In fact failure required two further storms at publication, and now requires one. The call is considerably harder than the page claimed when it was filed.
What is not being changed. The stated probability of 0.6 stays as published. Re-cutting a probability after discovering the premise was wrong would make the record meaningless, so the entry will be graded on 30 September at the odds it carried on the day it was filed, and the miss, if it comes, counts at full weight.
How it happened. The count was taken from a news summary during drafting and not checked against the National Hurricane Center record. Season counts are now verified against the NHC list before any storm-count entry is filed.
Frequently asked
Where does the count stand now?
Four through late August 2026: Arthur, Bertha, Cristobal and Dolly, with Edouard forming on 31 August for a total of five. None reached hurricane strength. The National Hurricane Center record is the reference. This page was first published with a count of two, which was wrong, and the correction is recorded on the page.
Why is the bar five?
The bar was set at five when the count was believed to be two. The count was in fact four at publication and five by 31 August, so a single further named storm before 30 September fires FAILED. The bar stands as published.
What happens to the earlier August checkpoint?
It grades separately on 31 August. Each checkpoint is its own entry so the season thesis accumulates graded evidence rather than one verdict at the end.
Read next
Published 27 August 2026. A dated forward call with its reconciliation condition stated in advance; the verdict will be recorded on this page and on the public board. Not financial, medical or personal advice.