Experimental · For fun only. Please don’t take these estimates seriously or treat them as official tornado ratings.
Start with the strongest damage indicator. Add supporting evidence through a smooth curve that approaches 260 mph. Stronger evidence always maintains or raises the estimate; confidence adjusts the added support.
B = strongest selected indicator
S = 0.04 × sum of max(indicator − 100, 0), excluding one leading indicator
P = (S + X-factor points + Calculation points) × Confidence
Final = B + (260 − B) × (1 − exp(−P / 85))
A custom contextual estimate. The 260 mph ceiling and support weights are model choices.
Custom estimate
--
Choose evidence values to calculate a contextual estimate.
Breakdown
House damage selection--
Tree wind anchor--
Ground wind anchor--
Vehicle wind anchor--
Strongest indicator · B--
Supporting indicators · S--
X-factor support--
Calculation support--
Confidence multiplier--
Adjusted support · P--
Available headroom · 260 − B--
Added contextual estimate--
Model range--
Range width depends on documentation and certainty. It is a model sensitivity band, not a statistical confidence interval.
Select evidence to see the calculation.
Confidence and range calculation
A score of 5 gives 1.000×. Scores interpolate linearly toward these minimum / maximum multipliers: documentation 0.940 / 1.032; radar 0.980 / 1.026; environment 0.975 / 1.030; certainty 0.910 / 1.018.
Range half-width = 4.5 + 0.75 × (20 − documentation score − certainty score), in mph. Apply it to the final estimate, keeping the range within 40–260 mph.
Matching evidence countsIndicators tied for the highest wind anchor still add corroborating support.
Missing means unknownHouse evidence is optional. At least one wind anchor is needed to calculate an estimate.
House bounds: Texas Tech EF Scale report, FR12 (A-3). Other wind anchors, point weights, the 85-point response scale, range width, and 260 mph ceiling are site-defined. This custom result is not an official EF rating.