Title of a talk (joint work with Frank Techel): “Fundamentals of ordinal natural hazard danger level forecast verification” discussing Brunswick’s Lens Model and introducing our own expansion of the lens model to apply to ordinal danger level forecasting.

Might be of interest to you @bweatherson.bsky.social

Problem solved and sense prevailed and we can travel with the dog.

Congratulations, Aidan, that’s a great accomplishment!

Could you add me too. Thanks.

Though this is up there too: ‘Insisting on balanced trade with every trading partner individually is bonkers—like suggesting that Texas would be richer if it insisted on balanced trade with each of the other 49 states, or asking a company to ensure that each of its suppliers is also a customer’

Agreed I think frequency is more suitable at the end of the percentage chance scale (close to 0 or 100) — you too easily round it

I also want to give a huge shout out to reviewer 2 -- the most amazingly valuable, charitable and extensive set of comments I have ever received. Thank you whoever you are!

Perhaps some folks will find it interesting, in particular our more general finding that end users give generally lower probability assessments in a frequency format compared to a percentage chance format.