Caplan win bets about the EU collapsing, but debate if he won on a technicality

Steyn made a prediction that the EU would collapse by 2020. Caplan then made a bet with him that no member would leave by 2020.

People claim he should concede because of Brexit. Caplan refuses.

when experts make barely incorrect predictions, they tend to plead bad luck (“I was almost right”); when they make barely correct predictions, however, they tend to plead great skill (“I was not almost wrong”). I will not do this. Although the UK remains in the EU today, it will leave on January 31, 2020. Therefore, I clearly got lucky. Or to be more precise, I first got unlucky (since David Cameron could have easily refused to call the Brexit referendum in the first place), and then I got lucky three times in a row.

Maybe the lesson is if a prediction has a basic thesis, like NATO will collapse. We should have many related predictions (countries withdrawing, military spending).