[jessriedel]

If the institutions' only methods for checking the insanity of men is experimental data, and that data only arrives on multi-decade timescales, then the institutions are broken. Furthermore, there are several intuitional changes physics could make to make better use of the meager data they have, but they do not. For example: forcing physicists to go on the record specifically and publicly about predictions (not just the one-off bet), and making hiring decisions based on it.

[mathattack]

Making hiring decisions based on bets may not work: time horizons are long and you may decrease bold experimental work. (There is a 1 in 100 that this works...)

I do agree about your underlying idea of having them put skin in the game. That forces better thinking in most fields though even financial markets are susceptible to irrational thinking.

[jessriedel]

You can make 1% predictions.

[mathattack]

Yes - an odds bet. But you need to make an awful lot of them to get one that hits. This is similar to why it is hard for employees to get the benefit of power law returns. There isn't enough career time to work for 100 companies. (Easier to invest in the outsize outcomes than work your way in)