It's heartwarming to see these people have fun while still sticking to serious work. Everything should be like that.

When it comes to science, only the British seem to be able do that.

Rovelli is Italian, and I think Pennington is American and Oppenheim is Canadian.

> a small amount of olive oil, balsamic vinegar, or wine

I think we can guess who write up the contract!

Crisps though!

> Should space-time be shown to be quantum, the loser will give to each of the winners, one ITEM* of their choice. If the alternative hypothesis is deemed to be correct, the losers will each give 5,000 ITEMS to the winner.

> * ITEM is defined to be a object of the winner’s choice, worth no more than 20 British pence on January 21st, 2021. Examples include some crisps, a bazinga ball, a small amount of olive oil, balsamic vinegar, or wine

The contract doesn't specify what kind of balsamic vinegar. Supermarket balsamic is just a mixture of colourings and flavourings. Proper solera balsamic is sipped as an aperitif, and is much too expensive for me to have ever tasted.

It doesn't say the amount either, just that each item must be worth less than 20 british pence.

So it can be one drop of "Aceto Balsamico Tradizionale di Modena DOP", which is the real deal. Prices start at around 50€ for 100mL, but it can easily go into the 100s if you want the more premium ones (25 years). A single drop would fit the 20 pence bill.

Why not have both: salt 'n' vinegar flavour crisps?

Interesting that three so obviously intelligent people could get the odds of the bet wrong: it’s not a 1:5000 bet, it’s a 2:5000 bet a.k.a. a 1:2500 bet.

I mean clearly if it's a mistake this is a matter of terminology not intelligence, but could you explain how it's wrong?

If A wins, B pays A $1. If B wins, A pays B $N. For the bet to be fair (zero expectation value), the probability that A wins needs to be p=N/(N+1) and that B wins needs to be 1-p=1/(N+1). My understanding of what "odds" are is the ratio of the outcome probabilities: p:(1-p) = p/(1-p) = N/1 = N:1. (This is why the "log odds" are log(p/(1-p)).)

Do they use a different notion of "odds" in sports gambling?

Turns out the only interesting thing is how sloppy my reading is (and how quick I am to blame others for that sloppiness…).

I simply missed the second “each” below:

> Should space-time be shown to be quantum, the loser will give to each of the winners, one ITEM of their choice. If the alternative hypothesis is deemed to be correct, the losers will each give 5,000 ITEMS to the winner.

Given this each, I'm surprised the headline isn't 10,000:1 bet.

Also, who wants 10,000 <$20 items? It's a cool payout, but if I wanted to spend ~$200,000, I wouldn't do it in 10,000 pieces.

It's not $20. It's 20 British pence - about 25 cents.... "Examples include some crisps, a bazinga ball, a small amount of olive oil, balsamic vinegar, or wine."

I think, reading it, it could be £1000 worth of wine. For example two cases of decent Gevrey Chambertin, I myself could bring myself to accept such a thing.

I don't think so. It's 5000 from each loser and there are two potential losers so 5000*2=10000 and 2:10000 is the same as 1:5000.