Inspired by this post:
https://pontifex.substack.com/p/the-snp-and-the-lib-dems-should-consider
I'll define an electoral pact as an agreement where the two parties agree to stand at most one candidate between them in at least 30 seats. (Scotland has 59 seats and I believe the SNP and the Lib Dems normally contest all 59.)
If other parties join the pact, that shouldn't change the resolution criteria, but feel free to ask questions to clarify if you think I've missed an edge case. (For example, if the SNP, Lib Dems and Greens all form a pact such that they only stand one candidate between them in 30 seats, and say the Greens stand a candidate in 3 of those seats such that SNP and Lib Dems only stand one candidate between them in 27 seats, this still resolves YES because those three seats where SNP/LD are only standing 0 candidates still count.)
If they have an electoral pact for fewer than 30 seats I will resolve PROB according to [number of seats in the pact] / [30] * 100%. If it is ambiguous how many seats are in the pact (for example if one party fails to stand a candidate in one constituency for another reason) I will use my best judgement. I think this is very unlikely. But be careful not to bet it up too high if a pact is announced and the number of seats affected is still to be decided. (Feel free to create a market for whether a pact with a smaller number of seats will be agreed if you think my choice of 30 seats is poor.)
If a pact is announced I won't necessarily resolve early. I'll wait until it's definite that they won't back out of it. (I don't think it will be necessary to wait for the election itself - I think at some point we will be able to find a list of candidates standing. Having said that, I don't intend to put a huge amount of effort into establishing candidate lists.) Similarly, if the leaders of both parties say "we will definitely not form a pact", that isn't enough to resolve early to NO. They could change their minds.