The trouble with most of these bets (which have been ubiquitous as long as I've been around) is that it is extraordinarily labor intensive to prepare the terms and definitions. For instance, there is a very good chance that within a year there will be a >1MB block on 'the longest proof-of-work chain'. But which chain is 'the'.
Clearly the longest proof-of-work chain built from Bitcoin's Genesis Block. The terms and definitions are trivially simple.
It's not simple at all of the blockchain forks because different groups consider different blocks to be 'valid.' This is a very real possibility and within this timeframe.
Regardless of anyone's definition of validity, one chain will be longer than the rest (it will contain more cumulative work). I am betting that this (longest) chain will contain a block greater than 1 MB in size by this time next year.
There is no ambiguity in that definition.
While I'm not on Peter R's side in the XT debate, he's correct about this. The terms of this potential "bet" are clear. In any case, even if there were some ambiguity, that's one reason there's a neutral third party in a 2-of-3 multisig.
Now for something completely different. Consider this modified version of "the bet": Peter R and anti-Peter R each put 1 BTC into a 2-of-3 multisig with a neutral third party. One year from now, the 2 BTC are paid out to Peter R on every chain that has had a block over 1MB. The 2 BTC are paid out to anti-Peter R on every chain that has not had a block over 1MB. This would actually give people a way to trade potential coins on potential multiple forks before the potential forks actually occur. It's a little surprising there aren't services offering something like this, given how adamant many people on different sides seem to be.
When a guy has been around as long as I (and is somewhat analytical by nature), one can spot certain things. Attacks by competent attackers are normally set up well ahead of time, and it's possible to see them coming.
One example which is classic is that big who-ha about 'block will fill and we are all going to die so we have to do something by {magic date}. The artifacts of this one continue to linger which is why we still see people who should know better and who were not taken for a ride by the XT/BIP101 attack continue to feel a subconscious need to 'do something'. Even if it makes no sense (e.g., a paltry 1x block size increase which will do absolutely zilch over what we have at 1MB.) The manipulation of the herd through psychological means is a high art and science and is extremely well developed by 2015.
Another one is the idea of the 'longest
valid chain.' The 'valid' part is very deliberately stripped out (though a lot of people who have no clue will parrot the words of those who do.) I'll tell you what is going on here: Even today it will be quite trivial to form a 'longest chain' which is of whatever structure TPTB desire. It's not a matter of generating certain blocks but rather of keeping blocks from being generated. This is trivial to do with injunction since blocks are generated at the pleasure of corporate service providers (especially network providers) and TPTB 'own the net.' Most of the real world who actually uses Bitcoin (e.g., Coinbase, TigerDirect, etc) are under the same sets of pressures.
The end-play of this 'longest chain' attack will be a bunch of people (such as Peter R) saying '
of course we don't like with the attack on the chain and it is terrible and all, but it's the longest chain(tm) so what can ya do?'
). 
