Questions on Nasty Events

Elsewhere on this site there is a debate going on about questions that relate to deaths, moral issues aside it is true that some work has been done on the value to political analysts of predictive markets.

I myself have created questions regarding terrorist interruption of the oil supply, it appears to my mind that my questions are faulty. If we want to help stop events through predictive markets we need to measure attempts, not successes, as thwarted attempts should be priced into a market for successful events as the same as if nothing happens, as that is how they pay off.

Nothing happens payout on no.
Event thwarted, payout on no.

Successful attack, payout on yes.

We need to move event thwarted into the same stable as event successful, and then the price in the market may give us a true measure of the likelihood of an attack, this information may be of use to analysts tasked with dealing with these events.

To help predict terrorist attacks we need questions that state, for example,

When will the next attempted (successful or otherwise) terrorist attack take place on a major oil refinery?

So far so good, but how do we judge attempts, rather than successes? Any suggestions?

NB: I am doing a PhD at Uni of York, UK in predictive markets, where I am a Teaching Fellow in Public Strategy.

A few stray thoughts...
An attack without bloodshed is still an attack.Death and injury are not always the purpose of the attack.

If authorities round up a couple of people who are basically flapping their gums saying "Let's blow something up somewhere" the media IMO have a tendancy to headline "terrorist plot foiled..." when really they could just as well be saying "Do I want to go fishing? Which spot shall I go to?". This to me doesn't count as an attempt.
If however action has begun to execute the attack then an attack has begun. It just remains to be seen if is thwarted, botched, aborted or successfull. By action I mean detailed planning, accumulation/production of materials and then on to the execution of the attack. After all an attack, successful or not, is not one isolated event (eg something going BANG) but a series of events that all must be undertaken in order for the end BANG to occur.
How do we judge attempts?
If the question read "When will the next terrorist attack take place on a major oil refinery? " I would argue that the question could be settled even if the attack was stopped in the planning stages. While the end objective of the attack might be to make the oil refinery blow up you can't say that no attack took place if it doesn't end up going ka-boom.
Same thing for the question you posed above. It's just states the obvious to my way of reading it. If however it was worded "When will the next successful terrorist attack take place on a major oil refinery?" that is then asking about a specific outcome of an attack and that is a different question entirely.

I think that the problem is not always in the questions that are being asked but in the preconceived ideas people bring with them when they look at the question. As with many other things, just because many people believe a thing to be true it does not follow that it is true.

i agree its all in the way a question is pharsed. the question can only have so many words. so you gotta read the background and settlement details CAREFULLY. when in doubt flag the question for clarifaction.

You could judge an attempt by someone being charged or arrested for one - it's usually highly publicised when there is a terrorist attack that has been planned without being executed.

If you just put that in the settlement details, it's a reasonable definition for a 'no' settlement.

But if being used for prediction purposes we want actual attempts, not a couple of guys talking about it in the pub/mosque and being charged, also if we are looking at specifics, such as the oil questions I asked then how do we allow for that.