In our last poll on 27 JUN, 61% of you predicted that CL would hit $154 in the front month before it hit $126, and only 9% of you predicted that CL would hit $126. Well, we've briefly broke that number today

I'm wondering which benchmarks we're using for the oil price, because Tapis closed at US$152.38 on 16Jul08, I didn't watch for it but it may have hit $154 the next day before this week's drop.

So rather than saying that the price of crude dropped to $126, I think it would be more accurate to say that since your last poll, the price of crude rose steadily from $126 to $154, but in the last week has traded in the $126-$154 range.

I think it might be worth making these polls on a regular schedule - every month, or every quarter. Any period chosen is arbitrary - 15Jun-15Jul is neither more nor less meaningful than 01Jun-01Jul - but having the polls at these random intervals turns it into a cherry-picking exericse. By bringing up a poll on a price high we overemphasise the rises, by brining it up on a low we overemphasise the drops.

For example, because you put the poll today at a low you're getting a poll result of around 40% saying it'll rise; had you put the poll a week ago when Tapis was near the $154 mark, I think you'd have got the old results of 60% or so saying it'd rise.

It'd be a bit like polling Americans on how they thought WWII was going after the battle of Monte Cassino, and then a week later after the landings at Normandy. By choosing when to do the poll you'd get different results.

It'd be better to have the polls at regular intervals.