If you're an 'expert determiner' and choose to go your own sweet way rather than obey instructions, then don't. Your decision is likely to be set aside
"Bingham!" yelled the teacher, "For the umpteenth time, I am telling you to do as you are told!" Even in short trousers, I never could fathom the command "do as you are told".

"Why?" I asked.

"Don't you get smart with me, young man."

"But, Sir, I only want to know why in principle I should do as I am told, and I only want to know why in the particular I should do as I am told." The answer was a clip round the ear.

The Court of Appeal is very like my old Mr Chips. The court says that if someone is appointed to carry out an "expert determination", that person will damn well "do as they're told".

It seems to me that what goes for an expert determiner goes for an adjudicator, an arbitrator and a judge. Go further; if you want a formal decision by any one of these dispute deciders to stick, make sure they do as they are told.

The expert determination that raised this question was to test the quality of a boatload of gas oil (Veba Oil Supply and Trading GmbH vs Petrotrade Inc ("The Robin") [2001]). The rules for testing required the application of a method called D1298 by an independent third party. Fail the test and the boat goes home unloaded, no messing about – the decision is final and binding. Now then, D1298 has a margin of error of seven thousandths of one per cent. Tiny, isn't it? So, when a test was done in Antwerp and there was a good result, the contractual specification was satisfied. It was just like an adjudication or arbitration. The boat sailed. The cargo was now sold on to the Lebanese ministry of oil on precisely the same terms. So, on arrival, the D1298 test was performed again. Snag is, this time, the gas oil failed the test. Oh dear. The cargo was sent away. What's up?

It was no business of the court to look at the effect the departure from instructions had on the end result. Courts hesitate to speculate

Here comes an odd twist. The first expert confessed that he had not used test D1298 – in other words, he hadn't done as he was told. He had used something called test D4052. In fact, in Antwerp they always do because D4052 is 10 times more accurate than what they regard as a lesser analysis. In court, the opponents agreed with each other that if the cargo passed D4052, it would pass D1298 with flying colours. But, said the complainant, the expert's decision was not final and binding, because the wrong testing method was used. Tell this to anyone in the world of commerce and almost all of them would fall about laughing. Do you really mean that the much better test was void simply because the specification called for the lesser one?

In court, it was argued that it didn't make any odds; put in a legal way, "there was no departure from instructions in a material respect". The job had been done properly. No damages were payable. The other side argued that it only had to show a clear departure from the instructions. It was no business of the court to look at the effect the departure had on the end result. Given what you have just read, I bet you're not surprised that they didn't want the court to look at the effect of going against instructions. I also bet you know what's coming next: the Court of Appeal said the decision-making expert was in error. His decision was set aside. He didn't do as he was told.

This is the way the court sees it. First, the contract lays down the procedures between the parties and the decision-maker. Such decision-making will frequently be pursued at speed. To ask what the ultimate effect is likely to be if the expert doesn't follow the rules is probably impossible to predict at the time he does it.

The effect of the change may be a long way down the line. Of course, it may be that the failure to follow instructions is so trivial and immediately obvious that it could make no possible difference to either party.

Having said that, the court is adamant that it is not its job to weigh the importance of a stipulation in a contract for third-party decision-makers. It is not often possible to say with confidence that "the departure from instructions could not have affected the result". Courts hesitate to speculate.