My Lords, unfortunately, I was not able to speak in the Second Reading debate, even though I was present, as I could not be certain that I could attend the whole debate. However, I assure the Committee that I intend to speak briefly and only to the amendment.
I write a football column each week for the Times, and have done for more than a decade. The column is concerned with the quality of football teams and helps readers to distinguish between noise and signal. A team that wins 50% of its games will not do so by regularly winning, then losing, winning then losing; it will do it in clusters. When statisticians assessed those clusters, they discovered that they are randomly distributed. In other words, a team will win a cluster but that may be just because it has a run of random results and then it will win, win, win, draw, draw, lose, lose in that cluster. There is a trophy that is awarded entirely for randomness: it is called the Barclays Manager of the Month. When a team’s cluster of wins coincides with a calendar month, its manager becomes Barclays Manager of the Month. In the following month, when that team loses its games, something occurs that is called regression to the mean, and the manager gets fired. This is a perfectly simple statistical concept that ought to be applied to the Bill.
It is not a very good idea to try to set a target in law, on which Parliament must report, that is attached to a single year’s variable data. It is much better to try to find a period that might represent some sort of significance over a long period of time. A single year cannot do that and five years can attempt to do that. Even five years is quite a short period but it is certainly a great deal better than a single year.
This is a technical objection to the Bill, even though others may think it goes to the heart of the argument for it, but the Bill would certainly be greatly improved if a simple concept of randomness was agreed so that we do not have a law for randomness in the same way that we have a cup for it.
12.30 pm