England, ICC Rankings and Central Contracts

Whilst I was at a bit of a loose end I found this, ICC historical test match rankings going back to 1952.

The graph below is England's ICC ranking points from 1952 to 2020, to simplify the exercise I've taken one ranking a year as at March  (or as close as I can get to March if no March results.)

What I was interested in was whether the ICC rankings could provide some insight into how effective the decision was to put England players onto central contracts in 1999, an issue I first looked at here. The year 2000 is marked by the vertical line on the chart,   I'll come clean and say I was expecting the evidence to show central contracts to have no impact on test performance.

But I have to admit the data shows something rather different.   The average ranking in the period prior to central contracts is 101 compared to 104 for the period after central contracts.  Given my complete lack of statistical know - how, I'm on shaky ground here, but I used various on line gizmos to give me a standard deviation for the data and from that went, via a z score, to a p score.  If I did all this correctly, and it's quite likely I didn't, I get a probability of 14% that the improvement post 2000 is purely due to random movements, or, to put it another way, an 86% chance the changes in 2000 led to improved performance.  If a scientist was looking at a p factor of 14% they'd probably conclude it was insufficient to prove the 2000 changes improved performance, but we are in a real world situation and I think we have enough (my bad stats excluded) to conclude central contracts probably had a beneficial effect.

Just a final note.  The changes were bigger than just the introduction of central contracts; there was also a shift to a two division county championship with all games played over 4 days and this might be another factor in improving performance post 2000.