Of course our away form is awful, I don't think anybody is denying that.Roosterbooster wrote:Small sample sizes only give significant results if the difference is vast. Normally you would expect to need a much larger sample size. And the larger the sample size the more valid the result. The tests are prone to error, particularly when the test isnt really designed for the question being asked, and when the sample is so small. But the suggestion is clear. Our away form is awful
My point was that it's not because we set up differently at home and away, it's because a) we're one of the worst teams in the league; b) all teams have worse results away from home; and c) we're at the bottom end of the expected distribution of results away from home (and probably at the top end at home).
Let's say we're expected to win one out of every 6 away games (maybe that should be worse, but I don't think we should be expected to win more given the quality of the oppositon in the PL), to go 17 without a win is hardly outside the realms of probablity and can be attributed purely to variance.