A colleague of mine wrote about match fixing in sumo wrestling
http://www.playingpasts.co.uk/articles/general/not-pushing-their-weight-corruption-in-sumo/ I know it's not obviously related but there is beginning to be more research into fixed games by academics. They can't prove anything but by using probability and patterns of behaviour, such as how many penalties give by a named ref to a particular team compared to the average, they can point to the suggestion that something isn't quite right.
None of us have enough stats, but wouldn't it be great if we could plot every incident in a game, link it to the competing teams and the referee to see if patterns emerge which are not consistent with averages? Sky could probably do this now, but probably wouldn't want to compromise the competition. No referee or footballer is likely to say he took part in a fix of any kind, but we know it's happened in the past. We also can work out whether some refs give more to certain teams than others. At the moment we all have a feeling that something's not right, perhaps it's time to gather the stats, use probability theory and comparisons?