simpathia wrote: The home field advantage has me a bit confused. I measured it last season for central europe top division and for division II.3. (I needed it for the ratings that I publish). Home field advantage for division II.3 was more than 15 points, for the top division only about 4.5 Any thoughts?

I think talent has a lot to do with it. There are bigger talent gaps between the top of Division teams and bottom of Division teams in Division II as opposed to Top Division. Therefore, when the better teams play at home, they will inevitably win by a higher margin potentially making it appear that the win-loss spread (15 points) is more than it actually should be. I think your Top Division 4.5pts spread is probably more accurate when ran among similar teams. This would mirror real life home field advantage which has been measured at between 3-5 pts.

Of course these numbers will always be more exaggerated when the better teams play the lower teams, but when calculated between 10 teams over a season, I would expect regression to the mean/Division average.

However, one thing is certain and that is the GI engine is incredibly unpredictable sometimes.