Worse, my field is data analytics and statistical modelling—I'm a member of the group which is comprised of some of the biggest blowhards and liars to ever grace humanity.Are you a salesperson for the company? Because that sounded like a marketing pitch.
I don't doubt there are analytic models out there, but they all cannot account for the x-factor variable(s) of the individual players (what's their fitness, are they distracted by the fans/weather, did they have a good constitutional that morning, are they having WaG issues,etc) during the free-form run of play.
I'm not trying to rag on you, but advanced stats for soccer are rather silly and self serving.
The variables you list are definitely legitimate (and always potentially problematic for results) but should never be a reason to avoid or discount statistical models as helpful tools (otherwise you'd have to throw out nearly all measures/metrics as problematic). All "good" models are merely the "least inaccurate" set of explanatory or predictive equations interpreting the "least bias" data available at this time. They are all wripe for destruction and reconstitution at any time as new, more accurate information and translations are discovered/created.
I rarely think tools are silly—I very regularly think the craftsman is, though, such as the armchair twit you mentioned in your original response. I also think the xG models are limited in very real ways (as highlighted in my original and subsequent posts) and hope they are improved.
