That makes sense, and thanks for doing the research, I’ve wondered that a few times now so glad to have it confirmed!
- Relatively tight game and one we in which we were definitely lucky to score 3 goals. - Noticeable from the passmaps how they really tried to overload Ngakia - Kayembes was worth 0.01 xG, Chaks goal was worth 0.03 xG and Rajs was worth 0.24 xG - We'd only created 0.17 xG in the first 75 mins of the match which is a concern.
Shows at 0.53 xG on FotMob for me, but 0.91 xGOT (basically the xG if the shot is on target). But yeah there can be a lot of variance between different sources for xG as each will use a different model which tracks different variables. But that's quite a big disparity: Both have Chaks goal as 0.03, Millwalls first is a bit of a spread as well - SB have it as 0.07 and 0.09 xGOT (compared to FotMobs 0.18 xG + 0.47 xGOT), Both are similar for Millwalls last goal - SB have it as 0.1 and 0.5 xGOT (Fotmob have it as 0.12 xG and 0.67 xGOT) So yeah, must be a big difference between the models to have such a large spread for Rajs goal (although SBs is generally regarded as the best!) There were some utterly terrible models when I first got started (that just tracked where the shot happened - so you've got a shot in the middle of the goal but with loads of players + keeper in the way that would give a pretty high xG value that didn't really pass the eye test), but now the best ones take into account GK position, positions of other players, height of the ball etc.
Great to have the stats back thanks @reids ! Also no surprise to me the XG for both teams. Got away with one really there, but I’ll take it!
Thanks for your obviously great analysis, but unfortunately, it's all double Dutch to me Keep up the good work, as I am sure there are many on here who are not as big a numpty as me However, if there is anyway of dumbing it down a bit for the likes of me, I am still interested in your analysis.
Going off the eye test, I don’t think Bachmann could have done much about the two goals, they were both hit hard, the first one was high and went in off the bar and the second one was in the corner so Fotmob sounds fairer to me
Had a quick look at how goal per game across all competitions has been trending under the Pozzos. Clear signs the plan is paying off.
I'm concerned some people are getting ahead of themselves. We slumped to an average 3.67 goals per game today
This is a follow up to the discussion at the end of last season about the "Drawford" phenomenon. I did some analysis over the summer, but never go to post it. I was prompted to post it now, as we have only one draw so far in a quarter of this season. I divided the teams in 9 leagues (4 English, Scottish, French, German, Spanish and Italian) by ranking their wins, draws and loses from most to least common. I identified 5 categories: "Dominating" Teams - Win, Draw, Loss "Upper/Middle" Teams - Win, Loss, Draw "Lower/Middle" Teams - Loss, Win, Draw "Dominated" Teams - Loss, Draw, Win Anomalies - Draw most common. Of the 138 teams, only 10 fell into the anomalies category. Drawford was one of them. You'll never guess who one of the others was. Having started by looking at the "Drawford Phenomenon" I was interested in how much the leagues differed in the balance between the five categories. For example Serie A has no teams in the "Lower/Middle" category. The summary table and data from three of the 9 leagues is presented below.
This reminds me of a thread we used to have which was updated on a weekly basis showing how we progressed each season with a side by side chart of previous seasons. I think it was first done by a member called David M Sawyer. Yourself and reids so some excellent charts, it's always a pleasure to read through them.
