Most percentages look normal with a steady climb. But that -4 rating difference is weird. A 61% win rate there might just be an anomaly.
I combined games from regular GM mode with some PVP matches. I left out ladder games since their win rates appeared unusually high.
For consistency, all my coaches were uniformly set at 90 for every stat. I’m still trying to figure out how exactly coaching influences these outcomes.
Hope you find this analysis interesting! Let me know your thoughts.
I’ve been tracking my own games too, and your findings match up pretty closely with what I’ve seen. The -4 rating difference anomaly is interesting though. In my experience, that’s often where you see crafty players pulling off upsets with unconventional strategies.
One thing I’ve noticed is that player morale can have a huge impact, especially in close rating matchups. I’ve had lower-rated teams go on hot streaks and take down much stronger opponents when confidence is high.
Regarding coaches, I’ve found that specializing in certain areas rather than having uniformly high stats can give an edge. For example, a coach with maxed offensive skills paired with a defense-heavy team can create some interesting synergies.
Have you looked at how home field advantage factors in? In my games, I’ve seen it swing results by 5-10% pretty consistently. Might be worth factoring that in if you expand the analysis further.
yo, great analysis! that -4 rating diff is wild, maybe those underdogs just got lucky? lol. i’ve noticed player form can make a huge difference too. like when my star striker’s on fire, we can beat anyone. have u looked at how injuries affect these stats? that could explain some upsets. keep up the good work man!
Your analysis provides valuable insights into the correlation between team rating differences and win rates. The data largely confirms the expected trend of higher-rated teams winning more frequently. However, the anomaly at the -4 rating difference is indeed intriguing and warrants further investigation.
One factor to consider is the potential impact of player skill and strategy, which can sometimes overcome rating disparities. Additionally, the sample size for each rating difference category varies significantly, which may affect the reliability of some percentages.
It would be interesting to see how these results compare across different game modes or seasons. Perhaps expanding the dataset or analyzing specific matchups could reveal patterns that explain outliers like the -4 rating difference.
Your decision to exclude ladder games and standardize coach stats enhances the consistency of the data. As you continue to explore coaching influences, it might be worthwhile to experiment with varying coach attributes to quantify their impact on win rates.