[82lsju]

I was trying to get Gemini and Claude to give me some exact win percentages on west coast teams playing early games on the east coast, and basically neither could do it unless I gave it a whole database of games myself. Pretty dissapointed in our machine gods TBH, they can be pretty lazy and stupid.

not college football but here's an analysis of NFL teams

https://maxmoacanin.substack.com/p/the-n...sadvantage

which asserts that after adjusting for home field advantage and relative team strength traveling east for a game is not a problem

Over the past 36 seasons, west coast teams did not underperform the model’s expectations when traveling east.

In the 494 games specified above:

• The west coast road had an average win probability of 44.1 percent, slightly lower than their 46.8 percent win rate. In other words, they won at a slightly higher rate than expected.
  
• They either matched or outperformed their point-spread benchmark 56 percent of the time (278 games).
  
• They covered the model’s point-spread benchmark by 1.61 points, on average.
  

And the results were similar for early kickoffs. Of the 494 games, 367 of them started before 2 p.m. EST. In those games:

• The west coast team won 46.6 percent of the time—slightly higher than the model’s benchmark, which gave them an average win probability of 43.8 percent.
  
• They covered the model’s point-spread benchmark spread by 1.35 points, on average.
  

[82lsju]

I was trying to get Gemini and Claude to give me some exact win percentages on west coast teams playing early games on the east coast, and basically neither could do it unless I gave it a whole database of games myself. Pretty dissapointed in our machine gods TBH, they can be pretty lazy and stupid.

not college football but here's an analysis of NFL teams

https://maxmoacanin.substack.com/p/the-n...sadvantage

which asserts that after adjusting for home field advantage and relative team strength traveling east for a game is not a problem

Over the past 36 seasons, west coast teams did not underperform the model’s expectations when traveling east.

In the 494 games specified above:

• The west coast road had an average win probability of 44.1 percent, slightly lower than their 46.8 percent win rate. In other words, they won at a slightly higher rate than expected.
  
• They either matched or outperformed their point-spread benchmark 56 percent of the time (278 games).
  
• They covered the model’s point-spread benchmark by 1.61 points, on average.
  

And the results were similar for early kickoffs. Of the 494 games, 367 of them started before 2 p.m. EST. In those games:

• The west coast team won 46.6 percent of the time—slightly higher than the model’s benchmark, which gave them an average win probability of 43.8 percent.
  
• They covered the model’s point-spread benchmark spread by 1.35 points, on average.