What's new
Fantasy Football - Footballguys Forums

This is a sample guest message. Register a free account today to become a member! Once signed in, you'll be able to participate on this site by adding your own topics and posts, as well as connect with other members through your own private inbox!

Simpson's paradox (1 Viewer)

Chase Stuart

Footballguy
This year, the Bucs are undefeated when playing teams with losing records. So are the Buffalo Bills.

The Bills, on the other hand, are winless when playing teams with winning records. But so are the Bucs.

Both Tampa and Buffalo have the exact same winning percentages against winning and losing teams this season. But Tampa has played 7 teams with losing records while Buffalo has faced 9 teams with winning records. Additionally, the Vikings are also undefeated against losing teams and winless against winning teams.

Cleveland? Like Buffalo, Minnesota and Tampa, they're undefeated against losing teams. But they've won 22% of their games against winning teams, so by that mark, they're doing better than those other three teams.

Is Cleveland better than Jacksonville? The Jags are 5-6, Cleveland 4-7. But Jacksonville has a .200 winning percentage against winning teams (Cleveland .222) and a .833 winning percentage losing teams (Cleveland 1.000). The same example occurs with the 49ers (4-7), who are winless against winning teams and are winning just two-thirds of their games against losing teams.

Cleveland and Jacksonville are the only pure example of Simpson's paradox affecting records in this fashion. Simpson's paradox "is an apparent paradox in which a correlation (trend) present in different groups is reversed when the groups are combined."

Here, the trends are:

Against winning teams, Cleveland (2-7) has a better record than Jacksonville (1-4).

Against losing teams, Cleveland (2-0) has a better record than Jacksonville (5-1).

When the groups are combined, Jacksonville (6-5) has a better record than Cleveland (4-7).

 
If the Browns would have sent Eric Wright to the moon prior to the season they'd be 7-4.
Don't know about Eric Wright but you could substitute Jake Delhomme into that sentence and it also correct. I didn't think it was possible but he is worse than he was last year. A 54.2 % passer rating is atrocious.
 
If the Browns would have sent Eric Wright to the moon prior to the season they'd be 7-4.
Don't know about Eric Wright but you could substitute Jake Delhomme into that sentence and it also correct. I didn't think it was possible but he is worse than he was last year. A 54.2 % passer rating is atrocious.
And he is starting this week. I've been lonely in the Pro-Mangini camp for some time now but this one has me perplexed.
 
so schedule has rewarded a bad team (tampa) and punished a bad team (buffalo)

Tampa is not as good as their record and Buffalo is not as bad as their record.

Got it. In the end though Buffalo actually wins due to have better draft picks next year

 
To address the point you were trying to make Chase, are the wins and losses against Cleveland and/or Jacksonville taken out before determining which teams had winning or losing records?

 
So what does this say about a team like the Washington Redskins -- 3-2 against teams with winning records, but 2-4 against teams with losing records?

 

Users who are viewing this thread

Back
Top