Bayesian stats predict world series
Ok, so I found this site:
http://spartan.ac.brocku.ca/~jyuen/MLB/mlbpred.html
and it's stats are based on Data up through September 10th, which is fine by me.
They have all of the playoff contenders essentially right (based on highest probability) except for:
1) Boston getting the East and MFY getting the Wild Card
2) They have Cubs (.35) slightly better than Astros (.32) for the NLWC
3) They have Oakland (.50) better than Anaheim (.47) for the West
Nevertheless, I like their theory. Leaves us with:
Round 1:
Boston (.65) beats Anaheim (.29)
MFY (.41) beat Twinkies (.34)
Cards (.67) beat Dodgers (.34)
Atlanta (.56) beats Roger (.15)
Pennant:
Boston (.43) beats MFY (.15)
Cards (.39) beat Braves (.28)
WS:
Boston (.24) beats St Louis (.22)
Any numbers that have good news are sound in my book.
-=>J
I can't argue with this.
Interesting that the Cards stay around 22% for the entire season...
How 'bout a quick primer on 'Bayesian'.
Then when the RS win the World Series, I'll know why my spam filter works.
Unfortunately, I think this all means that the more spam we get between now and Halloween, the less likely the Red Sox are to win the World Series.
Astros over Braves.
Dodgers over Cardinals.
Twins over Yankees.