Top Senate predictions

by: sleepleswithamoustache

Fri Jan 29, 2010 at 18:18:48 PM EST

Deviations	Predictor	Brown	Coakley	Kennedy
0	Actual	51.9	47.1	1
0.2	Ken Pittman	52	47	1
0.2	Paul “Cool Cal“ Ferraro	52	47	1
1.6	Mike “DD4RP” Rossettie	52	46.2	1.6
1.8	yankeepundit	51	48	1
1.8	Republican Ram Rod Radio	51	48	1
2.5	RMG prediction average	51	46.9	2.4
10.3	BMG prediction average	46.7	50.4	2.8

Congrats to Ken Pittman and Paul "Cool Cal" Ferraro. You really couldn't have gotten it any more dead on. Overall, RedMass got these predictions really close, and the RedMass average prediction was closer than any single BlueMass prediction.

Full listing (including BMG predictions in blue) below the fold, along with some possible reasons for the discrepancy.

Possible reasons for why RedMass did so well?

1) In a tossup race, BMG erred on the side of Coakley, RMG on the side of Brown. Race broke to Brown, RMG got lucky.

2) RMG better at predicting outcomes of elections.

or 3) (based on feedback from here) With nothing actually riding on the predictions, each candidate's supporters wants to motivate their base by saying their own candidate will win in a squeaker if they just GOTV enough.

Future data may help answer that question.