Sanders ran 13 delegates better than we predicted in the five states he contested: 4 delegates in Oklahoma; 4 in Massachusetts, 2 in Vermont; 1 in Colorado; 2 in Minnesota.
Clinton ran 15 delegates better in what seemed liked marginally contested states: 5 delegates in Alabama; 4 in Georgia; 3 in Tennessee; 2 in Virginia; 1 in Texas.
Oklahoma was the only state we had wrong when we predicted a Clinton victory.
We only incorrectly predicted the winner in seven contests: Oklahoma PLEO and CD 1, 2, and 4; Massachusetts CD 6 and 9; Texas SD 30 (the margin there is currently 57 votes.)
What follows are the specific contests we got wrong and the way we did. Correct predictions are not included.
ALABAMA
Clinton + 5 delegates over predictionPLEO
Prediction: 5-2 ClintonActual: 6-1 Clinton
CD 4
Prediction: 2-2 tieActual: 3-1 Clinton
CD 5
Prediction: 2-2 tieActual: 3-1 Clinton
CD 7
Prediction: 7-2 ClintonActual: 9-0 Clinton
AMERICAN SAMOA
PerfectARKANSAS
PerfectCOLORADO
Sanders + 1 over predictionCD 1
Prediction: 5-3 SandersActual: 4-4 tie
CD 3
Prediction: 3-3 tieActual: 4-2 Sanders
CD 7
Prediction: 3-3 tieActual: 4-2 Sanders
GEORGIA
Clinton + 4 over predictionCD 2
Prediction: 4-1 ClintonActual: 5-0 Clinton
CD 4
Prediction: 4-2 ClintonActual: 5-1 Clinton
CD 12
Prediction: 3-2 ClintonActual: 4-1 Clinton
CD 13
Prediction: 4-2 ClintonActual: 5-1 Clinton
MASSACHUSETTS
Sanders + 4 over predictionAt Large
Prediction: 11-9 ClintonActual: 10-10 tie
PLEO
Prediction: 7-5 ClintonActual: 6-6 tie
CD 6
Prediction: 4-3 ClintonActual: 4-3 Sanders
CD 9
Prediction: 4-3 ClintonActual: 4-3 Sanders
MINNESOTA
Sanders +2 over predictionCD 5
Prediction: 5-4 SandersActual: 6-3 Sanders
CD 8
Prediction: 3-3 tieActual: 4-2 Sanders
OKLAHOMA
Sanders + 4 over predictionPLEO
Prediction: 3-2 ClintonActual: 3-2 Sanders
CD 1
Prediction: 3-2 ClintonActual: 3-2 Sanders
CD 2
Prediction: 3-2 ClintonActual: 3-2 Sanders
CD 4
Prediction: 3-2 ClintonActual: 3-2 Sanders
TENNESSEE
Clinton + 3 over predictionCD 6
Prediction: 2-2 tieActual: 3-1 Clinton
CD 8
Prediction: 3-2 ClintonActual: 4-1 Clinton
CD 9
Prediction: 5-2 ClintonActual: 6-1 Clinton
TEXAS
Clinton +1 over predictionAt large
Prediction: 31-17 ClintonActual: 32-16 Clinton
SD 9
Prediction: 3-1 ClintonActual: 2-2 tie
SD 19
Prediction: 3-2 ClintonActual: 4-1 Clinton
SD 20
Prediction: 3-2 ClintonActual: 4-1 Clinton
SD 22
Prediction: 2-2 tieActual: 3-1 Clinton
SD 24
Prediction: 4-2 ClintonActual: 3-3 tie
SD 30
Prediction: 2-1 ClintonActual: 2-1 Sanders
VIRGINIA
Clinton plus 2 over predictionCD 4
Prediction: 4-2 ClintonActual: 5-1 Clinton
CD 10
Prediction: 3-3 tieActual: 4-2 Clinton
VERMONT
Sanders +2 over predictionCDS:
Prediction: 9-2 SandersActual: 11-0 Sanders
*Math mistakes.
As you might expect with so many delegates at play, there were also a few cases where our assessment of the situation was correct but slight errors were made in mathematical calculations. Fortunately for us this time, such mistakes largely cancelled each other out.
In Virginia, we simply added our individual totals incorrectly, wrongly listing a 65-30 overall split when the allocations we made actually added to 60-35. This led to a plus 5 error on Clinton’s behalf. In contrast, in Texas, we misapprehended the number of PLEO delegates at 24 when there were actually 29. That led us to list a 140-82 split when our separate totals should have produced a 145-77 total. This error was plus 5 on Sanders’ side.
Finally, in Oklahoma, we thought there were 39 delegates when in fact there were 38. Since we picked Clinton to win, our original estimate had her one delegate too high.
0 comments:
Post a Comment