We do these straw polls every two weeks for two reasons: 1) for fun, and 2) to get a read on the sentiments of the Daily Kos community.
Until two weeks ago, those two reasons worked hand-in-hand to get us a nice look into the Democratic primary race. We’d average 7-8,000 responses, and we’d end up with a nice snapshot of community sentiment. These polls may not be “scientific”, but I have no doubt they were accurate. Just watch how little results change after about 1,000 responses, maybe a point or two. It’s the MoE phenomenon at work.
But that was upended on its head two weeks ago. At 7,000 responses, the voting was tied 48-48. If the goal was to have fun and get a read of community sentiment, mission accomplished! But then the word went out in Bernie Sanders circles to freep the poll, and a couple of hours and 5,000 votes later, it was Sanders 65-32. (Just about every single vote after 7,000 went to Sanders.)
|
11/3 |
10/20 |
9/29 |
9/15 |
9/1 |
8/19 |
8/5 |
7/21 |
7/9 |
6/23 |
6/9 |
Sanders |
65 |
53 |
62 |
62 |
58 |
58 |
58 |
57 |
67 |
63 |
69 |
Clinton |
32 |
41 |
29 |
29 |
34 |
34 |
35 |
36 |
29 |
31 |
24 |
Other |
1 |
1 |
3 |
3 |
2 |
2 |
3 |
2 |
2 |
1 |
3 |
No Clue |
2 |
2 |
3 |
3 |
3 |
3 |
2 |
3 |
2 |
1 |
2 |
n= |
12.3K |
8K |
7.8K |
7.2K |
7.4K |
7.5K |
8.1K |
7.1K |
7.4K |
8K |
14.2K |
Now freeping—getting people around the web to vote on a web poll—isn’t against the rules (see reason #1). From a site standpoint, it’s great! We get new people at Daily Kos, and hopefully a few create new accounts. So I don’t mind it. Nothing stopping the Clinton crowd from doing their own freeping. But it does mean that reason #2 doesn’t apply to those final results. I’m going to go with the final numbers, because those are the rules and I’m sticking with them. But this community was 48-48 two weeks ago. Let’s see where things are this week.