Tuesday, October 19, 2010 at 6:44 pm by Darryl
The Marist poll gives Sen. Patty Murry (D-WA) a 48% to 47% lead over real estate property prognosticator and perennial political candidate Dino Rossi (R). The poll of 589 likely voters was taken by live interview from 14th to 17th October, and has a margin of error of 4%.
With this poll, Murray has now led in eight consecutive polls, all taken in October.
The Monte Carlo analysis using a million simulated elections gives Murray 561,018 wins to Rossi’s 426,991 wins. This poll gives evidence that Murray would have a 56.8% probability of beating Rossi in an election held now.
As with the other poll today, I’ll pool it with all the other October polls and do the same type of analysis. The nine polls give 7,368 “votes” of which Murray received 3,599 (48.8%) and Rossi received 3,361 (45.6%). The simulation analysis give Murray 975,800 wins to Rossi’s 23,478 wins. Murray, in an election now, would win with a 97.7% to Rossi’s 2.3% probability. This is ab very slightly lower probability than the 98% probability found from eight earlier October polls.
The Marist poll is the lowest spread (+1%) for Murray from a live-interview poll over the entire election season. (The Washington Poll in May had a +2% spread, otherwise, live-interview polls have been over +6%.) In other words, this poll goes against my argument about voter enthusiasm and robopolls versus live-interview polls discussed earlier today. Well, sort-of…the Marist poll has a relative small sample size, so that the margin of error overlaps most other live-interview polls—all but the Elway poll.