Here’s the game that’s being played all over the Fourth Middlesex senate district.
Given: 14,500 people will vote in the primary.
The vote will split as follows:
Burlington 1,800
Lexington 2,100
Billerica 2,400
Woburn 2,900
Arlington 5,300
Insert candidates,
split vote among candidates,
develop scenarios how a candidate can win the race.
Please share widely!
I looked at Tuesday’s vote in the Billerica special election, and applied that percentage of the primary vote to the other communities in the district.
just give the number of registered voters per town and go from there? seems odd to me to apply the voting rate from one town to other towns.
Past performance in primary elections is a better baseline predictor than the number of registered voters. Some towns are just plain better at bringing out voters in a primary election.
<
p>
Polling firms use this technique all the time. If your poll is to be predictive for the election, you need to calculate your data based on likely voters, not registered voters.
<
p>
A GOTV drive can change these numbers. A candidate can look at the numbers and say that he/she needs to raise voter turnout in certain towns or precincts in order to win. That goes into the equation when you are deciding if you can mount a winnable campaign.
i agree that past performance in a town is a useful predictor for that town. but you said you were going to use past performance of billerica and apply that percentage to other towns. not the same thing? is participation in billerica really a good indicator of participation in, say, arlington? i’m not so sure. the communities are quite different.
While there are many primary elections held in the district, under many different circumstances, there haven’t been many special elections. By using regularly scheduled primary votes for all communities, and using Billerica’s numbers from last Tuesday to get a special election effect, I came up with the best baseline that I could produce for every municipality in the district.
<
p>
Again, candidates will attempt to move these numbers to their advantage through a GOTV campaign, but if I were planning campaign strategy, I would be comfortable using these numbers.
Laurel's point is clearly a good one. It's obvious to most junkies that even recent special election data from Billerica has absolutely nothing to do with a different kind of special election in a community… especially transfer between Arlington and Billerica, which were only about 4,000 people apart in the 2000 census while Arlington was twice as diverse as Billerica. And I assume the numbers you use as likely voters: Billerica 2400 and Arlington 5300 indicates there has been substantial growth… that or a math error.
Obviously, planning for variables like an Arlington, Burlington and Woburn candidate would move the numbers further. I would love to see a better analysis of these numbers because I do think Pablo's point about margins winning the election is a good one.
If I were planning for a campaign I would hope the other candidates were using these numbers…
in the district from previous non-presidential, non-gubernatorial primaries? Those numbers would provide an interesting comparison for Pablo's estimates – an upper bound for total voter turnout for a special election primary (e.g. barring some burning, yet-to-emerge issue, we won't see more voters come out for a special election primary than for a regular off-year primary), plus they should give us an indication of turnout trends from the different towns.
OK, here are the primary results in the 4th from last time…
City or townPRTPGonsalvesHavernArlington 21212,0287,447Billerica 11112,1402,177Burlington 661,4841,685Lexington 661,1522,687Woburn 12123,4331,723Totals 565610,23715,719
I am guessign the turnout will be far lower than Pablo's estimate…significantly under 10,000. The majority of the vote comes from the Arlington, Lexington section of the district and the balance from Billerica, Burlington and Woburn. If Rep. Marzilli is the only candidate out of Arlington, he has a huge leg up on Reps. Natale and Murphy splitting the Woburn, Burlington, Billerica vote.
Billerica cast 4317 votes in this senate primary. It was a hot primary, with the open governor’s race drawing lots of interest.
<
p>
Billerica’s Democratic primary vote last Tuesday:
<
p>
Donoghue 764
Eldridge 90
Finegold 222
Miceli 248
Tsongas 1002
<
p>
TOTAL 2348
<
p>
There was lots of media play in the local press, but no home town candidate. I figure there will be less casual voters in the state senate primary, but a whole bunch of favorite sons who will be pulling out their troops. I think a turnout in the neighborhood of 55-60% of the 2006 primary is a pretty good mathematical model to use for predicting a baseline turnout.
State reps Marzilli from Arlington, Natale from Woburn, Murphy from Burlington. Peters from Lexington. Donnelly (retired Lexington firefighter). Mooney (former selectman) from Burlington. Anyone else? Is Gonsalves planning to run again?