Jake Rubinstein
jlrube.jakerubinstein.com
Jake Rubinstein
@jlrube.jakerubinstein.com
Splitting time between here and my X. Slowly migrating.

/ Survey Methods / random pictures of my cat / #f1 / US politics

Data/Polling/Quant methods at HIT Strategies.

More active on the birdsite with the same handle

https://linktr.ee/jlrube
I'm guessing someone on election Twitter had done this already, though I haven't seen it.

I'm planning to do a bunch more analysis but want to do it in a coherent way, not as random tweets
July 24, 2025 at 2:05 AM
Oh, this is great stuff. I'm glad to see it. I missed it when it came out.
July 23, 2025 at 2:13 PM
Yours had the batch elimination, mind does not. But yeah, I did miss yours
July 23, 2025 at 3:55 AM
After I'm done with my writeup, I'd be happy to work with you all to answer anything specific though!
July 23, 2025 at 1:51 AM
BOE published the raw CVR on their site this afternoon.

I'm undecided if I'm broadly sharing my cleaned version, but definitely won't until I've written the longer piece I'm planning.

But there isn't a technical reason you couldn't run something like that.
July 23, 2025 at 1:48 AM
After I'm done with my writeup, I can probably make my cleaned CVR available for something like this, even if I don't just post the whole file for anyone to use.
July 23, 2025 at 12:51 AM
Yeah, saving that for my longer write up!
July 23, 2025 at 12:23 AM
Yeah the CVR was posted at like 5. It's a pain to work with. Only reason I was so quick is I did it in 2021 and the spec didn't change

vote.nyc/page/electio...
Election Results Summary | NYC Board of Elections
vote.nyc
July 23, 2025 at 12:22 AM
@xkonstantinidis.bsky.social would you believe all the processing code for this is in stata 😂
July 23, 2025 at 12:11 AM
There is a lot more analysis to come and I'm planning to write a longer form article examining the CVR.

H/T to @benjrosenblatt.bsky.social
for his geography data which is vital create a comprehensive CVR.
July 22, 2025 at 11:46 PM
3) Ramos's Cuomo endorsement didn't help much, Mamdani received 13% of the transfers to Cuomo's 16%. 29% of Ramos transfers where exhausted.

4) Lander(!) was the largest beneficiary when Adams was eliminated, getting 31% of the transfers.
July 22, 2025 at 11:46 PM
2) DREAM worked, when Lander was eliminated in round 11, more votes were exhausted than transferred to Cuomo.

Mamdani received 56% of Lander's transfers, 23% were exhausted, and 21% went to Cuomo.
July 22, 2025 at 11:46 PM
Just about 20k of them
July 15, 2025 at 8:41 PM
No, there were only about 13k Wiley only voters.

More so that Wiley voters ranked other folks who had already been eliminated (e.g. Stringer, Morales, and Donovan)
July 15, 2025 at 5:13 PM
At the moment no, because I'm in the middle of reworking my code to prep for 2025. But NYCCFB reports there were about 125k bullet voters for mayor and about half of those were for Adams

This is an older table I ran on most common ballots which is also helpful context here.
July 15, 2025 at 4:44 PM
okay, fair, that is less boring than I was imagining
July 15, 2025 at 3:43 PM
Non RCV races would be boring as hell
July 15, 2025 at 3:40 PM
Turns out this data was revised by the NYT.
June 30, 2025 at 9:23 PM
Turns out NYT revised this data substantially. Deleting
June 30, 2025 at 9:23 PM