Conor Kelly

Sometimes I say things. I make a lot of charts

[Conor Kelly @CohoKelly]

@corncobanalysis Can we get an arizona turnout explainer?

[Conor Kelly @CohoKelly]

@NapervillePol How's it line up to 2022?

[Conor Kelly @CohoKelly]

@sixtysixwards Could you share some info on how you deal with conflicting info (e.g., differences in voter numbers, unrealistic ones)?

[Conor Kelly @CohoKelly]

@DjsokeSpeaking @JonahLandergan Oh yeah, that's a big bump on last days of primary. I was one of those!

Thanks

[Conor Kelly @CohoKelly]

@AntiMattersWX @DjsokeSpeaking Thanks! Would be interested to see a take on it. It's a pretty different story from 2020 in an all-mail state

[Conor Kelly @CohoKelly]

@JonahLandergan @DjsokeSpeaking Of course not. Just putting in the context of overall turnout in Dem-leaning states

[Conor Kelly @CohoKelly]

@norbert_d_d @DrSkywalker14 @blockedfreq Idk if there's any baked-in assumptions about how the D/R difference in early votes translates to actual margin (including independents), but it doesn't seem to be configurable in the tool

[Conor Kelly @CohoKelly]

@norbert_d_d @DrSkywalker14 @blockedfreq It's a cool site! My take is:

(1) The default setting has Rs winning E-Day vote 60-40 (that's a setting you can change)

(2) Assuming that the actual difference in raw votes from early vote is just the difference between R and D registered who voted early. Less clear on that

[Conor Kelly @CohoKelly]

@LPDonovan @lxeagle17 Correct answer

[Conor Kelly @CohoKelly]

@ECaliberSeven @votehub And 7/8 specials got bluer than 2020? But the combined total was redder. Is that right?

[Conor Kelly @CohoKelly]

@UnhitchingHour @umichvoter @votehub Totally

[Conor Kelly @CohoKelly]

@umichvoter @votehub What’s the benchmark you’re looking for here? No idea if this was impacted by redistricting. Could be a fun barometer of urban turnout/margin?

[Conor Kelly @CohoKelly]

@admcrlsn As someone who desperately hopes polls don’t miss in favor of Trump again, it is hard to argue against this (even with n=2) or dismiss out of hand

[Conor Kelly @CohoKelly]

@votehub Confirmed my suspicion that I've never lived outside of a liberal bubble before

[Conor Kelly @CohoKelly]

@votehub Had never been able to see precincts around my alma mater before (had been missing from similar NYT tool). Neat!

[Conor Kelly @CohoKelly]

@Ragnarok_1er @NateSilver538 I think Beacon/Shaw is Fox

[Conor Kelly @CohoKelly]

@Ragnarok_1er @NateSilver538 Oh, interesting. When sorting by influence the NYT poll from July shows up first, so I thought it must be missing. Thanks!

[Conor Kelly @CohoKelly]

@NateSilver538 Any reason your "Polls included in our model" table doesn't reference NYT/Siena from a few weeks ago?

It's also interesting to see the high rate of decay in influence for polls over time. Emerson poll today worth 10x polls from a couple weeks ago at a time when polling is scant