Latest / Decision Desk HQ Podcast / Episode 38: Our Forecast Shows A Tight Race for Congress

Transcript

Episode 38: Our Forecast Shows A Tight Race for Congress · Decision Desk HQ Podcast

• 0:00
  SPEAKER_00: Democrats have about a 38% shot of having both the House and Senate
• 0:03
  SPEAKER_00: and Republicans have a 37% shot of having
• 0:05
  SPEAKER_01: both the House and Senate.
• 0:07
  SPEAKER_01: We're not going to see a 30 or 40 seat swing here.
• 0:09
  SPEAKER_01: We're going to see a 10 or 15 seat swing.
• 0:12
  SPEAKER_01: And if we don't see a 10 or 15 seat swing, it could go either way.
• 0:16
  SPEAKER_02: There's still a very steep climb for Democrats for the Senate,
• 0:19
  SPEAKER_02: but it's not so steep when the environment is kind of has shifted in their favor
• 0:23
  SPEAKER_02: and is kind of more on their side of the field.
• 0:37
  SPEAKER_00: Welcome to the Decision Desk HQ podcast.
• 0:40
  SPEAKER_00: I'm your host, Geoffrey Skelley, the Chief Elections Analyst here at Decision Desk HQ.
• 0:46
  SPEAKER_00: You know,
• 0:46
  SPEAKER_00: we recently released our 2026 midterm election forecast for the US Senate and
• 0:51
  SPEAKER_00: the US House.
• 0:52
  SPEAKER_00: And the top line numbers suggest that both chambers are legitimately up for grabs.
• 0:57
  SPEAKER_00: In the House,
• 0:58
  SPEAKER_00: our forecast shows Democrats are about three and five favorites to take a majority
• 1:01
  SPEAKER_00: with the average outcome being a 225 to 210 edge for Democrats.
• 1:07
  SPEAKER_00: Which would be a net gain of 10 seats for them right now,
• 1:09
  SPEAKER_00: with Republicans having a 220 to 215 advantage.
• 1:12
  SPEAKER_00: And then in the Senate,
• 1:14
  SPEAKER_00: Republicans are also about three and five,
• 1:17
  SPEAKER_00: have about a three and five chance of winning the majority and keeping the majority
• 1:21
  SPEAKER_00: that they currently hold.
• 1:22
  SPEAKER_00: However,
• 1:23
  SPEAKER_00: with an average seat total of 50 seats for each party,
• 1:26
  SPEAKER_00: which would represent a three seat gain for Democrats,
• 1:29
  SPEAKER_00: but not quite enough for them to get a majority because Vice President J.D.
• 1:32
  SPEAKER_00: Vance would break ties.
• 1:34
  SPEAKER_00: But, and we'll talk about this some,
• 1:36
  SPEAKER_00: counterintuitively,
• 1:37
  SPEAKER_00: both parties actually have a very real chance of winning control of both chambers
• 1:41
  SPEAKER_00: to hold power.
• 1:42
  SPEAKER_00: So even though our forecast makes each party slight favorites,
• 1:45
  SPEAKER_00: the fact that they're only slight favorites creates a lot of opportunities
• 1:48
  SPEAKER_00: for either party to have power over both at the end of the day.
• 1:52
  SPEAKER_00: So in this episode, we're going to talk a little bit more about that.
• 1:55
  SPEAKER_00: We're going to talk about the ins and outs of the forecast, including how it works.
• 1:58
  SPEAKER_00: We'll get into a breakdown of some of the key Senate races.
• 2:01
  SPEAKER_00: We'll also talk a bit about the broader electoral picture and how that pertains
• 2:05
  SPEAKER_00: the Senate and the House.
• 2:07
  SPEAKER_00: And to do all this,
• 2:07
  SPEAKER_00: I am joined by two of my Decision Desk HQ colleagues,
• 2:11
  SPEAKER_00: Scott Tranter,
• 2:12
  SPEAKER_00: the Director of Data Science here at DDHQ,
• 2:14
  SPEAKER_00: and Maddisen Murrell, a Data Science Fellow who helps run the forecast.
• 2:18
  SPEAKER_00: So Scott and Maddisen, welcome to the Decision Desk HQ podcast.
• 2:22
  SPEAKER_01: Thanks.
• 2:22
  SPEAKER_01: I think we're underselling Maddisen though.
• 2:24
  SPEAKER_01: Maddisen is the one who looks at the forecast five times a day,
• 2:30
  SPEAKER_01: as well as like 3 a.m.
• 2:31
  SPEAKER_01: when we send her messages like what the hell is going on with California
• 2:34
  SPEAKER_01: or whatever.
• 2:36
  SPEAKER_01: And I believe Maddisen,
• 2:38
  SPEAKER_01: you even rebuilt a lot of the historical data as we restructured it this year,
• 2:42
  SPEAKER_01: right?
• 2:43
  SPEAKER_01: So basically as well.
• 2:45
  SPEAKER_02: Yeah,
• 2:46
  SPEAKER_02: I contributed to a lot of the training set and kind of what goes into the forecast
• 2:50
  SPEAKER_02: and getting that set up.
• 2:52
  SPEAKER_02: Yeah, I run it on demand basically.
• 2:56
  SPEAKER_02: So at least minimum once a day, unless something crazy happens.
• 3:01
  SPEAKER_02: So yeah, it's encoded into my brain a little bit at this point.
• 3:07
  SPEAKER_00: Well, I think to some extent, the election picture is encoded in us.
• 3:14
  SPEAKER_00: So it's gonna be really fun to talk about sort of what our forecast is saying
• 3:18
  SPEAKER_00: a way of illuminating what exactly is happening.
• 3:21
  SPEAKER_00: And I think also maybe Pushing back on some conventional wisdom, perhaps in some cases.
• 3:27
  SPEAKER_00: So, you know,
• 3:28
  SPEAKER_00: just to sort of start out with that,
• 3:30
  SPEAKER_00: let's talk a little bit about the overall electoral picture.
• 3:33
  SPEAKER_00: And so I'm going to bring up the Senate map.
• 3:37
  SPEAKER_00: So again, our current forecast here is 50-50 is sort of our average outcome.
• 3:44
  SPEAKER_00: And if you look at sort of a seat by seat breakdown, that's roughly where things are.
• 3:50
  SPEAKER_00: You know, we make Republicans slight favorites here with a 58% shot at winning a majority.
• 3:58
  SPEAKER_00: Maddisen,
• 3:59
  SPEAKER_00: you spend a lot of time looking at this and I was curious if you had sort
• 4:02
  SPEAKER_00: of thoughts on the,
• 4:05
  SPEAKER_00: what is maybe most responsible for,
• 4:08
  SPEAKER_00: obviously we know Republicans came into this with a lot
• 4:10
  SPEAKER_00: of red leaning seats they were protecting.
• 4:13
  SPEAKER_00: And so that made it maybe a little easier for them to hold onto the majority.
• 4:16
  SPEAKER_00: But why are they not more of a sure bet, you might say?
• 4:19
  SPEAKER_02: Yeah.
• 4:20
  SPEAKER_02: So I think a really big part of it is just that the environment isn't really
• 4:24
  SPEAKER_02: in Republicans' favor right now.
• 4:26
  SPEAKER_02: I think the generic ballot as it stands at the time that we're recording this is D plus 5.3.
• 4:32
  SPEAKER_02: Or something there, let's say D plus five.
• 4:35
  SPEAKER_02: And so that really is contributing to the fact that no matter,
• 4:38
  SPEAKER_02: you know,
• 4:39
  SPEAKER_02: there's still a very steep climb for Democrats for the Senate,
• 4:42
  SPEAKER_02: but it's not so steep when the environment is kind of has shifted in their favor
• 4:47
  SPEAKER_02: and is kind of more on their side of the field.
• 4:50
  SPEAKER_00: Yeah, Scott, I don't know if you have any sort of, yeah, please.
• 4:54
  SPEAKER_01: I mean, if you could see our Slack chats about this, right?
• 4:57
  SPEAKER_01: I mean, Maddisen’s been on the call.
• 4:58
  SPEAKER_01: Like this has been something that has kept me up from a forecast perspective for
• 5:03
  SPEAKER_01: the last year, because I guess as a podcast, we do long form, right?
• 5:07
  SPEAKER_01: Like there are less polls than there were two and four years ago.
• 5:11
  SPEAKER_01: The polls that are up public are more partisan.
• 5:14
  SPEAKER_01: And let's just be honest,
• 5:14
  SPEAKER_01: the partisan polls,
• 5:16
  SPEAKER_01: whether they're Republican or Democrat,
• 5:17
  SPEAKER_01: they tend to shade the generic ballot to where they want.
• 5:20
  SPEAKER_01: Whether that's on purpose or methodology or whatever it is, not for me to decide.
• 5:24
  SPEAKER_01: But what I've noticed is a huge variance We can get a generic ballot on one day
• 5:31
  SPEAKER_01: where it's D plus eight and another day run roughly in the same timeframe as
• 5:35
  SPEAKER_01: the D plus eight, we can get a D plus three.
• 5:38
  SPEAKER_01: And so we've spent a lot of time in our polling average trying to smooth this over
• 5:42
  SPEAKER_01: with the cubic spline.
• 5:43
  SPEAKER_01: And Matt is going to explain,
• 5:44
  SPEAKER_01: you know, we all came up with a technique for cubic splining to cubic splining.
• 5:49
  SPEAKER_01: So like out the generic ballot,
• 5:52
  SPEAKER_01: Is something that is worsened because it is such a huge input to the forecast.
• 5:56
  SPEAKER_01: If you move and I think you Skelley, you have this, you know, you work this out, right?
• 6:00
  SPEAKER_01: A D plus seven.
• 6:03
  SPEAKER_01: I'm sorry, D plus seven, D plus six, correct me.
• 6:05
  SPEAKER_01: It's one of my favorite graphics.
• 6:06
  SPEAKER_01: D plus seven,
• 6:07
  SPEAKER_01: D plus six is like the demarcation line for the Republicans potentially holding
• 6:11
  SPEAKER_01: the House.
• 6:12
  SPEAKER_01: And depending on which generic ballot you pick,
• 6:15
  SPEAKER_01: you could be D plus seven on one or D plus six on the other.
• 6:19
  SPEAKER_01: Obviously, we're fans of averages, but this can really affect.
• 6:22
  SPEAKER_01: It's one of the The major contributing factors to a forecast.
• 6:25
  SPEAKER_01: And so that's obviously,
• 6:27
  SPEAKER_01: you know,
• 6:27
  SPEAKER_01: almost like the forecast inside the forecast and being able to smooth
• 6:30
  SPEAKER_01: that out is important.
• 6:33
  SPEAKER_01: And also I think, look, whatever it is today, what is July 30th when we're doing this?
• 6:38
  SPEAKER_01: I think Maddisen,
• 6:39
  SPEAKER_01: what we've seen the variance over 30 days in generic ballot be two or three points,
• 6:44
  SPEAKER_02: which is pretty- Yeah,
• 6:45
  SPEAKER_02: I was just looking and right around when we launched,
• 6:47
  SPEAKER_02: it was around three or something like that.
• 6:50
  SPEAKER_02: So we've even jumped two points in the past two weeks or so.
• 6:54
  SPEAKER_02: Yeah.
• 6:55
  SPEAKER_02: And I think even back in June, it was at like D plus seven.
• 6:58
  SPEAKER_02: So it's been bouncing around quite a bit and mitigating that has definitely been
• 7:03
  SPEAKER_02: at least a topic of conversation on Slack, if not elsewhere.
• 7:06
  SPEAKER_00: Yeah.
• 7:07
  SPEAKER_00: Yeah.
• 7:07
  SPEAKER_00: Um, just to pause for a second, Scott, I think you meant like D plus three.
• 7:12
  SPEAKER_01: Yeah.
• 7:13
  SPEAKER_01: Six or seven.
• 7:15
  SPEAKER_01: But yes, this is the graphic I'm looking, the one you put up.
• 7:19
  SPEAKER_01: Simple but powerful.
• 7:20
  SPEAKER_00: Yeah.
• 7:21
  SPEAKER_00: So, so yeah, no, you're all good.
• 7:22
  SPEAKER_00: All good.
• 7:23
  SPEAKER_00: I get it.
• 7:24
  SPEAKER_00: Um, yeah.
• 7:24
  SPEAKER_00: So basically, yeah,
• 7:25
  SPEAKER_00: roughly around D plus three,
• 7:26
  SPEAKER_00: if you look at this,
• 7:28
  SPEAKER_00: uh, this, uh, you know, the, the, you know, Republicans are favored in the Senate.
• 7:33
  SPEAKER_00: Democrats a bit more favorite in the House,
• 7:35
  SPEAKER_00: but we've had sort of different kind of a range of Democratic leads recently.
• 7:38
  SPEAKER_00: And obviously,
• 7:39
  SPEAKER_00: the closer it gets to sort of that two and a half,
• 7:42
  SPEAKER_00: three point, the more it really is a pure sort of 50 50 proposition in the House.
• 7:47
  SPEAKER_00: And obviously,
• 7:48
  SPEAKER_00: in that situation,
• 7:49
  SPEAKER_00: the Senate would be better for the GOP or even better for the GOP.
• 7:53
  SPEAKER_00: But then that's how you maybe get into a situation where Republicans are more likely
• 7:57
  SPEAKER_00: to actually hold on to full control.
• 8:00
  SPEAKER_00: But, you know,
• 8:01
  SPEAKER_00: the flip side of that is, yeah,
• 8:02
  SPEAKER_00: we've gotten the last couple of days,
• 8:03
  SPEAKER_00: there's been, I think, some better numbers for Democrats on the generic ballot.
• 8:07
  SPEAKER_00: And, you know,
• 8:09
  SPEAKER_00: maybe you're getting more on the end where,
• 8:12
  SPEAKER_00: you know,
• 8:12
  SPEAKER_00: they're favored in the House,
• 8:14
  SPEAKER_00: not overwhelmingly,
• 8:15
  SPEAKER_00: but a little more comfortable than where they are even right now.
• 8:21
  SPEAKER_00: So, yeah,
• 8:22
  SPEAKER_00: and maybe that's a good segue,
• 8:24
  SPEAKER_00: actually, Maddisen, into talking a little bit about
• 8:29
  SPEAKER_01: How this all works exactly.
• 8:31
  SPEAKER_00: You know, what are the components to go into the forecast?
• 8:33
  SPEAKER_00: How are we sort of, what are maybe some decisions we made?
• 8:37
  SPEAKER_00: You know,
• 8:38
  SPEAKER_00: the inputs and then some of the things that we do
• 8:41
  SPEAKER_00: to affect how those inputs spit out a probability.
• 8:45
  SPEAKER_00: Cause we've run like a million simulations every time based on this model.
• 8:50
  SPEAKER_02: Yeah, so I can talk about that.
• 8:53
  SPEAKER_02: So I would say the inputs kind of like,
• 8:55
  SPEAKER_02: you know, the raw data that goes into it falls into like kind of two categories.
• 8:59
  SPEAKER_02: And it's stuff that we already know and stuff that changes every day.
• 9:02
  SPEAKER_02: Stuff that we already know is,
• 9:04
  SPEAKER_02: for example,
• 9:05
  SPEAKER_02: how certain seats or how certain congressional districts voted for in
• 9:09
  SPEAKER_02: the 2024 presidential election.
• 9:11
  SPEAKER_02: That's not something that's going to change.
• 9:13
  SPEAKER_02: So we already know is if the person running in that seat is an incumbent,
• 9:16
  SPEAKER_02: if that incumbent is A Republican or a Democrat,
• 9:20
  SPEAKER_02: if this is a midterm year, which it is, in case nobody knew, it is a midterm year.
• 9:26
  SPEAKER_02: Newsflash.
• 9:27
  SPEAKER_02: So those are kind of like, that's like the really, like the base of it, right?
• 9:31
  SPEAKER_02: That's all the historical data that we kind of bring in that doesn't,
• 9:34
  SPEAKER_02: you know, sits, it doesn't get touched for the foreseeable future.
• 9:41
  SPEAKER_02: And then there's the stuff that changes every day.
• 9:43
  SPEAKER_02: So we already talked about a big one, which is the generic ballot.
• 9:45
  SPEAKER_02: That's political environment.
• 9:47
  SPEAKER_02: Historically, we kind of based that on the congressional popular vote difference.
• 9:54
  SPEAKER_02: That's like our historical generic ballot.
• 9:56
  SPEAKER_02: We currently use our own homemade average for this.
• 10:01
  SPEAKER_02: And that gets smooth a little bit just to kind of dampen those day over day swings
• 10:06
  SPEAKER_02: that we've talked about that we've been seeing these past couple of weeks
• 10:09
  SPEAKER_02: with different polls coming in.
• 10:10
  SPEAKER_02: We also take in other polls.
• 10:12
  SPEAKER_02: So obviously on our website, you can see all of our public averages.
• 10:16
  SPEAKER_02: We take in a lot of other polls that might not be public,
• 10:19
  SPEAKER_02: but those are definitely getting factored into things because if there's a poll
• 10:22
  SPEAKER_02: on it,
• 10:22
  SPEAKER_02: we probably have it and it's probably affecting the forecast in some way,
• 10:26
  SPEAKER_02: shape or form.
• 10:27
  SPEAKER_02: We also take in,
• 10:28
  SPEAKER_02: and this is something that I'm pretty excited about this year,
• 10:30
  SPEAKER_02: is prediction markets.
• 10:32
  SPEAKER_02: And those, we take in both Kalshi and Polymarket and those get blended in.
• 10:37
  SPEAKER_02: Based on volume of, you know, dollars spent in each market.
• 10:41
  SPEAKER_02: And those also affect and kind of, you know, polls lag.
• 10:45
  SPEAKER_02: So we kind of put in the markets to kind of mitigate that inherent lag or bias
• 10:49
  SPEAKER_02: that might come with polling.
• 10:53
  SPEAKER_01: It's the prediction market stuff.
• 10:55
  SPEAKER_01: Again, it was the topic of our 9 a.m. data science call this morning as Maddisen will know.
• 11:00
  SPEAKER_01: And we'll know, we do Polymarket and Kalshi.
• 11:02
  SPEAKER_01: We do not take any money from them.
• 11:04
  SPEAKER_01: Those are my favorite tweets to read.
• 11:07
  SPEAKER_01: Which is why we can use both and average both and be intellectually honest.
• 11:11
  SPEAKER_01: Because if we took the money, I'm sure it would be, oh, you can't use the other person's.
• 11:16
  SPEAKER_01: But obviously, the use of prediction markets is a hot debate.
• 11:22
  SPEAKER_01: We decided to put it in because what we're looking for is signal.
• 11:27
  SPEAKER_01: And look...
• 11:29
  SPEAKER_01: How these prediction markets advertise themselves and who used them,
• 11:34
  SPEAKER_01: all that kind of stuff, that's a different topic.
• 11:37
  SPEAKER_01: What we are pretty confident in,
• 11:40
  SPEAKER_01: we looked at some literature to this effect, is they have signal.
• 11:44
  SPEAKER_01: And so if they have signal,
• 11:45
  SPEAKER_01: we think they should be included because that's what we're doing here.
• 11:51
  SPEAKER_01: Some of the internal discussions,
• 11:54
  SPEAKER_01: which I'm positive will be refined when we do the 2028 forecast,
• 11:57
  SPEAKER_01: is what is the threshold?
• 11:59
  SPEAKER_01: You know, how much money should a market have before we include it?
• 12:02
  SPEAKER_01: And then what kind of weighting should it have on the forecast?
• 12:04
  SPEAKER_01: And those are open questions that we may have picked the wrong calibration for those.
• 12:10
  SPEAKER_01: We admit that, right?
• 12:11
  SPEAKER_01: It's like all these forecasts.
• 12:13
  SPEAKER_01: You make decisions.
• 12:14
  SPEAKER_01: Um, and you make them with the best.
• 12:16
  SPEAKER_01: There is art to the science,
• 12:17
  SPEAKER_01: you know, if
• 12:18
  SPEAKER_00: there's math, but you have to still make some judgment calls.
• 12:21
  SPEAKER_00: Yeah.
• 12:21
  SPEAKER_01: Yeah.
• 12:21
  SPEAKER_01: Yeah.
• 12:21
  SPEAKER_01: And I, and I think,
• 12:22
  SPEAKER_01: you know,
• 12:23
  SPEAKER_01: I think we made,
• 12:24
  SPEAKER_01: and Maddisen can,
• 12:25
  SPEAKER_01: can illustrate, she was in a lot of these calls, um, going over with us.
• 12:29
  SPEAKER_01: I think we made reasonable.
• 12:32
  SPEAKER_01: You know, empirical based and where available academic research based decisions.
• 12:36
  SPEAKER_01: But look,
• 12:37
  SPEAKER_01: like all forecasts,
• 12:38
  SPEAKER_01: we're going to look at the end of the year,
• 12:39
  SPEAKER_01: see where it adds and see where we can improve upon it.
• 12:42
  SPEAKER_01: But as it sits now,
• 12:44
  SPEAKER_01: we're pretty happy with how the markets are responding because one of
• 12:46
  SPEAKER_01: the problems we have and, you know, this cycle is no different.
• 12:51
  SPEAKER_01: We have a lot of static variables and polls are not static variables,
• 12:55
  SPEAKER_01: but they're very, they're slow.
• 12:57
  SPEAKER_01: Sometimes they're more static than you want them to be because no one's pulling
• 13:00
  SPEAKER_01: the race, you know?
• 13:01
  SPEAKER_01: Exactly.
• 13:02
  SPEAKER_01: When Graham Platner drops out,
• 13:05
  SPEAKER_01: we have to wait a week or two to figure out that it's Troy Jackson and
• 13:08
  SPEAKER_01: whether there's polling or not.
• 13:09
  SPEAKER_01: And so there's signal to be gained from prediction markets.
• 13:14
  SPEAKER_01: And when we want to run this every day, it's kind of a big deal.
• 13:18
  SPEAKER_01: You know, and I haven't asked Maddisen this, so we'll learn something on here.
• 13:22
  SPEAKER_01: What's your take on how we, I mean,
• 13:24
  SPEAKER_01: I know you generally like the prediction market stuff,
• 13:27
  SPEAKER_01: but what do you think we may have gotten wrong on this?
• 13:29
  SPEAKER_01: Like when we sit down in December,
• 13:31
  SPEAKER_01: what do you think you're like, okay, we got to adjust this because of that?
• 13:35
  SPEAKER_02: Yeah, and I think it's what you talked about.
• 13:37
  SPEAKER_02: It's going to be like the dollar amount thresholds or...
• 13:41
  SPEAKER_02: It's going to be the percent weighting.
• 13:44
  SPEAKER_02: They currently sit at a maximum of 25% of that final probability value that we see here.
• 13:51
  SPEAKER_02: Maybe that's too high.
• 13:52
  SPEAKER_02: Maybe that's too little.
• 13:54
  SPEAKER_02: I think it's in a good spot right now.
• 13:57
  SPEAKER_02: But like Scott said, I think polls are a little bit slow sometimes.
• 14:02
  SPEAKER_02: And so the forecast and the model itself,
• 14:05
  SPEAKER_02: there's no way to code who Graham Platner is into a forecaster or take your pick
• 14:12
  SPEAKER_02: of candidate.
• 14:13
  SPEAKER_02: There's no way to kind of put character into a forecast.
• 14:16
  SPEAKER_02: There's no way to put shock and one news article drops here, another news article drops there.
• 14:21
  SPEAKER_02: There's no way to really put that into a clean data set in a way that's standardized.
• 14:26
  SPEAKER_02: And so we bring in prediction markets to kind of offset that a little bit
• 14:30
  SPEAKER_02: and especially cover races that might not even get polling.
• 14:34
  SPEAKER_02: Some Some races aren't worth pollsters' time of day,
• 14:38
  SPEAKER_02: and that's kind of the game that we have to play here.
• 14:41
  SPEAKER_02: And so we try to mitigate that with the markets.
• 14:43
  SPEAKER_02: And I'm sure that we'll have some happy thoughts and some not-so-happy thoughts
• 14:49
  SPEAKER_02: about this in December, but I am pretty happy with where it sits right now.
• 14:53
  SPEAKER_00: Yeah, I've brought up on screen just for folks at home from our forecast on votes,
• 15:00
  SPEAKER_00: our portal, which you should definitely check out if you haven't already.
• 15:03
  SPEAKER_00: I'm sure many people listening and watching have already done so,
• 15:06
  SPEAKER_00: but votes.decisiondeskhq.com or DDHQ Votes in Google and you'll find it.
• 15:14
  SPEAKER_00: I brought up the Texas Senate race forecast page where right now I should note
• 15:18
  SPEAKER_00: that Ken Paxton is a very slight favorite.
• 15:20
  SPEAKER_00: It's actually in our toss up range.
• 15:22
  SPEAKER_00: Ken Paxton, James Talarico here.
• 15:24
  SPEAKER_00: So close race and neither candidate has at least a 60% chance of winning.
• 15:29
  SPEAKER_00: So we deem that a toss up.
• 15:31
  SPEAKER_00: But right now,
• 15:33
  SPEAKER_00: basically the inputs that Maddisen was talking about,
• 15:36
  SPEAKER_00: about half of what the forecast is saying about Texas is coming from fundamentals,
• 15:39
  SPEAKER_00: which
• 15:40
  SPEAKER_00: might have to do with how red Texas has been historically.
• 15:43
  SPEAKER_00: But then polls,
• 15:44
  SPEAKER_00: the fact that Talarico has a slight edge in the polls is bringing in is about
• 15:47
  SPEAKER_00: a quarter of it right now.
• 15:49
  SPEAKER_00: And then markets,
• 15:50
  SPEAKER_00: because I know the Texas Senate markets are ones that are getting a
• 15:53
  SPEAKER_01: fair amount
• 15:54
  SPEAKER_01: of it.
• 15:56
  SPEAKER_01: There's a lot of action on those, so they're not too small to be maybe weighted a
• 16:01
  SPEAKER_00: bit less.
• 16:02
  SPEAKER_00: Right now, they're getting about 25% of the weight as well.
• 16:05
  SPEAKER_00: So I think that just serves as a nice example of sort of these three things
• 16:10
  SPEAKER_00: that are playing into why our forecast says what it does for each particular race.
• 16:15
  SPEAKER_00: And then you bring all that together.
• 16:17
  SPEAKER_00: And you run simulations based on those probabilities.
• 16:20
  SPEAKER_00: And you end up with the slight edge we give basically each party in the Senate, the House.
• 16:28
  SPEAKER_00: And I should say,
• 16:29
  SPEAKER_00: we should talk for a little bit about the House,
• 16:30
  SPEAKER_00: because we've talked a little bit about the Senate.
• 16:34
  SPEAKER_00: Democrats, slight favorites in the House, about three and five shot right now, 60%.
• 16:40
  SPEAKER_00: I would say that compared to a lot of other forecasts we're seeing,
• 16:43
  SPEAKER_00: this is,
• 16:43
  SPEAKER_00: I think,
• 16:44
  SPEAKER_00: a bit More bearish on the Democrats' chances than some of the other forecasts.
• 16:48
  SPEAKER_00: And I think,
• 16:49
  SPEAKER_00: you know,
• 16:50
  SPEAKER_00: part of that is that maybe our generic ballot is because we take in every poll.
• 16:57
  SPEAKER_00: I would say is a little, it's not quite as blue as what some of the generic ballots are showing.
• 17:02
  SPEAKER_00: But Maddisen, please,
• 17:04
  SPEAKER_00: that's my perception of why we're maybe not quite as,
• 17:08
  SPEAKER_00: there's also redistricting and other things taken into account.
• 17:10
  SPEAKER_00: But I know other forecasts are doing that too.
• 17:12
  SPEAKER_00: So I know that people have,
• 17:13
  SPEAKER_00: you know,
• 17:14
  SPEAKER_00: We've seen the replies on Twitter when we tweet out about the forecast of people,
• 17:18
  SPEAKER_00: Democrats being like, hey,
• 17:20
  SPEAKER_00: this is underselling us and Republicans being like,
• 17:22
  SPEAKER_00: yeah, that actually might be about right.
• 17:26
  SPEAKER_00: So I was curious actually on sort of your take on what we're finding.
• 17:31
  SPEAKER_02: Yeah.
• 17:32
  SPEAKER_02: Yeah.
• 17:34
  SPEAKER_02: Again, I kind of agree with what's going on here.
• 17:36
  SPEAKER_02: I do drink a lot of the Kool-Aid that we do produce here.
• 17:40
  SPEAKER_02: But I think part of it is because the generic ballot we take in everything and,
• 17:44
  SPEAKER_02: you know, law of large numbers is either working in or against our favor.
• 17:48
  SPEAKER_02: And there's no real way to know until November 4th how that's going to play out.
• 17:52
  SPEAKER_02: But I do really kind of I think that this is,
• 17:56
  SPEAKER_02: you know,
• 17:56
  SPEAKER_02: pretty spot or I'm not going to say spot on,
• 17:58
  SPEAKER_02: but pretty close to kind of how I imagine it.
• 18:00
  SPEAKER_02: And as we saw in that graph you had earlier,
• 18:03
  SPEAKER_02: if the generic ballot gets like D plus seven,
• 18:06
  SPEAKER_02: this Probability for Democrats gets to 70%,
• 18:09
  SPEAKER_02: which is probably a little bit closer to the range
• 18:11
  SPEAKER_02: that conventional wisdom would lead you to.
• 18:14
  SPEAKER_02: But also in discussions, I can't super speak to what was happening in 2018 because I was
• 18:23
  SPEAKER_02: I was not
• 18:28
  SPEAKER_02: super tuned in to the political climate of 2018,
• 18:31
  SPEAKER_02: but from what I've heard and from what I've learned,
• 18:35
  SPEAKER_02: it's similar, but what we're experiencing right now is a milder version of 2018.
• 18:41
  SPEAKER_02: And so as much as it is like the first midterm after Trump's election,
• 18:45
  SPEAKER_02: it's not as strong of an effect as we saw eight years ago.
• 18:51
  SPEAKER_00: Yeah, I think that is how I would sum it up too.
• 18:54
  SPEAKER_00: I mean, you know,
• 18:55
  SPEAKER_00: I think back because I was at 538 at the time and our polling average,
• 18:59
  SPEAKER_00: I think pretty quickly got close to D plus eight in the generic ballot,
• 19:03
  SPEAKER_00: like by about this point.
• 19:06
  SPEAKER_00: And then it kind of stayed there for the next like three months.
• 19:10
  SPEAKER_00: And this time around,
• 19:11
  SPEAKER_00: you know,
• 19:12
  SPEAKER_00: we're seeing,
• 19:13
  SPEAKER_00: we've been seeing sort of D plus four,
• 19:16
  SPEAKER_00: even down to D plus three, it's back to D plus five right now.
• 19:19
  SPEAKER_00: But that's not obviously on the same level as we saw in 2018.
• 19:25
  SPEAKER_00: And that could speak to a number of things.
• 19:26
  SPEAKER_00: I think you could talk about sort of what's going on with the two parties right now.
• 19:33
  SPEAKER_00: I think Trump in a way, the president is more of baked in as a reality.
• 19:38
  SPEAKER_00: The Republican party has been more fully sort of reshaped in his image.
• 19:42
  SPEAKER_00: And so folks that are voting Republican tend to be Uh, sort of more aligned with that.
• 19:47
  SPEAKER_00: And so people that have shifted into the Democrat camp, I mean,
• 19:50
  SPEAKER_00: we had people shift into the Republican camp clearly,
• 19:52
  SPEAKER_00: cause that's how Trump won in the first place in 2016,
• 19:55
  SPEAKER_00: but you had people shift into the Democrat camp.
• 19:57
  SPEAKER_00: But I think the other thing that's going on that could be a problem for Democrats
• 20:01
  SPEAKER_00: and the reason why maybe they're not able to maybe take advantage
• 20:04
  SPEAKER_00: of President Trump having an approval rating,
• 20:06
  SPEAKER_00: you know, close to 40%, is that we see this in polls.
• 20:11
  SPEAKER_00: People think the Democrats are actually a bit more extreme.
• 20:14
  SPEAKER_00: When push comes to shove,
• 20:16
  SPEAKER_00: and they ask, like, which party is more ideologically extreme right now?
• 20:21
  SPEAKER_00: At least a few more respondents in polls saying Democrats.
• 20:23
  SPEAKER_00: And so if you think about like what happened in 2022, that might be like a good example of this.
• 20:28
  SPEAKER_00: President Biden had an approval rating in the low to mid 40s.
• 20:32
  SPEAKER_00: Wasn't looking terribly good in that way from historical data points for Democrats.
• 20:37
  SPEAKER_00: But Republicans only ended up with a very small gain in the House.
• 20:40
  SPEAKER_00: And that's because there was maybe a countervailing wind, which was the Dobbs decision.
• 20:45
  SPEAKER_00: And that and also maybe the perception that Republicans were a bit more extreme at the time.
• 20:50
  SPEAKER_00: May have actually helped Democrats sort of reduce their losses in the House.
• 20:54
  SPEAKER_00: They actually gained a seat in the Senate.
• 20:56
  SPEAKER_00: So I see that and I see like what we're seeing here and I see a lot of potential
• 21:01
  SPEAKER_00: for this to be a bit more like 2022 than 2018.
• 21:06
  SPEAKER_01: Scott, I don't know if you had any thoughts on- It's interesting, right?
• 21:11
  SPEAKER_01: The 2022 to 2018 comparison I tend to buy into that having not been 14 in 2018
• 21:20
  SPEAKER_01: and lived through 2018 looking around what's going on.
• 21:24
  SPEAKER_01: I think the difference between ’18 and ’22 that we have...
• 21:31
  SPEAKER_01: What I was thinking about the other day,
• 21:37
  SPEAKER_01: I was like,
• 21:37
  SPEAKER_01: let's say we take the Texas gerrymander off the table,
• 21:40
  SPEAKER_01: we take the California gerrymander off the table and all the other ones, right?
• 21:44
  SPEAKER_01: And we just go into the map the way it is, what we're talking about now.
• 21:48
  SPEAKER_01: And I think what we would be talking about is probably a 55%,
• 21:54
  SPEAKER_01: 60% GOP hold on the house, I think, right?
• 21:59
  SPEAKER_01: Which would be kind of surprising Given the environment,
• 22:04
  SPEAKER_01: again, that would be a fun math actual exercise to do.
• 22:08
  SPEAKER_01: But if you take the gerrymandering off the table,
• 22:10
  SPEAKER_01: this race, I don't know, maybe not 50-55 Republican, but not 60-40 down.
• 22:14
  SPEAKER_01: It's much closer to 50-50, I think.
• 22:16
  SPEAKER_01: And so gerrymandering is one of those things that I think made this thing,
• 22:22
  SPEAKER_01: it basically kept it the same, but made there's fewer seats to fight over.
• 22:27
  SPEAKER_01: Right?
• 22:27
  SPEAKER_01: Like these huge swings, we're not gonna see a 30 or 40 seat, 30 or 40 seat swing here.
• 22:32
  SPEAKER_01: We're gonna see a 10 or 15 seat, 10 or 15 seat swing.
• 22:36
  SPEAKER_01: And if we don't see a 10 or 15 seat swing, It could go either way, right?
• 22:41
  SPEAKER_01: Like if, you know, this is all a range of probabilities.
• 22:43
  SPEAKER_01: If you scroll down,
• 22:44
  SPEAKER_01: you can see our simulations and, you know, we're huge fans of very fat tails, right?
• 22:50
  SPEAKER_01: Like, you know, that's the beauty of this.
• 22:53
  SPEAKER_01: There's so many different combinations.
• 22:54
  SPEAKER_01: And so I think that's what's different is,
• 22:57
  SPEAKER_01: is we are not going to see some big sweeping changes and there's enough,
• 23:04
  SPEAKER_01: there's not enough seats out there to see a 40
• 23:06
  SPEAKER_00: seat change.
• 23:08
  SPEAKER_01: And again, let's take 22 and 18 off the set, like we saw in 2006, right?
• 23:13
  SPEAKER_01: Like we saw in some of these other years where you did see this big thing.
• 23:19
  SPEAKER_01: Just less room for variance, or I'm sorry, less room for huge movements back and forth.
• 23:25
  SPEAKER_00: Absolutely.
• 23:25
  SPEAKER_00: I think that's right.
• 23:26
  SPEAKER_00: And I think the gerrymandering wars of this cycle have made that even more likely,
• 23:32
  SPEAKER_00: or sort of reducing the Or at least pushing a few more seats out of the conversation
• 23:38
  SPEAKER_00: that would have been in the conversation as potentially competitive.
• 23:43
  SPEAKER_00: And so in that sense, you are maybe narrowing the range of outcomes.
• 23:48
  SPEAKER_00: Now we have the fat tails,
• 23:50
  SPEAKER_00: but just it becomes,
• 23:52
  SPEAKER_00: I think,
• 23:53
  SPEAKER_00: easier to imagine a small gain for Democrats or something closer to the status quo,
• 23:59
  SPEAKER_00: which is how the GOP manages to hold onto the House with very small shifts.
• 24:05
  SPEAKER_00: Um, across the map working out,
• 24:07
  SPEAKER_00: you know,
• 24:07
  SPEAKER_00: to roughly break even, uh, and, and so Republicans would hold on in that case.
• 24:12
  SPEAKER_00: Uh, so maybe just to, to talk a bit more about, let's like shift over to the Senate again.
• 24:19
  SPEAKER_00: I just wanted to talk a little bit about some of the states,
• 24:21
  SPEAKER_00: cause obviously,
• 24:21
  SPEAKER_00: you know,
• 24:22
  SPEAKER_00: we can talk about things at, at, at,
• 24:24
  SPEAKER_00: you know,
• 24:25
  SPEAKER_00: 30,000 feet,
• 24:26
  SPEAKER_00: but people are gonna be interested in sort of the individual races,
• 24:28
  SPEAKER_00: uh, and, and what's happening.
• 24:31
  SPEAKER_00: And so I thought actually one way of getting at this would be to look at,
• 24:39
  SPEAKER_00: if I can bring it up here on screen,
• 24:44
  SPEAKER_00: just sort of a summary of the seats in question here.
• 24:48
  SPEAKER_00: And if it will ever load, here we go, data wrapper.
• 24:53
  SPEAKER_00: I kind of summarized The sort of the key seats in the Senate here.
• 24:59
  SPEAKER_00: And then we can talk about each of them or,
• 25:01
  SPEAKER_00: you know,
• 25:01
  SPEAKER_00: some of them a little more in depth,
• 25:04
  SPEAKER_00: but essentially it looks like there's sort of 10 seats that are,
• 25:08
  SPEAKER_00: that neither party has really greater than an 80% chance of winning.
• 25:12
  SPEAKER_00: And so that basically leaves, so Republicans currently in the majority, 53, 47.
• 25:19
  SPEAKER_00: They got 31 seats they hold that are not up this cycle.
• 25:22
  SPEAKER_00: They're 15 where they got an 85% or greater or 80% or greater chance, basically.
• 25:27
  SPEAKER_00: And that's 46.
• 25:29
  SPEAKER_00: So at that point, they need four more seats to get to 50.
• 25:32
  SPEAKER_00: And then Vice President J.D.
• 25:33
  SPEAKER_00: Vance breaks ties in their favor.
• 25:35
  SPEAKER_00: On the flip side, Democrats, they're 34 seats they hold that are not up in this cycle.
• 25:40
  SPEAKER_00: They're 10 where they're more clear favorites.
• 25:42
  SPEAKER_00: And that's 44.
• 25:44
  SPEAKER_00: So then they have to find seven because they need an outright majority.
• 25:47
  SPEAKER_00: Tie obviously goes to the Republicans.
• 25:49
  SPEAKER_00: So they got to find seven of these 10.
• 25:52
  SPEAKER_00: They got to win seven of them in order to win a majority.
• 25:55
  SPEAKER_00: And like right now,
• 25:56
  SPEAKER_00: you know,
• 25:56
  SPEAKER_00: I think you can see if you're looking at this,
• 25:59
  SPEAKER_00: like Republicans somewhat more clear, more likely to hold on to Florida.
• 26:03
  SPEAKER_00: So maybe that gets them up to 47.
• 26:05
  SPEAKER_00: And then Democrats are favored,
• 26:07
  SPEAKER_00: not quite as heavily,
• 26:08
  SPEAKER_00: but not too far away,
• 26:10
  SPEAKER_00: favored in Georgia,
• 26:11
  SPEAKER_00: New Hampshire,
• 26:12
  SPEAKER_00: in North Carolina,
• 26:13
  SPEAKER_00: which would be sort of the best chance Democrats have a flipping a seat.
• 26:16
  SPEAKER_00: So if you added three to that and you're up to 47,
• 26:19
  SPEAKER_00: 47, and then everything is down to sort of those final six seats.
• 26:24
  SPEAKER_00: And in most of those,
• 26:26
  SPEAKER_00: four of them,
• 26:27
  SPEAKER_00: the GOP is at least very,
• 26:28
  SPEAKER_00: very slightly favored,
• 26:29
  SPEAKER_00: but they're all kind of I'm actually not sure which side Michigan is on right now of
• 26:34
  SPEAKER_00: the toss up line.
• 26:35
  SPEAKER_00: Cause again, 60% is our, our toss up line, but it's like right by it.
• 26:39
  SPEAKER_00: So it's either very slightly lean dim or toss up and,
• 26:43
  SPEAKER_00: you know, mathematically 59.9 and 60.1 are not really different.
• 26:49
  SPEAKER_00: Uh, so, um, uh,
• 26:51
  SPEAKER_00: in that sense, it's,
• 26:52
  SPEAKER_00: it's the six seats where I think everything is,
• 26:55
  SPEAKER_00: you know,
• 26:56
  SPEAKER_00: are most likely to come down to in terms of which party will have the majority.
• 27:00
  SPEAKER_00: And obviously Most of those are Republican-held seats,
• 27:02
  SPEAKER_00: but it does seem like,
• 27:05
  SPEAKER_00: and maybe this is just timely because we've got Michigan this coming Tuesday,
• 27:09
  SPEAKER_00: but Michigan looms as the one Democratic-held seat that does seem to be really,
• 27:14
  SPEAKER_00: really on the table.
• 27:16
  SPEAKER_00: And I forget what the Republicans,
• 27:18
  SPEAKER_00: I believe now,
• 27:19
  SPEAKER_00: if they won Michigan and we didn't know anything else about this
• 27:23
  SPEAKER_00: About what's happening.
• 27:25
  SPEAKER_00: Maddisen question.
• 27:26
  SPEAKER_00: This is Maddisen.
• 27:28
  SPEAKER_00: I'm scrolling
• 27:31
  SPEAKER_02: through my code notebook as we speak to find it.
• 27:35
  SPEAKER_00: Well, actually I pulled it up.
• 27:38
  SPEAKER_00: Okay, perfect.
• 27:39
  SPEAKER_00: You said the data earlier.
• 27:40
  SPEAKER_00: So I have a chart here and this is basically just like if certain flips happened.
• 27:46
  SPEAKER_00: So if Republicans want a Democratic seat or Democrats want a Republican seat,
• 27:50
  SPEAKER_00: what are like the chances?
• 27:53
  SPEAKER_00: And so actually I think I meant to change this to 58% as the baseline
• 27:56
  SPEAKER_00: for Republicans, but ignore that typo.
• 28:00
  SPEAKER_00: The point is 58% right now for Republicans.
• 28:03
  SPEAKER_00: They won Michigan and we do nothing else about any of the other outcomes.
• 28:07
  SPEAKER_00: They win 86% of the time in the Senate.
• 28:10
  SPEAKER_00: And I think that's a really striking number because maybe that really puts,
• 28:14
  SPEAKER_00: you know, just how important it is for Democrats.
• 28:17
  SPEAKER_00: If they're going to have a path to a majority, they almost have to hold on to Michigan.
• 28:21
  SPEAKER_00: I mean, you know,
• 28:22
  SPEAKER_00: there are other things that have to happen for them,
• 28:24
  SPEAKER_00: but they can't lose what they have, right?
• 28:27
  SPEAKER_00: I mean, that's really the tough spot.
• 28:29
  SPEAKER_00: Maddisen, I know you've looked at a lot of these scenarios.
• 28:32
  SPEAKER_00: So what do you make of that?
• 28:35
  SPEAKER_02: Yeah, I mean,
• 28:36
  SPEAKER_02: the thing with 50-50 is that it seems it kind of acts like it's up in the air
• 28:41
  SPEAKER_02: in terms of overall seat count.
• 28:42
  SPEAKER_02: But if you think about it,
• 28:44
  SPEAKER_02: the path to a Democratic Senate is a much steeper hill than a Republican Senate.
• 28:51
  SPEAKER_02: The Republicans just have to hold on.
• 28:52
  SPEAKER_02: They just have to keep what they have,
• 28:55
  SPEAKER_02: can't lose anything,
• 28:56
  SPEAKER_02: might maybe pick up one or two,
• 28:58
  SPEAKER_02: but they pick up one or two,
• 29:00
  SPEAKER_02: they're in a much better place than If they were just holding on,
• 29:03
  SPEAKER_02: even if they just hold on, they win.
• 29:05
  SPEAKER_02: But for the Democrats, they have to pick up four seats and hold what they have.
• 29:10
  SPEAKER_02: And I ran some of these numbers for Geoffrey earlier,
• 29:12
  SPEAKER_02: but in scenarios where they have to hold or when they have to grab seats,
• 29:16
  SPEAKER_02: they basically have to every time take North Carolina, Alaska and Maine.
• 29:22
  SPEAKER_02: Those are pretty much if the Democrats want to hold the Senate,
• 29:25
  SPEAKER_02: those are pretty much non-negotiable for their chances.
• 29:28
  SPEAKER_02: Right.
• 29:29
  SPEAKER_02: And then after that, you can take your pick of Texas, Ohio and Iowa.
• 29:32
  SPEAKER_02: Um, and one of those playing in, but they have to hang on to what they have.
• 29:36
  SPEAKER_02: I don't think George is personally, I personally don't think George is really going anywhere.
• 29:40
  SPEAKER_02: Um, similar situation with New Hampshire,
• 29:42
  SPEAKER_02: but Michigan, especially with how the primary is going to go this upcoming Tuesday.
• 29:46
  SPEAKER_02: Um, we'll, we'll, we'll see, but they, they have to hold their ground and gain.
• 29:53
  SPEAKER_00: Yeah.
• 29:53
  SPEAKER_00: And, and so obviously,
• 29:55
  SPEAKER_00: you know,
• 29:55
  SPEAKER_00: I think when we think about the Michigan race, um,
• 29:58
  SPEAKER_00: they're The polling's been a little all over the place,
• 30:01
  SPEAKER_00: but I think the thought is,
• 30:02
  SPEAKER_00: look, at the end of the day, Abdul El-Sayed is the more progressive candidate.
• 30:06
  SPEAKER_00: If he does win the nomination,
• 30:07
  SPEAKER_00: which seems the most likely outcome at this point for the Democrats,
• 30:10
  SPEAKER_00: not a given, but he is leading in the polls.
• 30:13
  SPEAKER_00: He's up 51-42 in our average right now of the Democratic primary race against Haley Stevens.
• 30:19
  SPEAKER_00: So if he does win, he is the more ideologically extreme candidate.
• 30:23
  SPEAKER_00: And we have seen candidates like that not do as well in the past,
• 30:26
  SPEAKER_00: especially in sort of highly competitive states like Michigan would be.
• 30:31
  SPEAKER_00: You can think of plenty of examples on the Republican side of the aisle.
• 30:34
  SPEAKER_00: Like a Cary Lake in Arizona in 2024.
• 30:36
  SPEAKER_00: I mean, President Trump carried that state by five and a half,
• 30:39
  SPEAKER_00: but Ruben Gallego won the Senate race.
• 30:41
  SPEAKER_00: And it wasn't by like the skin of his teeth either.
• 30:44
  SPEAKER_00: So you are looking at a situation where,
• 30:47
  SPEAKER_00: okay, it's going to be a real test if El-Sayed does win of sort of that proposition.
• 30:53
  SPEAKER_00: Like can an ideological, more,
• 30:55
  SPEAKER_00: I should say comparatively here,
• 30:57
  SPEAKER_00: ideologically extreme candidate win in a key swing state.
• 31:01
  SPEAKER_00: So that's really, I think, the thing that everything hinges on with that Democratic contest.
• 31:05
  SPEAKER_00: But you pointed out another state there, a couple others, Alaska and Maine.
• 31:09
  SPEAKER_00: But obviously Maine's been in the news a ton for obvious reasons.
• 31:14
  SPEAKER_00: It seems like if Democrats are able to flip that,
• 31:18
  SPEAKER_00: which on paper is their most favorable seat,
• 31:22
  SPEAKER_00: It's one that Kamala Harris carried handily of any of the seats held by Republicans.
• 31:27
  SPEAKER_00: It's the only seat Republicans hold in the Senate that was won by Harris,
• 31:31
  SPEAKER_00: in a state that was won by Harris.
• 31:33
  SPEAKER_00: Their chances would improve not knowing anything else.
• 31:36
  SPEAKER_00: It would improve to 34%,
• 31:38
  SPEAKER_00: or sorry,
• 31:39
  SPEAKER_00: Republicans would go from being a 58% chance of winning to 34 if they lost Maine.
• 31:45
  SPEAKER_00: And maybe that speaks to just how,
• 31:46
  SPEAKER_00: as you were saying,
• 31:47
  SPEAKER_00: it's like one of these states that Democrats kind of have to win.
• 31:50
  SPEAKER_00: Cause they gotta find,
• 31:51
  SPEAKER_00: they gotta,
• 31:52
  SPEAKER_00: at the end of the day,
• 31:52
  SPEAKER_00: they're probably gonna have to win a state like Iowa or Ohio to win this thing.
• 31:57
  SPEAKER_00: So they gotta get the other ones too along the way.
• 32:00
  SPEAKER_01: I mean, they gotta have an inside straight, right?
• 32:02
  SPEAKER_01: If we're gonna use analogies and they can't,
• 32:04
  SPEAKER_01: they can't lose anything on the, on any of the, on any of the turns.
• 32:10
  SPEAKER_01: It's interesting.
• 32:11
  SPEAKER_01: So Maddisen, you think Georgia is no problem for Ossoff, right?
• 32:15
  SPEAKER_02: I'm not gonna say no problem,
• 32:17
  SPEAKER_02: but I don't think it's a big, I think, I don't think it's a big problem.
• 32:21
  SPEAKER_01: Gotcha.
• 32:23
  SPEAKER_01: And that one's interesting.
• 32:24
  SPEAKER_01: If you go back to your previous slide, I just want to hone on that one from a model perspective.
• 32:29
  SPEAKER_01: We have that as an R plus two state, obviously.
• 32:35
  SPEAKER_01: That is Georgia's estate just 10 years ago when
• 32:41
  SPEAKER_01: in a different political world,
• 32:42
  SPEAKER_01: Georgia was not as competitive of a state.
• 32:46
  SPEAKER_01: In fact,
• 32:46
  SPEAKER_01: if you go back 15 years,
• 32:48
  SPEAKER_01: it was a solid red as maybe Texas is perceived today or perceived back then.
• 32:52
  SPEAKER_01: And so it's interesting how it's moved.
• 32:55
  SPEAKER_01: And when you look at it, I believe we have it.
• 32:58
  SPEAKER_01: If the Dems keep Georgia and they flip North Carolina and they flip Maine,
• 33:07
  SPEAKER_01: what they're at about an 80, 85% chance of winning the Senate.
• 33:10
  SPEAKER_01: Is that right, Maddisen?
• 33:12
  SPEAKER_02: Should be roughly around there.
• 33:13
  SPEAKER_02: Yeah.
• 33:14
  SPEAKER_01: Yeah.
• 33:14
  SPEAKER_01: And so like, this is how much, you know, all these seats are correlated.
• 33:20
  SPEAKER_01: And I think the margins are so small,
• 33:23
  SPEAKER_01: there is no room for Hey, this is a pretty close election.
• 33:29
  SPEAKER_01: Depending on how it goes,
• 33:30
  SPEAKER_01: I think either side is going to be able to say we swept it and we had a wave.
• 33:34
  SPEAKER_01: Now, I don't know that the numbers are going to hold up that way.
• 33:37
  SPEAKER_01: Like I said, I don't see 30 or 40 seats flip in the House.
• 33:42
  SPEAKER_01: But if the Democrats end up with 227, 225, they're going to call it a wave.
• 33:47
  SPEAKER_01: And those of us who remember like 06 and some of the other states are like, that's not a wave.
• 33:51
  SPEAKER_01: You won, maybe a solid win, but not a wave.
• 33:53
  SPEAKER_01: And same in the Senate, right?
• 33:54
  SPEAKER_01: Like, if the Democrats take over the Senate,
• 33:57
  SPEAKER_01: they win,
• 33:58
  SPEAKER_01: they flip Georgia,
• 33:59
  SPEAKER_01: they flip North Carolina,
• 34:05
  SPEAKER_01: and they keep Michigan, what, they're going to end up at, what, 51, 52?
• 34:08
  SPEAKER_01: Is that right?
• 34:10
  SPEAKER_01: Again, solid win, but that's not a mandate.
• 34:14
  SPEAKER_01: We're not looking at 55, 57 seat majorities like we did in the early 2000s.
• 34:19
  SPEAKER_01: And I think that's the underscore on this,
• 34:21
  SPEAKER_01: at least when I look at the forecast,
• 34:22
  SPEAKER_01: is there's not a lot of middle ground outcomes.
• 34:25
  SPEAKER_01: There's a lot of outcomes where I think both sides are going to claim some sort
• 34:30
  SPEAKER_01: of mandate,
• 34:30
  SPEAKER_01: but the numbers don't necessarily match that,
• 34:32
  SPEAKER_01: which will be interesting going into presidential.
• 34:36
  SPEAKER_00: Yeah, Maddisen actually ran some numbers for me about sort of what are
• 34:39
  SPEAKER_00: the most common scenarios that
• 34:42
  SPEAKER_00: our forecast has for if Democrats do manage to flip
• 34:46
  SPEAKER_00: the Senate, sort of what combination of states do they tend to get the most?
• 34:51
  SPEAKER_00: And obviously, again, Maine, North Carolina, Alaska tend to pop up in every single one.
• 34:58
  SPEAKER_00: And then actually the most common,
• 35:00
  SPEAKER_00: and it was only 2% of cases, so I don't want to like overstate it.
• 35:04
  SPEAKER_00: But they actually get, uh,
• 35:07
  SPEAKER_00: six of these, of the, of,
• 35:09
  SPEAKER_00: of the Republican held seats,
• 35:11
  SPEAKER_00: um, that are sort of the, of, of the seven on this list.
• 35:14
  SPEAKER_00: So basically everything except Florida.
• 35:17
  SPEAKER_00: And that sort of speaks to maybe their high end issues.
• 35:19
  SPEAKER_00: Like in that scenario,
• 35:20
  SPEAKER_00: they're, they're usually winning Michigan are usually holding onto Michigan.
• 35:24
  SPEAKER_00: So they're, they're getting to 53.
• 35:26
  SPEAKER_00: And so that's like kind of their best case scenario, but that could happen.
• 35:30
  SPEAKER_00: If we end up with one of the results that's sort of on
• 35:32
  SPEAKER_00: the slightly more democratic leaning of the simulations
• 35:36
  SPEAKER_00: that our forecast is spitting out,
• 35:38
  SPEAKER_00: alternatively,
• 35:39
  SPEAKER_00: you could have a situation where Democrats, yeah,
• 35:42
  SPEAKER_00: they hold on to Georgia and New Hampshire,
• 35:44
  SPEAKER_00: but Republicans,
• 35:46
  SPEAKER_00: and maybe it's actually there's no net change because Republicans managed to hold on
• 35:50
  SPEAKER_00: to Maine.
• 35:52
  SPEAKER_00: They lose North Carolina, but they flip Michigan.
• 35:54
  SPEAKER_00: And then it's like, oh, well, 53-47 on the other end.
• 35:56
  SPEAKER_00: So And in that scenario,
• 35:59
  SPEAKER_00: maybe they're holding onto the house and maybe this is a good time
• 36:01
  SPEAKER_00: to bring up something that I think is a really fun aspect of some
• 36:06
  SPEAKER_00: the data we have here,
• 36:08
  SPEAKER_00: which is essentially, Scott was pointing out that all these things are correlated.
• 36:13
  SPEAKER_00: Well, that means that there are actually a lot of cases,
• 36:17
  SPEAKER_00: in fact,
• 36:17
  SPEAKER_00: about three and four, where one party gets the majority at the end of the day.
• 36:23
  SPEAKER_00: And so even though I think a lot of talk is about how Democrats are slight favorites
• 36:27
  SPEAKER_00: in the House, so maybe they win the House.
• 36:29
  SPEAKER_00: Republicans are slight favorites in the Senate.
• 36:31
  SPEAKER_00: Maybe they hold on the Senate.
• 36:32
  SPEAKER_00: And it's so sort of the top line would suggest a split.
• 36:36
  SPEAKER_00: But there's so many simulations and because of the correlation
• 36:40
  SPEAKER_00: across these simulations that in a lot of the outcomes in about three and four,
• 36:44
  SPEAKER_00: either Democrats get both houses or Republicans get both.
• 36:48
  SPEAKER_00: And so I'm going to bring that up.
• 36:50
  SPEAKER_00: Uh, as a, because actually right now.
• 36:53
  SPEAKER_00: The mathematical hot take head scratch.
• 36:56
  SPEAKER_00: It is.
• 36:56
  SPEAKER_00: It is.
• 36:57
  SPEAKER_00: Well, it's a little counterintuitive, I think to people.
• 36:59
  SPEAKER_01: I think the first time we saw it, we're like, oh, let's check the math.
• 37:04
  SPEAKER_00: Yeah.
• 37:04
  SPEAKER_00: So, you know,
• 37:05
  SPEAKER_00: right now there is a,
• 37:08
  SPEAKER_00: a 21% chance of a democratic house and a Republican Senate,
• 37:12
  SPEAKER_00: which tends to be,
• 37:12
  SPEAKER_00: you know,
• 37:13
  SPEAKER_00: I think is so roughly a,
• 37:15
  SPEAKER_00: a one in five shot at the thing that I think people might fundamentally think is
• 37:19
  SPEAKER_00: the most likely outcome.
• 37:20
  SPEAKER_00: And then there's a very small chance like three, four,
• 37:23
  SPEAKER_00: I think it adds up to about 25 across the two of them with decimal places considered
• 37:30
  SPEAKER_00: of a Republican House but a Democratic Senate.
• 37:33
  SPEAKER_00: That's pretty unlikely.
• 37:35
  SPEAKER_00: But Democrats have about a 38% shot of having both the House and Senate
• 37:39
  SPEAKER_00: and Republicans have a 37% shot of having both the House and Senate,
• 37:43
  SPEAKER_00: which yes, adds up to three fourths.
• 37:45
  SPEAKER_00: So I think the...
• 37:49
  SPEAKER_00: The fact that you have sort of these,
• 37:53
  SPEAKER_00: the most likely outcome is that one party will have both is,
• 37:56
  SPEAKER_00: I think,
• 37:57
  SPEAKER_00: you know,
• 37:57
  SPEAKER_00: says a lot about the correlation across outcomes and might surprise people, I think.
• 38:04
  SPEAKER_00: Maddisen,
• 38:04
  SPEAKER_00: obviously,
• 38:05
  SPEAKER_00: since you've been looking at these numbers, I mean, what do you make of that?
• 38:08
  SPEAKER_00: Because it does seem counterintuitive to people.
• 38:10
  SPEAKER_02: Yeah, it seems counterintuitive.
• 38:12
  SPEAKER_02: And even when I was looking at these and kind of setting this part up,
• 38:16
  SPEAKER_02: I also was like, oh, obviously a split building is going to be what kind of happens.
• 38:23
  SPEAKER_02: But then as you think about it, let's take the House, for example.
• 38:27
  SPEAKER_02: So the House is currently leaning towards the Democrats by about 60%,
• 38:31
  SPEAKER_02: which means the Republicans have 40% chance of taking it.
• 38:36
  SPEAKER_02: And we're going to round that 38 and 37 that we see here also to 40 for the sake
• 38:39
  SPEAKER_02: of this example.
• 38:41
  SPEAKER_02: If you take that 40% chance that the Republicans take the House,
• 38:45
  SPEAKER_02: if they take the House, they're taking the Senate with them.
• 38:48
  SPEAKER_02: There's pretty much like no scenario where you're going to, again,
• 38:51
  SPEAKER_02: as we talked about,
• 38:51
  SPEAKER_02: like a Republican House and a Democratic Senate, that's nearly impossible.
• 38:57
  SPEAKER_02: I would be very shocked to see that and,
• 38:59
  SPEAKER_02: you know,
• 39:00
  SPEAKER_02: You know,
• 39:01
  SPEAKER_02: if that happens and I'll take the consequences that I have by predicting that
• 39:04
  SPEAKER_02: that won't happen.
• 39:05
  SPEAKER_02: But I don't think that that's going to happen.
• 39:07
  SPEAKER_02: And so that's how you kind of get to the 40 percent that they get both.
• 39:10
  SPEAKER_02: Right.
• 39:11
  SPEAKER_02: And it works on the flip side, too.
• 39:12
  SPEAKER_02: So with the Senate, Democrats have a 40 percent chance of taking it.
• 39:17
  SPEAKER_02: If they take the Senate, as we've talked about, that is a steep hill to climb for them.
• 39:20
  SPEAKER_02: They're taking the House almost, you know, hand over foot.
• 39:24
  SPEAKER_02: If they have the Senate, the House is close behind, if not right there with them.
• 39:28
  SPEAKER_02: So I think that that's kind of the way that I've been framing it,
• 39:31
  SPEAKER_02: at least for myself and as I've been talking about it outside.
• 39:33
  SPEAKER_02: It's just kind of these probabilities reflect the correlation between two chambers.
• 39:38
  SPEAKER_02: We don't see a lot of,
• 39:39
  SPEAKER_02: you know, split voting much these days as maybe, you know, 10, 20 years ago.
• 39:44
  SPEAKER_02: And on top of that,
• 39:45
  SPEAKER_02: it kind of reflects,
• 39:46
  SPEAKER_02: you know, each party's, let's say, each chamber's least likely outcome.
• 39:50
  SPEAKER_02: And what follows.
• 39:52
  SPEAKER_00: And to your point,
• 39:53
  SPEAKER_00: it's like the fact of the matter is that it's not particularly unlikely that
• 39:59
  SPEAKER_00: the underdog in either chamber could win that chamber.
• 40:02
  SPEAKER_00: So you really are creating a lot of possibilities here for the same party
• 40:07
  SPEAKER_00: to actually end up winning both because the underdogs aren't significant.
• 40:11
  SPEAKER_00: They're not huge underdogs by any means.
• 40:13
  SPEAKER_00: To your point,
• 40:14
  SPEAKER_00: I think if they were both 80-20 Republicans,
• 40:17
  SPEAKER_00: 80-20 in the Senate and Democrats, 80, 20 in the house.
• 40:19
  SPEAKER_00: Yeah.
• 40:20
  SPEAKER_00: They'd be talking about it, about it roughly 20% or 40, I guess what?
• 40:23
  SPEAKER_00: Yeah.
• 40:24
  SPEAKER_00: You know, 40% chance, 20% chance.
• 40:27
  SPEAKER_00: Yeah.
• 40:27
  SPEAKER_00: Let me do the math there.
• 40:28
  SPEAKER_00: 20% chance of,
• 40:29
  SPEAKER_00: of each party having full control,
• 40:32
  SPEAKER_00: which would be, you know, less than half of overall outcomes across the board.
• 40:38
  SPEAKER_00: So, which I think might be how people have viewed things at times in this cycle.
• 40:44
  SPEAKER_00: And yet our forecast suggests that is,
• 40:48
  SPEAKER_00: you know, Underselling the possibility that you're going to get a pretty decent result
• 40:54
  SPEAKER_00: for one party and they're going to end up getting both as a result of it.
• 40:58
  SPEAKER_01: Yeah.
• 40:59
  SPEAKER_00: So let's cycle over the House, talk about the Senate for a little bit here.
• 41:07
  SPEAKER_00: In terms of,
• 41:08
  SPEAKER_00: I thought something that would be fun to talk about with the House And
• 41:14
  SPEAKER_00: so I don't keep you guys here forever was to think about like,
• 41:18
  SPEAKER_00: obviously the house Democrats are somewhat favored.
• 41:20
  SPEAKER_00: There are a lot more seats in the house.
• 41:22
  SPEAKER_00: So it's sometimes a little bit harder to like wrap your head around.
• 41:25
  SPEAKER_00: Like there are a lot of different possible combinations.
• 41:30
  SPEAKER_00: Even if,
• 41:30
  SPEAKER_00: if we sort of know the,
• 41:31
  SPEAKER_00: the broad cross section of seats that will be highly competitive,
• 41:35
  SPEAKER_00: which is not a ton.
• 41:36
  SPEAKER_00: Cause I know that's part of the reason why sort of there's a narrower band of,
• 41:39
  SPEAKER_00: of likely outcomes in the house.
• 41:42
  SPEAKER_00: But I think something that will be helpful for people is when they think about
• 41:45
  SPEAKER_00: on election night, you're trying to make sense of what's going on in the house.
• 41:48
  SPEAKER_00: I think what's really helpful is to pick out a handful of seats and to think,
• 41:53
  SPEAKER_00: okay, If this seat goes, what does that suggest based on our forecast?
• 42:00
  SPEAKER_00: Where could things go from here?
• 42:02
  SPEAKER_00: And so this isn't just because I have East Coast bias as a native Virginian.
• 42:07
  SPEAKER_00: I mean, I do,
• 42:10
  SPEAKER_00: but it's more just those will be the places we get results from first
• 42:14
  SPEAKER_00: on election night.
• 42:15
  SPEAKER_00: And I think a really interesting example to sort of start out with is
• 42:21
  SPEAKER_00: the Virginia second congressional district,
• 42:24
  SPEAKER_00: which until all the redistricting was basically,
• 42:27
  SPEAKER_00: was one of like more or less the median,
• 42:29
  SPEAKER_00: what was close to being the median seat in the house.
• 42:32
  SPEAKER_00: And this is one where right now the Democrats are favored.
• 42:37
  SPEAKER_00: Let me actually pull this up on screen so people at home can see it.
• 42:40
  SPEAKER_00: It's this light blue guy right here.
• 42:44
  SPEAKER_00: So right now Democrats slightly favored in the second congressional district in Virginia.
• 42:51
  SPEAKER_00: Maddisen, you sent me the numbers on this.
• 42:55
  SPEAKER_00: This is a state that closes at 7 p.m.
• 42:56
  SPEAKER_00: If we have results early in the night there,
• 42:59
  SPEAKER_00: and we know that Democrats right now are about 60% shot to win the House,
• 43:03
  SPEAKER_00: their chances, they win about 73% of the time in cases where they win the second.
• 43:08
  SPEAKER_00: So it's like,
• 43:09
  SPEAKER_00: I think it's a good example of, okay,
• 43:12
  SPEAKER_00: this is a seat obviously the Democrats probably have to win
• 43:14
  SPEAKER_00: because that's still leaving about a one in four shot for Republicans to win
• 43:17
  SPEAKER_00: the House.
• 43:18
  SPEAKER_00: But it does suggest that As the night is going on,
• 43:22
  SPEAKER_00: if Democrats were to flip this, that their chances are improving.
• 43:25
  SPEAKER_00: Conversely,
• 43:25
  SPEAKER_00: if Republicans are actually in a position to potentially hold this,
• 43:29
  SPEAKER_00: that would seem to bode poorly for Democrats.
• 43:36
  SPEAKER_00: So thinking about this,
• 43:38
  SPEAKER_00: I think the other two examples that we wanted to bring up are on paper harder
• 43:44
  SPEAKER_00: for Democrats and maybe a little bit better for Republicans.
• 43:47
  SPEAKER_00: And one of those is the first district,
• 43:49
  SPEAKER_00: which I think if I recall correctly,
• 43:51
  SPEAKER_00: you sort of told me it was the,
• 43:52
  SPEAKER_00: if Democrats won this,
• 43:54
  SPEAKER_00: where Don Davis's district got redistricted,
• 43:56
  SPEAKER_00: but he's running,
• 43:56
  SPEAKER_00: he's trying to hold onto it,
• 43:58
  SPEAKER_00: the Democratic incumbent,
• 43:59
  SPEAKER_00: if he were to win it,
• 44:01
  SPEAKER_00: that would actually be a very,
• 44:03
  SPEAKER_00: or at least an even better sign for Democrats at the start of the night.
• 44:07
  SPEAKER_01: Here's the question I got for you, Geoffrey, on the House.
• 44:09
  SPEAKER_01: What's your bellwether?
• 44:12
  SPEAKER_01: What's your bellwether district?
• 44:13
  SPEAKER_01: And I'm going to have the same question for Maddisen, right?
• 44:15
  SPEAKER_01: Like what are one or two districts on election night?
• 44:19
  SPEAKER_01: Obviously now we're going to get really East Coast biased because no one's going
• 44:21
  SPEAKER_01: to say like California 22, which is what I would say.
• 44:26
  SPEAKER_01: Like what California in here?
• 44:28
  SPEAKER_00: Yeah, yeah.
• 44:29
  SPEAKER_00: Maddisen is too.
• 44:31
  SPEAKER_00: That's true.
• 44:31
  SPEAKER_01: I'm outnumbered two to one.
• 44:35
  SPEAKER_01: What are the seats you're looking at?
• 44:38
  SPEAKER_01: And possibly is the seat you're looking at a bellwether,
• 44:41
  SPEAKER_01: do you think it's not the forecast is under or overselling it?
• 44:46
  SPEAKER_01: Where do you think the mismatch is, if there are any?
• 44:50
  SPEAKER_00: I think for me,
• 44:52
  SPEAKER_00: it's gonna be some of these seats in places like Ohio and Iowa,
• 44:56
  SPEAKER_00: because those are also the turf where the Senate could be decided.
• 45:00
  SPEAKER_00: So, you know, for instance, let me just zoom in on this.
• 45:03
  SPEAKER_00: My wife is from this district, so she'll appreciate me bringing this up.
• 45:08
  SPEAKER_00: I was first- Wow.
• 45:09
  SPEAKER_01: Mr. Skelley's from Iowa.
• 45:12
  SPEAKER_01: Okay.
• 45:12
  SPEAKER_01: That is- Yeah.
• 45:14
  SPEAKER_00: Iowa's first congressional district and my in-laws still live in Iowa.
• 45:19
  SPEAKER_00: Uh, it is, uh,
• 45:21
  SPEAKER_00: you know,
• 45:22
  SPEAKER_00: this seat is basically the extremely blue county,
• 45:25
  SPEAKER_00: the bluest county in the state,
• 45:26
  SPEAKER_00: Johnson County, and then a lot of mostly red hinterland, uh, around it.
• 45:32
  SPEAKER_00: And so on election night,
• 45:34
  SPEAKER_00: it's sort of a question of how red was,
• 45:36
  SPEAKER_00: was the,
• 45:37
  SPEAKER_00: the more rural parts and just how blue was Johnson County where Iowa city,
• 45:40
  SPEAKER_00: that's where the university of Iowa is college town.
• 45:43
  SPEAKER_00: Uh, and, uh,
• 45:46
  SPEAKER_00: For me right now,
• 45:47
  SPEAKER_00: this is one where Marionette Miller-Meeks,
• 45:48
  SPEAKER_00: the Republican incumbent, has been in a number of very close calls over the years.
• 45:53
  SPEAKER_00: I think she won by like six votes in 2020.
• 45:56
  SPEAKER_00: Anyway, a lot of tight races over the years for her.
• 46:01
  SPEAKER_00: And right now she's a slight underdog in our forecast.
• 46:05
  SPEAKER_00: But this is the kind of seat that Democrats,
• 46:08
  SPEAKER_00: I think, have to win if they're going to win back the House is to win.
• 46:12
  SPEAKER_00: It's a somewhat right-leaning seat.
• 46:15
  SPEAKER_00: Not overwhelmingly,
• 46:16
  SPEAKER_00: but if they are getting a swing in their direction and knowing
• 46:21
  SPEAKER_00: that Miller-Meeks has been a weaker than average incumbent,
• 46:24
  SPEAKER_00: you know,
• 46:24
  SPEAKER_00: and this, Christina Bohannan, this is her third try, so she can't win it this time.
• 46:30
  SPEAKER_00: The Democratic nominee, that would be telling as well.
• 46:33
  SPEAKER_00: But I think to me, that's like an example.
• 46:36
  SPEAKER_00: And I could pick an Ohio one too,
• 46:38
  SPEAKER_00: like Marcy Captor trying to hold on there,
• 46:41
  SPEAKER_00: incumbent in a red-leaning seat,
• 46:43
  SPEAKER_00: you know,
• 46:43
  SPEAKER_00: it's like,
• 46:44
  SPEAKER_00: If Democrats are able to hold on or flip seats like this,
• 46:47
  SPEAKER_00: it's like,
• 46:48
  SPEAKER_00: you know,
• 46:49
  SPEAKER_00: I think that they're on their way probably,
• 46:51
  SPEAKER_00: but if Republicans are able to beat a captor or hold on to a first,
• 46:56
  SPEAKER_00: it gets tougher because then as the election night is going on,
• 46:59
  SPEAKER_00: we are getting to California and then,
• 47:01
  SPEAKER_00: and we don't,
• 47:01
  SPEAKER_00: maybe won't know for a few days one way or the other how things panned out.
• 47:07
  SPEAKER_00: I guess, Maddisen, what is there one you're looking at?
• 47:13
  SPEAKER_02: I think the ones you pointed out are really good,
• 47:14
  SPEAKER_02: especially when you consider,
• 47:15
  SPEAKER_02: you know,
• 47:15
  SPEAKER_02: poll closing time and that's something that I'm,
• 47:17
  SPEAKER_02: you know,
• 47:18
  SPEAKER_02: actively looking into,
• 47:19
  SPEAKER_02: right, is just how safe are the early closes versus not, right?
• 47:25
  SPEAKER_02: Because, I mean,
• 47:25
  SPEAKER_02: just looking at this, there's a lot of lighter colors as we go right on this map.
• 47:30
  SPEAKER_02: Um, then there are darker colors.
• 47:32
  SPEAKER_02: Um, so that's, I think that you, that's a really good point.
• 47:35
  SPEAKER_02: And the ones you brought up like Virginia two,
• 47:37
  SPEAKER_02: uh, North Carolina one, and I believe it's Florida 14, right?
• 47:41
  SPEAKER_02: Those are some early closes.
• 47:43
  SPEAKER_00: Yes.
• 47:43
  SPEAKER_00: Yes.
• 47:44
  SPEAKER_00: Tampa, Tampa seat.
• 47:46
  SPEAKER_02: Yeah.
• 47:46
  SPEAKER_02: Those are some early closes that I think,
• 47:48
  SPEAKER_02: I think might,
• 47:49
  SPEAKER_02: you know,
• 47:49
  SPEAKER_02: kind of give us some,
• 47:51
  SPEAKER_02: some semblance of sing signal of how the night's going to go
• 47:54
  SPEAKER_02: while California finishes counting in the weeks to come.
• 47:58
  SPEAKER_01: I'm going to lean in.
• 47:59
  SPEAKER_01: Maddisen took one of mine.
• 48:01
  SPEAKER_01: Virginia 2.
• 48:01
  SPEAKER_01: Go to the Virginia 2 forecast for me.
• 48:03
  SPEAKER_01: And here's what I'm looking at.
• 48:05
  SPEAKER_01: Virginia 1 will be the same.
• 48:06
  SPEAKER_01: Same old Pennsylvania 10.
• 48:07
  SPEAKER_01: You look at our probability.
• 48:09
  SPEAKER_01: 72% on the Dem side.
• 48:10
  SPEAKER_01: But I like Virginia because they tend to count fairly efficiently and fairly predictably.
• 48:14
  SPEAKER_01: It'll be early in the night.
• 48:15
  SPEAKER_00: What a great state.
• 48:16
  SPEAKER_00: What can I say?
• 48:17
  SPEAKER_01: Yeah, yeah, yeah, yeah.
• 48:20
  SPEAKER_01: Click on projected vote share.
• 48:23
  SPEAKER_01: This is a close one, right?
• 48:25
  SPEAKER_01: Like,
• 48:25
  SPEAKER_01: and this is where you can make an argument if you're looking at our vote shares
• 48:28
  SPEAKER_01: and you're looking at our probabilities.
• 48:30
  SPEAKER_01: And again,
• 48:31
  SPEAKER_01: the math makes sense,
• 48:32
  SPEAKER_01: but like a generic ballot D plus eight, this isn't a 52-48 seat.
• 48:41
  SPEAKER_01: A generic ballot D plus two, this is also not a 52-48 seat in the other way, right?
• 48:47
  SPEAKER_01: And we're talking about four points that can move it either way in a congressional.
• 48:51
  SPEAKER_01: And so these are the races,
• 48:53
  SPEAKER_01: even if the Democrat wins this one,
• 48:56
  SPEAKER_01: but only wins it by a point and the margins lower,
• 48:58
  SPEAKER_01: that's going to tell me something, right?
• 49:00
  SPEAKER_01: Same with Pennsylvania 10,
• 49:01
  SPEAKER_01: the Scott Perry seat, again, a probability one, but I'm looking at the margins.
• 49:06
  SPEAKER_01: Those are the types I'm looking at.
• 49:07
  SPEAKER_01: It's not necessarily your wins, but what are the margins on some of these close ones?
• 49:10
  SPEAKER_01: Because that will tell me how,
• 49:12
  SPEAKER_01: you know,
• 49:12
  SPEAKER_01: Iowa 1 will go, Marcy capped her seat, certainly some of these California ones.
• 49:19
  SPEAKER_00: Yeah, and actually,
• 49:20
  SPEAKER_00: I think actually one of the, just,
• 49:21
  SPEAKER_00: you know,
• 49:22
  SPEAKER_00: one of the things that is fun about this particular year is that as we get results
• 49:29
  SPEAKER_00: from some of these states, Pennsylvania has sort of three...
• 49:35
  SPEAKER_00: Maybe four actually seats that I think are all worth watching on election night
• 49:41
  SPEAKER_00: and how close or how well each of the parties is doing
• 49:44
  SPEAKER_00: in them could actually tell you a lot about the larger story.
• 49:47
  SPEAKER_00: So just as example,
• 49:48
  SPEAKER_00: you brought up 10th district,
• 49:50
  SPEAKER_00: you got Scott Perry,
• 49:51
  SPEAKER_00: maybe underperforming the Republican baseline district a little bit.
• 49:55
  SPEAKER_00: You know,
• 49:55
  SPEAKER_00: he's a Freedom Caucus guy,
• 49:57
  SPEAKER_00: maybe a bit right to the right of the district,
• 49:59
  SPEAKER_00: even though it is a, it's kind of a slightly right-leaning seat.
• 50:02
  SPEAKER_00: Janelle Stelson only narrowly lost to him in 2024.
• 50:04
  SPEAKER_00: She's back for another try, the Democratic nominee.
• 50:07
  SPEAKER_00: So that one is the one pure toss-up that we have right now.
• 50:11
  SPEAKER_00: But then you go up into the Northeast,
• 50:14
  SPEAKER_00: Scranton District, Robert Bresnahan is the incumbent there, won the seat in 2024.
• 50:21
  SPEAKER_00: Paige Cognetti is the Democratic nominee.
• 50:23
  SPEAKER_00: It's like a 60-40 kind of race right now in our probabilities.
• 50:26
  SPEAKER_00: Just down the way,
• 50:28
  SPEAKER_00: seventh district,
• 50:28
  SPEAKER_00: Bob Brooks is a slight favorite now over the incumbent, Ryan McKenzie there.
• 50:33
  SPEAKER_00: And then in the first district,
• 50:34
  SPEAKER_00: right out in the suburbs of Philadelphia,
• 50:36
  SPEAKER_00: you've got Brian Fitzpatrick,
• 50:37
  SPEAKER_00: who's like the great overperformer among the Republican candidates out there in
• 50:42
  SPEAKER_00: a seat that's very 50-50 politically.
• 50:46
  SPEAKER_00: It was basically 50-50 exactly in the 2024 presidential race.
• 50:49
  SPEAKER_00: But Fitzpatrick comfortably won it.
• 50:51
  SPEAKER_00: It's like,
• 50:52
  SPEAKER_00: In a world where he's in danger,
• 50:54
  SPEAKER_00: you know,
• 50:54
  SPEAKER_00: maybe you're looking at a better environment for Democrats,
• 50:56
  SPEAKER_00: but I think it's fun to have four seats in one state that are,
• 51:00
  SPEAKER_00: now maybe you as a Californian, Scott,
• 51:02
  SPEAKER_00: are used to this,
• 51:03
  SPEAKER_00: but it's nice when you see some other states that have a kind of a sizable number
• 51:07
  SPEAKER_00: of battleground seats that you can maybe tell a larger story even just looking
• 51:13
  SPEAKER_00: at sort of the four there.
• 51:14
  SPEAKER_00: Iowa's got three that are interesting.
• 51:16
  SPEAKER_00: For instance, Wisconsin's got a couple.
• 51:19
  SPEAKER_00: So, you know, I think Pennsylvania... Shocker, Pennsylvania, it's really important!
• 51:23
  SPEAKER_00: And hasn't that been the Decision Desk HQ line forever?
• 51:26
  SPEAKER_00: Because Brandon Finnigan, Pennsylvania.
• 51:29
  SPEAKER_00: The Pennsylvania thing for as long as I've known him.
• 51:34
  SPEAKER_00: Everything comes down to Pennsylvania.
• 51:36
  SPEAKER_00: And the house,
• 51:38
  SPEAKER_00: what happens in the house will weigh in part,
• 51:40
  SPEAKER_00: will depend in part of what happens in those four seats in Pennsylvania.
• 51:45
  SPEAKER_00: Well, guys,
• 51:46
  SPEAKER_00: I don't want to keep you much longer,
• 51:47
  SPEAKER_00: but I didn't know if there were sort of any closing thoughts you had
• 51:49
  SPEAKER_00: about I don't know, the forecast.
• 51:52
  SPEAKER_01: Look, I got to ask the fellow Californian,
• 51:54
  SPEAKER_01: what is the California race, federal or otherwise, that you're looking at, Maddisen?
• 52:01
  SPEAKER_02: Oh, gosh.
• 52:02
  SPEAKER_02: Let me pull up a map really quickly.
• 52:11
  SPEAKER_02: I'm not from Pelosi.
• 52:12
  SPEAKER_02: I'm from California 10, so Mark Desaulnier.
• 52:15
  SPEAKER_02: You're from a much more politically entertaining part of the state than I am.
• 52:20
  SPEAKER_02: Put it that way.
• 52:22
  SPEAKER_02: Yeah, I think a lot of the stuff,
• 52:25
  SPEAKER_02: basically all the like redistricted stuff I think will be interesting and how
• 52:29
  SPEAKER_02: that plays out,
• 52:30
  SPEAKER_02: especially like middle of the middle of the state and Orange County and all
• 52:34
  SPEAKER_02: that good stuff will be kind of fun to watch.
• 52:36
  SPEAKER_02: And I know you've said California 22,
• 52:39
  SPEAKER_02: not that it's actually going to decide the forecast,
• 52:41
  SPEAKER_02: but California 40 will be interesting regardless.
• 52:44
  SPEAKER_01: Yeah.
• 52:46
  SPEAKER_02: There's no way that a Democrat will win that one,
• 52:47
  SPEAKER_02: but it'll still be fun to watch how it comes down.
• 52:50
  SPEAKER_00: Yeah, yeah.
• 52:51
  SPEAKER_00: Actually,
• 52:52
  SPEAKER_00: speaking of California,
• 52:53
  SPEAKER_00: I think I saw, well,
• 52:57
  SPEAKER_00: I think we've been talking a lot about like,
• 52:59
  SPEAKER_00: okay, how's the election night going, right?
• 53:00
  SPEAKER_00: Like, what are some of the possibilities?
• 53:03
  SPEAKER_00: And obviously talking about the East Coast,
• 53:05
  SPEAKER_00: but we're not going to know the outcome of the House until we sort of start
• 53:09
  SPEAKER_00: to see numbers in from California, and then that could take a bit.
• 53:12
  SPEAKER_00: I think it is worth noting that there is, it is quite possible.
• 53:16
  SPEAKER_00: And we kind of,
• 53:17
  SPEAKER_00: I think alluded to this,
• 53:18
  SPEAKER_00: but it's true,
• 53:19
  SPEAKER_00: at least in some cases for both the Senate and the House,
• 53:23
  SPEAKER_00: that it could be days actually before we know,
• 53:25
  SPEAKER_00: there are scenarios where we're waiting on California counts or we're waiting
• 53:30
  SPEAKER_00: on like Alaska Senate race to decide majorities.
• 53:34
  SPEAKER_00: And for us to be able to actually project them,
• 53:37
  SPEAKER_00: Alaska in particular,
• 53:38
  SPEAKER_00: if it's that Senate situation could be a while because between male voting,
• 53:42
  SPEAKER_00: a lot of remote communities, Slow count and rank choice voting.
• 53:48
  SPEAKER_00: A lot of things could come into play there if it's basically 50 Democrats
• 53:52
  SPEAKER_00: and 49 Republicans and we're waiting to see
• 53:54
  SPEAKER_00: if Republicans are gonna get their 50th seat or Democrats are gonna get their 51st.
• 53:57
  SPEAKER_00: On the House side,
• 53:59
  SPEAKER_00: it is possible that we might just be waiting on a couple
• 54:04
  SPEAKER_00: of these California seats like the 48th and the 22nd to see whether
• 54:09
  SPEAKER_00: or not Democrats have actually won a majority or not.
• 54:13
  SPEAKER_00: You know,
• 54:13
  SPEAKER_00: I don't know if it's going to be a full on election week,
• 54:15
  SPEAKER_00: but that possibility is out there and people should keep that in mind.
• 54:22
  SPEAKER_00: Yeah.
• 54:23
  SPEAKER_00: Cool.
• 54:23
  SPEAKER_00: Well, with that, Scott and Maddisen, thank you so much for joining me to talk about this.
• 54:30
  SPEAKER_00: Again, this is Scott Tranter,
• 54:32
  SPEAKER_00: the head of data science here at Decision Desk HQ and Maddisen Murrell,
• 54:37
  SPEAKER_00: who is a data science fellow who has been working Very closely
• 54:41
  SPEAKER_00: on our forecast model.
• 54:44
  SPEAKER_00: And I wanna thank Stephen Stover for producing this episode and Eric Cunningham
• 54:49
  SPEAKER_00: for his aid as well.
• 54:50
  SPEAKER_00: And also just would like to applaud the entire Decision Desk HQ team
• 54:56
  SPEAKER_00: The forecast model,
• 54:58
  SPEAKER_00: both the model and the visualization of it, I think is really tremendous.
• 55:03
  SPEAKER_00: So you should definitely go check it out at votes.decisiondeskhq.com, DDHQ Votes.
• 55:10
  SPEAKER_00: It is, I think just, it's, look,
• 55:14
  SPEAKER_00: you can waste a lot of time clicking through
• 55:18
  SPEAKER_00: all these races and looking over sort
• 55:22
  SPEAKER_00: of the possibilities and it's, It looks great and so highly recommend.
• 55:29
  SPEAKER_00: So with that,
• 55:31
  SPEAKER_00: I want to thank everybody for watching and listening
• 55:33
  SPEAKER_00: and I will see you again next week.