Anna well here's the national average what we got here is all of the recent major polls nationally what is the sort of composite when you average them together Harris with a three-point Advantage nationally over Donald Trump now if you're a Democrat the good news besides the fact that you're ahead is that's an improvement for Democrats from where things were before Harris became the candidate Joe Biden almost all year was trailing Donald Trump so that's changed Harris ahead by a few points there now if you're a republican what you can take Salis from here is this is not unfamiliar territory for Donald Trump take a look here at the last two elections 2020 2016 the polling at this point coming out of Labor Day beginning the fall rush in 2016 Hillary Clinton LED on average by Five Points of course Donald Trump won in 2016 and Joe Biden had an even bigger lead on average Labor Day 2020 Donald Trump didn't win that election but certainly in the Electoral College he came this close to doing so Anna so Trump has run from behind before certainly so yeah Trump's running behind according to the polls uh we got Mark Mitchell on and the question um first of all I want Mark to explain what the term suppression poll means and I want to ask him whether there's been any time in recent history where you've seen a more aggressive of use of suppression polls than now hey Mark welcome back to the War Room sir happy to be here yeah suppression polls it looks like there's a lot of them right now so suppression poll is a poll that somebody puts out they probably do some dirty tricks to it and the idea is to drive a negative narrative there's been a lot of really great examples although I have to say this year is giving 2020 a run for its money but a really good example not that long ago Fairley Dickinson did a poll that showed that Harris was beating Trump nationally by one point and then further in the poll they asked a bunch of questions about gender and race on purpose to see if it impacted how people responded to the matchup question and then it showed that Harris was leading by seven and then immediately the hill took that headline Harris plus 7 number and ran with it as though that was the actual race and there's no way and another really great example that yeah go ahead hang on let me just for the audience so they get the lingo down I think that's called a push question oh is that what they call that when okay so you do the initial where is the race and then you you ask a bunch of questions that are going to basically make it more favorable to whoever you're whoever you're pimping and then the push qu so rather than do the matchup they did the push question right 100% which is why you put matchups in the beginning of the poll one of a million dirty tricks that pollsters can do but there was there was a suppression poll right there in that clip if you go to Real Clear Politics aggregate which I think they use very even-handed standards to determine who they let into their tracker uh I think it's a pretty fair representation of all the polls that are out there they show Harris leading by 1.8 but In that clip they show Harris's average is actually plus three over Trump so already they've done something selective there to get to Harris plus three in their segment just the framing of how they put that segment together was really scummy and dirty but there was a poll out um ABC ipsos um that had Harris leading by I think four in a two-way six in a multi-way I don't know it was a lot right pretty close to Harris plus three but if you look in the internals there they have favorability number by can they had Harris with a plus three Net favorability and Donald Trump with like a negative 24 this is the man that received more votes than any Republican in history and you're telling me in this 20124 poll he's been around for a decade everybody knows who he is there's not many people who are undecided about Donald Trump you're telling me that he's got a negative he's got only a 33% favorability rate and yet is somehow taking 46% of the vote there's absolutely no way so I just there you show me polls that have Harris plus three plus four plus five they are doing something what let me just just as a as a Layman what that would reveal is oversampling of of Democrats right wouldn't that reveal that many flavors of that and this one's tricky because it's hard to spot and luckily they did disclose this but I can go out and Sample Republicans and Democrats and then wait the poll to party but the problem is is that there are Democrats that support Trump there's Republicans that don't and then there's Independence and so I can measure their 2020 vote and then wait for that but even then there's Republicans that just don't have a very favorable impression of trump but they're still going to potentially vote for him somehow ABC got a ton of those people and so all these things have ramifications for the numbers but it's like they pulled almost entirely never Trump Independents and Republicans listen in that poll we have Donald Trump's favorability in our numbers like 80% among Republicans so let's just assume 90% you know with the Democrats let's assume 100% Republicans are about a 30 electorate that gets you to 33% Trump favorability so that means like in their independent sample Trump has like a z% favorably it just makes absolutely no sense if I got results like that I would really be questioning what happened in my poll because if I was not being disingenuous like there is a real problem with the way that we sample and of course in our numbers everybody who's been following us knows this Harris has always been a little bit less favorable than Biden Biden's been less favorable than Trump and Trump isn't super favorable but he's gotten 49 50 5 1% and Biden was higher than Trump until he had his presidential term right and it got a lot lower all of that to me makes internal consistent sense and you don't need to be a pollster to rip apart these things go download the cross tavs if you're one of the pollsters that isn't putting the cross tabs behind the pay wall and look for consistency and that's where you'll start to find these tactics don't need again a lot of lot of lingo here a lot of lingo here um give give the the listeners a an example of a cross tab where you're I think it's something like U Democrats on abortion or Young Folks on forward policy something like that what what's a cross Tab and and what how would a cross tab reveal shry you're never going to know 100% if a poll is crooked there is no amount of disclosure because even if I H had their data I don't know if they're spiking data points they could just be throwing out extra Trump supporters but the most that you're going to find is in a document called the cross tavs which is pretty much the deliverable of a poll of course things are in therein yeah yeah the margin of error the sampling methodology it's going to have some stuff like that um but what it's also going to show is not just the Topline numbers which is like nationally Trump versus Harris or nationally every single question they should be every question they ask sometimes they don't disclose them all but then it's going to show how each single demographic that they have data for in the poll answer the questions that is where all of the bodies are buried because I can look at by party who says they're going to go Trump versus Harris and the most important column is independence and so my Independence right now tell me they're going for Trump by five to 10 point margin depending on the poll which is a really major reversal from in our polling in 2020 when Trump lost Independence but if you if you see a mainstream media pollster right now that hits Harris up they're probably showing Trump losing to Harris among Independents and that to me especially in the swing States is going to be the key to this election and who's right and who's wrong and in my opinion if you're New York Times it's really hard to get Trump supporting Independence to respond to your poll I think that might be the key to what's happening of our story again maybe we're just out in right field we're going to be toally I think over cycle over cycle we've been proven right but the 18 to 39y olds the Hispanic voters and even the black voters have been going more and more Trump what's the lag time uh but now in in the trade so what can you EXP if a poll comes out today it's not reflective of today it's reflective of some time prior to today how many days or weeks um is it usually uh the lag you just pointed out scumbag pollster tactic number 37 which is sit on your data okay now every pollster is going to disclose ostensibly honestly the exact days the sample was taken and then they maybe sit on it for a week or two and then drop it and that was a lot of what was happening in August because everybody was sampling in the very beginning of Harris's candidacy when she got a lot of positive press and then dumping the polls two weeks later to drive a Kamal mum narrative um but we're the only people that do this we started doing it this cycle if Denver wants to throw up the chart we collect data every night and to show that we are being as transparent as possible we show you what happened literally last night so this is data that was taken you know 7 to 9:00 last night and everybody else they sort of go hey I'm going to pull today and they go into the field and then they pull and then they wait two weeks and so there might be an absolutely huge like two three weeks is a massive amount of time in this cycle because if you're getting data from 3 weeks ago it's the DNC bounce right yeah so what we showed so Mark how many hang onang how many people do you sample a night and then how do you decide when and how to aggregate them to issue kind of a a broader poll you sample like two you get 200 responds a thousand or what what do you do every night sample size is about 360 yeah it's about 360 people a night that go into that chart so each individ night is about a plus or minus 5 margin of error anybody who studied statistics knows that that's a wide range however the polls are going to tend to be close to the population mean and so you really can look at the trends and that's why I put it out there because there's all these narratives about the race shifting or momentum and if you look at that chart it really kind of doesn't show it that much it shows Trump again there's swings in there it shows Trump with a pretty twoo lead over time and because we're not trying to drive narratives every Thursday we just take an aggregate of the last five business days what about the argument okay let's say let's say that you've got the the one of the only honest polls out there you show Trump in the lead by two you know the Carville theory is that all these other polls are off by three so you know if if if ABC shows Trump trump down by three and you got him up by two uh what do we glean from that yeah look at look at the real Flair politics aggregate it's Harris 1.8 I think based on what we saw across the industry in 2016 to 2020 I think it is reasonable to assume the entire mainstream polling industry leans about three points left and probably really didn't correct that okay and if you add three points to negative 1.8 you get essentially exactly what I do which is Trump winning nationally in popular vote I got it I got it yeah yeah yeah yeah he he's never been there before right all right brother you keep doing what you're doing how can uh the posi stay AB breast of particularly of this daily poll where do you want them to go uh go to Twitter at rasor Poole I started posting about two months ago too markor r _ Mitchell the conversation is happening on Twitter and we have a lot of stuff about this race coming out just today a lot of new polling about how Harris isn't going to be able to distance herself from the failures of the Biden Administration I think it's a pretty important story so love to have you guys follow us all right man always a pleasure posi check mark Mitchell and his page out it's a nightly adventure with Ras musen