Ricochet is the best place on the internet to discuss the issues of the day, either through commenting on posts or writing your own for our active and dynamic community in a fully moderated environment. In addition, the Ricochet Audio Network offers over 50 original podcasts with new episodes released every day.
My 2018 Midterm Prediction Model
Inspired by @HeavyWater‘s challenge to come up with my own midterm prediction model, I’ve come up with a fairly simple one that anyone can use. I encourage you to plug in the relevant numbers based on the following algorithm.
I realize that reasonable people can disagree about how many of POTUS’s tweets qualify as “crazy” in any given week, so I trust you to use your own discretion. I gave him three this week for the tweets in which he rubbed salt in the wounds of Puerto Rico over last year’s hurricanes. I thought that was ill advised, and he paid the price in my model. Others may disagree.
Try it and report your results. Right now, I’m at D + 2.4.
- Take my current bank balance.
- Subtract my bank balance one year ago.
- Divide the difference by my bank balance one year ago.
- Throw darts at a picture of Justin Trudeau (because it helps me focus).
- Subtract Trump’s current daily rate of crazy tweets (averaged over the past week).
- Multiply that number by 100. That’s the true Republican polling (dis)advantage.
Please copy your Edie, at your bank, on this article. I was trying to calculate the index, and I must have spent 30 minutes trying to get her to give me your current bank balance. She finally did, but she said I would have to pay a fee of 35.00 to get your balance from a year ago because all that data has been transferred to floppy disks.
I’m glad I bank with a behemoth like Chase. Try using your own bank information.
Hahaha! Works as good as any of ’em.
Oh, and love that pic of Justin. What an idiot.
Uh oh, it looks like some Republican-held battleground districts could be holds, based on my model.
Reuters:
Of course, their bank balance gains have to outweigh their suburbanite overreaction to crazy tweets.
I think you need to add a term to take into account the number of times Nancy Pelosi has spoken in public in the last week. It doesn’t matter the subject – no one can figure that out anyway
Your method is far superior to that of Nate Silver . . .
I’d say all the handwringing about Trump’s tweets is ridiculously overwrought, the average voter pays little attention to that. It’s strictly the preserve of pundits and media obsessives. What does concern me a bit is the endless media hysteria, if someone is getting a steady diet of that anytime they turn on the news how does that affect them? I’m a bit worried that it could drive some independents to just throw up their hands and stay at home or worse vote for Democrats hoping that divided government will cause everyone to just calm down.
Even in the worst of the Bush years the press wasn’t this relentlessly partisan, it has to have some effect doesn’t it? I suppose we shall see in November.
Most people who read this have more money in their bank account than they did a year ago, amIright? I used that metric instead of annual income, because it kind of adjusts for inflation. If the price of stuff you buy increases by more than your income, then your bank balance is gonna go down. Also, people who just took a large, unexpected financial hit are probably wondering “what good economy, you Republican jerks!?” Those who’ve been doing really well financially but are really sensitive to tweets that don’t adhere to the Marquess of Queensbury rules are probably voting Democrat or if they were otherwise inclined to vote Republican will spend Election Day on the fainting couch.
The “average voter” and “independents” don’t vote in the midterms. Only hardcore partisans like us do.
This exactly, the anti-Trump media echo-chamber is the reason that I weighted crazy tweets so heavily, but you’ll notice that positive, admirable tweets get Republicans no points, because they get ignored.
Also, we need to remember how much these crazy tweets traumatize suburban Republicans who are so crucial in swing districts. Let’s just hope that the huge income gains in those swing districts will compensate for the crazy tweet factor.
Thanks, I will be giving updates from my prediction model periodically until Election Day. Let’s face it, Nate Silver was way off in 2016. I’ll put my common sense, back of the envelope model up against his Hollywood/Manhattan model and see who gets closer on Election Day.
Yeah, except Nate Silver really was not off in 2016 on aggregated polls. Trump won about the national percentage he was predicted to win. The only reason he won was by razor-thin margins in very particular states, that tipped the race electorally in a way that no one foresaw. I’d be really careful about crowing over this particular topic. Things look really bad for Republicans, and we’d better take it seriously.
Absolutely, everyone needs to vote!
Republicans going off at 5 to 1? That’s nonsense on stilts. Notice that Nate Silver won’t make book on those odds. Predictit has it closer than that.
Right now, Nate Silver’s House model has the House going to the Democrats and Silver’s Senate model has the Senate staying in Republican hands (which includes the possibility of a Senate split 50 to 50 since Mike Pence would be the tie breaking vote).
Right, but his odds on the House are way off. Would you really give anyone 5 to 1 odds on Republicans keeping the House? To clarify, I bet $1, and if Republicans hold the House I cash my ticket in for $6 ($5 profit). Predictit is under 3 to 1 on that bet. Silver is closer to the actual line on the Senate, but he’s still biased in favor of the Democrats on the Senate. Let’s be clear, Silver will not take the action on his odds. He will under no circumstances send me $6 if I bet $1 on a Republican hold, and Republicans hold. So it’s hard to take him too seriously.
In 2016, 538 overestimated Democrat support. I’m saying now that in 2018 538 is overestimating Democrat support. That’s not crazy talk.
But in 2012 Nate Silver nailed it almost perfectly. So, we’ll have to see what the results are like in a little over 7 weeks.
It’s not like Silver is just pulling numbers out of thin air. He’s a math nerd and he uses all kinds of data, including a formula that determines how reliable a poll is, whether a poll generally is biased toward one party or another, and then reviews his model to see how well it would have performed in previous elections and then he refines his model further.
Also, if you have the patience, you can read the web page on 538 where Silver explains his model in detail.
I’m not saying I would bet my bank account on his predictions. But that’s because I don’t like gambling money I can’t afford to lose.
As for 2016, when others gave Hillary Clinton a 98 percent change of winning, even before election day Nate Silver explained why those other models weren’t factoring in the possibility that all of the polls were wrong in the same direction.
I’m not saying Nate Silver is Moses. I’m just saying I’ll put more value in his election model than any other model available.
However, people who just want to “go with their gut” are welcome do to that.
And other times he doesn’t nail it.
Polling these days is not meant to reflect opinion, rather it is to sway it. IMHO, all this “blue wave” polling and subsequent talk is designed to discourage anyone from voting Republican – why bother voting if Nate Silver guarantees you’re going to lose?
Here’s a few things to remember:
I frankly worry when I see folks on the right bashing polls. There are many solid, substantial reasons to see a blue wave coming. It’s not a guarantee. But I take little solace in pointing out unreliable polls in an upset that no one really saw coming. 2016 was the aberration, not the myriad of accurate polls before and after.
Nate Silver isn’t guaranteeing that anyone will win or lose.
For example, if you look at 538’s analysis of the race in the California 48th congressional district, his web page shows that the Democrat candidate, Harley Rouda, has a 66.5 percent chance of winning and the Republican candidate, Dana Rohrbacker, has a 33.5 percent chance of winning.
In Tennessee’s US Senate race, 538 gives the Republican candidate, Marsha Blackburn, a 69.8 percent chance of winning and the Democrat candidate, Phil Bredesen, a 30.2 percent chance of winning.
And Silver isn’t just “going with his gut.” His methodology is public, for everyone to look at and comment on.
When folks talk about Trump tweets I always wonder if they are talking about Trump tweets (Donald J. Trump @realDonaldTrump ) or his Pesidential tweets (President Trump@POTUS). Might make a difference for your “crazy tweet” calculation.