Sabato: Bet on a Republican Senate

[Miles]

Article.



The most notable change here was CO from Tossup to Lean R.

[IceSpear]

Let's hope two of Larry's "lean R" ratings are just like his "lean Rs" for Virginia and Florida in 2012.

[Lief 🗽]

Let's hope two of Larry's "lean R" ratings are just like his "lean Rs" for Virginia and Florida in 2012.

He said Virginia was a lean R in 2012?

brb printing out a picture of larry sabato's face and throwing it in the garbage

[KCDem]

Alaska is LEAN REPUBLICAN?!     

More evidence that Loony Larry doesn't know what he's talking about.

[IceSpear]

Let's hope two of Larry's "lean R" ratings are just like his "lean Rs" for Virginia and Florida in 2012.

He said Virginia was a lean R in 2012?

brb printing out a picture of larry sabato's face and throwing it in the garbage

Still not as bad as Rothenberg though, where Obama won literally every single state in the "pure toss up" category: http://rothenbergpoliticalreport.com/ratings/president

[Lief 🗽]

Well Rothenberg is a useless coward. What's the point of predicting an election if you leave 1/5 of the electoral votes in the toss-up column?

[Vosem]

Sabato's probably the best of the mainstream predictors, but he's not foolproof and his habit of giving every race a 'Leans' at the end means he's made more wrong predictions than others. Recently they've been favorable to the GOP, but this hasn't always been so: he projected Florida for Kerry in 2004. (He got all but 2 states correct, flipping FL and WI, resulting in his projection of a 269-269 election that year).

EDIT: Looking back through the archives, I want to note that Sabato had a phenomenal House projection in 2004: he only got one election wrong, which was projecting Baron Hill be reelected against Mike Sodrel. (Hill would end up beating Sodrel in the end, in a rematch 2 years later). Other than that, he got 434 races correct.

[Attorney General, Senator-Elect, & Former PPT Dwarven Dragon]

Let's hope two of Larry's "lean R" ratings are just like his "lean Rs" for Virginia and Florida in 2012.

He said Virginia was a lean R in 2012?

Unlike Rothenberg and Cook, who leave a bunch of races in their toss-up category, Sabato makes a prediction for every single race by election day. This is one of the things I really like about him because it allows for true analysis of how accurate his modeling is, as opposed to the other two who can claim 'complete correctness' as long as their toss-ups go roughly equally to each party and they don't make any 'high-risk' advantage assignments.

[Nichlemn]

Let's hope two of Larry's "lean R" ratings are just like his "lean Rs" for Virginia and Florida in 2012.

He said Virginia was a lean R in 2012?

Unlike Rothenberg and Cook, who leave a bunch of races in their toss-up category, Sabato makes a prediction for every single race by election day. This is one of the things I really like about him because it allows for true analysis of how accurate his modeling is, as opposed to the other two who can claim 'complete correctness' as long as their toss-ups go roughly equally to each party and they don't make any 'high-risk' advantage assignments.

Fun fact - you can put every single race in the tossup column and barring a wave  (which you can excuse yourself before, because in waves "one party wins a lot of the tossups"), you'll always be roughly right, especially for the House.

[IceSpear]

Let's hope two of Larry's "lean R" ratings are just like his "lean Rs" for Virginia and Florida in 2012.

He said Virginia was a lean R in 2012?

Unlike Rothenberg and Cook, who leave a bunch of races in their toss-up category, Sabato makes a prediction for every single race by election day. This is one of the things I really like about him because it allows for true analysis of how accurate his modeling is, as opposed to the other two who can claim 'complete correctness' as long as their toss-ups go roughly equally to each party and they don't make any 'high-risk' advantage assignments.

Which is funny since in Rothenberg's "toss ups", Obama won 100% of them.

[Nichlemn]

Let's hope two of Larry's "lean R" ratings are just like his "lean Rs" for Virginia and Florida in 2012.

He said Virginia was a lean R in 2012?

Unlike Rothenberg and Cook, who leave a bunch of races in their toss-up category, Sabato makes a prediction for every single race by election day. This is one of the things I really like about him because it allows for true analysis of how accurate his modeling is, as opposed to the other two who can claim 'complete correctness' as long as their toss-ups go roughly equally to each party and they don't make any 'high-risk' advantage assignments.

Which is funny since in Rothenberg's "toss ups", Obama won 100% of them.

That's not unexpected as Presidential results tend to be quite closely correlated among states, so the uncertainty often disproportionately favours one candidate.

Still, it is hard to falsify a prediction after the fact. Even Nate Silver's favoured method, whether your predictions are calibrated correctly, can still be horribly wrong if your calibration errors cancel each other out (as they do on in my "rate every race a tossup" example).

[SWE]

Sabato eliminated the Tossup category for House

[SPC]

Aside from the pure tossups (of which Rothenberg has 11), the two only disagree on NH-1.

[Bandit3 the Worker]

Can I literally bet on it? I mean that literally. I want to be able to at least be compensated if the election goes wrong.

[ElectionsGuy]

Sabato eliminated the Tossup category for House

The only one that I seem to strongly object to is NY-1, that for sure should be leaning D.

[Lief 🗽]

Is CA-31 really only Lean D? I thought it was supposed to be an easy pick up.

[Senate Minority Leader Lord Voldemort]

Is CA-31 really only Lean D? I thought it was supposed to be an easy pick up.

Democrats have been ahead in the polls, but not really a commanding lead. It will be a pick up, but it'll be a narrow Democratic win. It's just a factor of low turnout, the incumbent Democrat in 2016 should have a much easier go of it.