Straight Dope Message Board

[Captain Lance Murdoch]

Originally Posted by Bricker

I just took my PredictIt winnings from Gianforte's victory and bought HANDEL.GA06.2017 shares at 42 cents.

That is: I predict, with not just a voice but real world money, that Handel beats Ossuff.

Of course, the reason Handel was at 42 cents in the first place was that other people betting with real-world money were predominantly going the other way.

[D'Anconia]

Originally Posted by iiandyiiii

Yes, talking about facts is ignorant, unless those facts are the electoral college.

I'm genuinely puzzled by how you don't seem to understand that the popular vote is, inherently, irrelevant.

Therefore, predictions about it, accurate or not, are also irrelevant.

[CarnalK]

Originally Posted by D'Anctonia

I'm genuinely puzzled by how you don't seem to understand that the popular vote is, inherently, irrelevant.

Therefore, predictions about it, accurate or not, are also irrelevant.

You are most certainly confused. Talking about predictions when discussing a predictions website is pretty much the whole discussion. "But not the popular vote prediction", you say, because Hillary lost so get over it morons!

[iiandyiiii]

Originally Posted by D'Anconia

I'm genuinely puzzled by how you don't seem to understand that the popular vote is, inherently, irrelevant.

Therefore, predictions about it, accurate or not, are also irrelevant.

Irrelevant to what? This discussion covers more topics than just the electoral college results.

[Bricker]

Originally Posted by Captain Lance Murdoch

Of course, the reason Handel was at 42 cents in the first place was that other people betting with real-world money were predominantly going the other way.

True.

But that's the other great benefit of real-world stakes: there is a win-loss record to observe.

[John Mace]

Originally Posted by iiandyiiii

Irrelevant to what? This discussion covers more topics than just the electoral college results.

True. It tells us that HRC was very popular (compared to Trump) in CA and NY. I therefore predict that the Democratic candidate will win both states in 2020. I'll go so far as to say that will happen even if Trump isn't the GOP candidate!

[iiandyiiii]

Originally Posted by John Mace

True. It tells us that HRC was very popular (compared to Trump) in CA and NY. I therefore predict that the Democratic candidate will win both states in 2020. I'll go so far as to say that will happen even if Trump isn't the GOP candidate!

The discussion (or this side-discussion) was actually about the prediction skills of Nate Silver's 538 website.

[DSYoungEsq]

Originally Posted by iiandyiiii

The discussion (or this side-discussion) was actually about the prediction skills of Nate Silver's 538 website.

Yes, and the issue is this: is the ability to successfully predict an irrelevant result meaningful when it comes to discussing the success of that website in making predictions about the 2016 United States Presidential and Congressional elections?

Personally, I don't see how that can be answered as yes or no. Watching people play tennis with the two is getting redundant around here.

[Dahnlor]

Meanwhile, Karen Handel says at last night's debate, "I do not support a livable wage."

[Chronos]

The ideal predictor would always predict some result with a 100% chance, and then be right 100% of the time. There are no ideal predictors, and so real-world predictors must make predictions with less absolute odds, such as 2 to 1. And if a predictor makes a number of 2 to 1 predictions, and all of them end up going to the more-likely outcome, then that predictor is a pretty bad predictor. Out of all of the 2 to 1 predictions that a predictor makes, they should be "right" on about two thirds of them, and "wrong" on about one third of them.

[Fotheringay-Phipps]

Originally Posted by DSYoungEsq

Yes, and the issue is this: is the ability to successfully predict an irrelevant result meaningful when it comes to discussing the success of that website in making predictions about the 2016 United States Presidential and Congressional elections?

Personally, I don't see how that can be answered as yes or no. Watching people play tennis with the two is getting redundant around here.

It's pretty obvious that iiandyiiii is completely correct about all this.

The ability to predict the popular vote is the essence of all polling predictions. The national popular vote does not directly determine the winner of the presidential election, but it's not as if modelers aren't doing their very best when they try to predict it. They are, and the extent to which they've been successful is a valid indicator of their success at this basic and fundamental polling skill.

[iiandyiiii]

Originally Posted by Fotheringay-Phipps

It's pretty obvious that iiandyiiii is completely correct about all this.

I think I'm going to have to frame this.

[Fotheringay-Phipps]

I call 'em as I see 'em ...

[Left Hand of Dorkness]

Originally Posted by Chronos

The ideal predictor would always predict some result with a 100% chance, and then be right 100% of the time. There are no ideal predictors, and so real-world predictors must make predictions with less absolute odds, such as 2 to 1. And if a predictor makes a number of 2 to 1 predictions, and all of them end up going to the more-likely outcome, then that predictor is a pretty bad predictor. Out of all of the 2 to 1 predictions that a predictor makes, they should be "right" on about two thirds of them, and "wrong" on about one third of them.

You know this correction, but this is the Straight Dope, where nothing is too pedantic to point out .

I predict that there's only a one in three chance that Bob will take an umbrella to work tomorrow, and Lakaia has only a one in three chance of taking an umbrella to work tomorrow, and Bill and Linda and Jiminez and Jo all have only a one in three chance of taking an umbrella to work tomorrow.

If all of them take an umbrella to work tomorrow, that doesn't make me a pretty bad predictor. It just means that the one-in-three event happened.

We should be careful about dinging polling agencies--and statisticians--for missing a lot of results when the true results were both within the margin of error and also were related to one another.

[Fotheringay-Phipps]

Originally Posted by Bricker

I just took my PredictIt winnings from Gianforte's victory and bought HANDEL.GA06.2017 shares at 42 cents.

That is: I predict, with not just a voice but real world money, that Handel beats Ossuff.

I wonder if you can share the basis for this confidence.

I'm aware that Price won by a big margin, but Price was a multi-term incumbent. That's not a good predictor for the margin by a newcomer. (Trump won by about 1%.)

I would be pleased if the Republican won, but that's not a reason to assume that will happen. Looks to me like a toss-up or Ossuf is a slight favorite, in line with the betting and the polls. I'm curious as to the basis for confidence that these are off.

[Little_Pig]

AJC poll: Ossoff opens lead over Karen "Keep 'em Barefoot and in the Kitchen" Handel in Georgia’s 6th.

[BigAppleBucky]

Originally Posted by Fotheringay-Phipps

. . . Ossuf . . .

What is it with Republicans and spelling? They use "Democrat" Party instead of Democratic Party. It's Ossoff.

***
How much did Handel's "I don't believe in a living wage" gaffe hurt her chances?

[Snowboarder Bo]

PredictIt won't let me register; says the service is not available in my area. I suppose it's because the site is basically gambling and I live in Las Vegas, so.... um...

[asahi]

The race, as predicted, is tightening.

http://www.politico.com/story/2017/0...-handel-239650

I suspect that Handel will probably get just enough votes to win in a squeaker. It'll come down to turnout, and when you factor in the historical and demographic advantages that a republican enjoys in this particular district, it's hard to bet against it. However, Ossoff is well-funded, not running as a scary liberal, and probably riding strong anti-Trump sentiments.

I think Trump needs this win badly. If Handel loses then I think it's the first solid piece of evidence that #neverTrumpers can use to hit the eject button before the real election next year. Even if Handel wins, I think the Democrats have to feel encouraged that they were again competitive in a GOP stronghold. But the Democrats still need a message -- just hating Trump isn't enough. In fact, if the Democrats do somehow manage to get Trump forced out of office and Pence takes over, they'd better damn well have a campaign strategy ready.

[The Other Waldo Pepper]

Originally Posted by asahi

I think Trump needs this win badly. If Handel loses then I think it's the first solid piece of evidence that #neverTrumpers can use to hit the eject button before the real election next year. Even if Handel wins, I think the Democrats have to feel encouraged that they were again competitive in a GOP stronghold.

Actually, that last bit makes me wonder about the first part: the other elections since Trump took office, the ones where the Republican endorsed by Trump won: were you thinking -- during each of those -- that "Trump needs this win badly"; and did you think that "the Democrats have to feel encouraged" after the losses?

[asahi]

Originally Posted by The Other Waldo Pepper

Actually, that last bit makes me wonder about the first part: the other elections since Trump took office, the ones where the Republican endorsed by Trump won: were you thinking -- during each of those -- that "Trump needs this win badly"; and did you think that "the Democrats have to feel encouraged" after the losses?

I actually did feel that way about the Montana race, though to a lesser degree. But I think this race is different because of how typically conservative this voting district is and also because Trump's popularity seems to be now in a slow but steady decline. We're possibly getting to a point where Republicans might start to question Trump's value to them and begin to imagine life without him.

[John Mace]

I'm not sure that a race where neither of the candidates is an incumbent tells us much about how the majority of the races are going to play out, since most of those will involve incumbents. Those Congresscritters have an amazing capacity of getting themselves re-elected.

[Left Hand of Dorkness]

Interesting article in The Nation.

Its thesis, roughly, is that we're too focused on a few high-profile races; while we look at those, we're missing the larger story of a bunch of Sanders-inspired progressives winning primaries and special elections at the local level around the country, both in solidly Democratic districts and even in places with traditional Republican majorities.

 While DC pundits have kept a reasonably close watch on congressional special elections in the districts won by Trump—and have seen signs of political movement— some of the clearest signals are coming from special elections for seats in the state legislative chambers that will redraw congressional district lines after the 2020 Census. Progressive Democrats running in historically Republican districts in New Hampshire and New York won breakthrough victories in May. “Republicans should absolutely be concerned: Two Republican canaries died in the coal mine yesterday,” GOP political consultant William O’Reilly said after the results were announced. He explained that “Trump voters and other Republicans simply didn’t show up, and voters from the left did.”

(The article has a lot more examples, but it lacks statistics; I'd be interested to see stats)

[RTFirefly]

Interviewer: "I guess I want to know why you think gay parents are not as legitimate as heterosexual parents."

Karen Handel: "Because I don't."

Well, that settles that, I guess.

But Bricker's got his money on Handel. Where your treasure is, there will your heart be also.

[Lance Turbo]

Over the last six hours or so this race has swung from about 58/42 Ossoff to even. Am I missing a body slam or something? Swings like this usually happen for a reason.

[Bricker]

Originally Posted by RTFirefly

But Bricker's got his money on Handel. Where your treasure is, there will your heart be also.

My money is on Handel because I predict, with real-world consequences for me if I am wrong, that she will win.

In other words, this is about being right: being accurately able to predict the outcome of real world events, as opposed to offering empty predictions based on desired outcomes as opposed to actual real world considerations.

The SDMB thrives on the model of a bunch of left-leaning people telling each other solemnly that the left-wing outcome will happen. And if it does, they congratulate each other; if it does not, they incur no ridicule or stigma for incorrect predictions. They simply move on to the next discussion. Trump wasn't going to win; then the Stein recount was going to fix things; then the "Hamiltonian electors," would give the election to Hillary anyway (or to an acceptable Republican); then Congress could refuse to count the electors votes. All of these were discussed in serious, extensive threads over days; all of them vanished after they were no longer possible.

All of them shared this characteristic: they were expressions of hopeful, wishful desire, not sober predictive analysis. And while some of the threads certainly acknowledged that the scenarios might not happen, others -- especially the Trump election itself -- brooked no dissent.

All of this to say: I am writing this in advance of the Georgia election. I may be wrong. I think my analysis, especially at 42 cents to win a dollar, was a good investment. But I don't claim it as a certainty. What I do say is that I was willing to commit to an actual. measurable real-world loss as a consequence if I erred.

[Lance Turbo]

Now things have swung to a slight Handel advantage.

Bricker, you could get out right now with like 115% ROI if you felt like it. Based on a similar conversation regarding MT I'm pretty sure you won't.

[Left Hand of Dorkness]

Originally Posted by Bricker

My money is on Handel because I predict, with real-world consequences for me if I am wrong, that she will win.

In other words, this is about being right: being accurately able to predict the outcome of real world events, as opposed to offering empty predictions based on desired outcomes as opposed to actual real world considerations.

The SDMB thrives on the model of a bunch of left-leaning people telling each other solemnly that the left-wing outcome will happen.

Oh, spare us. This is about your enjoying gambling, no more, no less. If you're correct, you made a bet that panned out; it doesn't indicate any sort of ability to accurately predict the outcome of real world events, any more than if I predicted a die would roll above a 3 and it rolled a 5 I'd show an ability to predict dice rolls accurately.

Of course people predict things the way that they hope they'll turn out, often. You do too. That's not interesting. What's interesting is analysis.

[Left Hand of Dorkness]

Nate Silver's analysis:

The “takes” you’ll read about the special election in Georgia’s 6th Congressional District on Tuesday night are probably going to be dumb. A close outcome in either direction would be consistent with what we know about the political environment. So if either Democrat Jon Ossoff or Republican Karen Handel wins narrowly, it will be portrayed as a more important predictive signal than it really is. A blowout result would be a bigger deal. But even then, Georgia 6 is a slightly unusual district, and the election would be one data point among many.

Worth reading, because after this paragraph he goes into why it's an interesting data point.

[E-DUB]

Maybe the swing is attributable to a "backlash/sympathy vote" following the baseball practice shooting.

[Bricker]

Originally Posted by Lance Turbo

Now things have swung to a slight Handel advantage.

Bricker, you could get out right now with like 115% ROI if you felt like it. Based on a similar conversation regarding MT I'm pretty sure you won't.

Yeah, I'm actually more tempted here. I got in at 42 and I can sell at 52. But I think I'd rather ride it.

[Bricker]

Originally Posted by Left Hand of Dorkness

Oh, spare us. This is about your enjoying gambling, no more, no less. If you're correct, you made a bet that panned out; it doesn't indicate any sort of ability to accurately predict the outcome of real world events, any more than if I predicted a die would roll above a 3 and it rolled a 5 I'd show an ability to predict dice rolls accurately.

Of course people predict things the way that they hope they'll turn out, often. You do too. That's not interesting. What's interesting is analysis.

I've been wrong. But not too often.

I'd argue that I'm right often enough to claim a better-than-random ability to accurately predict the outcome of events on which I take a position.

[Snowboarder Bo]

Originally Posted by Bricker

I've been wrong. But not too often.

I'd argue that I'm right often enough to claim a better-than-random ability to accurately predict the outcome of events on which I take a position.

I live in Las Vegas. Everyone I know who gambles regularly will tell you that they are ahead, winning more than they lose. They just "have a feel for it" or "a knack". They have a talent that others do not, they will say, and so it isn't really gambling.

None of them are correct.

[galen ubal]

xkcd's take on the matter.

[Dr. Strangelove]

Originally Posted by Left Hand of Dorkness

If all of them take an umbrella to work tomorrow, that doesn't make me a pretty bad predictor. It just means that the one-in-three event happened.

That's pretty much what Chronos was saying.

It's important to note the distinction between dependent and independent events. If we predicted that 100 people each have a 1/3 chance of bringing an umbrella, and that these were independent, and then all 100 brought an umbrella, then we really did have a bad predictor.

But umbrellas are not independent events. Most likely, the 1/3 chance we're predicting is really the chance of rain, and if it rains everyone brings an umbrella; otherwise not. There's no problem with the 1/3 prediction here unless we got it wrong over a long period of time--in either direction.

At any rate, your final point is correct. It's dumb to ding Silver for only giving Trump a 30% (IIRC) chance of victory. In fact he deserves a lot of credit for correctly handling the correlation between states, which gave Trump much better odds than if one assumed no or little correlation.

[Left Hand of Dorkness]

Originally Posted by Bricker

I've been wrong. But not too often.

I'd argue that I'm right often enough to claim a better-than-random ability to accurately predict the outcome of events on which I take a position.

So have a cookie. What do you want, for us to all take your predictive powers seriously?

If I want fortune-telling, I probably still have a Tarot deck lying around somewhere. Analysis is what's interesting. Not analysis of your predictive powers, analysis of the current political scene.

And yes, there's plenty of bad analysis on this board from the left, just like there's plenty of bad analysis on this board from the right. The worst analysis is still more interesting than bragging about bets you've placed.

[asahi]

Originally Posted by Lance Turbo

Over the last six hours or so this race has swung from about 58/42 Ossoff to even. Am I missing a body slam or something? Swings like this usually happen for a reason.

New polling data perhaps. If it's within the margin of error, which I think it is, I think you'd have to give the edge to Handel. As I recall she ran against something like five other conservative or right-leaning challengers while Ossoff rocked the Bernie / Hillary vote. What the recent polling data probably means is that a lot of supporters of those other right-leaning candidates probably needed some time to get up, dust themselves off, and think about how much they really wanted to support Handel. I have some relatives and friends on FB who live in Metro Atlanta, and just reading my feed, I've noticed a slight uptick in conservative interest in this race. I do realize this is hardly scientific but if that's any indication, I think there's mounting conservative urgency to keep the seat, if only for symbolic reasons. And for that reason, I'd have to agree with Bricker that she'll probably keep the seat for the GOP, but just barely.

Even so, my position is that this seat is far more critical for Republicans than Democrats. In fact, regardless of the outcome, the Democrats have already shown that the GOP is increasingly vulnerable. Losing in the primary was probably bad optics but Handel was still the odds-on favorite to win the run-off. However, she has really struggled to gain the advantage that was anticipated, and if the Republicans lose tonight, then it would be a major shock. I agree that the Republicans could still easily recover from this, but at the same time, it's also the sort of event that could trigger soul searching and infighting.

Having said that, the Democrats still have a very long way to go before being able to compete with Republicans where it really matters, which is at the district and state-level. I'm seeing some indications that they're 'getting it' but there's still not a real platform or message that they're selling. Bernie Sanders, as much as I've criticized him in the past, is the one progressive who actually has a message. I think Perez needs to work with Sanders to figure out how to adopt and embrace some parts of his broad message while being open to working things out with moderates on the right. Not an easy line to walk on I'm afraid.

[yellowjacketcoder]

Well, I voted today, since I'm in that district.

There was a line to vote (only 5 people and it went in about 30 seconds) which is unheard of for special elections and REALLY unusual for runoffs. Probably has something to do with the fact that I've had canvassers come to my house 4 times in the last week. I feel bad for those folks in Ohio during presidential elections now.

[Shodan]

Originally Posted by Left Hand of Dorkness

Oh, spare us. This is about your enjoying gambling, no more, no less. If you're correct, you made a bet that panned out; it doesn't indicate any sort of ability to accurately predict the outcome of real world events, any more than if I predicted a die would roll above a 3 and it rolled a 5 I'd show an ability to predict dice rolls accurately.

You need to read up on probabilities.

A pattern of predictions is different from random chance, if the predictions turn out to be true at a rate higher than random chance.

Of course people predict things the way that they hope they'll turn out, often. You do too. That's not interesting. What's interesting is analysis.

The analysis of people who turn out to be wrong is less interesting, and less useful, than the analysis of people who turn out to be right.

The SDMB consistently over-estimates the likelihood of any Democrat winning any election. Their predictions, therefore, are less interesting than Bricker's, because they are more likely to be wrong.

'Just because you were right doesn't prove anything.' Yes, it does. It proves that your analysis was more likely to be useful than an analysis that says Trump/Bush/Walker/Handel/the Republican is going to lose.

I understand the desire for analysis based on wishful thinking. How did that kind of analysis work out in 2016, and what makes you think it will work out differently in 2018?

Regards,
Shodan

[Left Hand of Dorkness]

Originally Posted by Shodan

You need to read up on probabilities.

Super interesting, dude--thanks!

Meanwhile, a Trump PAC is running for Handel by using a quote from Obama out of context, in a repulsive manner.

[Fotheringay-Phipps]

ISTM that there are two separate issues being discussed.

The fact that one guy is willing to put his money where his mouth is while others are not might mean that the one guy is genuinely more confident that his predictions are reality-based, while the others may be willing to talk smack but sense at some level that they're biased. Or it might just mean that the one guy is less risk averse than the others. I do think the broader prediction markets have value in that in aggregate they will tend to filter out the partisan wishful thinking. But it's hard to say anything definitive about any small group of people.

The second issue is whether the guy betting is in fact a better predictor than the others. This has a bit of overlap with the above "confidence" issue but is not very heavily correlated. And it would depend on the guy's record over a longer term, rather than the mere willingness to bet.

I would suggest a sort of game where all the participants can bet notional amounts of money might be a good way to settle the matter. There's no real money at stake, so the risk-averse people can play, and you could track results against each other over the long term.

[ElvisL1ves]

Originally Posted by yellowjacketcoder

There was a line to vote (only 5 people and it went in about 30 seconds) which is unheard of for special elections and REALLY unusual for runoffs.

Do you have a robust early- and absentee-voting system? How has that turnout been?

[Lance Turbo]

Karen Handel voted today (presumably for herself). Jon Ossoff did not because he doesn't live in the district.

Some people have tried to make this fact a campaign issue but I think it's overblown. Ossoff grew up in the district but lives just outside of it now. Handel grew up in Maryland, but has lived in the district for something like 15 years.

All that said there is a small twisted part of me that hopes that Handel wins by exactly one vote as a result of this quirk.

[Little_Pig]

This political ad brought to you by the Principled PAC.

[E-DUB]

Left Hand of Dorkness wrote: "Meanwhile, a Trump PAC is running for Handel by using a quote from Obama out of context, in a repulsive manner."

What else is new?

And Sanders, on whose behalf the baseball shooter was a volunteer, isn't even a Democrat.

[HurricaneDitka]

Originally Posted by Little_Pig

This political ad brought to you by the Principled PAC.

'Don't help put Pelosi in power' and attacks to that effect, are pretty potent for conservatives. I'm curious if there's ever been a Republican Speaker of the House (or minority leader) that was similarly polarizing? Has anyone from the GOP been such a focal point for attacks, to the extent that he regularly appeared in attack ads for other congressional candidates' opponents?

[Left Hand of Dorkness]

Originally Posted by Fotheringay-Phipps

ISTM that there are two separate issues being discussed.

The fact that one guy is willing to put his money where his mouth is while others are not might mean that the one guy is genuinely more confident that his predictions are reality-based, while the others may be willing to talk smack but sense at some level that they're biased. Or it might just mean that the one guy is less risk averse than the others.

Or it might mean that he enjoys gambling. I enjoy roleplaying games, but if I statted up Ossoff as a fifth-level cleric, that wouldn't make my analysis more accurate (although it might be more interesting, I admit).

I would suggest a sort of game where all the participants can bet notional amounts of money might be a good way to settle the matter. There's no real money at stake, so the risk-averse people can play, and you could track results against each other over the long term.

Fun game; I have no interest in it, but if others do, we are fortunate enough to have a Thread Games forum where it'd fit right in.

[HMS Irruncible]

Originally Posted by asahi

Even so, my position is that this seat is far more critical for Republicans than Democrats. In fact, regardless of the outcome, the Democrats have already shown that the GOP is increasingly vulnerable.

Can we just retire this line of thinking already? Losing a scattering of national elections by a slimmer-than-expected margin doesn't mean jack. It means that the strategy of hoping for popular backlash hasn't yielded any useful results in a useful timeframe. It will be a long slow slog back out of the wilderness. Sorry for the rant, I'm just tired of seeing progressive energy expressed in ways that seem wasteful and ineffective to me.

I voted for Ossoff and I hope he wins, but it's frustrating that this inexperienced nobody has become a standard-bearer for "The Resistance". Who's the next random dude we throw our weight behind when he gets defeated?

[HurricaneDitka]

Originally Posted by HMS Irruncible

Can we just retire this line of thinking already? Losing a scattering of national elections by a slimmer-than-expected margin doesn't mean jack. It means that the strategy of hoping for popular backlash hasn't yielded any useful results in a useful timeframe. It will be a long slow slog back out of the wilderness. Sorry for the rant, I'm just tired of seeing progressive energy expressed in ways that seem wasteful and ineffective to me.

I voted for Ossoff and I hope he wins, but it's frustrating that this inexperienced nobody has become a standard-bearer for "The Resistance". Who's the next random dude we throw our weight behind when he gets defeated?

You seem very pessimistic about Ossoff's chances today. I'd say he's got a decent chance to win. Why the negativity? Or is that just my perception?

[Left Hand of Dorkness]

Originally Posted by HurricaneDitka

You seem very pessimistic about Ossoff's chances today. I'd say he's got a decent chance to win. Why the negativity? Or is that just my perception?

From my perspective, it seems a complete tossup to me. I'll be thrilled if he wins, but I worry about putting too much symbolic weight on it, for a couple of reasons. First, as others have said, Democrats need to strengthen local and state positions as a priority. Second, putting this much weight on a tossup/underdog race is really risky.

That said, one takeaway has already happened, both here and in Montana: districts that should be comfortably Republican are not nearly so comfortable. I don't know that we've had a lot of special elections so far in districts that were previously tossup districts; have we? Those elections are the ones that will, in my mind, be much more significant if Republicans hold on to.

Montana and Georgia were "dog bites man" stories that looked like they might turn into "man bites dog" stories. Montana turned out to be a "dog bites man" story after all. If Georgia goes the same way, that's not super-significant; "dog bites man" is the narrative we should expect.