Ant-Man and the Wasp: Quantumania | February 17, 2023 | Competing with Eternals on RT, Competing with BvS on box office legs

[Factcheck]

48 minutes ago, Factcheck said:

I think this site doesn't count those who don't watch horror movies when it comes to tracking general awareness of horror movies. So awareness and interest are limited to those who watch horror movies. For Ant-Man, awareness and interest are of those who watch action, adventure, sci-fi, and superhero movies, reaching a much broader audience. This explanation could be given.

Haha. I think I've solved this quorum puzzle. This is the reason why John Wick 4 has much more awareness and interest than Ant-Man 3 but will do less than it.

[Menor the Destroyer]

3 minutes ago, Factcheck said:

Haha. I think I've solved this quorum puzzle. This is the reason why John Wick 4 has much more awareness and interest than Ant-Man 3 but will do less than it.

That is not the methodology they've ever described. Their awareness and interest numbers purport to be out of their whole sample.

[Factcheck]

8 minutes ago, Menor Reborn said:

That is not the methodology they've ever described. Their awareness and interest numbers purport to be out of their whole sample.

If I don't watch horror movies at all, then does it matter whether I'm aware of Scream 6 or not?

Edited January 27, 2023 by Factcheck

[Cooper Legion]

5 minutes ago, Menor Reborn said:

That is not the methodology they've ever described. Their awareness and interest numbers purport to be out of their whole sample.

Unclear who one would calculate an interest out of those who were a no on awareness.  I mean I guess you I can immediately provide a description or play a trailer during the polling, but the latter takes a while and the former introducing an unpleasant amount of variance to the interest score via exactly how you craft the description provided. 
 

Anyway, quorum is just VERY imprecise. If it’s your only data available to predict off your ranges are going to be like “good chance 70-140 (no guarantee though).”   
 

That’s fine, sometimes that’s better than nothing, but the idea that it has a ways to go for 76 to be safe is detached from the reality of the situation, just reading more into one specific (noisy) indicator than can actually be read.

 

[Menor the Destroyer]

7 minutes ago, Favorite Fearless Legion said:

Unclear who one would calculate an interest out of those who were a no on awareness.  I mean I guess you I can immediately provide a description or play a trailer during the polling, but the latter takes a while and the former introducing an unpleasant amount of variance to the interest score via exactly how you craft the description provided. 

 

I was imprecise in the wording. Yes, they don't track interest for those who aren't aware. But what I meant was that I don't see anywhere that they exclude responses based on level of interest in genre. Something like that would theoretically show up in the interest score anyway.

[Factcheck]

10 minutes ago, Menor Reborn said:

I was imprecise in the wording. Yes, they don't track interest for those who aren't aware. But what I meant was that I don't see anywhere that they exclude responses based on level of interest in genre. Something like that would theoretically show up in the interest score anyway.

I can explain this. Let's say Scream 6 has 60% awareness; now 60% awareness doesn't mean that these 60% people watch horror movies. Let's say that out of this 60%, 50% don't watch horror movies at all, so actual or meaningful awareness will be 30%.

[Verrows]

3 hours ago, M37 said:

It’s a common problem in tracking and polling to have difficulty reaching some groups of people. If the actual demo breakdown is close, then it won’t matter much, but when a film’s audience skews one way or the other, it will throw off the relative ratios; that’s why Thor is higher in Quorum numbers than films with bigger OWs. 

 

A whole lot of tracking underestimations  can be traced to higher audience share among non-white or otherwise less reliably tracked audience (and vice versa), like TGM’s strength in non-coastal regions. Basically, we’re only seeing half of the picture and extrapolating with the presumption the other half will be “normal”, when it quite often isn’t

I don't get it. Why would people who don't have white skin provide less reliable tracking???

[Porthos]

20 minutes ago, Verrows said:

I don't get it. Why would people who don't have white skin provide less reliable tracking???

 

Lower social trust = harder to poll. 

 

Sometimes.

 

A variation of this was seen in the 2020 US Elections polling where it is thought one reason why some polls were off is that some conservatives/folks leaning right wing being reached via poll would refuse to answer the phone/poll because of Big Evil Media Being Out To Get Them.

 

(in other words, culture not genetics)

 

More generally, if a group is harder to reach via various methods (including how and when they're contacted), it stands to reason that might show up in polling unless one is very careful about methodology.  Or even if one is very careful about methodology as only so much one can do with chaotic samples.

 

Edited January 27, 2023 by Porthos
decided to generalize one of my points

[Verrows]

13 minutes ago, Porthos said:

 

Lower social trust = harder to poll. 

 

Sometimes.

 

A variation of this was seen in the 2020 US Elections polling where it is thought one reason why some polls were off is that some conservatives/folks leaning right wing being reached via poll would refuse to answer the phone/poll because of Big Evil Media Being Out To Get Them.

 

(in other words, culture not genetics)

 

Plus just having polling geared to middle/upper class respondents (time when they're contacted and how they're contacted).  If a group is harder to reach via various methods, it stands to reason that might show up in polling unless one is very careful about methodology.  Or even if one is very careful about methodology as only so much one can do with chaotic samples.

 

I guess that makes sense. I just wonder what about movie polling would be an issue specifically for non white people. I mean, we (as in our multicultural, diverse population) are pretty streamlined nowadays in terms of pop culture trends and accessibility to these things.

 

If anything I figured the discrepancy for something like this would be more socioeconomic.

[Porthos]

22 minutes ago, Verrows said:

I guess that makes sense. I just wonder what about movie polling would be an issue specifically for non white people. I mean, we (as in our multicultural, diverse population) are pretty streamlined nowadays in terms of pop culture trends and accessibility to these things.

 

If anything I figured the discrepancy for something like this would be more socioeconomic.

 

We're staring to get into the weeds of polling issues here, but yes socioeconomic factors are absolutely one consideration (as part of how/when to reach someone).  That it is getting harder and harder to reach people via traditional polling methods and that the newer methods (like internet sampling) has a whole host of issues with it, another.

 

But it's not just "negative response bias" at play but positive response bias as well.  Remember when I said one reason polls might have been off in 2020 was because of conservatives perhaps not responding as much?  There's an equally compelling theory that some polls might have been off because liberals were more gung-ho to respond to polls and that skewed samples, even when controls were attempted.  And/or people just being stuck at home during COVID shutdowns having nothing better to do than answer the phone and vent when being polled (this might explain some particularly large outliers in Wisconsin, for instance, which was in the middle of a really bad 'rona surge when the final polls for the 2020 election were being done). 

 

On the flip side, in 2022 it could be that conservatives (juiced by all the talk of the Red Wave) were more willing to answer the phone and crow that they were finally gonna get Brandon stick it to Biden.

 

Ultimately if a group is more willing to tell people that "Yes, they really really look forward to something" for whatever reason, that can put a thumb on the scale in the opposite direction.

 

All of this is to basically say: Polls are tools and some tools are better than others in different situations.

[M37]

2 hours ago, Verrows said:

I don't get it. Why would people who don't have white skin provide less reliable tracking???

Yes, as @Porthos said it’s not about race per se, it’s more about reachability (which is why I also gave the TGM example), and that can include socioeconomic status. Teens are often a blind spot too (see Missing, Gentleminions). 

 

It’s not in any way a critique of any specific group, only an observation that sampling only gets you so far, and can miss - or overstate - depending on how the eventual audience breaks down 

[Verrows]

Thanks for clarifying @Porthos and @M37. Makes a lot of sense.

[Saul Goodman]

 

[fabiopazzo2]

1 hour ago, Saul Goodman said:

 

Not bad

[DAJK]

Anecdotal, but hype for this seems to be spiking on Tiktok at just the right moment. Normally I would say "sure, but social media hype doesn't equal real-world hype" but we all saw what happened with Minions.

[eddyxx]

On 1/27/2023 at 4:49 PM, Porthos said:

 

We're staring to get into the weeds of polling issues here, but yes socioeconomic factors are absolutely one consideration (as part of how/when to reach someone).  That it is getting harder and harder to reach people via traditional polling methods and that the newer methods (like internet sampling) has a whole host of issues with it, another.

 

But it's not just "negative response bias" at play but positive response bias as well.  Remember when I said one reason polls might have been off in 2020 was because of conservatives perhaps not responding as much?  There's an equally compelling theory that some polls might have been off because liberals were more gung-ho to respond to polls and that skewed samples, even when controls were attempted.  And/or people just being stuck at home during COVID shutdowns having nothing better to do than answer the phone and vent when being polled (this might explain some particularly large outliers in Wisconsin, for instance, which was in the middle of a really bad 'rona surge when the final polls for the 2020 election were being done). 

 

On the flip side, in 2022 it could be that conservatives (juiced by all the talk of the Red Wave) were more willing to answer the phone and crow that they were finally gonna get Brandon stick it to Biden.

 

Ultimately if a group is more willing to tell people that "Yes, they really really look forward to something" for whatever reason, that can put a thumb on the scale in the opposite direction.

 

All of this is to basically say: Polls are tools and some tools are better than others in different situations.

Why did no one tell me that Steve Kornacki had a BOT account?

[Cooper Legion]

5 hours ago, DAJK said:

Anecdotal, but hype for this seems to be spiking on Tiktok at just the right moment. Normally I would say "sure, but social media hype doesn't equal real-world hype" but we all saw what happened with Minions.

The tiktok whisperer

 

I am no expert (I am the exact opposite of an expert) but I would think that now is actually a bit early for an ideal spike? Or is it like, the beginnings of a spike which may take long enough to develop that it lasts 2.5 Wks?

Edited January 30, 2023 by Favorite Fearless Legion

[M37]

5 hours ago, eddyxx said:

Why did no one tell me that Steve Kornacki had a BOT account?

You joke, but there is a lot of overlap between the process of entertainment tracking and political polling, and both would be well served to learn from each other. The latter is much more refined, but only gets samples every 2 years to check their process, while the former has many more opportunities and data samples but also a much higher variance with a frankly weaker process 

 

The recent polling “misses” can be explained by understanding tracking misses for film, how a “hidden” audience - one that is undersampled and/or underweighted - shows up in larger than typical numbers and boots a film’s gross.  

 

Somebody who could bridge the gap between the two could really refine the precision (and prob make gobs of money) for forecasting both 🤔

[MovieMan89]

I think the possibility that this one will underperform needs to be highly weighed. This has always been the least successful sub-franchise of MCU and Phase 4 on the whole killed a ton of the momentum for MCU. This is not Ant Man 2 days where MCU was riding the ultimate popularity snd goodwill high. I would not be shocked with an OW somewhere between the first and second’s OWs. WOM needs to be incredible I think for this to be any kind of breakout. 

[Menor the Destroyer]

46 minutes ago, MovieMan89 said:

I would not be shocked with an OW somewhere between the first and second’s OWs.

Get ready to be shocked then.