US Election

Discussions about serious topics, for serious people
User avatar
jimbob
Dorkwood
Posts: 1018
Joined: Mon Nov 11, 2019 4:04 pm
Location: High Peak/Manchester

Re: US Election

Post by jimbob » Sun Jun 21, 2020 1:42 pm

discovolante wrote:
Sun Jun 21, 2020 1:40 pm
KPop fans are...the future?

https://www.vogue.co.uk/news/article/k- ... rump-rally
I must admit over the last few days, I have developed an affection for KPop fans.
Have you considered stupidity as an explanation

EACLucifer
Snowbonk
Posts: 456
Joined: Fri Dec 13, 2019 7:49 am
Location: Behind you

Re: US Election

Post by EACLucifer » Sun Jun 21, 2020 2:31 pm

discovolante wrote:
Sun Jun 21, 2020 1:40 pm
KPop fans are...the future?

https://www.vogue.co.uk/news/article/k- ... rump-rally
I hope not. Cultish devotion to manufactured idols isn't a healthy trait.

User avatar
Little waster
Snowbonk
Posts: 522
Joined: Tue Nov 12, 2019 12:35 am
Location: About 1 inch behind my eyes

Re: US Election

Post by Little waster » Sun Jun 21, 2020 3:11 pm

discovolante wrote:
Sun Jun 21, 2020 1:40 pm
KPop fans are...the future?

https://www.vogue.co.uk/news/article/k- ... rump-rally
I suppose it should have been obvious that we’d reach Peak 2020 on Midsummers Day.
Shamelessly recycling old jokes since 1952.

User avatar
Vertigowooyay
Clardic Fug
Posts: 156
Joined: Mon Nov 11, 2019 1:25 pm

Re: US Election

Post by Vertigowooyay » Sun Jun 21, 2020 3:17 pm

Oh my, this is not what you can describe as the return of a leader triumphant.

https://twitter.com/projectlincoln/stat ... 18625?s=21
Calm yourself Doctor NotTheNineO’ClockNews. We’re men of science. We fear no worldly terrors.

User avatar
lpm
Dorkwood
Posts: 1291
Joined: Mon Nov 11, 2019 1:05 pm
Location: IMPEACH AND EXTERMINATE

Re: US Election

Post by lpm » Sun Jun 21, 2020 4:11 pm

That's awful sad.

Watch where his feet land with each step. For most of it you can't actually see his feet, but you can tell where they land.

This event feels like a real turning point. Maga crowds losing interest.
I'll miss him after he's died in the pandemic

FlammableFlower
Snowbonk
Posts: 417
Joined: Mon Nov 11, 2019 1:22 pm

Re: US Election

Post by FlammableFlower » Sun Jun 21, 2020 4:16 pm

Vertigowooyay wrote:
Sun Jun 21, 2020 3:17 pm
Oh my, this is not what you can describe as the return of a leader triumphant.

https://twitter.com/projectlincoln/stat ... 18625?s=21
Looks more like the aftermath of a rough night on the tiles than a mega "sellout" rally...

User avatar
Bird on a Fire
Stummy Beige
Posts: 2605
Joined: Fri Oct 11, 2019 5:05 pm
Location: with the birds

Re: US Election

Post by Bird on a Fire » Sun Jun 21, 2020 4:27 pm

He's always alone, isn't he. No friends. Must be hard.
now I'm falling asleep and she's calling acab

User avatar
jimbob
Dorkwood
Posts: 1018
Joined: Mon Nov 11, 2019 4:04 pm
Location: High Peak/Manchester

Re: US Election

Post by jimbob » Sun Jun 21, 2020 4:48 pm

Bird on a Fire wrote:
Sun Jun 21, 2020 4:27 pm
He's always alone, isn't he. No friends. Must be hard.
As far as I can see, to know him is to despise him, so I'd guess that the most cutting memes are circulating amongst his own staff too.
Have you considered stupidity as an explanation

User avatar
jimbob
Dorkwood
Posts: 1018
Joined: Mon Nov 11, 2019 4:04 pm
Location: High Peak/Manchester

Re: US Election

Post by jimbob » Sun Jun 21, 2020 5:22 pm

I'm not sure if his ten minute complaint about WaterGait was too thrilling for his crowd.

https://twitter.com/janemarielynch/stat ... 16289?s=20

I've seen people look less bored when waiting for a delayed train.
Have you considered stupidity as an explanation

FlammableFlower
Snowbonk
Posts: 417
Joined: Mon Nov 11, 2019 1:22 pm

Re: US Election

Post by FlammableFlower » Sun Jun 21, 2020 5:57 pm


User avatar
lpm
Dorkwood
Posts: 1291
Joined: Mon Nov 11, 2019 1:05 pm
Location: IMPEACH AND EXTERMINATE

Re: US Election

Post by lpm » Sun Jun 21, 2020 7:13 pm

Little waster wrote:
Sun Jun 21, 2020 11:17 am

Ha the jokes on you libtards, masterful trolling by Trump by tricking the voting public into thinking he’s a feeble-bodied as he is feeble-minded. We are heading towards a landslide re-election clearly.

PICK IT UP! PICK IT UP!
Perfect.

Trump 2016: Rants about emails.

Crowd: LOCK HER UP! LOCK HER UP!

Trump 2020: Looks nervously at a glass of water.

Crowd: PICK IT UP! PICK IT UP!
I'll miss him after he's died in the pandemic

User avatar
Vertigowooyay
Clardic Fug
Posts: 156
Joined: Mon Nov 11, 2019 1:25 pm

Re: US Election

Post by Vertigowooyay » Sun Jun 21, 2020 8:47 pm

lpm wrote:
Sun Jun 21, 2020 4:11 pm
That's awful sad.

Watch where his feet land with each step. For most of it you can't actually see his feet, but you can tell where they land.

This event feels like a real turning point. Maga crowds losing interest.
I wonder how long it takes for a cult of personality to crumble once the cracks become this obvious? He rambled. His supporters were bored. He tried to get a cheer by drinking a glass of water. I think a lot of his base will have finally recognized a flailing, weak man.

If Mike Pence isn’t having discussions about taking the ticket, even at this late stage, I’ll be quite surprised. It’ll come to nothing but I bet he’s reached out in the administration.
Calm yourself Doctor NotTheNineO’ClockNews. We’re men of science. We fear no worldly terrors.

User avatar
Little waster
Snowbonk
Posts: 522
Joined: Tue Nov 12, 2019 12:35 am
Location: About 1 inch behind my eyes

Re: US Election

Post by Little waster » Sun Jun 21, 2020 9:33 pm

I'm not sure I can quite deal with such an unexpected turn of events.

I mean we were all reassured by this guy that Trump would be the healthiest president ever, and you can tell he knows what he is talking about by just looking at him.

Image
Shamelessly recycling old jokes since 1952.

EACLucifer
Snowbonk
Posts: 456
Joined: Fri Dec 13, 2019 7:49 am
Location: Behind you

Re: US Election

Post by EACLucifer » Mon Jun 22, 2020 6:15 am

It's easy to mock. Very easy, in fact. But the next rally is set for Arizona, where covid is surging.

This won't just hit those attending, and given people are travelling from out of state, could well harm people right across the USA.

Image

User avatar
Brightonian
Fuzzable
Posts: 247
Joined: Mon Nov 11, 2019 3:16 pm

Re: US Election

Post by Brightonian » Mon Jun 22, 2020 7:57 am

Interesting Twitter thread about the effect of K-pop fans etc. on the Trump rally data mining operation: https://twitter.com/aetherlev/status/12 ... 91587?s=19

User avatar
lpm
Dorkwood
Posts: 1291
Joined: Mon Nov 11, 2019 1:05 pm
Location: IMPEACH AND EXTERMINATE

Re: US Election

Post by lpm » Mon Jun 22, 2020 9:06 am

I don't understand that thread?

If you have a database of 1,000,000, of which 900,000 are trolls, you can still send 1,000,000 emails asking for cash. Your 100,000 target will still get them. The Trump begging requests aren't particularly sophisticated - they just constantly bombard with emails and texts asking for money.

Once you get a single donation, you're away, you get stuck into the data and try to maximise the revenue stream from the sucker.

You can send bots out into the database. If they fail to link up with a real human's Facebook account, they fail. If they succeed, you get a return and can start asking for cash. If you don't care about the percentage of hits - and I don't see why you would - you still get the benefit of the absolute number of real responses.
I'll miss him after he's died in the pandemic

User avatar
Little waster
Snowbonk
Posts: 522
Joined: Tue Nov 12, 2019 12:35 am
Location: About 1 inch behind my eyes

Re: US Election

Post by Little waster » Mon Jun 22, 2020 9:13 am

Parscale's defence is fairly underwhelming.

Basically Trump's inability to get a handle on the pandemic combined with the civic unrest fuelled by Trump's promotion of racist police brutality and support for neo-Nazis has meant people didn't feel safe enough to even leave their homes to celebrate what a thoroughly wonderful job Trump has done as the law'n'order MAGA POTUS.

You can see how he got his human psychology "genius" tag.
Shamelessly recycling old jokes since 1952.

User avatar
Gfamily
Dorkwood
Posts: 1353
Joined: Mon Nov 11, 2019 1:00 pm

Re: US Election

Post by Gfamily » Mon Jun 22, 2020 9:25 am

lpm wrote:
Mon Jun 22, 2020 9:06 am
I don't understand that thread?

If you have a database of 1,000,000, of which 900,000 are trolls, you can still send 1,000,000 emails asking for cash. Your 100,000 target will still get them. The Trump begging requests aren't particularly sophisticated - they just constantly bombard with emails and texts asking for money.

Once you get a single donation, you're away, you get stuck into the data and try to maximise the revenue stream from the sucker.

You can send bots out into the database. If they fail to link up with a real human's Facebook account, they fail. If they succeed, you get a return and can start asking for cash. If you don't care about the percentage of hits - and I don't see why you would - you still get the benefit of the absolute number of real responses.
I'm not a data analyst, but the value of the data isn't just the 100K identities that are true Trumpers. It's that from the details of the 100K they can get a profile of what they like, who they follow and what they respond to.

That level of data for the 100K who responded to Tulsa can then be used to identify others with similar profiles and they can be targetted with specific messages to reinforce and amplify their likely predispositions before they get directly asked for money. That widens the field from the 100K that responded, to the 30 or 40 million who are like them. In agricultural terms, the data is used as fertiliser to grow the crop before it is harvested.


It's the ability to target the messages to people like the 100K (while keeping the targetting hidden from the rest of the radical left us) is what has been wrecked (it is claimed) by the 900K false responses.
My avatar was a scientific result that was later found to be 'mistaken' - I rarely claim to be 100% correct

User avatar
bjn
Snowbonk
Posts: 517
Joined: Wed Sep 25, 2019 4:58 pm

Re: US Election

Post by bjn » Mon Jun 22, 2020 10:39 am

EACLucifer wrote:
Mon Jun 22, 2020 6:15 am
It's easy to mock. Very easy, in fact. But the next rally is set for Arizona, where covid is surging.

This won't just hit those attending, and given people are travelling from out of state, could well harm people right across the USA.

Image
Which is another reason why the pathetic turnout in Tulsa is a good thing. I’ve seen figures putting the number of “real” attendees below 2000, with the bulk of the people there being paid to attend for one reason or another.

Hopefully the pattern will repeat in Arizona, with far fewer people attending than was expected. Ideally, the things should just be cancelled, as Biden has done, but obviously that’s not going to happen.

User avatar
jimbob
Dorkwood
Posts: 1018
Joined: Mon Nov 11, 2019 4:04 pm
Location: High Peak/Manchester

Re: US Election

Post by jimbob » Mon Jun 22, 2020 11:10 am

lpm wrote:
Mon Jun 22, 2020 9:06 am
I don't understand that thread?

If you have a database of 1,000,000, of which 900,000 are trolls, you can still send 1,000,000 emails asking for cash. Your 100,000 target will still get them. The Trump begging requests aren't particularly sophisticated - they just constantly bombard with emails and texts asking for money.

Once you get a single donation, you're away, you get stuck into the data and try to maximise the revenue stream from the sucker.

You can send bots out into the database. If they fail to link up with a real human's Facebook account, they fail. If they succeed, you get a return and can start asking for cash. If you don't care about the percentage of hits - and I don't see why you would - you still get the benefit of the absolute number of real responses.
Isn't the main problem that they have about 6000 less journos etc genuine responses.

And these are almost certainly the keenest Trump fans, so they probably already have their information.
Have you considered stupidity as an explanation

tom p
Catbabel
Posts: 625
Joined: Mon Nov 11, 2019 1:14 pm
Location: the low countries

Re: US Election

Post by tom p » Mon Jun 22, 2020 12:18 pm

lpm wrote:
Mon Jun 22, 2020 9:06 am
I don't understand that thread?

If you have a database of 1,000,000, of which 900,000 are trolls, you can still send 1,000,000 emails asking for cash. Your 100,000 target will still get them. The Trump begging requests aren't particularly sophisticated - they just constantly bombard with emails and texts asking for money.

Once you get a single donation, you're away, you get stuck into the data and try to maximise the revenue stream from the sucker.

You can send bots out into the database. If they fail to link up with a real human's Facebook account, they fail. If they succeed, you get a return and can start asking for cash. If you don't care about the percentage of hits - and I don't see why you would - you still get the benefit of the absolute number of real responses.
She explains it in the reply to let them eat cupcakes. I, and they, initially clearly thought the same as you.
Turns out there's more to it than you might expect, including higher costs just for sending 10x the emails you otherwise would have.

bagpuss
Fuzzable
Posts: 270
Joined: Tue Nov 12, 2019 12:10 pm

Re: US Election

Post by bagpuss » Mon Jun 22, 2020 1:15 pm

tom p wrote:
Mon Jun 22, 2020 12:18 pm
lpm wrote:
Mon Jun 22, 2020 9:06 am
I don't understand that thread?

If you have a database of 1,000,000, of which 900,000 are trolls, you can still send 1,000,000 emails asking for cash. Your 100,000 target will still get them. The Trump begging requests aren't particularly sophisticated - they just constantly bombard with emails and texts asking for money.

Once you get a single donation, you're away, you get stuck into the data and try to maximise the revenue stream from the sucker.

You can send bots out into the database. If they fail to link up with a real human's Facebook account, they fail. If they succeed, you get a return and can start asking for cash. If you don't care about the percentage of hits - and I don't see why you would - you still get the benefit of the absolute number of real responses.
She explains it in the reply to let them eat cupcakes. I, and they, initially clearly thought the same as you.
Turns out there's more to it than you might expect, including higher costs just for sending 10x the emails you otherwise would have.
I'm not entirely sure I agree with the logic of her argument there. Much of the additional cost mentioned is the overhead of creative, test design, funnel management, data analysts' time, etc, etc. And 10x the number of email addresses doesn't equate to 10x the cost of those things. The data analysts are going to have a massive headache trying to clean the data, that's for sure (my week is suddenly looking a whole lot better knowing that I won't be among them), but the other things won't multiply up very much.

That said, there are definitely additional costs that aren't always obvious. For example, software for handling large scale email campaigns often has variable license costs banded based on the number of email addresses stored in the system. Adding another 900k email addresses that are probably all completely new to you as they're not already in your target base could very well take you up to the next band and cost you money.

There are subtler impacts too - such as sender reputation. This can affect whether your emails actually make it to the inboxes of your intended recipients, and one thing that can adversely affect your reputation score is the number of emails you send out. The number of bounces you get is another - so if those K-Pop fan added emails are not real, their bounce rate is going to go sky high. Yet another is how many of your recipients mark your emails as spam. So I hope that all those K-Pop fans who used genuine email addresses will mark every email they get from the Trump campaign as spam, which will than affect the likelihood of any true Trump supporter actually receiving his emails.

ETA: the bit about bounces

User avatar
dyqik
After Pie
Posts: 1932
Joined: Wed Sep 25, 2019 4:19 pm
Location: Masshole
Contact:

Re: US Election

Post by dyqik » Mon Jun 22, 2020 2:19 pm

I think the major thing is about modelling your addressees for targeting and tuning ads. If all their data is now contaminated with very large numbers of K-pop fans and Gen-Zers who aren't reachable, then what the data says their sample are interested in, and concerned about, may be counterproductive for engaging their actual target audience of uncertain but persuadable potential Trump voters, and Trump supporters who are unreliable voters. Note that these are two separate groups, who would be advertised at separately and differently, and Trump needs both.

I'm not convinced that a large subsample of K-pop fans is that hard to deal with in the data, mind.

User avatar
Woodchopper
Dorkwood
Posts: 1418
Joined: Sat Oct 12, 2019 9:05 am

Re: US Election

Post by Woodchopper » Mon Jun 22, 2020 2:24 pm

dyqik wrote:
Mon Jun 22, 2020 2:19 pm
I think the major thing is about modelling your addressees for targeting and tuning ads. If all their data is now contaminated with very large numbers of K-pop fans and Gen-Zers who aren't reachable, then what the data says their sample are interested in, and concerned about, may be counterproductive for engaging their actual target audience of uncertain but persuadable potential Trump voters, and Trump supporters who are unreliable voters. Note that these are two separate groups, who would be advertised at separately and differently, and Trump needs both.

I'm not convinced that a large subsample of K-pop fans is that hard to deal with in the data, mind.
I agree. It would though be amusing if Midwest pensioners were targeted with K-pop themed ads.

bagpuss
Fuzzable
Posts: 270
Joined: Tue Nov 12, 2019 12:10 pm

Re: US Election

Post by bagpuss » Mon Jun 22, 2020 2:33 pm

dyqik wrote:
Mon Jun 22, 2020 2:19 pm
I think the major thing is about modelling your addressees for targeting and tuning ads. If all their data is now contaminated with very large numbers K-pop fans, then what the data says their sample are interested in, and concerned about, may be counterproductive for engaging their actual target audience of uncertain but persuadable potential Trump voters, and Trump supporters who are unreliable voters. Note that these are two separate groups, who would be advertised at separately and differently, and Trump needs both.

I'm not convinced that a large subsample of K-pop fans is that hard to deal with in the data, mind.
While that's definitely the bigger issue on the face of it, I'm not sure it's all that much of a problem when it comes down to it. For this issue, there are 3 relevant scenarios for each of those K-pop email addresses:
1. email address is either fake or no profiling information can be obtained for other reasons. Result: no impact on profiling
2. lots of profiling information can be obtained. Result: given the big difference between K-pop fans and Trump's normal supporter base, very probably easy to identify as K-pop fan and remove from the data and thus no impact on profiling
3. some profiling information can be obtained but not enough to identify confidently as K-pop fan. Result: probably not enough information to be able to do useful profiling with either so little to no impact on profiling

Really, the only impact for profiling, therefore, is the additional time and cost of dealing with the additional data and weeding it out. Which is more than zero and thus definitely has an impact, but I don't think it will have much effect on the final profiling work.

If there was less of a distinctive difference between K-pop fans and the typical Trump supporter base, mind you, it would be a much harder job. A similar number of people who are, say, middle-aged white Americans but who are Democrat supporters or even non-Trumpian Republicans, would be a much bigger problem to weed out of the data.


ETA: I've realised that this may all be nonsense when it comes to stuff like Facebook ad targeting, etc, as I've never dealt with that and I don't know enough about how it works. If the marketer has no visibility of the targeting mechanism and the profile of the users but it's all done automatically based on the email addresses and the information the 3rd party (eg Facebook) has access to then of course there's nothing the marketer can do to weed them out and everything I've written above is irrelevant.

Post Reply