What is the standard of proof?
Is it up to the same standard as the ranges given for a full charge on Tesla's?
Elon Musk said his bid to acquire and privatize Twitter "cannot move forward" until the social network proves its claim that fake bot accounts make up less than five per cent of all users. The world's richest meme lord formally launched efforts to take over Twitter last month after buying a 9.2 per cent stake in the biz. He …
"Is it up to the same standard as the ranges given for a full charge on Tesla's?"
Or when Solar City (chaired by Elon) stated on a filing that they had money to fund operations for the next year when they were nearly in breach of funding covenants with one of their creditors and knew they'd have significant shortfalls for the foreseeable future.
I read somewhere that twitter based their analysis on a sample of 100 accounts.
Where they carefully selected accounts or some sort of random sample?
I’d expect their sample to be 100k or 1m not 100.
Would be hilarious if he pulled out and didn’t have to pay that $1b, he’d normally be out of pocket after selling those Tesla shares but given how things are going he’ll likely be quids in.
A bigger sample changes the accuracy of the result, not the result itself as such. If the accuracy inherent in 5 in a sample size of 100 (which I think is sqrt(5), about 2.2%) is adequate for regulatory or commercial purposes, there is no need for a larger sample.
Of course, had they sampled 100 cases and came up with a big number (eg 95%), the answer would be both inaccurate and alarming.
Musk is making himself look naive...
That's a measure of precision, not the accuracy*. Also, the standard deviation of a Poisson distribution is skewed, not noticeably for a large number but very significantly for a small count so although the figure can't be much less (you can't have a negative count) it can be larger.
This sentence caught my attention: "Still, he claimed the internal estimates for the past four quarters have reported spam users make up less than five per cent of all accounts.". If they have estimates for the past four quarters it suggests more than a one-off sample was taken. Does it mean one sampling per quarter? One per month? One per day? And how do they decide what's a bot? And does it matter is a spammer is a bot or a human?
There's a reason why scientific reports explain how figures are obtained. You'd think that with the amounts of money at stake financial reports would be held to the same standard.
* The difference between accuracy and precision: Imagine you have a 100m surveyor's tape measure. It's marked off in metres, centimetres and millimetre. The last allows you to measure to the nearest mm. That's the precision. But the tape is a cloth tape and it's stretched by 1% so your measurement of 100m to the nearest mm is in fact in error by 1m. That's a failure of accuracy.
a corollary to what you say about precision and accuracy - increasing the number of samples tends to 'average out' accuracy problems and CAN result in better precision, kinda like how "oversampling" works.
So rather than 100 random samples, how about 10,000 random users gathered from various locations, asked to participate in surveys in some cases (but not all), and things like that?
Bots and spammers clearly lower the value of Tw*tter. I have seen the results of questionable filiings prior to acquisitions before. One company got shafted with $millions in excess obsolete inventory, valued at purchase price (not what it was worth) but then they TURNED AROUND AND BURNED HOSTILE A TAKEOVER with the SAME THING. yes it WAS the early 90's... and the ivory tower most definitely got their "golden parachutes" out of the deal,
My guess: Elon is aware of such tactics
But the tape is a cloth tape and it's stretched by 1% so your measurement of 100m to the nearest mm is in fact in error by 1m. That's a failure of accuracy.
...If you don't know about the degree of stretch! I have heard of very old measurement practises that involved a table of humidity vs paper shrinkage, to correct measurements according to the prevailing weather. That was surprisingly accuracte, apparently!
I think we're all agreed that the accuracy of the measurement depends on how many times it's been taken. That might be apparent if it's been included in multiple filings down the years; if it's slightly different every time, but within spitting distance of 5% (and NOT picking the same 100 accounts every time, of course), then it would be reasonable to conclude that the 5% figure is reasonably sound.
Interestingly, the Ordnance Survey here in the UK made its mapping surveyors repeat each measurement several times, starting from scratch. Apparently, they could tell when someone (up a hill, in winter, in the pouring rain, in the midst of a howling gale, freezing their brass monkeys off, with a heavy yet delicate theodolite, a broken pencil, some soggy paper and their lunch has been eaten by the local sheep) was not doing it properly, or was making up the variations, presumably from the distribution of the data. Humans aren't good at random.
There's some really useful info available about tax fraud which gives insight into how humans mess up with made up numbers and make their 'fudging the books' really obvious, I suspect it would apply to the OS surveyors, and indeed most other data.
Of course, it could also be used to avoid detection...
Musk claims that Twitter uses a sample size of a hundred accounts. Twitter claims that it's quite a bit more complex than that. It's hard to assess as an outsider as Twitter has the benefit of having lots of backend data to use in its analysis.
For instance, without access to backend data it's hard to know whether an account is active or not. There will be quite a few accounts of people who log in daily and follow topics and hashtags, DM with other users, click on ads etc. (and are therefore considered 'monetizable' by Twitter) but that rarely post themselves. To outsiders those accounts look dead, Twitter knows they are not.
Similarly there might be accounts that look perfectly normal to us but where Twitter knows the account is part of a batch of accounts that are harmless but that they know all come from the same IP address at a university research lab and so might not be actual 'monetizable' users.
Also, there should be a lot more consideration for the fact that the set of 'non-monetizeable' accounts are not homogenous at all. For instance, El Reg has a Twitter account. It has over 100,000 followers so clearly many people believe the account adds value to their experience on Twitter. That means that accounts like that add value to Twitter as a company too. Yet, it makes no sense to consider these types of accounts as 'monetizable' as the people/teams operating such an account will not respond to ads. Twitter will not include them in their count of 'Monetizable Monthly Active Users' but will, at least internally, not lump them together with spam accounts, political bots etc.
All in all this means that Twitter has the ability to do much more accurate analysis than any outsider ever has. Whether Twitter Inc. are truthful in their statements or not is a separate matter but analysis by outsiders, which can be very interesting and insightful, will likely always suffer from a greater inaccuracy than Twitter's internal data.
Also, whatever the accuracy of what is reported to the SEC doesn't actually matter, so long as it is consistent over a reasonable period of time. What actually matters is profit / loss, revenue / costs. Consistent reporting of fake accounts allows someone to make a comparisons against previous financials, and perhaps then make a reasonable stab at predicting future financials when the fake-account count changes.
It may be that there is no correlation between the fake account count and the financials, in which case it matters not one jot. It may be that there's a good correlation, in which case it's a useful tool as-is. But it could also be a self-fulfilling proficy. If the fake-account estimate is believed, and it goes down, Twitter's advertising revenue might go up due to enthused ad customers. And vice versa. Separating out cause and effect could be a tricky problem, as the only way of knowing is if the ad customers' product sales figures alter in correlation with the amount of ad spend they have on Twitter, but it's unlikely that they can measure that correlation very well.
Changing the methodology, sample size, etc may improve the theoretical accuracy, but there's likely a stage at which further improvement becomes pointless and uninformative. If that stage is reached, and there is still no correlation between financial performance and the improved fake account count estimate, then there's no point reporting it in the first place.
Someone else can probably say whereabouts on that scale of possibilities the current data actually sits.
Indeed, the potential pay off of USD 1 billion if things don't go through is a lot less than the loss in value of his Tesla stock that is being used to guarantee loans.
The SEC should intervene because interventions like this during supposed due diligence quite clearly move the stock price by interested parties.
I wouldn't count on it. The SEC has quite a few tools at their disposal and uses them, some times quite effectively, too. Unfortunately, in the case of share price manipulation they usually have a weak case.
In order to get a conviction the SEC require intent, not just an appearance of manipulation that could be coincidental too. It means that the SEC would need to find a smoking gun in the form of an email from Musk to someone else stating something along the lines of "Let's buy Twitter shares, then pretend to buy Twitter to inflate its stock, and then sell them off again". Musk may be erratic but he is not enough of a fool to put those things in writing.
Also, if anything, so far Musk has lost out big time. He bought a considerable chunk of Twitter shares and then managed to drive the price down with his antics, at a loss to his own portfolio. Also, as he has already signed the purchase contract at a higher price, he has now committed himself to buying the company for more than it's currently valued. Other Twitter shareholders would sue if he breaches the purchase contract.
The Guardian: What are Elon Musk’s options in the Twitter takeover deal?
“Also, Twitter’s disclosure to the US Securities and Exchange Commission with regard to bots is sufficiently hedged that a buyer should reasonably know that it might be higher than 5%.”
Morgan Ricks, professor of law at Vanderbilt Law School, adds: “Nothing that has happened so far comes anywhere close to a material adverse effect. Musk’s comments about fake accounts are an irrelevant sideshow.”
I'm not sure collapsing the share price is of any help to Musk as he has already signed the purchase contract at the higher price. Twitter can sue Musk if he doesn't pay up.
I wouldn't at all be surprised if Musk ended up with industrial quantities of egg on his face, without owning Twitter but with a damaged reputation for doing business. Also, any future company or its shareholders that are approached by Musk for a takeover will require additional safeguards and collateral to protect them from his erratic behaviour.
It also impacts Musk's personal wealth and costs him a couple of billion. Not just the $1 billion dollar fine if the breaks contract but he also owns a considerable chunk of Twitter shares that have lost value because of the shambles. Oh, and Tesla shares are down too.
In the early days, Twitter was touted as a human readable event feed and for some of the accounts I follow it's just that - automated news/weather/traffic that doesn't have a human directly typing tweets but are of value. So at what point is it spam? If my local hobby shop automatically posted new products and I followed that feed, is that spam? It's spam to some, not me. If UK members of parliament repeat the same text automatically to boost signal, is that spam? It's a human pressing the button but the text is copied is just like a bot.
I don't really care about what rich people do with their money or how they choose to behave but I am concerned that a tool I find handy might be cut to shreds so that I have to find another way!
I've moved from Twitter to Mastodon (the decentralized alternative to Twitter) about two months ago and love it.
It's amazing how some technical differences (the deliberate lack of an equivalent to 'Quote Tweets' for instance, or the absence of algorithms that push for clashes, pile-ons and radicalization to drive engagement metrics) make for a much more fruitful experience. You can have polite debate and disagreements without instant shouting matches or accusations of being Nazi or woke. I have rediscovered how nice it can be to exchange ideas with people you don't know personally.
Give it a shot, though make sure to spend the first week following a bunch of people (even if you unfollow them again a week later, at least they will have helped you discover more interesting accounts) because there is no algorithm pushing you to follow the usual suspects so it might appear empty at first. If there are certain Twitter accounts you would miss, such as your local hobby shop, then there are ways to have their tweets fed to your Mastodon timeline.
> So at what point is it spam?
Strange question. It is spam if you didn't request it. If you are indeed interested in receiving offers for counterfeit meds and shady get-rich-quick schemes, it's not spam for you, it's valuable information. So, if you mass-mail subscribers it's a mailing list, if you mass-mail unsuspected random victims it's spam. Easy.
I agree that it works for email but how does that work for Twitter? If you are seeing posts retweeted by people you follow or appear in your suggestion feed because the Twitter algorithm has decided you would be interested, how do you judge the originator is a spam bot?
> If you are seeing posts retweeted by people you follow or appear in your suggestion feed because the Twitter algorithm has decided you would be interested
I'm no Twitter user, but I think you answered your own question: "People you follow" are your choice (no matter what they post), but "Twitter algorithm suggestions" are by definitions spam: Unrequested and pointless nose, probably trying to sell you something.
“…Musk objected to: he felt that figure should be a lot higher.”
So if Musk originally made his offer in the belief that the figure was “a lot higher” than 5% then what is he complaining about? Even if the 5% figure is wrong and should be higher then he’s still getting what he expected
This makes no sense at all and sounds like a straw argument
Why aren't the American financial securities not all over this guy again for shitting around with share prices on twatter again, only this time re a big mouth, not so cunning stunt purchase of the actual twatter itself?
Can't someone please just ban his account!
I for one hope the alternative platform(s) get some serious traction and exposure and shove his new baby the way of the NFT.
I mean, it's just a public version of texts on t'web!
With his behaviour driving the share price down by about $10 a share, that means he's going to end up paying about $8 billion over the odds if he completes the deal at the offer price. And there I was thinking that NFTs were an expensive piece of shit.
This is surely heading for an eye-wateringly expensive court case,when he tries to weasel out of the $1 billion breakage fee, despite not bothering with due diligence.
No matter whether he completes the deal or not, he is headed for a very expensive court case with the SEC. A previous case about stock manipulation via Twitter left him legally required to refrain from such behaviour, the last few weeks have broken that agreement big time. The SEC haven't said anything yet but government wheels turn slowly, they will eventually flatten him like a steamroller.
"A previous case about stock manipulation via Twitter left him legally required to refrain from such behaviour, "
That's specifically related to Tesla since Elon's Twitter account is designated as an official source of information regarding Tesla. I doubt that his commentary outside of that is subject to any restrictions (or restraint).
Let's say that Elon doesn't spend all of his money and leverage all of his assets buying Twitter and just pays the attorneys to weasel out of the violation of the NDA and the breakage costs. The next business he tries to "disrupt" is far less likely to take him seriously and any smart investor is not going to put themselves out buying stock to reap a nice windfall. It has to be kept in mind that Twitter wasn't planning on selling or taking the company private. This would be a hostile takeover plain and simple. I expect that Twitter would have been much happier if Elon had just continued stirring up controversy and driving people to Twitter so they can be sure to absorb the blather first hand as soon as Elon posted it.
Is there any point in buying Twitter and taking is private? Tesla investors seem to think is too much of a distraction at a time when Tesla is rapidly dialing down to only 2 car models that are very similar and very expensive while having saddled themselves with a proprietary charging network that they can't monetize without modifying it to work with other EV's and payment schemes. At the same time, SpaceX is betting the company on Starship and Starlink both of which may turn out to be complete boondoggles as some analysts think they will (me included, though I'm not an analyst). With international tensions and Russian pulling out of ISS, the contract to supply cargo and transfer astronauts may come to an end. Between ISS and Starlink, that's the huge percentage of Falcon 9 launches. Wouldn't $20bn go a long way towards getting the Roadster, Cybertruck and Semi into production?