196 comments

[ 3.1 ms ] story [ 244 ms ] thread
This is an incredible story. 349! Everybody wins!
For everybody who was saying "Wikipedia is complete, there's nothing left to write about" yesterday, as far as I can there is not a word about this on the whole site, certainly not on the Pepsi page. Time to get to work!

https://en.wikipedia.org/w/index.php?search=Pepsi+349&ns0=1&...

Edit: I stand corrected, there is a single paragraph in here. Nothing standalone/linkable though.

https://en.wikipedia.org/wiki/1992_in_the_Philippines

Is this something that should be on the Wikipedia Pepsi page?

While it is an interesting story I don't really think it is the kind of thing I'd be looking for.

There's acres on eg. Crystal Pepsi and New Coke, neither of which (AFAIK) caused any people to die.
Really? You don't think a corporations blunder that took lives and led to riots is deserving of inclusion?

I can say Pepsi certainly doesn't want it there, your motivations are interely baffling though.

I think it was a technical blunder and I don't buy into the idea that they're responsible for the actions of individuals in a riot.

If someone upset that they (or more likely someone else) should have won a prize takes actions that lead to the death of another ... I don't believe Pepsi is responsible for that.

It still seems like a very interesting part of history whether they were at fault or not. They have made dumb mistakes on multiple occasions if you remember the Harrier Jet give away they blundered and someone came up with the points. So I believe this is worthy of mention on the Pepsi Wikipedia page it’s part of the business success and hardships. Maybe there will be more. Maybe not.
So then let people come to that conclusion on their own after they've read the facts. The Wikipedia content shouldn't imply fault. It just needs to document what actually happened.
Interesting, a company's history is about the only reason I'd visit their Wikipedia page.
What about on the Filipino Wikipedia? It's better to translate rather than start from scratch. I think it would be an excellent new English article with a mention on the Pepsi article, if there are enough sources to claim that it's significant enough.
Controversial stuff, even well documented controversial stuff, gets scrubbed from Wikipedia all the time.

Check out https://en.wikipedia.org/wiki/RAND_Corporation

You would suppose a think tank directly responsible for such policy gems as :

• Mutually assured destruction

• The Missile Gap

• Planned Shrinkage

• tax cuts for the rich and service cuts for the poor

• Systematic Torture of enemy soldiers in the Vietnam war

• Deployment of Agent Orange

• Planned military coups of a number of third world nations

• Their employee (Daniel Ellsburg) leaked the pentagon papers

And all manner more of cold war controversies and excesses would be better covered, but any attempt to introduce them to the article are almost immediately cleaned from that page (check the edit history).

The history of the RAND corporation is documented, but you aren't going to find that information on Wikipedia.

There is very clearly a PR organization sitting on top of that article scrubbing and sterilizing all the dirty stuff RAND has been involved in.

You can link to edits. Why don't you show us some of the ones you're talking about?
I started hunting through the edit history which is pretty involved to see if I could find some examples. I notice that MAD is mentioned in the section "RAND Corp".

I gave up then!

I looked at the edit history and couldn't find such edits in the last three or so years. Do you have any particular edits you're thinking of?
>> For everybody who was saying "Wikipedia is complete, there's nothing left to write about"

What?!? This is one of the stupider things I’ve heard from this community.

Not only do I find it very suspect that ‘everything has an article’, but these people are aware of new things being discovered and invented, no?

People think Wikipedia is complete???

I could add stuff in my field of specialty constantly. I just don't want to jump through all the hoops that come with being a wikipedia contributor, so I keep all my notes on a private wiki.

This selfish
Perhaps it is the wikipedia administration, moderation, and editing team that have put up such barriers who are selfish. It is selfish to expect others to jump through your hoops to contribute advanced technical knowledge out of the kindness of their hearts.
Just a reminder that you're only assuming that the information that arkades would like to add is even accurate. (No offense to them, I'm sure it is, just pointing out that the process of editing that encyclopedia is burdensome for a reason).
No offense taken. Weeding out for accuracy is a fair and important concern, and there's no reason to assume a random person on the internet is the expert they believe themselves to be.

Though I'd point out that unless wikipedia has a team of people whose expertise equals or at least vaguely approximates my own (or that of any expert that might contribute), I have to raise the question of whether that's really what they're filtering, or simply something that superficially looks like accuracy to someone unfamiliar with the material.

It also assumes that the inaccuracy of the current articles is less than the inaccuracy of articles under a hypothetical alternative moderation policy. That might be true, or it might not - chase away enough contributors and you're left with one-sided, stale information. I imagine that's fine for an article on the simpsons, but maybe not great if you're looking up leukemias.

And then you end up running into a moderator who just refuses to accept factual well-sourced additions based on their own incorrect knowledge. I'm still bitter from my last attempt to contribute.
I do not contribute to wikipedia because I dislike the politics there but I often browse it and see content in technical articles that helped me in the past magically disappearing, or wiki articles getting deleted along with their history because the moderators were too clueless about a certain topic.
Do you contribute to Wikipedia? I have, and to Wikibooks, and it can be a thankless task when you run into an admin who is more interested in formatting than content and more interested in the enforcement of petty rules than in helping the author get it right.
sounds like contributing to open source
Indeed, Wikipedia arguably has a deletionist bias[0].

[0]: https://www.gwern.net/In-Defense-Of-Inclusionism

This is largely what's kept me away. I'm comfortable adding piles of sources and citations to pass through some sort of Wikipedia Peer Review-Lite, but if I put in the hard work of creating a repository of knowledge for free and just see it go away...

I've tried contributing once or twice, and my work was erased like marks on the beach, promptly and without explanation. That was enough to ensure I don't do it again.

I have been editing Wikipedia for more than 5 years now. I usually get little push back for the edits I add. I try to keep every sentence cited and use my best judgement with the POV of sources. The worst experience I've had was people deleting content that they viewed through an ethnic chauvinist lens, or someone who strongly believed a statement with little documented evidence (both on the same article.) I keep seeing this viewpoint on HN, but I really thing editing is pretty smooth for most people. There are problems - the visual editor is only visible to those logged in, for example, and there are a good number of rules. But, in general, being bold has been rewarding.
I've noticed that corporate history tends to be a blind spot for Wikipedia. I suspect part of the reason for this is corporate PR departments have more resources than volunteer encyclopedians.
Interesting story, but the clickbait headline is not needed here. Story ends with:

> At the end of everything, Pepsi’s total combined losses, between physical, legal, and brand equity costs, would top $20M. Their market share would plummet and take years to rebound.

The "$32 billion" seems to come from the amount of bottles who would win the prize, but Pepsi didn't actually pay out the promised amount, so it didn't cost them $32 billion in the end.

Don't think this is really that clickbaity - the typo amount accounted to 32B, and the outrage caused by that non-existent 32B lead to riots in which people died, in other words, deadly riots.

sometimes things are legitimately outrageous and worthy of an emotionally charged headline.

It is clickbaity. There was no "typo" (Pepsi said it was a "system error") and there was nothing that cost them $32B. The only true parts of the headline is "Pepsi" and "caused deadly riots", which I agree with you, is outrageous and should never have happened. But we don't have to lie about the content of a story in order to make this outrageous, it's already outrageous enough with the facts.
Pepsi's typo didn't cause 32B of loss to them, but by some measure, there was a tiny, brief window where the country thought it was 32B richer. And this was before the extensive use of the internet, so it probably took weeks for people to discover that they shouldn't have bought that car/house/whatever.

So yes, Pepsi's typo didn't cost them 32B, but 32B of economic value was instantaneously created, and then, after some time, destroyed.

Was its total economic cost 32B? No, but was 32B a reasonable number to state? Sure. Pepsi's 20MM mistake over 32B caused riots seems a lot more wordy, no?

If it was only $20M, they probably just would have paid it, avoiding riots. What caused riots was their unwillingness/inability to pay the $32B they had inadvertantly promised.

I think the $32B is appropriate in the headline, it's the number that they accidentally promised, and could not deliver on, causing the riots.

But yes, the headline is written in a way to attract clicks, as all headlines are these days. I might have written it slightly differently. I don't think it's especially misleading.

> But yes, the headline is written in a way to attract clicks, as all headlines are these days.

Except that the attracted readers wouldn't lierally click on anything until news moved to the web, that's been the point of headlines forever.

It's not a lie though. All the title says is $32B of wealth was accidentally advertised (in this case, prize money based on numbers in a bottle caps). The headline doesn't say it cost Pepsi that amount, it just says the typo (ok, that bit isn't strictly accurate) was for the amount of $32B. A typo isn't the same thing as a payment.

Anecdotally the article was pretty much what I expected from the title: a large amount of wealth was accidentally promised (I assumed by physical property worth $32B and where the quantity was accidentally an order of magnitude more) and Pepsi, like all successful businesses, wormed out of delivering. As it turns out the story was more absurd than the headline lead me to believe but largely it was accurate to my assumptions based on the title.

Moving on from the title: this article did make me wonder what would have happened to the Philippine economy had Pepsi paid out (assuming they had the money to do so). With 800,000 people suddenly winning enough to buy a luxury house, I bet that would have caused house prices to rise massively and likely did all sorts of other massively disruptive things to the economy. Any other theories on this "whatif" scenario?

I mean, I don't like Pepsi and understand people being pissed, but going out to protest with hundreds of other 'winners' when its obvious this must be a big mistake is pretty ridiculous. The fact that Pepsi regained market share only shows that its easier for people to bust shit up then boycott a soft drink.
Unless OP changed the title, there is nothing that infer that Pepsi lost $32Bn in the process. Just that the size of the mistake was $32Bn.
This critique is unnecessarily pedantic. If I issue you a $200 check and (through whatever error) it ends up being written for $200,000,000, we would fairly call it a "$200MM typo" despite the fact that it would bounce.

The headline delivered exactly the story it promised. I was surprised I had never heard about it before.

Eh, I think it's debatable and not pedantic. I literally clicked on the link because I inferred the headline to mean that Pepsi made a typo that somehow cost them $32B.

And re: your example, I would not call that a "$200MM typo", I would call it a typo. If I intended to write a check for $200 and later when it was cashed found I had accidentally wrote it for $220, I would call that a $20 typo.

How's "Pepsi's mistake cost innocent lives by an exploding grenade"? Technically correct, but sounds even more horrendous and catastrophic, while technically 100% correct. If your argument is that the title is trying to be click-baity they could've probably done a lot worse even while staying with the strict facts. I really don't think the author meant to clickbait.
I inferred the headline to mean what it actually did, and was surprised to find comments here like yours. I would even argue that is the true literal meaning.

The fact that it caused riots was especially a giveaway. If Pepsi had actually paid out after making their mistake, why would there be riots? Yes I can imagine situations (e.g. they only paid out after there were deadly riots) but it still leads you to the right interpretation.

BTW I agree that your example is a $20 typo, but that seems unrelated. I would still call it a $20 typo if you then didn't cash in that cheque.

Taken a whole, I can see what you mean about the article title. I think this is a just bad title, but titles are hard (:

> I would still call it a $20 typo if you then didn't cash in that cheque.

Interesting. To me calling something a "$XX [typo|mistake|error|etc.]" means that the typo/error/mistake resulted in a loss of $XX. Any other way to read that feels confusing.

> If I intended to write a check for $200 and later when it was cashed found I had accidentally wrote it for $220, I would call that a $20 typo.

So my example was a "$199,999,800 typo"? Can't we just round up?

The relevant difference between our examples is not the amount. It's that my example check was cashed and therefore cost me $20. Your example check bounced and didn't cost you anything. If that bounce caused your bank to charge you a $10 fee, I'd call it a $10 typo.
It was a $32B typo that only cost them $20M
Hey there (I'm the author of the piece). I was about to make this exact point. Perfectly said.

As a writer, creating a headline is always a balancing act. You want to drive interest and create curiosity but you also don't want to mislead readers because that will burn them and they'll never read your stuff again.

I felt I delivered on that curiosity. But the internet is a big place. Someone will always find a bone to pick.

I'm new to this site. My Medium stats said I was getting a sudden avalanche of traffic from here so I created an account just to address the initial comment of this thread.

Anyways, thanks for reading everyone.

I had fun reading and thought the headline was fine, thanks!
No, we wouldn't, exactly due to the known outcome.
You'd think maybe as they were being printed you'd record the actual number being printed on the side somewhere and audit that every once in a while...
(comment deleted)
Great piece but click bait title and featured image is Hong Kong democracy protests...?
Clickbait, yes. To be fair, though, the top image is frequently stock photos, and they probably just did a search for riot or something and picked the first result.
I would love to know what the bug was, sadly we'll probably never know. I imagine the engineer responsible for `NEVER 349` made a mistake that ended up flipping a single bit, to the `ALWAYS 349` position...
800,000 bottles ? This mightve been intentional
a Coca-Cola operative infiltrated in the factory took care of it
my guess : they setup a machine that prints the successive numbers inside the bottle caps and there is no way to skip a number, they had to remove them afterwards.. someone simply forgot to do that.
Making 800,000 extra bottle caps and removing them later seems like a plan that wouldn't have left the drawing board. Especially with the risk of fraud from these particular caps.
Wouldn't surprise me if they had a little piston or something to push off every Xth bottle but there was an off-by-one error so they wound up pushing off every bottle with 348 printed instead, or perhaps they had to restart the line once and no one told the engineer to adjust for it so the place was off after that.
It sounds like the bottling factory printing all the other numbers simply missed the memo on not printing 349.
So many possibles.

Complete speculation / personal bias theory:

I'm going to go with human process error such as the instructions weren't communicated to the right people at the bottling companies and never reached the person who would enter `NEVER 349`.

Classic case of someone communicating a thing and assuming "everyone reads my memos / should do what I say" and not recognizing the complexity of such communication or that maybe there's a proper process to follow.

I run into that a lot.

That wasn't a computer error.
It sucks for whoever got the “real” 349 bottles.
I wonder if this is a third-world thing, soda contests for cash.

I remember when I was 12 in Bolivia, we had literal Coca Cola scratch off cards and had to collect bottle caps with the numbers to make a lot of cash. Kids would trade bottle caps, rumors floated about fulanito having the rare cap and he would sell it for the right amount, etc.

Been in the states for a few years now and never heard of something like that here. Then again I haven't had soda in the same amount of time.

We used to have those kinds of bottle cap for cash contests in the US, I think it was most common in the late 90s and early 00s.
When it debuted, McDonald's "Monopoly" game was a huge deal and people would eagerly open their tickets to see if they won a big prize. If they won it was usually a small fries or another ticket, but sometimes it was some money and one person won $1M.

After a while the luster wore off and it faded into the background. I'm sure it boosted sales in its first few years quite a bit but now you hardly hear about it. I'm actually surprised it still happens.

I think there is a documentary out now about how that whole thing was gang controlled haha. Havent watched, not sure if i believe, but wouldnt be that far fetched.
It's still a thing with Tim Hortons' "Roll up the rim to win" campaign, where you can win money, cars, barbecue grills, dougnnuts, or coffee when you buy a coffee.

Actually, it was cancelled this year, because millions of people handing pieces of coffee cup rims to workers isn't good when there's a global pandemic going on.

It wasn't cancelled, it was just moved to their app. When you scanned your card/phone you would get credits to "roll up". I only realized when I opened the app (instead of using mobile wallet) and saw I had 25 "roll ups" to use.
It's not a physical thing though. It's just another form of virtual points.
I think it might just be a 90s thing. I remember similar promotions in Ireland when I was a kid, but they don't seem to be a thing anymore.
Moral of the story - if you are a big multinational company you can literally get away with murder and even increase market even with the most botched marketing campaign selling sugar water
(comment deleted)
> you can literally get away with murder

Pepsi killed anyone? That's not what the story says at all. Accidental death is not the same as murder.

Wrong sugar-water vendor.

https://en.wikipedia.org/wiki/Sinaltrainal_v._Coca-Cola_Co.

(edit HN is botching the URL. There's a '.' character on the end)

A trailing period in URLs in written prose is always ambiguous. Is it the end of the sentence, or part of the URL? If a trailing period is legal, what do you do if a URL is at the end of the sentence? Leave a space between them? It's fine for HTML because you control where the a tag goes, but auto-linking like HN does can't distinguish between the two cases.
HTTP and e.g. markdown has solved this with angle brackets. I tried that (didn't work), and putting a question mark at the end (also didn't work).

What did work was putting two full-stops, but that was so ambiguous that I left it as-was. Wikipedia has a redirect in place so obviously HN isn't the only one with problems.

Pepsi didn't kill anyone as far as that story goes.
> if you are a big multinational company you can literally get away with murder

Pepsi's cockups are no doubt the catalyst of the events that unfolded but can you say that they are responsible for someone murdering someone else during what happened? are any actions the protestors undertake in anger the responsibility of Pepsi?

Shared culpability is a common thing in the law. A get away driver is not directly robbing a bank, but they still go to prison.
I think that's way too distant a comparison.

Pepsi made a mistake that disappointed some folks, nobody lost anything because of what Pepsi did directly, Pepsi wasn't assisting anyone in committing a crime, they weren't out to commit a crime / help anyone commit crime like you describe.

not the same as a get away driver is still part of the organised action.

I would say a better, or atleast much more controversial example is this - you cannot hold BLM movement responsible for the people that loot and burn, for exactly the same reason pepsi isn't accountable - because these actions (promises made -> not delivered -> riot started) are causated, but they are not correlated.

On the other hand - imagine you are a director at Pepsi managing this promotion. Imagine you personally sit down and send an email(or a courier-delivered letter I suppose) to the head of whatever factory is bottling your product, saying "under no circumstances print the number X on the bottle caps". Then the factory prints number X on 800k bottles.

Why is or should Pepsi be responsible for this? If I send an email to someone explicitly telling them not to do something, and they do it anyway - why should I be responsible for it? Unless the bottling factory is their own, but the article doesn't specify.

That's why executives of big cos get paid big money. They are held accountable (or should be anyway) for the actions of their company. The buck stops with them so to speak.

If the thing is really a big deal, as in this case, maybe the exec should do more than just "send an email" and assume everything is fine.

Leadership is always responsible. They chose to work with the factory. Where is the vetting? The QA?

Sending an email doesn’t absolve one of responsibility...

Never knew!

Nowadays in the Philippines, Pepsi is the dominant brand, Mountain Dew is the #1 beverage in the Country and FEMSA, the bottler for Coke, which is the same as it's Mexican and South American bottler had to divest and sell it's assets back to Coke USA because it was so unprofitable because it didn't properly plan for Sugar futures.

So, Coke is #2 in the country and there was a while where it was unavailable in Metro Manila, even in BGC/Taguig..where it's corporate head quarters are.

(comment deleted)
TLDR; via Wikipedia [1]

> [May 25, 1992] As part of PepsiCo's local promotion titled Number Fever, it was announced that the person who possessed the Pepsi bottle cap with the Number 349 is eligible to claim the 1 million peso prize. About 800,000 were eligible for the prize instead of an intended single winner. Several protests and bombings followed after Pepsi refused to award the 1 million prize to thousands of bottle cap holders and said that a computer glitch caused the incident. [2]

[1] https://en.wikipedia.org/wiki/1992_in_the_Philippines#Events

[2] https://archive.seattletimes.com/archive/?date=19930726&slug...

Is it just me or others also think that medium has really spoilt our reading experience on the web... they are putting anything behind paywall, without author's consent...

This is not what we wanted Medium to become.... Please stop publishing to Medium.

There was a similar glitch in a contest Kraft held back in the late 1980s.

https://apnews.com/aa740ca43955c874dadb0394afc22940

TLDR almost every piece was a winner instead of just a few rare ones.

I managed to mail in a few pieces before the packages got recalled, I remember being in multiple class action suits and getting multiple waves of settlement checks and coupons for free food over the next few years. Obviously I was not going to get 4 vans, 12 bicycles, and 10 skateboards.

Contests now regularly use the "Kraft Clause" to avoid this problem.

https://www.lexology.com/library/detail.aspx?g=c742f572-f3e2...

"“If due to a printing, production or other error, more prizes are claimed than are intended to be awarded for any prize level, the intended prizes will be awarded in a random drawing from among all verified and validated prize claims received for that prize level. In no event will more than the stated number of prizes be awarded.”

It’s great to be powerful, you can completely fuck some shit up and walk away with a shrug. Pepsi should have been held to pay the $32 billion and if they didn’t have it, go bankrupt and become the property of the creditors.
Reading the judges' conclusions I have to wonder if Pepsi bribed some people to get away with the not guilty verdict...
What strikes me as food for thought is the stupidity and wickedness of the riots. I understand the rage if you own one of the few winning tickets of a lottery and the organization doesn't want to pay out. But in this case they knew that it had to be a mistake, as they all had the winning ticket. Were they really thinking that Pepsi should pay $40k to each and all of them because they all possessed the same identical plastic cap? Would they not realize that receiving all almost $20 was a larger collective gain?

And more in general, how do you govern a country where the population exhibits this kind of inability to see the larger picture and the point of view of the opponents in a dispute?

Ask that about Canadians, who trashed Vancouver when the Canucks lost a damned hockey game.
You have to imagine people in poverty and a sum that is not only poverty-ending, but probably allowing you to live in relative luxury, probably allowing you to start a business to have a better source of income. Assuming you're an American not living in a leaky tin-roofed cardboard box in 2020, sure, for you $40,000 is not that sum of money, but in early 1990's Philipphines it probably was...
You can simply multiply that sum by 100 or 1000 and ask yourself how would you react if the same happened to you. In this order:

1) you buy several Pepsi

2) you discover you won 40 million

3) you discover everyone else won 40 million

4) it is explained to you that there was a mistake and you can get 100 dollars as a refund.

5) you start burning cars and throwing bombs together with all the others because each really really wants his 40 million.

Do you think you would go from 4 to 5? Do you think your society would work if it contained a large percentage of people who would go from 4 to 5? That's what I was trying to point out.

No you can't multiply it by a number and say it's comparable...

If I don't win 40 million, I'll still have a warm apartment, a well paid job, a car, electronic toys, and a vague plan of a leisure trip to Iceland after this whole pandemic blows over. Once in a while I'll say "Fuck, at one point I thought I was a multi-millionaire, but it was a mistake!", and I'd be able to distract myself by, I don't know, playing Xbox or going out drinking overpriced cocktails with friends.

Meanwhile a lot of the Filipinos probably thought they'd get to no longer live in a cardboard box in a slum, no longer wear ratty clothes, worrying about work and food... And we don't know how many rioted, presumably many cried quietly after having their dreams quashed..

Yeah, ok. So what did they want? Pepsi Cola paying $40k to every person in the country? Really?
There is no way to make them happy. Pepsi clearly didn't have $32B. The only conclusion you can make from this situation is that these people clearly didn't want to be happy.
I'm sorry they weren't all clear-headed zen monks who could accept the loss of their presumed liberation from poverty (in some minds maybe even a return to poverty?) without releasing any anger.

If only the whole world was like you right, I'd bet it'd be an easier to place to govern, right?

> stupidity and wickedness of the riots.

There was indeed stupidity and wickedness but it started with Pepsi, when they ran an unlicensed lottery and then fucked it up. By advertising the $40k prize for the winning number and then reneging on the prize, Pepsi engaged in false advertising, and breaking a contract. If they ever pulled that in the US they'd face a hefty FTC penalty (or at least would have done at the time; today, who knows?)

> Would they not realize that receiving all almost $20 was a larger collective gain?

You come off as out of touch and insensitive with that comment. A "larger collective gain" means fuck-all to people living in abject poverty. Each of the winners, in their mind, lost $38,980 - more money than they could hope to save up in a lifetime. In poor countries the competition for everything is cutthroat and there's often no collective "national spirit". People will help their friends, neighbors and family and that's about it.

Why don't you tell me how you'd feel if you won a $100 million lotto jackpot (after playing for years) and then 2 days later someone told you "oopsies our bad, here's $200."

> Each of the winners, in their mind, lost $38,980 - more money than they could hope to save up in a lifetime.

Yes. Then as soon as they meet all the people that have made the exact same win they must realise something is wrong, right?

> Why don't you tell me how you'd feel if you won a $100 million lotto jackpot (after playing for years) and then 2 days later someone told you "oopsies our bad, here's $200."

Sure. I'd feel enraged, but once I'd discovered that everyone else won the same lotto jackpot it would be totally clear it was a mistake and it makes no sense at all to claim it. I'd be happy to have the jackpot divided in equal parts among all those who bought the winning ticket (the alternative being having a new ticket issued for free and a new turn played- so much less chances to gain anything).

It's more likely that the "winners" would sue the lotto company for false advertising, breach of contract, and mental distress, and force a much larger settlement - possibly 1000x what was offered ($200k in my example). They might even force the company into bankruptcy. But that option isn't available to poor people in places like the Philippines.

The "multiple winners" thing "makes no sense" to you because you're looking at it from Pepsi's perspective. Not from the perspective of someone who thought their life was going to be transformed, and now it's not. Their thought process is more like "It's not my problem if the people running the lottery screwed up. I won fair and square according to the rules that were set out. They have to pay and that's that."

Pepsi marketed the heck out of this thing, gained a ton of market share, and would stand to make lots of money even after their ultimate $20 million bill. But instead they reneged on a contract by blaming it on a technical error. I don't understand why you're so eager to let them off the hook for this - don't contracts mean anything?

> the alternative being having a new ticket issued for free

There are plenty of other alternatives actually, all the way to "Pepsi actually pays the $32 billion, in installments". But that's not realistic - they'd probably just exit the Philippines market instead.

Here's a more realistic alternative that might have gone over better: Pepsi says, "We get that a lot of you bought Pepsi because of this promotion and we feel terrible. To make it right, we're upping the prize pot to equal all of our sales in the Philippines this year. After all, the only reason we made most of those sales was because of the prize, so we're gonna forfeit those."

1996 in Argentina, Coke did a promotion where promised "two caps for a Ramones concert ticket" (we are huge fans here). Of course they didn't can keep with this promise and disturbs occurred:

https://www.youtube.com/watch?v=ciq9ONHj4bc (spanish)

Contemporaneous reporting:

https://www.washingtonpost.com/archive/business/1993/07/29/a...

The article refers to a similar incident in Chile:

> The Philippines Senate Committee on Trade and Commerce accused Pepsi of "gross negligence" and noted that Pepsi was involved in a similar fiasco in Chile just a month before the 349 incident.

I can't find any details about the Chile incident, but I did find another Pepsi-related giveaway that was newsworthy, this time in the U.S. where they worried about overloading the telephone system:

https://www.nytimes.com/1991/01/25/business/the-media-busine...

> As part of its Diet Pepsi commercials during the telecast of the Super Bowl this Sunday, Pepsi had planned to give $1 million to each of three randomly chosen callers to a toll- free telephone number that would be shown during the commercials. But last night the company scrapped the idea.

> But the company announced last night that it had canceled the promotion after discussing it with the Federal Communications Commission. "We did not want to do anything that would have even the slightest chance of disrupting our nation's ability to communicate," said David Novak, executive vice president for marketing and sales.

> Telephone industry experts had estimated that as many as 50 million calls could flood the nation's telephone network during the promotion.

> The number “349” was the $40,000 winning number. > Pepsi had explicitly told its vendor factories not to print this number at all.

Sounds dangerous to decide the winning number in advance. Makes it easier for a vendor or someone else to cheat.

But I suppose they wanted short numbers.

Worse yet, they told them, “don’t print this number”! However thd vendor misunderstood and mistakenly printed them as the low value payout number, so 800,000 of them...

It’s almost like when in a high stress situation you tell someone “don’t press the red button” or on a tight rope “hey don’t look down and lose your balance”

How else would you define there to be only two winning numbers then? If you define the winning number afterwards, there’s no way telling how many numbers of those would’ve been in circulation.
Make the winner number outside the normal printed range.
This is so obvious - what are the downsides? I mean, it could be easily guessable that the bottlecap is valuable, ruining the surprise.

Ultimately they didn't test. Or confirm the vendor's test. It's all Pepsi's fault for a mission critical bug.

I mean...that's what they tried to do...
Make it a non-number?
Use a CSPRNG to produce the tokens and keep a list of the outputs. Select uniformly/randomly from this list for payouts.
Perhaps they should have instructed the manufacturers to not include numbers which were multiples of, say, 15, but to use all other numbers uniformly.

Random audits (with expensive penalties) could have detected whether the manufacturers had complied with this, without making it too obvious to contestants that the winning number would be a multiple of 15.

The winning bottles could then be manufactured to contain some specific multiple of 15, while still allowing the organisers some freedom to choose which one.

Should have made the winning number outside the range of printed numbers instead of being 'skipped'.
It's not a lottery where people pick their own numbers. As long as you can't find out what it is without opening the bottle, it didn't have to be a number at all; it could just say "you have won" where all the other bottle caps say nothing.

Though, making it non-forgeable might be trickier.

"You have won" caps would disappear mysterious from the factory floor.

You need a system where no one can identify a winner until after the caps have been printed, and the soda has been bottled and shipped.

yeah exactly; if I'm with a vendor factory and hear, "Whatever you do, don't print any 349s!" you can damn well bet I've convinced myself what the winning number is.
Would the world have been better off if Pepsi had essentially gifted itself to the people of the Philippines?