Thousands in Norway told they had won life-changing sums in lottery error
41 comments
·June 30, 2025mailund
martinald
Somewhat sarcastically, I could see this being TDD, with a typo in the test for this :).
dagw
The error wasn't made in any of the 'important' paths that dealt with real winnings or real money (which I assume are heavily tested) but 'only' on the notification on the mobile app. Apparently if you logged into your account on their website you would see the correct amount. So I'm guessing the bug in this case might have been something as simple as typo in a string formatting template, that wasn't tested.
Toritori12
Also many life-changing amounts in the same batch should prob trigger some alarms.
konsalexee
Definition of testing straight into the production, "what could go wrong"
Neywiny
I just want to make sure I'm understanding, this means when they say somebody expected to get x euros, really they won (x/100)/100 euros? Once to get it back to what it was before math, a second time to do the math correctly?
xg15
That was my understanding as well. So the person who thought they had won £138,000 did really get £13.80.
jagermo
relevant Reno 911 clip
gsky
Test failures for the greater good
Double_a_92
Why did the CEO leave because some dev made a typo? That wasn't necessarily caused by some error in the strategy of the company...
pjc50
This is one of those cultural transitions which is difficult for people on the other side to understand; it belongs to the forgotten era of personal honor. These days one would simply lie on TV, or hire a PR firm to do that for you, and of course put the blame on the lowest individual that can be found. Repeat when more mistakes are made, because low level employees are disposable accountability shields. (See, for example, the UK post office/Fujitsu Horizon scandal)
But it used to be the case that leaders were expected to take responsibility for the culture and systems underneath them, rather than just taking as much of a salary as the business will bear from it.
That is, if a low level employee makes a significant money-handling mistake on this scale, that's a systems failure. There should be checks and testing and a software development culture which makes this kind of error unlikely. This is what was lost with "move fast and break things". After all, it's only other people's money.
(edit: it seems not to have been an actual money-handling error, but a notification error. Still fairly serious in terms of angry customers)
tialaramex
In the public enquiry for the Horizon scandal we saw a pattern that happens over and over where extremely well paid and supposedly high performing executives suddenly remember they were idiots and had no idea about anything so it's not their fault.
I believe we should impose a statutory burden on boards for such outfits to establish that their executives are not idiots, so that when inevitably this happens again the executive is obliged to agree either they are an idiot who couldn't have known they were doing awful things but, they also lied over and over in this mandatory paperwork - and so they go to jail for lying, or, they weren't an idiot after all, they're guilty.
We use this same approach for drink driving here. You can let us take the sample, proving you were hopelessly drunk, then you get banned for drink driving, or, you can refuse the sample, we can't prove you were drunk but you refused and the same penalty applies for refusal.
I'm sure some people will protest - why should I have to put myself at jeopardy, surely I should be allowed to do crimes without consequence? OK, well lets try it out, if every CEO job goes vacant once such a rule comes into effect I guess we'll have discovered every single one of you is a crook, which is good to know. But I suspect instead we'll have no trouble finding candidates and this was simply mispriced - we won't fix the criminality but now those responsible will sometimes end up behind bars at least.
leokennis
Simplified, but if CEO's want to make the big bucks when their developers create some great new feature, they should also stop making the big bucks (resign) when their developers mess up.
Svip
Because CEOs should take responsibility for the public errors of companies? If they had had proper testing procedures in place, this mistake would have been caught long before it reached the public. Clearly they did not, and the buck stops at the CEO.
Double_a_92
But how is taking responsability just disappearing after you messed up?
squigz
What does taking responsibility look like to you?
leakycap
> If they had had proper testing procedures in place, this mistake would have been caught long before it reached the public
I don't know a CEO outside of smaller strictly-tech companies who would have any familiarity or direct involvement with testing procedures.
This is a weird head to roll for a developer's typo.
Svip
I am not asking the CEO to be familiar with the details of programming or with whatever other technical minutiae a company is forced to deal; they can – however – demand insurance that their systems are properly tested, even if they don't understand the details of those procedures.
CEOs are the public faces of companies; with which comes higher pay but more responsibility. Conversely, employees shouldn't be sacked for honest mistakes, since they are supposed to be faceless. If an employee makes a mistake at a company, and that company does not have procedures in place to protect against that mistake; then it's the company's problem, not the employee's. And when it's the company's problem, it becomes the CEO's.
frereubu
The CEO should set the organisational culture. I would argue that - alongside the fact this isn't the first issue of this kind by the sounds of things - the fact that a serious error like this could be caused by a developer making a typo reflects very badly on the culture of the organisation.
ochrist
Quality Assurance is the responsibility of management. In PRINCE2 e.g. it is stipulated to be organized by the steering group.
It other words: Management (steering group or CxO level) should ensure there is a QA organization to handle this. I don't know if this was the case here or if somebody just 'tested in production'.
tossandthrow
Resignation definitely does not seem like the right direction, and is likely a panic reaction.
That said. The ceo is responsible regardless of their knowledge - in tech heavy companies, like this apparently is, they'd usually appoint a trusted cto for things like this.
But failure always run up the chain of command, and ofcause it should. It is after all why CEOs are paid well. They are responsible.
detaro
"Figure out who won what" is kind of the core business function of a lottery, obviously upper management is responsible for ensuring that happens correctly. Not necessarily through direct involvement, but through ensuring the right people and processes are in place.
If they aren't responsible for that, what else would a lottery CEO be responsible for?
Dildonics4All
[dead]
Rygian
It's the CEO's job to organize their company in such a way that a dev's fat finger does not lead to such large-sized consequences.
HelloNurse
The "strategy of the company" must include correct prize computations, and it failed.
DanielHB
CEO puts a lot of pressure in delivery
CEO doesn't want to invest the resources to match the delivery demand
Quality goes down, safety and security standards are lowered
Also known as the Okta-effect
pavlov
The Norway lottery is not a private business, it’s a state-owned gambling monopoly.
Hence the CEO is practically a political appointee. And in politics, you want to see heads roll. Probably the Prime Minister told her to resign.
v5v3
State owned lottery. So politicians get blamed, and the CEO will be fired as the politicians need to show they are 'tackling the issue'.
That was always implicitly in the job description for any public sector CEO role.
It won't affect the ceo's job prospects in any way, they will pop up in a new gig soon enough.
saltvedt
Norsk Tipping has recently gotten fines (46MNok and 36MNok) from the gaming authorities for other issues:
https://www.vg.no/sport/i/KMAxP4/varsler-gigantbot-til-norsk...
https://www.nrk.no/innlandet/norsk-tipping-far-bot-pa-36-mil...
> “The amount has been multiplied by 100, instead of being divided by 100.”
I'm very interested to know if there is any coverage of automated tests at all for this system? This seems like the kind of mistake that would be picked up immediately with a bare minimum test suite.