Hacker News Reader

Comment by setr | Hacker News Reader

Its accidentally $82,000. If you allow such a bug to be honored, then potentially it could manage a near-infinite payout.

Such bugs becomes not a costly mistake but instead existentially threatening

And we certainly don’t have a software industry capable of supporting that environment

View on HN

Hedge funds seem to automate things in an environment where certain bugs are existentially threatening. They hire doctors of math and get their code correct. It's not impossible, just expensive.

Right. If you make system creation more expensive, the house needs to collect more juice.

I think most sports betting consumers here prefer the current, lower cost, more diverse system with a slightly higher error rate.

I'd reckon most people complaining here don't even bet on sports, but want to grandstand against bookies. I can understand that I guess.

I'm not going to comment on moral or legal responsibilities of bookies and gambling companies, because I am uneducated on both subjects, but I found your use of the term "collect more juice" really interesting. It seems odd to relate pulling revenue from fruit to pulling revenue from people's incorrect bets. Is that an industry term?

https://en.wikipedia.org/wiki/Vigorish

This isn't a nonprofit charity. It's a bookie. They decided to offer automated betting systems, in the hopes of winning money from their customers. They should stand behind those same systems when they don't work in their favor.

If you can't afford the errors you should not be in the business (but they can easily afford this error).

I don't think so. The legal system in Europe honors good faith.

Luckily for the better New Jersey is not in Europe.

Oh yeah, it is - I just wanted to offer an alternative to your views, and added a real-world example.

We do too have software that can be that reliable. A aviation and space flight are quick examples. Gambling houses probably don't want to pay the cost of that software, but it could be done.

It wouldn't just be bookies paying that cost. It would be forwarded to the customers placing bets. I'd prefer to keep the costs low, since nobody is going to die and things aren't going to blow up. The worst that will happen is someone thinking they scored a huge payout and not getting it.

Sure, but in this case it’s not an infinite payout. And the courts will care about that more than an engineering hypothetical, i would wager.

The software industry absolutely does have ways to prove a much higher level of code correctness than Paddy Power is demonstrating.

Best use humans then. Otherwise there is no trust.

You know they did not inform people of errors the other way.

The best part is they did use humans, this bet was placed in person.

> The bettor, who identified himself to News 12 New Jersey as Anthony Prince, placed the wager over the counter at the sportsbook at the Meadowlands Racetrack

Reply on news.ycombinator.com