Leopardstown - 14:00 - Christmas Hurdle - Voler la vedette

[pdupre1961]

I don't see why Betfair couldn't have said all that yesterday, preventing lots of the 'What are they hiding?' speculation.

Betfair's position is plausible and reasonable IMHO. However, I'd like to know why they needed to cancel all in-play bets, and not just those placed after the ghost money entered the market...

Jeff

What are you talking about Jeff,

There's nothing new here.

[Iron]

They went into more detail than previously.

BTW, I was also impressed that the statement came from the guy who will soon be acting CEO, and that he expressed his sadness about what had happened. It's nice to see a bit of humility from time to time...

Jeff

What are you talking about Jeff,

There's nothing new here.

[pdupre1961]

They went into more detail than previously.

What is new...

[Iron]

There could be nothing new.

It's a Friday afternoon at the end of the year, I'm thinking of having a few drinks, and I really can't be bothered comparing and contrasting this statement with the previous ones!

Jeff

They went into more detail than previously.

What is new...

[superfrank]

I still don't understand how one bet managed to evade the exposure limit and get placed... software doesn't work like that - it crashes from time to time yeah, but the same code doesn't just occasionally produce a freak result.

[Iron]

Surely it's no different to what happens with home computers from time to time, when they do something silly like stopping responding to the keyboard for no apparent reason, or a game you're playing inexplicably crashes...

Jeff

software doesn't work like that - it crashes from time to time yeah, but the same code doesn't just occasionally produce a freak result.

[mugsgame]

I have received an email from the Betfair CEO with an explanation of VDV incident.

On face value this appears to be what he says it is.

I accept his explanation, and draw a line under this weirdness.
It's a shame, I would have been happy to give them my 50% of my winnings too.

[superfrank]

that's very different and also not what happened.

they are saying that a piece of software that normally performs a check on the exposure limit before accepting the transaction suddenly let just one transaction through without the check out of billions of transactions that must have gone through the same code. it doesn't happen.

[pdupre1961]

Exactly Superfrank,

Complete bollocks from Betfair...

[freddy]

Im no computer expert, but surely there can be a bug
that can lay hidden for years, just waiting for the right or wrong conditions however rare to come into play.

Happens with all software does it not ?, thats why there are so many patches and updates to correct problems as they are gradually exposed, when extensive tested in the real world by the general public.

[Iron]

Computer errors happen. Maybe there was something significant about the transaction number of the bet or the precise amount that was being put through that caused the normal filtering process to be suspended, or fooled Betfair's servers into thinking that the account had sufficient funds.

And what's the alternative hypothesis? That Betfair placed the lay themselves, and had planned to pocket the winnings if the horse had lost? That doesn't seem very plausible. For a start, Betfair would have gotten more bang for their buck by laying way below 29. And they'd risk seriously damaging their reputation amongst the public at large, which is one of the biggest assets they have.

Jeff

they are saying that a piece of software that normally performs a check on the exposure limit before accepting the transaction suddenly let just one transaction through without the check out of billions of transactions that must have gone through the same code. it doesn't happen.

[pdupre1961]

Too much of a co-incidence that the stake (size of bet) was 2^32 / 2 (2 to the power of 32, divided by 2).

32-bit overflow...

£21,474,836.48 = 2147483648 pence

[Iron]

Maybe, because of the significance of that number in computing, something caused the customer's computer to come up with that number.

I think all we can say about the person behind this is that they're either a computing genius or a poor programmer!

Jeff

Too much of a co-incidence that the stake (size of bet) was 2^32 / 2 (2 to the power of 32, divided by 2).

32-bit overflow...

£21,474,836.48 = 2147483648 pence

[Krzysztof]

We had to void both the win and place markets and, to be clear, we would have voided the race regardless of who had won and whether or not we had suspended these markets sooner. This is because the bet in question had significantly impacted both markets.

Isn't BF creating new rule here?

What if bet exceeded for some reason liability but by being much smaller hadn't significantly impacted the market - would it be void then?

I guess not because of the rule:

However in the event that we process an offer for a bet or the acceptance of a bet in an amount outside the applicable thresholds, such bet will nevertheless stand.

Shall the difference in amount of money matched be the measure of decision that the bet will stand or be void?

Everyday someone impacts a market significantly by having 'fat fingers'. Does it mean their bets should be void too?

[gutuami]

1. 2147483647 is a prime number and this is the greatest at present known to be such.
2. 2,147,483,647 is also the maximum value for a 32-bit signed integer in computing. The appearance of the number often reflects an error, overflow condition.
3. (214) 748-3647 is the sequence of digits represented as a United States phone number and is the most common phone number listed on web pages

source: http://en.wikipedia.org/wiki/2147483647_(number)