Retrospective - API Release To Prevent Minimum Bet Abuse - 19th June

[oldBloke]

You can place 0.8 stake amount at 1.01, watch my video:

https://www.youtube.com/watch?v=Xf5ysUiBzPU

Sorry, was this in response to my post?

I was looking at a lay stake of 0.07 gbp placed @ 1.18

[WTPooh]

Sorry, was this in response to my post?

I was looking at a lay stake of 0.07 gbp placed @ 1.18

It's impossible because this bet will return 25% more than it ‘ought’ to. Read the first post of the thread.

[oldBloke]

It's impossible because this bet will return 25% more than it ‘ought’ to. Read the first post of the thread.

Yes, that's what I worked out as well.

Was just pointing out that the odds listed don't work all the time.
Some price/size combinations will just NOT be allowed

But that makes trying to hedge awkward.

[WTPooh]

Yes, that's what I worked out as well.

Was just pointing out that the odds listed don't work all the time.
Some price/size combinations will just NOT be allowed

But that makes trying to hedge awkward.

You need to check your bet size before hedge and adjust it if it's necessary.
In your case you can adjust size to 0.06 (liability 0.01) or 0.13 (liability 0.02).

[WellDoneSoft]

I suggest the following solution: https://youtu.be/y8GaCe60BqA

[Sansa]

So you can lay £0.01 @ 1.8, but that is 25% more so by the BDP definition that should be invalid.

[geoffw123]

Hi

Odds liab rounded % error
1.81 0.0081 0.01 23.46
1.80 0.0080 0.01 25.00
1.79 0.0079 0.01 26.58

I think they just worded their update a little loosely, as you said lay £0.01 @ 1.8 is right on the change over point, they seem to let an exact 25.00% through as valid. Up a tick would be OK too, but down a tick would fail

I wonder why they chose to fix this low value stake abuse in this way. It broke a lot of code, and it is a fiddly fix for everyone to work around it. I am suprised that they didnt tweak their rounding code on the Server, so these edge cases rounded in their favour. It wouldnt have broken everyones code and would have made them a few pennies !!

[BFairUser]

I think Neil was not at work in past days.
Someone tried to do something good, but it has been created a problem.
Issue will be fixed, in an exaustive way (as usual) soon.

Thanks Neil for your work, and thanks to all BDP team for daily engagement and work.

[LetsGo]

Been running for 12 years then an urgent fix to a "problem" that has always been there.

Can we get access to the "Cash out" API?

[bfexplorer]

Been running for 12 years then an urgent fix to a "problem" that has always been there.

Can we get access to the "Cash out" API?

Cash out web api must work on the same rules, as it uses the same api internally as rest api is built on.

[Hefi]

Had someone already implemented a workaround-algorithm for this issue? It is one thing to find the nearest good size/stake combination of the target bet, but we also have to move the initial bet (from above the minimum stake) to our target bet (by checking that we don't place/cancel the order to an 'unfair' bet).

[geoffw123]

As you say, you cannot now blindly place the dummy £2 bet at odds of 1.01, as that can fail when you try the subsequent cancel operation.
I couldnt see a very trivial fix for this so I had to add a calcMinOddsToPlaceDummyLayBet(double layStake) function
Then starting at odds of 1.01 iterate around calculating the true profit and the rounded profit and the percent difference between them, if it fails move up to the next tick and try again, return the lowest odds that passes the new -20/+25 % criteria.
Use the returned odds to place the intial £2 bet at.

Of course that doesnt allow for the scenario where it might return odds of 1.50 say, and the price is already down to 1.05.
You will have to ask Betfair how to workaround that one. They would likely just say tough, you cant.

Regards

[Cymatic]

Because when during this "small bet placement" operation if anything goes wrong you need to cancel original minimal + small bet stake, and not leaving it on market as your software and bet angel does. Because then of course user could take actually bigger loss than with it original bet opening trading session.

The current version does not leave a stranded bet. Anyway, I think the purpose of this topic is to hear about the new Betfair rules relating to rounding abuse, not to start comparing software products. With best wishes.

[bfexplorer]

The current version does not leave a stranded bet. Anyway, I think the purpose of this topic is to hear about the new Betfair rules relating to rounding abuse, not to start comparing software products. With best wishes.

I mentioned that because you have got bug in software leaving your clients to make unnecessary bigger loss on trading bets, that is really bad.The same bug had Bet Angel, other vendors forums did not report such problems.

Can it be done better?

Well of course, if not developers of trading app make unit tests for each critical code, then at least betfair during certification process could discover such problems in software.

That is my point in the previous reply. To the rest, we all need to use api according to betfair T&C.

[oldBloke]

I think they just worded their update a little loosely, as you said lay £0.01 @ 1.8 is right on the change over point, they seem to let an exact 25.00% through as valid. Up a tick would be OK too, but down a tick would fail

This is not always the case ( at the moment )

£0.10 @ 1.16 is also right on the change over point, but it gets rejected.

Hoping betfair will correct or clarify this soon