This is when us manual dinosaurs sit back and read with glee. No not really Contrarian. But it must give you botters some sleepless nights worrying about these sorts of problems I would presume.
This is when us manual dinosaurs sit back and read with glee.No not really Contrarian.But it must give you botters some sleepless nights worrying about these sorts of problems I would presume.
Opportunitie to profit abounding then ? For those with quick fingers and nimble minds. Too late in the day for me though alas. Slowing down time for me.
Opportunitie to profit abounding then ?For those with quick fingers and nimble minds.Too late in the day for me though alas. Slowing down time for me.
Yeah I know trevh. But it was primarily established for botters who like to trade automatically ? . Was it I don't know ? But I agree high volume manual bettors, particularly in play, presumably also make necessary and valuable use of the API.
Yeah I know trevh.But it was primarily established for botters who like to trade automatically ? . Was it I don't know ?But I agree high volume manual bettors, particularly in play, presumably also make necessary and valuable use of the API.
Did it in fact migrate over to sports betting from x-games in the first place ? I mean how would have the original non BF mkt. makers on Xgames proceeded with an API ? Impossible I would say, with a new game every couple of minutes.
Did it in fact migrate over to sports betting from x-games in the first place ?I mean how would have the original non BF mkt. makers on Xgames proceeded with an API ?Impossible I would say, with a new game every couple of minutes.
Persistent intermittent issue for days now. Maybe it's just one dicky server in the farm or a network issue but I keep getting slow responses, around 1 in 20 requests take 100ms or more longer than expected.
Persistent intermittent issue for days now. Maybe it's just one dicky server in the farm or a network issue but I keep getting slow responses, around 1 in 20 requests take 100ms or more longer than expected.
Perhaps the lesson here for all auto-bot users is make sure your software has good error trapping and can litigate correctly all possible retun calls to get better position outcomes.
There are a number of apps that can fall over when bet-placing when not fore-filling the bet objective through implicit procedure structure in relation to retuned information.
Perhaps the lesson here for all auto-bot users is make sure your software has good error trapping and can litigate correctly all possible retun calls to get better position outcomes.There are a number of apps that can fall over when bet-placing when
BFF - if anyone knew what you were on about you'd be dangerous.
I just received a reply from Betfair to confirm that the lousy data on Saturday 10th was due to a "single server failures" and of course they are "working to ensure any such future failures are identified and rectified as quickly as possible so that they have minimal impact".
Apart from the "implicit procedure structure" approach of BFF, does anyone have any ways to catch a server problem? I haven't seen any updates on http://service.betfair.info/.
BFF - if anyone knew what you were on about you'd be dangerous.I just received a reply from Betfair to confirm that the lousy data on Saturday 10th was due to a "single server failures" and of course they are "working to ensure any such future failur
Simple way is to keep track of total matched volume, if it goes down and a non-runner hasn't just been declared then you are probably looking at old data from a wonky server.
Simple way is to keep track of total matched volume, if it goes down and a non-runner hasn't just been declared then you are probably looking at old data from a wonky server.
If you program the API yourself you can easily tell the returned data is incorrect if it does not comply to what you expect to get,then you take appropriate action to normalize your betting activities, trying to do this via a third party app on the fly would cause many problems, the good news is that server problems are rare but to assume servers done ever go wrong would be bad planning.
If you program the API yourself you can easily tell the returned data is incorrect if it does not comply to what you expect to get,then you take appropriate action to normalize your betting activities,trying to do this via a third party app on the fl
That will work for GetMarketPrices, but what about the other calls. I can't see that there is a simple way of distinguishing an outdated call from, say, GetMUBets. The only way, I think, of doing this is to have the program retain data over a period of time, so that it can know, for instance, that a bet that was previously recorded as matched is no longer showing in the return from GetMUBets, and therefore the data must be outdated.
Fred,That will work for GetMarketPrices, but what about the other calls. I can't see that there is a simple way of distinguishing an outdated call from, say, GetMUBets. The only way, I think, of doing this is to have the program retain data over a pe
a lot can stray out of range when engaging in SOAP operations as you are ultimately relying on a second party to function the data beyond your control.
a lot can stray out of range when engaging in SOAP operations as you are ultimately relying on a second party to function the data beyond your control.
My comments are only from a programmers knowledgs, not up to that stage yet in my programming of the api and your comments about sometimes not knowing if the return call is outdate4d has been noted
My comments are only from a programmers knowledgs, not up to that stage yet in my programming of the api and your comments about sometimes not knowing if the return call is outdate4d has been noted
a logical observation though of the problem is how do users know that the returned call is out-dated as if a human can tell this, then a computer would be able to do the same if programmed.
a logical observation though of the problem is how do users know that the returned call is out-dated as if a human can tell this, then a computer would be able to do the same if programmed.
Well, as I said before, it is quite easy to set up the program to distinguish outdated calls if one provides it with a memory of previously returned data. And that is obviously how we, as human observers, can see immediately that old data is being returned. The point is, though, that this is a pain. For lots of applications, you don't ordinarily want/need to be collecting the data as the program receives it. You just act on it and then move on.
Well, as I said before, it is quite easy to set up the program to distinguish outdated calls if one provides it with a memory of previously returned data. And that is obviously how we, as human observers, can see immediately that old data is being re
Lots of extra checks and balances on code have had to be added these week for API progs. Mine have been failing all over the place. Website been running like clockwork though.
Lots of extra checks and balances on code have had to be added these week for API progs. Mine have been failing all over the place. Website been running like clockwork though.
I think I would have to add a call to GetMarketTradedVolume to the cycle and look out for a decrease (not accompanied by a non-runner being declared). Contrarian, that would also work for you if you spot a problem with the traded volume you can guess that your GetMUBets data is unreliable. Tracking a change in the matched bets is tricky because something could become void or settled.
Maybe Betfair will decide to refund commission on markets where their servers send inaccurate information?
Good tip from Fred!, thanks.I think I would have to add a call to GetMarketTradedVolume to the cycle and look out for a decrease (not accompanied by a non-runner being declared). Contrarian, that would also work for you if you spot a problem with the
Personally I just mainly count errors when I place bets, if I receive too many ominous responses such as UNKNOWN_ERROR, INTERNAL_ERROR, or BET_IN_PROGRESS it's time to stop. I don't actually bet enough on each individual bet to be too concerned about having to know my exact position.
Personally I just mainly count errors when I place bets, if I receive too many ominous responses such as UNKNOWN_ERROR, INTERNAL_ERROR, or BET_IN_PROGRESS it's time to stop. I don't actually bet enough on each individual bet to be too concerned about
That's a conservative approach, trouble for me is that I generally leave the bot running during the day or overnight and so don't want a five minute server wobble to mean an early bath. And my margins are pretty tight so I do care about whether I got matched or not.
That's a conservative approach, trouble for me is that I generally leave the bot running during the day or overnight and so don't want a five minute server wobble to mean an early bath. And my margins are pretty tight so I do care about whether I got
It's awkward. You take an early bath or expose yourself to risk. Another option if you see a wobble is to take a 5 minute nap. If you've taken 6 naps in the last half hour then it's time for bed!
It's awkward. You take an early bath or expose yourself to risk. Another option if you see a wobble is to take a 5 minute nap. If you've taken 6 naps in the last half hour then it's time for bed!