I've stuck together a series of images on this race. Now we just have to speculate on how it happened?
Warwick- 15:00 - 10/01/2015
Here is my opener for for ten.
It should be impossible for this to happen unless there is a method / exploit that Betfair have that allows somebody to go into the book outside of the normal manner?
Was it a display issue or something else? I know that trying to get matched would only drop the order in at the lowest price. So the display was wrong but the matching process 'correct' somehow.
I lost money by trying to trade against the prices on display.
It should be impossible for this to happen unless there is a method / exploit that Betfair have that allows somebody to go into the book outside of the normal manner?
Was it a display issue or something else? I know that trying to get matched would only drop the order in at the lowest price. So the display was wrong but the matching process 'correct' somehow.
I lost money by trying to trade against the prices on display.
My data capture wasn't on the race, the only time during the day it wasn't!
I switched to the market when the 14:55 at Punchestown was about to start and spotted it straight away. So I'd imagine the timing would be 14:55 to around 15:05, in there somewhere.
I switched to the market when the 14:55 at Punchestown was about to start and spotted it straight away. So I'd imagine the timing would be 14:55 to around 15:05, in there somewhere.
-
Wainwright
- Posts: 108
- Joined: Mon Jan 23, 2012 1:10 pm
I have the data for this from 10 mins before the start - the problem was already there at that point and persisted until the market was turned in play.
The data returned from the API seems to be OK apart from the glaring error that best AvailableToBack odds are stuck at [email protected]. Every call returns that combination until it went inplay at 15:00:46, at which point it seems to sort itself out.
Traded volumes seem fine so it looks like there was no problem with matching, so maybe an issue with market data server caching somehow.
Haven't seen this exact problem before, but have seen problems with strobing caused by individual servers not updatating properly - maybe some new variation of that ?
The data returned from the API seems to be OK apart from the glaring error that best AvailableToBack odds are stuck at [email protected]. Every call returns that combination until it went inplay at 15:00:46, at which point it seems to sort itself out.
Traded volumes seem fine so it looks like there was no problem with matching, so maybe an issue with market data server caching somehow.
Haven't seen this exact problem before, but have seen problems with strobing caused by individual servers not updatating properly - maybe some new variation of that ?
Backwardation and contango are not applicable here, they concern spot v future prices.
This was closer to a crossed market - http://www.investopedia.com/terms/c/crossedmarket.asp
I have seen it before, there was a rouge £5 in the I'm A Celeb market this year.
This was closer to a crossed market - http://www.investopedia.com/terms/c/crossedmarket.asp
I have seen it before, there was a rouge £5 in the I'm A Celeb market this year.
-
Wainwright
- Posts: 108
- Joined: Mon Jan 23, 2012 1:10 pm
Some pre & post off order book samples.
- the data shown is the time, runner id, then the three best back prices with amount available then the three best lay prices/amounts, as returned from API.
Pre-off;
14:50:06:190 RP_8498565 2.16~247.25~2.12~414.44~2.1~1016.09~2.14~626.7~2.16~1980.13~2.18~39.12
14:55:01:506 RP_8498565 2.16~247.25~2.1~51.75~2.08~653.19~2.12~650.79~2.14~1106.26~2.16~5625.72
14:57:54:552 RP_8498565 2.16~247.25~2.02~2527.72~2~3092.71~2.04~2194.93~2.06~995.9~2.08~693.41
14:59:39:921 RP_8498565 2.16~247.25~1.97~3225.65~1.96~1975.84~1.98~217.36~1.99~546.04~2~2446.84
15:00:44:412 RP_8498565 2.16~247.25~1.94~395.52~1.93~91.92~1.95~765.93~1.96~413.7~1.97~479.91
- all pre-off order book data from 10 mins out has the "stuck" [email protected] best back price
In-play;
15:00:46:844 RP_8498565 1.88~309.54~1.86~22.08~1.83~6.25~1.95~56~1.96~26.7~1.99~39.04
15:00:58:212 RP_8498565 1.84~357.78~1.83~400~1.82~70.03~1.86~212.88~1.87~8.82~1.88~55.14
- once in play, rogue value cleared and all data as would be expected
- the data shown is the time, runner id, then the three best back prices with amount available then the three best lay prices/amounts, as returned from API.
Pre-off;
14:50:06:190 RP_8498565 2.16~247.25~2.12~414.44~2.1~1016.09~2.14~626.7~2.16~1980.13~2.18~39.12
14:55:01:506 RP_8498565 2.16~247.25~2.1~51.75~2.08~653.19~2.12~650.79~2.14~1106.26~2.16~5625.72
14:57:54:552 RP_8498565 2.16~247.25~2.02~2527.72~2~3092.71~2.04~2194.93~2.06~995.9~2.08~693.41
14:59:39:921 RP_8498565 2.16~247.25~1.97~3225.65~1.96~1975.84~1.98~217.36~1.99~546.04~2~2446.84
15:00:44:412 RP_8498565 2.16~247.25~1.94~395.52~1.93~91.92~1.95~765.93~1.96~413.7~1.97~479.91
- all pre-off order book data from 10 mins out has the "stuck" [email protected] best back price
In-play;
15:00:46:844 RP_8498565 1.88~309.54~1.86~22.08~1.83~6.25~1.95~56~1.96~26.7~1.99~39.04
15:00:58:212 RP_8498565 1.84~357.78~1.83~400~1.82~70.03~1.86~212.88~1.87~8.82~1.88~55.14
- once in play, rogue value cleared and all data as would be expected
This is not new, I have seen this happen before as well, off the top of my head twice. Ive always just passed it off as a problem my end and accepted it. One particular time I was on another vendors software, so I feel its not a Bet Angel problem.
Very odd, has to be something to do with cross matching? Here is the txt file which contains a basic parsing of marketBook..
https://www.dropbox.com/s/o6x9cwwyw81yv ... 0.txt?dl=0
And a quick viz to help view it..
https://public.tableausoftware.com/view ... zHome=no#1
https://www.dropbox.com/s/o6x9cwwyw81yv ... 0.txt?dl=0
And a quick viz to help view it..
https://public.tableausoftware.com/view ... zHome=no#1