TPD Feedback and Issues

[jamesedwards]

Tamps Bay race 2: no TPD data all race

James, - Do you get your quota increased for these races where TPD data isnt present?

I don't know, I've not been keeping track. It is of no impact to me as I won't be using my full quota this month anyway.

[SomeDude]

So... just got completely screwed at Sam Houston. Layed a runner that showed well behind (>15m) in-play. Turns out that runner was leading. Logs are crystal clear, so it appears transponders screwed up or data just wildly wrong. I've said it before and I'll say it again... TPD is NOT accurate sometimes. I'm not talking about small margin of error type wrong; I'm talking about wildly, completely wrong to the point that it looks like someone is manipulating the data deliberately. The type of wrong that makes me want to ask for regulatory investigation. This could be a good service, but they need to sort out these types of problems. Oh well, at least I know have more data points around which I can build controls. Buyer beware.

[TPD-Zone]

Sorry to hear about the Sam Houston experience and the US offerings generally, which are clearly still causing problems. Rest assured that this has been escalated internally and the Developers will be all over this. We will report back soonest.

[TPD-Zone]

An update re Sam Houston; there were "coupled" races in this meeting. You don’t often see this but in the US, horses in a given race that have the same owner are often ‘coupled’ for betting purposes. So if you bet on either of the coupled horses you will win if either horse wins. Obviously on Betfair, the converse applies to laying either of said horses. Investigations today have highlighted that TPD attributes are most likely being duplicated for coupled horses.

A fix will be applied soonest and the meantime we would obviously advise avoiding coupled races, particularly if using automation. You can spot a coupled race in that Betfair will number them 1, 1A and 2, 2A if another occurrence in the same race.

[TPD-Zone]

For info the only coupled races that I can see tonight; Laurel Park 18:09 and 19:40.

[LinusP]

An update re Sam Houston; there were "coupled" races in this meeting. You don’t often see this but in the US, horses in a given race that have the same owner are often ‘coupled’ for betting purposes. So if you bet on either of the coupled horses you will win if either horse wins. Obviously on Betfair, the converse applies to laying either of said horses. Investigations today have highlighted that TPD attributes are most likely being duplicated for coupled horses.

A fix will be applied soonest and the meantime we would obviously advise avoiding coupled races, particularly if using automation. You can spot a coupled race in that Betfair will number them 1, 1A and 2, 2A if another occurrence in the same race.

Fairly sure this isn’t true in regards to horses being coupled on Betfair.

[SomeDude]

An update re Sam Houston; there were "coupled" races in this meeting. You don’t often see this but in the US, horses in a given race that have the same owner are often ‘coupled’ for betting purposes. So if you bet on either of the coupled horses you will win if either horse wins. Obviously on Betfair, the converse applies to laying either of said horses. Investigations today have highlighted that TPD attributes are most likely being duplicated for coupled horses.

A fix will be applied soonest and the meantime we would obviously advise avoiding coupled races, particularly if using automation. You can spot a coupled race in that Betfair will number them 1, 1A and 2, 2A if another occurrence in the same race.

Thanks for replying. Yes, I can see now that this a coupled race and your explanation matches the log data.

As at the end of the race, here's what my logs show for the winner, Lets Go Alamo, right at the end of the race:

Lets Go Alamo: runner_position = 5
Lets Go Alamo: calculated_runner_position = 5
Lets Go Alamo: calculated_velocity_rank = 1
Lets Go Alamo: percentage_remaining = 0.855
Lets Go Alamo: distance_remaining = 8.6
Lets Go Alamo: distance_behind_leader = 15.2
Lets Go Alamo: velocity = 15.42

The runner_position is as reported by TPD. The calculated values (calculated_runner_position and calculated_velocity_rank) are my own, based on TPD data. The reason I do this is that I've found that TPD quite often reports the runner position incorrectly (the winner is often reported as runner 2), even if the GPS position is accurate within MoE.

The runner position is wildly wrong, as the runner shown "Lets Go Alamo" had already crossed the finish line at this point, despite being shown as in 5th position, 15.2 behind the leader. In going back and checking what you're saying, I can also see that Capt St Jacques was reporting duplicate data with Lets Go Alamo for the whole race. In checking, the runners have the same trainer and owner.

Will the fix ensure that each runner reports its own data, regardless of coupling?
Coupling aside, TPD's error cost me money I never would have wagered had the data been correct (had Alamo shown actual position, I wouldn't have wagered); can I at least get a credit on my TPD service?

[ShaunWhite]

TPD's error cost me money I never would have wagered had the data been correct (had Alamo shown actual position, I wouldn't have wagered); can I at least get a credit on my TPD service?

They will if you promise to send them a cheque everytime you win money due to data errors.

Data errors are a zero sum game, if they aren't then that's an opportunity.

[SomeDude]

TPD's error cost me money I never would have wagered had the data been correct (had Alamo shown actual position, I wouldn't have wagered); can I at least get a credit on my TPD service?

They will if you promise to send them a cheque everytime you win money due to data errors.

Data errors are a zero sum game, if they aren't then that's an opportunity.

Yeah... it's not zero sum when they can control how/when the errors happen. And you're right, it's definitely an opportunity, just not for the paying customers.

[captainbang]

No position of runner showing for me in the 16:00 race at Lingfield. No issues with any other race at Lingfield today that I noticed but I wasnt glued to the screen so could have been more issues.

[TPD-Zone]

I have just received word from the Developers that the issue that caused loss of data in the 16:00 has been identified and fixed for tomorrow's races.

[jamesedwards]

I have just received word from the Developers that the issue that caused loss of data in the 16:00 has been identified and fixed for tomorrow's races.

Any update on the 1A issues in USA please?

[captainbang]

I have just received word from the Developers that the issue that caused loss of data in the 16:00 has been identified and fixed for tomorrow's races.

Thanks

[gstar1975]

No position of runner showing for me in the 16:00 race at Lingfield. No issues with any other race at Lingfield today that I noticed but I wasnt glued to the screen so could have been more issues.

Capture456.PNG

Surely you could just order the grid in the distance to finish then this would display the runners in positional order.

[jamesedwards]

I have just received word from the Developers that the issue that caused loss of data in the 16:00 has been identified and fixed for tomorrow's races.

Also all Gulfstream turf races still no stride, distance or leader data. Just speed and coordinates.