TPD (Total performance data)

A place to discuss anything.
Post Reply
jamesg46
Posts: 3769
Joined: Sat Jul 30, 2016 1:05 pm

I noticed some have already put forward there desire for TPD to become a feature of BA. So I thought, well it was recommended starting a new thread about it. A few things come to mind,

Is it a possibility that BA is considering?
How much interest is there in having it?
If it were to become part of BA, what do you see as a reasonable price to pay?
How much do people think it will be of use & why?
If it were to be put into such an advanced trading software, do we anticipate increases in the inplay trading activity?

I think the scope for automation with TPD in BA would be awesome.

Love to hear your thoughts & interests on the tech/topic.
Last edited by jamesg46 on Sun Apr 11, 2021 9:57 pm, edited 1 time in total.
User avatar
megarain
Posts: 2040
Joined: Thu May 16, 2013 1:26 pm
Contact:

It's my understanding the daily cost to buy the feeds from TPD is in the region of 35 quid + Vat per track per day.

Some Betfair accounts may get it cheaper.
PeterLe
Posts: 3715
Joined: Wed Apr 15, 2009 3:19 pm

The feeds from TPD are not a magic bullet. if you are not already a very successful in play trader this in isolation won’t turn you into one IMO
Betfair have a very limited number of licences that they have allocated out, but you have to demonstrate a track record and if they do grant you a licence, you have to make sure you put them to good use.
As Mega has mentioned, you can deal direct with TDP rather than via Betfair if you want to
Regards
Peter
User avatar
LeTiss
Posts: 5386
Joined: Fri May 08, 2009 6:04 pm

PeterLe wrote:
Sun Apr 11, 2021 10:48 pm
The feeds from TPD are not a magic bullet. if you are not already a very successful in play trader this in isolation won’t turn you into one IMO
Regards
Peter
Totally agree
User avatar
wearthefoxhat
Posts: 3205
Joined: Sun Feb 18, 2018 9:55 am

The only experience I've had with TPD has been it's collaboration with Betmover. The way they incorporate their software to intepret TPD, in my view, isn't quite there yet. The way Bet Angel has their IP trader set up @20ms, is. All we need now is BA, to somehow, join the two together.

Can TPD be displayed at a 100% no delay?
Will there be extra costs incurred?
Is it worth being sold as a stand alone product branded by BA?

Not all racecourses are covered yet by TPD, I believe they plan to include more over time.

In-Play trading has increased overall from around 20% to 30% of the pre-off market. (conservative view)

The fact that BA are interested in using TPD is exciting for the future of all IP traders.
PeterLe
Posts: 3715
Joined: Wed Apr 15, 2009 3:19 pm

wearthefoxhat wrote:
Mon Apr 12, 2021 8:06 am
The only experience ....
The thing is, unless you get a licence from betfair, the cost from the feed could run into hundreds a pounds a day (esp as we approach summer). Are there many in play traders willing to pay > £5K a month for the feed!
My honest and truthful advice is to fully exploit BA, there is plenty of low hanging fruit first. I have my own bespoke written bots these days (and Liam's excellent Bflw and Flumine) , but i still use betangel for my test and dev systems and even some live strategies.
Regards
Peter
arbitrage16
Posts: 532
Joined: Tue Feb 14, 2017 7:27 pm

wearthefoxhat wrote:
Mon Apr 12, 2021 8:06 am
The only experience I've had with TPD has been it's collaboration with Betmover. The way they incorporate their software to intepret TPD, in my view, isn't quite there yet. The way Bet Angel has their IP trader set up @20ms, is. All we need now is BA, to somehow, join the two together.

Can TPD be displayed at a 100% no delay?
Will there be extra costs incurred?
Is it worth being sold as a stand alone product branded by BA?

Not all racecourses are covered yet by TPD, I believe they plan to include more over time.

In-Play trading has increased overall from around 20% to 30% of the pre-off market. (conservative view)

The fact that BA are interested in using TPD is exciting for the future of all IP traders.
be interested to hear more about your experiences with betmover, cheers.
User avatar
The Silk Run
Posts: 901
Joined: Mon May 14, 2018 12:53 am
Location: United Kingdom

My experience of TPD is not current but we sourced their software direct from the provider, not from Betfair due to their manipulation of latency. And of course this product is owned by former Betfair employees !!!
It' not that expensive if you specialise in in-play horse racing, and your race reading skills are GOOD, that is key here, the software is only an aid ...
It worked out far cheaper than renting trading boxes, and employing rail runners, that was the motive.
I personally didn't use it much, it was my colleagues. I was using Bet Angel in-play Trader with a LIVE SIS link that was tremendous with fantastic returns.
User avatar
wearthefoxhat
Posts: 3205
Joined: Sun Feb 18, 2018 9:55 am

arbitrage16 wrote:
Mon Apr 12, 2021 10:50 am
wearthefoxhat wrote:
Mon Apr 12, 2021 8:06 am
The only experience I've had with TPD has been it's collaboration with Betmover. The way they incorporate their software to intepret TPD, in my view, isn't quite there yet. The way Bet Angel has their IP trader set up @20ms, is. All we need now is BA, to somehow, join the two together.

Can TPD be displayed at a 100% no delay?
Will there be extra costs incurred?
Is it worth being sold as a stand alone product branded by BA?

Not all racecourses are covered yet by TPD, I believe they plan to include more over time.

In-Play trading has increased overall from around 20% to 30% of the pre-off market. (conservative view)

The fact that BA are interested in using TPD is exciting for the future of all IP traders.
be interested to hear more about your experiences with betmover, cheers.

On balance it's okay. I can see there's been a lot of work put in to help visualise/predict the in-running process.

Betm.png

The pace graph is really good. (top right). It helps quickly identify if there's a slow/fast pace during the race and that alone is worth knowing for front runners/held up types. In the example on their site, that race is faster than average. Ideally they need to be sandwiched somewhere in between. My decision would be on the lines of laying to back the front runner (s) if other factors line up too.

The sticking point IMO, is the bet placement. Bet Angel's IP-trader has nailed it, where you can click left/right of the selection when you need to enter a trade/bet.

For me, it's about getting the whole picture and reacting quickly to the way the IP race develops.

IMO, Bet Angel is a long way ahead with all the other automation than can help a trader manage an exit after the entry, in a fast moving environment.
You do not have the required permissions to view the files attached to this post.
User avatar
jamesedwards
Posts: 2234
Joined: Wed Nov 21, 2018 6:16 pm

I tried this for a while but I found GPS accuracy to be a big problem. IMO it is very useful as a guide to what is happening, but can't be used as fact. Roughly 10-20% of the races I tried it for the software had the wrong horse in front at the finish line, for example.
User avatar
ShaunWhite
Posts: 9731
Joined: Sat Sep 03, 2016 3:42 am

I think it was the case that early on you just got 2d co-ordinates and didn't even have the course layouts mapped to refer to. It seemed pretty maths heavy to get much out of it. eg you needed speed and a bearing otherwise just the 'fastest' horse could be heading across the fields.
jamesg46
Posts: 3769
Joined: Sat Jul 30, 2016 1:05 pm

jamesedwards wrote:
Mon Apr 12, 2021 7:18 pm
I tried this for a while but I found GPS accuracy to be a big problem. IMO it is very useful as a guide to what is happening, but can't be used as fact. Roughly 10-20% of the races I tried it for the software had the wrong horse in front at the finish line, for example.
I'm not an inplay trader but I sort of get the impression (maybe wrongly) from those that are, don't see it as a thing they must have... which speaks volumes to me.

The odds obviously reflect what is happening on course & the inplay trader tool it seems is a much better representation. Feedback of opinion every 20ms is much better than feedback of position every 300ms.

I do like the others features though, such as par time, how each runner is doing in comparison to the fav, the visual of the track (rough position too), how many hurdles/furlongs are left. Could be a seperate add on, like soccer mystic/tennis trader or maybe an addition to the tool that's already there. I don't know what's in the pipeline and if people are that bothered either... I'd imagine inplay traders probably already have such knowledge hard wired in to their brains.
User avatar
ShaunWhite
Posts: 9731
Joined: Sat Sep 03, 2016 3:42 am

I always assumed in-play experts took their cues from how a horse looked and moved rather than raw position/current speed stats. I'm not horsey but even I can tell the difference between one that looks knackered and one that's cruising. The GPS data would be more use to a blind bot than an eyes-on trader but I haven't heard of any that see a big enough gain to pay for it.

GPS feels a bit dated and clumsy, it can't be long before more sophisticated video analysis turns up once the cost of entry comes down. Gait analysis which gets used in the Far East more than face recognition is probably the stepping stone.
User avatar
Derek27
Posts: 23468
Joined: Wed Aug 30, 2017 11:44 am
Location: UK

ShaunWhite wrote:
Mon Apr 12, 2021 9:48 pm
I always assumed in-play experts took their cues from how a horse looked and moved rather than raw position/current speed stats. I'm not horsey but even I can tell the difference between one that looks knackered and one that's cruising. The GPS data would be more use to a blind bot than an eyes-on trader but I haven't heard of any that see a big enough gain to pay for it.

GPS feels a bit dated and clumsy, it can't be long before more sophisticated video analysis turns up once the cost of entry comes down. Gait analysis which gets used in the Far East more than face recognition is probably the stepping stone.
Pace can be quite deceptive to the eye. It's a disadvantage to be too close to an overly strong pace (unless it suits the horses running style) or giving significant ground in a slowly run race. When several horses are pulling and fighting their jockey it gives the impression that the pace is slow but that's not always the case, so current speed/times would be useful if you've done the maths and know what sort of pace you're expecting.

I've never used GPS much before but I went out for a run last week. When I looked at the GPS map on my smartphone I couldn't believe how accurate it was! It claimed that I took a short cut and ran through a few front gardens but other than that the line was bang on the pavement. :)

You won't want to know what my average speed was! :oops:
User avatar
jamesedwards
Posts: 2234
Joined: Wed Nov 21, 2018 6:16 pm

ShaunWhite wrote:
Mon Apr 12, 2021 9:48 pm
I always assumed in-play experts took their cues from how a horse looked and moved rather than raw position/current speed stats. I'm not horsey but even I can tell the difference between one that looks knackered and one that's cruising. The GPS data would be more use to a blind bot than an eyes-on trader but I haven't heard of any that see a big enough gain to pay for it.

GPS feels a bit dated and clumsy, it can't be long before more sophisticated video analysis turns up once the cost of entry comes down. Gait analysis which gets used in the Far East more than face recognition is probably the stepping stone.
One of the main selling points is latency vs video. It is supposed to be less than half a second behind real time which beats normal video hands down. As video technology improved and latency reduces this will of course become less relevant.
Post Reply

Return to “General discussion”