Opinions on TPD - is it worthwhile?

[Euler]

To put that into context, the original pricing was £35 per meeting, so it's really cheap compared to that. The commercial license fees are way higher as well.

Just choose the lowest priced subscription and have a play.

[Dave64]

There still seems to be very little information or examples available generally about TPD and the ways in that it can be used. I've been dabbling with automated IR on the UK horseracing markets over the past month (having previously spent many years trading IR profitably back in the day) and to the naked eye it seems to be tied-down pretty tightly with regards to price moves and the eventual result. Has it become a battle of who has the sharpest analytical mind and/or automated routines? I'm interested to see what it can do but my question is how long does it take to get to grips with the functionality and how many have tried and succeeded implementing automated strategies versus how many have tried and failed to create anything vaguely profitable?

[tico]

There still seems to be very little information or examples available generally about TPD and the ways in that it can be used. I've been dabbling with automated IR on the UK horseracing markets over the past month (having previously spent many years trading IR profitably back in the day) and to the naked eye it seems to be tied-down pretty tightly with regards to price moves and the eventual result. Has it become a battle of who has the sharpest analytical mind and/or automated routines? I'm interested to see what it can do but my question is how long does it take to get to grips with the functionality and how many have tried and succeeded implementing automated strategies versus how many have tried and failed to create anything vaguely profitable?

That's an interesting question .I tried it for a few months and quite honestly found it too distracting,and on many occasions unreliable.I believe the unreliability issues have all been cleared up now but for me,I suppose I am set in my ways of race-reading .What made me give up with it ,was that on several occasions all the indications according to TPD where that horse A was going to win whilst my visual analysis was saying horse B was going to win...I always went on my own observations and won out over TPD.
The thing is sometimes TPD can be stunningly accurate (gobsmackingly so!!) ,but the problem here is knowing when.I think the simple reality here is despite the hype ,TPD will get it wrong more times than it get's it right...otherwise we would all be millionaires.As an aid I can see it's advantages and perhaps long term perhaps it could be beneficial once it's peculiarities are mastered.But like you I question the lack of info available and I cant recall seeing any major set of videos winning using TPD.I think Betmover have some but that was just a one off and I don't know if they have put more up.
As regards automation ,that's a tricky one ,I don't like automation ,but there are some real experts on here who do...but then you are up against the old problem...who will release a winning automated strategy .
What would be a fascinating exercise is putting somebody up who is totally unfamiliar with traditional form analysis and ask them just to use TPD.Would they win out over the traditionalists??
Regards
Tico

[Euler]

IMO, people who say it's inaccurate aren't using it correctly.

It's simply the most stunning addition to racing that has ever happened. But of course nobody who uses it is going to come onto a forum and just say, "Here you go, here's my edge."

Just think it through. Look at what you're getting and put it to use.

For example, lots of people have pushed the lay the field strategy for many years in play. Now you've got actual data that you can use against that strategy.

[jamesedwards]

Most people expect TPD to be a latency advantage and then get disappointed with the accuracy. Latency is only in-line with Betfair RMG streams and the drones, and positional data up to a half a length out, so using it to smash 1.01 on the winner is not going to work.

TPD is a data advantage. Multiple attributes for every horse updating 5 times a second means you pull out some crazy value as the race progresses towards the final stages. Personally I focus on manual in-play trading and mostly use TPD to help me with this. But I also have TPD automations running in the background that make me low 5 figures per year on their own. It picked me out a 999/1 winner a few weeks ago. viewtopic.php?p=390263#p390263

[tico]

IMO, people who say it's inaccurate aren't using it correctly.

It's simply the most stunning addition to racing that has ever happened. But of course nobody who uses it is going to come onto a forum and just say, "Here you go, here's my edge."

Just think it through. Look at what you're getting and put it to use.

For example, lots of people have pushed the lay the field strategy for many years in play. Now you've got actual data that you can use against that strategy.

Hmm
To say that this is the most stunning addition to racing that has ever happened is IMHO a wild exaggeration .It is an addition and an aid , but astronomically significant it certainly isn't .
I think your final observation about laying the field more or less proves my point ,there are 101 ways of laying the field prior to the running of a race.If prices go in your favour you will win ,it is a fixed and controllable scenario ,but once in running !!! I daresay the algorithm doesn't exist that could cope with that .

[ShaunWhite]

,but once in running !!! I daresay the algorithm doesn't exist that could cope with that .

So you're saying that knowing their relative positions and relative speeds isn't useful? Without that all you have is price, and there's no way to know if a price move relates to what's actually happening on course, or if it's just a punter cutting losses or others overreacting to a commentary error. Trading IS about the difference between prices and actual liklihood. Everyone knows the first, TPD is a big step towards knowing the second. Not ideal, but better than decades of nothing.

[Dave64]

Certainly a most interesting debate and I value all of your feedback. Obviously as a potential new user I'd like to hit the ground running when it comes to trialling the software and it's integration with BA. Where is the best source of info (as well as on here) with regards to it's usage, amount and types of data that it provides etc? Additionally do you guys who use it spend any time studying horse form or running styles or can it be equally effective used as a standalone piece of software helping to guide you in the right direction? This is particularly important to me as ideally I don't want to spend hours each day with my nose in the Sporting Life or watching videos as it feels as if I've already spent half of my life doing that

[Euler]

IMO, people who say it's inaccurate aren't using it correctly.

It's simply the most stunning addition to racing that has ever happened. But of course nobody who uses it is going to come onto a forum and just say, "Here you go, here's my edge."

Just think it through. Look at what you're getting and put it to use.

For example, lots of people have pushed the lay the field strategy for many years in play. Now you've got actual data that you can use against that strategy.

Hmm
To say that this is the most stunning addition to racing that has ever happened is IMHO a wild exaggeration .It is an addition and an aid , but astronomically significant it certainly isn't .
I think your final observation about laying the field more or less proves my point ,there are 101 ways of laying the field prior to the running of a race.If prices go in your favour you will win ,it is a fixed and controllable scenario ,but once in running !!! I daresay the algorithm doesn't exist that could cope with that .

It is singularly the most important thing that's happened to racing. It's completely transformative.

On your last paragraph you're saying that not knowing where the horses are, how fast they're running, and whether they're gaining or losing ground doesn't help you perform that strategy better?

[tico]

IMO, people who say it's inaccurate aren't using it correctly.

It's simply the most stunning addition to racing that has ever happened. But of course nobody who uses it is going to come onto a forum and just say, "Here you go, here's my edge."

Just think it through. Look at what you're getting and put it to use.

For example, lots of people have pushed the lay the field strategy for many years in play. Now you've got actual data that you can use against that strategy.

Hmm
To say that this is the most stunning addition to racing that has ever happened is IMHO a wild exaggeration .It is an addition and an aid , but astronomically significant it certainly isn't .
I think your final observation about laying the field more or less proves my point ,there are 101 ways of laying the field prior to the running of a race.If prices go in your favour you will win ,it is a fixed and controllable scenario ,but once in running !!! I daresay the algorithm doesn't exist that could cope with that .

It is singularly the most important thing that's happened to racing. It's completely transformative.

On your last paragraph you're saying that not knowing where the horses are, how fast they're running, and whether they're gaining or losing ground doesn't help you perform that strategy better?

each to his own I suppose ,but I would rate Oisin Murphy as far more useful

[Dave64]

I've trialled TPD over the last few weeks and have monitored/traded 150 races, a small sample I know, and here are my findings so far:

• The raw data (and there's quite a bit of it) needs to be taken with a big pinch of salt! For instance, the 'Position in the race' metric is often inaccurate, even when horses are a few metres apart (irrespective of the stage of the race). I don't use the 'Distance behind leader' information but I'd guess that this could also be inaccurate at times as well. I was expecting this data to be much more accurate tbh.

• The metrics such as CE, VE, VF, etc in their raw form need a lot of understanding and to be used effectively need to be analysed either alongside what you are actually seeing in the race itself or by filtering/enhancing through automation, especially at the start of the race.

• The data itself has many anomalies caused by various factors (some of which can be baffling to understand).

• The information available about how to use TPD is scarce. Now I'd expect there to be videos all over YouTube of people showing how easy it is to win £200 per day sitting at home (and selling a course or instructional videos) but either it's not happening or the people who are doing it (if any) are keeping things close to their chest. There's plenty of people showing how to trade on BF but none mentioning anything about TPD?

Despite this I'm about to sign up for another 450 races I'm forever the optimist and can see the potential of using TPD as an aid to both manual and automated trading.

Manual trading: the information presented can be overwhelming (especially in the shorter races). Trying to combine what your eyes are telling you along with the TPD data and then making a bet/trade requires a lot of brain power and can also be quite stressful (which is something to consider health-wise in the long-term).

Automated trading: at this point in time I think that this is the way to go. I think that the speed advantage and distancing from emotion offered by automation can outweigh the occasional not so accurate data from TPD. Hopefully with further use and analysis I'll be able to filter out any erroneous info in order to enhance the data that TPD is providing.

I'm planning to have my head buried in BF/BetAngel/TPD etc over the next few days and will report back with my findings ]

PS Any feedback or observations are more than welcome - Dave

[lotora]

Fair play, Dave, that sounds like a very sensible assessment after 150 races. I’d treat TPD as another useful input rather than gospel, especially the position data. James (jamesedwards), who is clearly a very successful trader, makes most of his money from manual in-play trading, whereas I do all my TPD trading automatically. For automation, I’d keep the rules simple at first, log every trigger, and add filters one at a time, so you can see which metrics genuinely add value and which are mostly noise. I’ll be interested to hear what you find after the next 450 races.

[Dave64]

Thank you Lotora

It's always helpful to get feedback and James has also provided some useful advice, so thank you. I already have a shed load of info to work on and I'm currently trawling through the data as well as trading for small stakes using my current V1.0 setup

Here's hoping that you all have a great and profitable very busy Saturday

[Euler]

I think it's invaluable. I have expressed that opinion further up the thread, but I don't use it for specific data, and therefore I'm not reliant upon it being absolutely 100% accurate, which it can't be anyway. I think using it to get a better understanding of how a race is run has been completely invaluable for me.

I think the problem with YouTube videos on TPD is that a lot of the videos you see on YouTube are just from people who are trying to make a fast buck one way or another. They just don't invest in things. They're trying to simplify things too much, and it doesn't matter to them anyway because they're probably not doing it for real anyway.

To use something like TPD properly, you have to invest a bit of time and effort. I think that falls very firmly in the remit of Bet Angel users, but not the mass market.

[tico]

I think it's invaluable. I have expressed that opinion further up the thread, but I don't use it for specific data, and therefore I'm not reliant upon it being absolutely 100% accurate, which it can't be anyway. I think using it to get a better understanding of how a race is run has been completely invaluable for me.

I think the problem with YouTube videos on TPD is that a lot of the videos you see on YouTube are just from people who are trying to make a fast buck one way or another. They just don't invest in things. They're trying to simplify things too much, and it doesn't matter to them anyway because they're probably not doing it for real anyway.

To use something like TPD properly, you have to invest a bit of time and effort. I think that falls very firmly in the remit of Bet Angel users, but not the mass market.

Aha......what about you doing a vid on a days racing based on TPD data .As you have pointed out lots of people on youtube probably don't do it for a living,whereas you are superdooperhenrycooper when it comes to trading....go on my main man...go for it
kind regards
Tico