Bookmakers, Big Data and The Future

[LinusP]

So its £150,000 for a year's subscription...

[ShaunWhite]

Faster information and 'big data' only benefits those with the resourses, intellect and capital to expolit it and therefore it can only be bad news for the vast majority of punters who aren't in that position. They'll be lead to believe this info is giving them an egde when it's actually benefitting their adverseries more. People don't mind losing a few quid on the races for a bit of entertainment, but if they feel ripped off they'll stay away.

Absolutely this is almost certain to happen.

+1 My thoughts too

I often feel like the Grinch on here, just a miserable sceptical contrarian that always manages to see the downside.

Mr & Mrs Average don't like complicated...fact. They don't even bother with betfair and this just makes racing look even more complicated and nerdy to the uninitiated who are the marketeers target.

The sucessful betting business model is well proven, either big returns for small stakes (eg the lottery or scratch cards) or give them something to care about (eg Grand National or England in the world cup). Winning £2 for a £10 stake in running because you've calculated your horse gains 3mph over fences going anticlockwise, won't set the industry alight.

[Dallas]

I often feel like the Grinch on here

Just as long as you don't try and steal Christmas - its my favourite time of year

[LeTiss]

Shaun White, speaking sense.........whatever next

[ShaunWhite]

Shaun White, speaking sense.........whatever next

It's just the law of averages mate, spout off enough and occasionally some of it makes sense. I'm a deft exponent of the infinate monkey method.

[Euler]

My reading of it is that this data is primarily to make the sport a bit more interesting in the same way that data does for Tennis and footy. Though I'm sure applications will be found for betting, trading and training purposes.

I'd be really keen to get my hands on it!

[mathsnerd]

If this data is made available for everyone it could be very exciting, although I doubt it is going to be free for those wanting it without a delay.

I have already started looking into positioning using GPS and it isn’t easy! Datum’s, ellipsoids and apparently the earth isn’t actually round...

I am quite new to the forum (been a lurker for a while) but have recently been using GPS data to try and get an edge in-play. I too have been through the trials of turning that GPS data into a 2D graphic to display position but I have found many inaccuracies in this stuff. Generally it works well but I have come across many races where "my" result from GPS is nowhere near accurate. Could be a bug in my code as that can never be ruled out but I would be really interested to hear about other people's experiences with using the data from the point of view of its accuracy.

[Euler]

Sporting lot - Sports trading workstation, used to use Turtrax data many years ago.

I found a video on it, dated 2006: -

https://www.youtube.com/watch?v=kG7WMn3k_m4

[mathsnerd]

Sporting lot - Sports trading workstation, used to use Turtrax data many years ago.

I found a video on it, dated 2006: -

https://www.youtube.com/watch?v=kG7WMn3k_m4

That is very interesting. I had no idea this tech was around in racing as far back as 2006. I would imagine that it would have significantly improved since then but some of the odd data I get sometimes makes you wonder. I can understand how data can be inaccurate but from what I have seen the data seems at times to exhibit systematic or calibration errors. I would assume (perhaps naively) that the errors in x and y coordinates would be independent and follow a bivariate Gaussian distribution (so a stationary runner might exhibit radially symmetric jitter about the true position) but this does not seem to be the case.
As stated though, you can never be certain that code does not contain bugs so would be very interested in the thoughts and findings of other people using this data in regard to accuracy.

[ShaunWhite]

Sporting lot - Sports trading workstation, used to use Turtrax data many years ago.

I found a video on it, dated 2006: -

That's a blast from the past, STW was the first trading software I used.

[mathsnerd]

I have attached an image of one of my own automated applications to illustrate what I have been trying to talk about. Right at the close the eventual winner is seen to be in 4th place. This was a tight finish for sure but at the point at which the runner that GPS thinks has won crossed the line, it locates the actual winner as 3.8 metres behind. I have omitted this on the display as there are some metrics that I don't want to be visible. This is just one example of a systematic error in positions that I have seen from the GPS. Is anyone else actually using GPS? Or has anyone used it and abandoned it due to the kind of errors like the one presented here? The race in question was on 31st January 2020

[iambic_pentameter]

Gentle bump from me for this thread.

Not especially tech-minded but would be interesting to see this discussion develop as GPS could potentially be a game-changer?

Iambic

[mathsnerd]

Gentle bump from me for this thread.

Not especially tech-minded but would be interesting to see this discussion develop as GPS could potentially be a game-changer?

Iambic

I was really hoping to engage with some GPS players on here as I have had zero luck finding anyone through other avenues. I am starting to think that the first rule of GPS club is "you don't talk about GPS club". Alternatively the errors like the one I mention above has simply put people off using it.

[iambic_pentameter]

Be interesting to see if people like Mega_Rain, Spreadbetting or PeterLe have any input on this.

Iambic

[xitian]

There are definitely people making good use of the GPS API. You should join the betfairlightweight Slack group if you want to get in touch with some of the people. Just do a google for betfairlightweight and you’ll find Liam’s github and link to the group.

Re: inaccuracy of GPS data, I’ve spoken to my account manager about it before and he basically said they’re constantly trying to improve the technology, but essentially the data is provided as is, and users are expected to factor in margins to account for data inaccuracies. There are people doing well from using the feed, so my advice would be to stay away from betting at the finish line, or at least manage your risk appropriately.