New Bloomberg Article on Bill Benter

[Ted Craven]

'The Gambler who Cracked the Horse Racing Code'

Bill Benter's exploits as a high-level pro-bettor in Hong Kong and North America - then and now.

https://www.bloomberg.com/news/features/2018-05-03/the-gambler-who-cracked-the-horse-racing-code

Ted

[Rverge]

(Pittsburgh boy a yinzer)

[gandalf380]

Fascinating article about solving horse racing in Hong Kong using algorithms. It was also interesting to see William Ziemba quoted. He hung out with the RDSS crowd at Saratoga last year.

[Rverge]

i did/do remember Bill Bender. some videos on you tube i viewed a while ago.

[Ted Craven]

Fascinating article about solving horse racing in Hong Kong using algorithms. It was also interesting to see William Ziemba quoted. He hung out with the RDSS crowd at Saratoga last year.

Also with Bill V and me at Breeders' Cup last year. I also spent a week with him in Vancouver earlier this year.

Ted

[Rverge]

Also with Bill V and me at Breeders' Cup last year. I also spent a week with him in Vancouver earlier this year.

Ted
:cool:

[mick]

Here's the report mentioned.

[atkinsrr]

great article Ted...seems like I read it a few years ago but don't remember where I found it....

[wigopher]

Thanks mick for posting that report. It is interesting as I have been dabbling into the world of R and Python for my work. I am at the absolute beginners level, but I've been trying to think what data I would need to put into this in order to project out winning characteristics.

Not sure if anyone has tried to play around with the outputs of RDSS to do some modeling with the data? Just curious to hear what they found to be the results.

[Ted Craven]

Much work has been done and posted on modeling in RDSS. Check out the video and discussions on exporting the full RDSS factor set (i.e. everything on all Analysis screens) to Excel files for use in the Excel MasterTemplate4.


Bill V has done some good reporting on factor hit rates over thousand of races and posted some of that in the RDSS 2.1 Forum (not yet ROI by factor ranks, but that's coming soon I think).


Bill Lyster has posted some Excel workups of his from a few years ago with ROI and hit rates from some of the high-level factors (also in the RDSS 2.1 Forum).


Mick has done some modeling work in Access - not sure where that might be posted.



I'm sure you could do a lot of work using this raw (and filtered) data in R and Access, with Python manipulation if needed, though the data is available by race already, awaiting your study!


Good luck!


Ted

[mick]

Thanks mick for posting that report. It is interesting as I have been dabbling into the world of R and Python for my work. I am at the absolute beginners level, but I've been trying to think what data I would need to put into this in order to project out winning characteristics.

Not sure if anyone has tried to play around with the outputs of RDSS to do some modeling with the data? Just curious to hear what they found to be the results.


You're welcome. It isn't an easy read but it's the basis for what many of the big betting syndicates (aka "whales") are doing. This is an oversimplification but it's the statistical analysis of factors and assigning them "weight" depending on their power in predicting the outcome of races. BTW that's exactly what Ted is doing with RDSS, especially the Rx factors.

Bill V, Bill Lyster, Lt1, Mitch44, rmath, myself and others have done modeling with the RDSS data and you'll find the results sprinkled throughout Pace and Cap.

If you're thinking about building a RDSS database for research purposes, please let me make two suggestions. First, I would wait until Ted's new release, which should be available later this month. It may have some additional information that you want to capture.

And second, after you've downloaded the mutuel prices and exported the race to Excel, copy the data for ALL the horses in the race to the Master Template, not just the winner's data (or the place horse's or the show horse's). There's a tab titled "All" and that's what it's for. If you don't have all of the data for a race, you won't be able to determine the Impact Value for a specific factor or a combination of factors. I learned this the hard way, after spending countless hours building databases for win and place horses only.

After you've collected several hundred races, you can export the data to Access or OpenOffice Base and run queries. Or you can use R or Python to build a "front end" to do that.

Please feel free to send me a PM and we can discuss this is more detail.