Tipsters are a bit like Ratings. They are someone's Opinion. I know of one tipster in NSW, that tips, the first 4 in the early market every day and get's paid for it. If the fav has a good day, he has a good day. You are right, Tipsters are hot and cold, just like horses, how does a horse, beaten 32lens in 3 starts, win the next start at 100/1?

Ratings are the same, someone's Opinion. If you take a set of rules and give each rule a value, you come up with the best horse, 100 rated. If it is the best horse, it will be the fav. So why not just give the fav 100 points, that's my opinion! Saves a lot of work.

As we all know, backing fav's will send you broke. Bookies and Corporations have already drop the prices on these horses, way below the chance factor.

So how do we win, well it is your own opinion in the end. The algorithm you come up with, is a start. Allocating point to varies Rules, to start your algorithm, could look like this, Market 50 points, 50 to the fav, then 40,30, 20,10 and 5 for the rest. Win in this Class, 30, Place in this Class, 15 points.
And we continue with a few other Rules for Points. Now we have our Opinion or Ratings.

That's fine BUT! There is not one Computerized Algorithm that will stand the test of time, In My Opinion. I have been doing ratings for 25 years, millions of test and they all fail the test of time, they go great for a Month or two, then 0/7 the next day, 1/8 the day after, WHY! Horses are no machines, they can be moody, for bid, the Jockey has a bad day, he let's the horse go back to last and wait's for gaps on the rail, that never appear and so on.

I do my for analysis with all these area's in mind. Look at the Jockey, is he hot or not? Is the horse is going up in distance and down in class? Blinkers First Time? These and other form factor's I use to determine my Opinion and value I'm getting for that Opinion, makes up my mind, whether to wager or not. Not Tipsters.

I am currently undertaking a linear regression model, but need to add some factors, most importantly up or down in class or same class. With 23 and a bit years of data, I think I can definitely create something with a long term profit for backing or loss for laying. Blinkers first time, gelded etc is not something I have a record of, but still, I think I can come up with something using pure mathematics. In form jockeys and trainers win, but they also lose and that's where I think the profit is. For Example September Run in the William Reid was way over the odds at $17.00, she'd finished 3rd in the Group 1 Newmarket at Flemington beaten only 0.6 lengths and was going over the same distance at the Valley from a Good4 to a Good3.

__________________
RaceCensus - powerful system testing software.
Now with over 407,000 Metropolitan, Provincial and Country races!
http://www.propun.com.au/horse_raci...ng_systems.html
*RaceCensus now updated to 30/09/2024
Video overview of RaceCensus here:
http://www.youtube.com/watch?v=W821YP_b0Pg

I have no doubt you could CP, You would have to pick your races though.

No Restricted races, 4 year old maiden races, Class 1 or BM races <58.

Your data would show that, as would 4 or more first up runners in the race, they are hard races to gauge their fitness.

That's why I say, Tipsters are a no go. They are trying to select the winner in every race. Computerized Algorithms are the same. AAP, Sky, Racenet, the list goes on. Computerized Algorithms are a one touch button for ease of application. Someone in the office, comes up with these Algorithms, then the youngest staff member, presses the button, then format's the selection results.

There is no way you can select enough winners, in all races to be profitable, in an algorithm.

Yes CP, laying those fav in the poor races and betting on the good races, you could make a profit long term, I agree but if we are talking about Tipsters, they mostly publish the winning hopes in every race.

I cant be too hard on tipsters as they don't know what the odds will be at race time so I they can only tip what is going to be the most likely winner and that is not surprisingly the favourite or second favourite.

I wouldn't use them in a systematic method though as they are not a controllable variable. You don't know when they are really putting in the work vs just going through the motions (or copying from somewhere else).

A rating is consistent though (providing you can talk to the person who creates it and they let you know when and if it changes). But there are still times that these can change and you are not told but its less changes then a tipsters picks.

And that's fair enough too UselessBetter. Tipsters usually do their tipping day's before the race, in some cases. The changing track conditions, can catch them out too.

Yes, they mostly tip fav's and that makes sense, as that would be the horse, with the best form. We know, that fav's win on average, around 34%. So, for this example, we are going to look for an average of 3 winners on a 8 race meeting. Let's call this system, "True Fav's". Say we take the Tipsters top 3 selections, we then do an analysis on them 3 only.

We are looking for a Top Jockey, a fit horse in form and a horse that has come in for good support, $4.00 into $2.80 for example. No odds on. Now we have a "True Fav".

A horse supported in the market, good form and a leading Jockey on board.
Hopefully, we get 3 bets at that meeting, all over $2.00, if not, the next meeting we may get 4 Selections?

Yes, this example is, at the pointy end of the market, but it can be a way of reducing the work load by, letting the Tipster and Market find the Qualifiers, then, you decide the best Jockey out of those Qualifiers and look at some form to confirm it's Selection. Weeding out the rubbish can and will increase your SR, to a level that makes this type of system profitable.

Another away of using Tipsters??

Shaun, is it possible to scrape data from Sky's new tipsters site, in Python?

Or does anyone know if it can be done in Python?

I have a couple of theories I'm testing, but it is very time consuming. Sat, 15 meetings, takes forever manually.

Any help would be appreciated.

Unstable

This will get you the meetings, just adjust date
https://api.beta.tab.com.au/v1/tab-...urisdiction=VIC

Copy the whole page and paste in to this website to make it easier to read.
https://jsonformatter.curiousconcept.com/

On the meetings page you will find other links like.

https://api.beta.tab.com.au/v1/tab-...urisdiction=VIC

And on that page you will find links to form.

https://api.beta.tab.com.au/v1/tab-...urisdiction=VIC

You just need to play with the dates, all the sky info is there like what they show in the panels, not sure about tipsters, i haven't looked.

__________________
One Drive

"If the corporates are treating you poorly , just go elsewhere."
"If they need you , they will soon find out."
"If you need them , you will soon find out." --moeee
_______________________________________________

Brilliant, thanks Shaun

Unstable