Focus Ratings is a Uk and Irish Horse Racing ratings service designed to aid you and help you narrow the field so that you can concentrate on the real contenders. Our completely computerised analysis system selects the top three rated horses from each race and sends out the ratings every day at 10:00 a.m.  Wonderful results and an excellent strike rate.  Use Focus Ratings to win more money and make more profits from British and Irish Horse Racing.  The only horse racing system you'll ever need.

Place Betting Strategy – Refined

mad_scientistPlace Betting Strategy

So, I have done some more work on the place betting strategy.

I'm doing this slowly so that the whole process can become a bit educative.

Now, I know that you guys know more about horse racing than I ever will but...

I bet (not that I'm a betting man) that I know more about Focus Ratings (and the way the numbers are created) than most of you.

Yesterday and today I took a look at some other filters.

This has given me a decent strategy but it only gives me 430 selections a year.

That's far less than I wanted; I wanted that "Holy Grail" - 3 bets a day and only 2 losing days a week but,

Never a losing week!

The basic rules of the system were...

1). Place bet (on Betfair) the top three rated horse which has the highest yield (confidence multiplied by forecast starting price) of the top three rated horses where...

2). There are 8 runners or more.

That's it; that's the whole system.

I then updated the system to remove Handicap races; That was a good idea.

I've created a new spreadsheet which lists all qualifying horses/races for 2014.

You may see this at... 2014refined.xls

Now, this spreadsheet just lists all top on yield horses (from the top three rated horses) in non-handicap races for 2014.

You'll see that there are a number of tabs on the bottom of the sheet. The first shows the raw data.

I'll talk you through the other tabs.

Runner Analysis

The first thing that I did was to split the races into Runner Bands which equate to the Betfair payout policy on place bets plus an additional one to show races with 20 or more runners.

Now, at this point I do have to say that there's an inaccuracy in my data (and I'm being honest and admitting it up front.)

You see, the number of runners that I record is the actual number of starters whereas Betfair pay out on the number of declared runners.

Thus, if, in an eight horse race, one horse becomes a non-runner, Betfair pay out for 3 places even though there were only 7 horses in the actual race.

However, this works in our favour as it slightly underestimates the profits; in other words, we make more money that I actually show.

So...

It can be seen that the races with 16 or more runners are profitable, the races with only 5 to 7 runners are unprofitable and the majority of races, those with 8 to 15 runners are about break even.

From this analysis I shall be discarding races with less than 8 runners.

SP Band Analysis

The next thing that I looked at was how the ROI varies with the Forecast SP that I generate at 9am every morning.

It should be borne in mind that my Forecast SP is a made up number and you'll never see it in the wild. That's not to say that it's a bad number (it's more accurate the the Racing Post's forecast SP 82% of the time) but it's a metric that only you (as members) are able to use.

It's also interesting that it refers to what I anticipate the Bookmakers Starting Price will be but I'm still using it as a metric for a Betfair Place betting strategy. However, even if it's the wrong number to use, it's consistently the wrong number to use and there is logic for using it here.

From this analysis, I determined that the following price bands were profitable...

Runners Price Band Races 16 to 19 12.0 to 40.0 33 16 to 19 4.0 to 8.0 23 20 Plus 12.0 to 22.00 11 20 Plus 2.0 to 8.0 9 8 to 15 10.0 to 18.0 268 8 to 15 25.0 Plus 78

During 2014 there 422 qualifying races.

Yield Band Analysis

I then repeated the whole exercise but I use the Yield Band as a metric.

This showed some interesting results but...

I decided that as Yield is a combination of 2 unique metrics (Yield is my Bookmakers Forecast Price multiplied by Confidence) and, as such, it was better to stick with just one metric.

Sometimes you can go down the wrong alley but it's worth doing so anyway.

LTO Analysis

One of the members suggested that I look at the Last Time Out figure and, although he is making profits by using this metric, I couldn't see any real advantage to further complicating a system that seems to work quite well on its own.

RC0 Analysis

Finally, I looked at RC0.

If you want to know what RC0 is, it's probably best to point you to the following posts...

My first post... http://focusratings.com/news/eureka/

My follow up post... http://focusratings.com/news/eureka-revisited/

And a system that I built based upon RC0... http://focusratings.com/strategies/free-system-rc0-makes-it-work/

It's basically a different way of looking at the ratings; some text from one of the posts explains it as...

RC0 – Good Races/Bad Races

I have managed to come up with three different levels of this Good Race/Bad Race thing.

These are expressed as RC0.

RC can be 1, 2 or 3.

RC0 = 1

As an example…

Let’s take an 10 horse race where it’s pretty obvious that the winner is going to be one of 6 of those horses.

I know that because those 6 horses are the ones that I’ve found easier to rate;

They have form,

They have had success (going wins, class places and other stuff like that.)

Because of this, the 10 horse race now becomes a 6 horse race.

And it’s far easier to pick a winner from a 6 horse race than from an 10 horse race.

This will give an RC0 of 1 (the best.)

RC0 = 2

Taking the above race…

If I can only rate 4 horses from that race, you might think that I’ve turned a 10 horse race into a 4 horse race and…

To a certain extent, I have.

However, there are 6 unrated horses and those are wild cards.

And, as such, a danger.

For races where I can only rate less than 55% of all runners…

I set RC0 to 2 (the second best.)

RC0 = 3

And, strangely…

Taking the above race…

If I can rate more than 8 horses from that race, you might think that I’ve got it right but….

To a certain extent, I have…

But…

I’ve actually turned a 6 horse race into a 8 horse race and…

For races where I can rate more than 88% of all runners…

I set RC0 to 3 (the third best.)

Statistics

I’ve just done a quick query on the Builder and the results are…

Top Rated Horse with any RC0 – Strike Rate = 26.08%

Top Rated Horse with an RC0 of 1 (the best) – Strike Rate = 31.68%

Top Rated Horse with an RC0 of 2 (2nd best) – Strike Rate = 29.31%

Top Rated Horse with an RC0 of 3 (3rd best) – Strike Rate = 23.64%

Overview

RC0=1 is where I can rate between 88% and 55% of all horses in a race.

RC0=2 is when I can rate less than 55% of all horses in a race (but more than 39%.)

RC0=3 is when I can rate more than 88% of all horses in a race.

By looking at the RC0 values I've decided that all races with 16 or more qualify (taking the other filters into account) but, for races with only 8 to 15 runners), I'll only include races with an RC0 of 1.

Summary

So, from my analysis, I have come up with the following strategy (please note that this is a Strategy and not a System - I'll explain that in my conclusion)...

Place bet on Betfair (back) the...

1). Top 3 rated horse that is top on Yield (of the top three rated horses)

2). In non-handicap races

3). Where there are 8 or more runners

4). Where the horse has a Forecast Starting Price (my unique forecast starting price) as follows...

Runners Price Band Races
16 to 19 12.0 to 40.0 33
16 to 19 4.0 to 8.0 23
20 Plus 12.0 to 22.00 11
20 Plus 2.0 to 8.0 9
8 to 15 10.0 to 18.0 268
8 to 15 25.0 Plus 78

5). If the race has between 8 and 15 runners then it must have a RC0 of 1.

That's it; 5 simple rules. I'll produce a PDF to show those rules and I'll also consider creating a daily PDF with any relevant selections.

So far today (from the first page of the ratings - 12th June 2015) there has only been one selection - it was a horse called Kodimoor who came 3rd in the 14:00 at York. Betfair paid 11.49. So, that's about 1,000% profit on the day - perhaps a "Stop at an Obscene Profit" rule is required.

Conclusion

Now, this is just my interpretation of my view of the numbers. You guys know far more about horse racing than I ever will; you may be able to interpret the numbers better - you may be able to come up with more profitable strategies.

I now need to do the following....

1). I need to revisit the 2014 data to show the results with my 5 strategy rules.

2). I need to apply those rules to 2015 data to see where we're going on this.

3). I need to write a post about monthly variances for systems.

This is a real issue with me.

It constantly worries and annoys me when I see people creating systems which assume that a Flat Turf race is the same in April as it is in September (and similar.)

Or that a Novice Hurdle in November has any relationship to a Novice Hurdle in March - they're totally different races (even at the same course, over the same going and with the same horses.)

And, what about a Novice Hurdle in June?

I have one member who has a really good little system but she actually has 8 versions of it for different times of the year - thus, she has 8 systems instead of one all encompassing one.

Hey, I'm not being altruistic here; if you come up with a system that is profitable over the year but then find that you lose money 3 months of the year, you may cancel your subscription and then I'll be struggling to feed my insatiable appetite for Red Wine and Smelly Cheese!

4). I really do need to write a post about the difference between Systems and Strategies.

Now, I know that I've done a couple of Captain Sensible posts (http://focusratings.com/articles/captain-sensible/ and http://focusratings.com/strategies/selection-filtering/) but I really do want to try and drive home the point.

My feeling is that you can have a System where you blindly back the selections or...

You can have a Strategy based upon a system but where you apply some human logic to the selections.

In other words, you can have a great and profitable system but, it becomes a strategy if you look at the selections and say to yourself...

"Actually, my system recommended that I back Dennis the Donkey today but, I've just had a look at him and I've noticed that he's only got 3 legs and hasn't won a race since 1892 - perhaps I ought to give him a miss."

That's the difference between a system and a strategy.

I'll probably also waffle on about System audits (on a monthly basis) - systems go wrong, things change, you need to keep on top of that.

My final word....

I've come up with something that looks good to me; it's not quite what I wanted (3 bets a day and all that) but it does look interesting.

I'll be producing the new spreadsheets tomorrow but...

Please don't think that this is the final word - you guys may well be able to come up with something better - I'd love to hear if you do!

Anyway, it's only 1 hours and 32 minutes until Wine O'Clock (not that I'm counting) and I have to get prepared for that.

These things are important, you know, especially if you live in France!

I hope that there will be a Morning News tomorrow - I've done some updates that I'm dying to tell you about.

Just a reminder...

The new spreadsheet is available at... 2014refined.xls

Anyway, until then,

As always...

My kindest regards

keith-eckstein1

Leave a Reply