Give me a B Give me an F

[peter]

Racing systems, its systems galore facts and figures and more –  for me there flying like sparks around my tethered head….gathering greatly until its off to bed!

Still on with the show – I’m feeling quite upbeat today. God knows why, but hey you’ve got to take it while its <br>on offer haven’t you. Serotonin bottled, what a great idea. Oh what a bloody night that will be…

You know maybe, just maybe I will give my fellow members an approach to gain a decent profit from the game. This is part one of a lengthy, troublesome…yet delightful escapade.

You may recall my Beaten favourite stint, well its back, but this time it may well just blinken work…

Right follow these rules. There not the easiest and I’m sure there not the greatest…but I feel the system has great potential.

<br>1…Beaten fav, lto and running over Turf or All-Weather.

2…Weight last time out, 9st 7lbs or less with a maximum field of 16 runners when defeated.

3…Today’s race must be a HANDICAP.

4…13 declared runners or less in today’s race.

5…In first two of the Racing Post betting forecast priced 2/1 to 15/2 inclusive.

6…Selection must be in the top 4 of the weights. (all joint weights count as one) for the CURRENT race.

7…Do not back fillies in open races (against all sexes) though consider all other types of handicap races.

8…Only bet from May 1st until December 31st each year.

76 winners from 217 selections showed an average yearly profit of 60 points, from two years study. An efficient 50% lsp % was recorded and although this was 1995 and 1996 data, I feel the worth of this system still holds true.

Best of luck with whatever system you may follow and good luck in running.

[Bricoman]

Hi Peter,

Good luck with this one, any chance of putting up the selections so we can monitor how you are doing and possibly save us all from a lot of hard work.

Regards

Bricoman

[peter]

<br>Yeah I will try and show the selections and maybe even back em myself from the 1st of june. Until then have great end of month party…I’m sure you will….

[stevedvg]

Peter

I tried running this through RSB for the years since 96 (ie 6 years 97-02) but I think either I misunderstood the conditions or they’re not clear as I got different results for 95 and 96.

For those years I got:

114 wins from 369 bets with a LSP % of 27.11%.

Restricting it to all age handicaps (ie 3yo+) it gets closer to your numbers with

80 wins from 237 with a 38.2% profit.

Anyway .. for 97-02

All races: 240 wins from 966 with LSP% 0.80%.

All aged: 126/548 LSP% -5.48%

Looking at the 8 conditions, I would say that:

(1) and (3) define the basic system (backing BF which run in hcaps).

(5), (6) and (7) are the sort of conditions I put into systems on a regular basis (though, personally, I usually use SP rather than FP).

However, I would be interested in how you came up with the other 3 conditions.

Did you come up with them before putting it through RSB or are they just back-fitted to the 2 years data you were using?

Anyway, I hope this helps and you can be pleased that the system (as written) produced a profit over the following 6 years.

Steve  

[stevedvg]

Actually, I made a mistake.

I missed out the "beaten" part of beaten favourites and included all horses that were favourite LTO whether they won or lost (does this mean I created a winning system?).

So .. trying again for 1995 & 1996. I get 67 out of 200 with LSP% 50.08%

(fewer bets than you but I’m not exactly sure where I’m being more restrictive than you)

For 97-02:   111 out of 475 LSP% 4.53%.

Should be able to improve on that using the exchanges, so let’s say 5% profit wouldn’t be an unreasonable target.

Sorry for the errors in my  previous post.

Steve

[peter]

Hey your all right Steve, we all make mistakes mate!

Thanks for the interest, time and energy that you have invested on my behalf.

The initial results did seem impressive and I thought they would simmer down somewhat. Still nice to see an overall lsp there, even though it is marginal. But then again the game is the same, (that rhymed mate!) you have to take what you can get I guess.

I’m unsure how you define backfitting – please tell me…

You know 4% thats pretty damn poor to be frank. A profit can be shown however so we musn’t grumble, must we.

I wonder if you could try and gain a better lsp, with the technology you have at hand, even if it does move away from the bearings of the initial system. Who cares if theres a profit involved.

At present I can’t take all this info in. I will work more tomorrow or today even! its getting a bit late plus the Matrix has just stormed my mind. Theres some awesome stunts in there…Classic!

I would like to clear up any misinterpretations, I’m lucky I was just going to throw the cds out but forgot, so they still could be used if required. Then again their on my file anyway… I tell you what I’ll do…thinking about it. I will relay the criteria exactly as stated and we will see where we can go from there…If your still interested in the plan of course.

If you are we can look forward to gaining plenty of winners!

All the best. Hope to speak to you soon.

[stevedvg]

Peter

I’ll have a go at explaining what back fitting is (this wont be easy for either of us):

I would define it as "refining a system to achieve maximum results in the test data by using rules based on coincidence rather than logic"

I’ll give an example:

Say you are using RSB and have 5 years data to work with.

Now, imagine you have an idea for a system and keep refining the rules of the system until it shows a nice profit over these 5 years.

You could end up with a rule such as:

"Not running at Hamilton, Kempton and Southwell (AW)"

This might have helped profits along in the test period, but could you expect these courses to continue to underperform in the others in the future?

Probably not. As it would be difficult to come up with an explanation as to what exactly these 3 courses have in common that separates them from all the other courses

(anyone with a smart alec reply, please post it – it would be a laugh)  

So that rule would be based on coincidence, not  logic.

Are you following along?

Obviously, the rule I just described is ridiculous and is being used as an exaggeration.

However, the same question applies to "more   believeable" rules

eg

7 furlongs or less

Good to firm or firmer

8 runners or more

I use all of these 3 rules in different systems I have but only when they "make sense" within that system.

Personally, I like to be able to explain every rule in my systems. The explanation might be b*****ks
, but I like to be able to justify them in some way.

That way, I can have an expectation of future profit.

Ideally, I like to, where possible, split my data into 2. I create the system with one half of the data and test it with the other half.

Pretty restrictive if you only have 2 years’ data, though.

Normally, when I create a system, I’m looking for a situation where punters underestimate the chances of a horse.

That way, I can use the idea (together with rules I use regularly such as SP, sex, age, weight etc) to decide which rules are consistent with the idea (and, therefore, can be expected to work in the future) and which are just a reflection of what occured over a short period in the past (and, therefore, may well not be reproduced once I’ve got my money down).    

For example, with beaten favourites, I might think as follows:

"Beaten favourites running in handicaps win more often than other horses. So I’ve got a decent strike rate. That would improve if I use my normal SP range (for hcaps 6/4-15/2). Yep, good strike rate but still a loss."

Then I’d look for situations where punters would be put off beaten favourites (ie where the odds offer value).

Some ideas of the top of my head (not tested) would be

switch from pro jockey to claimer

stepping up in class

AW to turf or vice versa

Lost in maiden LTO and now running in hcap

All of these situations are ones where I could come up with an expanation of why punters might be put off the horse.

Anyway .. I hope this helps and I hope (but I’m in no way certain) that my explanation of backfitting makes some sense to at least some of those reading this.  

If you can turn a profit using some of the ideas I gave you, all the better.

If you decide to go down the systems route with your punting, I would seriously recommend "Winning Without Thinking" by Nick Mordin. It gives a very good exaplanation of how to think about systems.

He’s 20 times better at explaining all this than I am.

Steve

[peter]

Mate your good! I think I got that!

Still interesting reading…and one which defines an area, that I’ve thought about about before… the games a big puzzle – and we’re just sorting out the pieces are’nt we!

Prediction is the name of the game. Trying to find an area that will reep consistent, reliable performance and attain decent long term profits could never be easy…Or could it?

With the technology at are disposal now the worlds our Oyster – as they say!

From the Beaten favourite approach I relayed, the results were indeed backfitted…to gain optimum performance.

The reasoning behind their success could only be speculated without further research – at a price I’m sure.

Too much information can be a bad thing…Don’t you think?

Its difficult to focus and feel committed (for me anyway!) when theres so many pathways to choose and choices to make…

In addition systems can be used in different ways. The automated selection they provide (like Zero Tolerance today) could be used as a pointer, a guide to consider a possible bet. Thats where its gets difficult though! And you have to consider whats the point of any given system if its not automated.

An area I’m thinking about at the moment is reducing my outlay. This is why the games so hard you play with one damn important commodity – Money. I think its best as in most speculation areas to be cautious, prudent and objective. Removing the emotion side of the game can help and is often recommended by the so called experts.

Anyway where were we Beaten favs. Yes thats it. Feel free to forward a BF system – or similar that you think might work well in future.

If your interseted I have a collection of systems that I could relay to you. Then you can tell me what you really think of them!  In regards to the profits that were shown in the past and the likelyhood of profit in the future.

[stevedvg]

Peter

From your reply I think you got that, too.

You make some interesting points about the general approach to using racing systems.

I’ll address a couple from my personal perspective and usage (just my opinion remember)

(1) Betting system qualifiers blindly or using them as pointers.

My own approach is to bet a qualifier unless there’s a good reason not to.

Repertory was an example of one I didn’t bet on (ran Sandown yesterday).

According to my records, it had run in pattern class in the UK 24 times and had never won. At 10 years old, I didn’t see that record being broken.

With older, exposed horses, I look for clear performance patterns or class barriers. With younger or less exposed horses, I typically bet all of them.

The only other thing I use to chuck out a horse is the draw (you still see horses going off at short odds from terrible draws – even on courses where the draw bias is consistent and well-known).

The horses I throw out rarely win (yes – I’ve started thinking about laying systems). I’ve kept records to make sure. That way I’m not kicking myself when one does.

(2) The emotion side of the game

This is what got me into systems as a basic approach. I had spells where I kept losing and they became self-perpetuating as doubts in my ability grew and grew.

I’d end up not betting horses and they’d win. I’d back others and they’d lose.  

In hindsight, my analysis wasn’t that bad but my betting and staking decisions were screwing everything up.

Now things are systemised, the emotional side is much less important.

If there is a lean spell, I know that it’s just one of these things and that my methods are good enough that they’ll turn things around.

So it’s just a case of keep going. Keep making the bets and the law of averages will do the rest.

(obviously takes a lot of faith in my systems)

I’ll decline your invitation to share systerms. I’m very wary of letting people in on my systems.

There’s a picture in my mind of newspaper tipsters scouring the internet for systems. And then undermining the value of them by tipping the qualifiers to their readers.  

I reckon that all newspaper tipsters use systems of one sort or another.  Whether it’s systems like your BF system or keeping ratings and tipping the highest rated horse in certain situations.

After all, there were 28 flat races yesterday. How does someone properly analyse that number of races in a day?

They probably analyse the highest profile races but, for the rest, there has to be a systemised approach being used.

Steve

[peter]

Yeah Steve,

I feel pretty new to the game when we’re talking here…yet I’ve been betting for 3 or 4 years now.

Still knowledge shared can be a useful tool, I’m still learning pal.

I guess I won’t be changing your mind set on sharing systems, which is a shame, though I can well understand the privacy issues. I could give you my e-mail address if you wanted. Anyway up you…

I printed off a fantastic set of pointers a few weeks back about betting – you may have seen it – The Top 10 betting mistakes, from Mathematician racing via the Racing – Index. Its a top site the Index isn’t it.

The factors there may well back up a few theories or help re-affirm beliefs. Who knows…

I’ve always wondered how the form rating compilers mark up a race. I’m guessing the usual array of factors, class, form, weight, fitness and so forth play a part.

I think the top two out there may well be Formcast from the Daily Mail and Postmark in the Post itself.

The two classifiers do very well. What I’ve wondered about recently is when your system highlights a horse and its not top rated by them – then in effect your betting against there systematic criteria that forwards – what – a 27% strike rate from the Postmark at the last count.

Still nobody can be right all the time and on many occasions the winner completely bypasses their top rated selections.

Say there a system out there – in theory – that marks up every race in the Racing Calender. (Adrian Massey ratings gets close to that stipulation)  And their average strike rate is 26% flat 32% jumps – on many occasions your mark up for the race would differ from there’s. In effect you are consistently, though perhaps intermittently betting against another proven system that at some point is going to take your money…

I may be going on a bit here but can you see my point on this one Steve. Its possible a system of ours could clash with a 10 year proven system with a strike rate of 48% and we would be in complete ignorance.

Knowledge is key and a clear mind set helps too.

What do you think Steve…

[stevedvg]

What do I think?

I think it’s irrelevant. :)

And I’ll explain why.

Imagine, there was a system out there that produced a single selection in every single race, had a stike rate of 48% and produced a massive profit. And has done so consistently over the last 10 years.  

(this is a fair reflection of the hypothetical scenario you proposed, right?)

We’ll call this "the Mega system"

Let’s say, I come up wih a system and I’ve tested it against a  period where the Mega system was working.

If my system produces a profit which can be expected to continue into the future, what relevance does the Mega System have to that?      

There will be many times when I would have been better off backing the Mega System qualifier than betting my system’s qualifier. That’s for sure.

There will probably also be times when I back the same horse as the Mega System recommends.

And there will be the majority of races where the Mega system horse loses and sometimes my system will win.

What the Mega System does won’t affect the expected long term profits of my system because any effect already existed in the creation phase.<br> <br>(try, for example putting "favourite", "PM selection" into the above 3 situations instead of "Mega System" and it might become more understandable).

If the system works, it works. It might not work as well as the "Mega System", but it still works.

As for Postmark, I think they are amazing.

I’ve said it before and I’ll say it again, I truly believe that if I was the only person with access to PM ratings, I’d make a bundle just blindly backing their top-rated selection.

However, they are still wrong in the vast majority of races (73% over jumps and 79% on the flat) and they still don’t return a long-term profit as people know how good they are (eg 11.35 LSP loss over the flat 00-02).  

So the winner bypasses their top rated selections the vast majority of the time.

Remember, you don’t have to be perfect in this game, you just need an edge that’s big enough to turn a decent profit.

Peter, but sometimes I get the impression that you’re looking for the Mega System: the holy grail that can be used in every race.

My own approach is to look at the different factors that affect a horse’s chance in a race and find the circumstances where one is over-bet or under-bet to such a degree that it can be exploited to make a profit.

For this reason, I have systems which only apply to Group races, others for 2yo selling races, some for sprint handicaps for older horses and others for middle distance handicaps for 3yos.

A factor which is overbet in a group race may be underbet in a selling race and might be correctly factored into a 3yo handicap.

For me, it’s a case of finding out which factors can be used to find value in each scenario.

And when it comes to profits, as the Postmark losses show, value beats strike rate every time.

What do you think?

Steve

[peter]

I think values gotta beat strike rate, in the importance stakes anyway.

Some sacrifice in sr is often needed to hail the decent priced selections and make a profit in the game.

But for me strike rate is important, not just for profit inducement but for the confidence game…I only back single win you see and I always want as many winners as I can get.

A high win sr gives a high place sr too – well it should do – I say in complete bliss cause I don’t know the facts fully in that case – but presume it to be true…God that was a long sentence!

Well anyway…while gaining a decent strike rate, say 30% theres 70% losers out there and a mass of placed horses that just missed the grade…

This a great oppourtunity, to cut your losses – at least systematically – because the odds are in your favour aren’t they…Just how you do it is open to opinion to question…I have factors that I feel determine a certain bet and if the horse doesn’t have those credentials then its open to question and possible abandonement.

Steve…I’m wondering what specifics you rate as the greatest threats to your selections winning…See I have never considered draw before and personally I feel weight can be discounted in most cases…I place more emphasis on form ratings and fitness – to name just two!

Theres so many God damn factors out there…I find initial thoughts are often the best though but its great when you’ve really done your best with knowing the form and the horse has gone to win in good style…Your might against the bookies…In yer face…bam…and its a winner….Yepeee!

Like when I backed Moon Ballad – still theres some highs and some lows – in my case anyway – You can’t get out for nowt!

I think your doing well there focusing on the specifics i.e different types of races…they certainly have different factors that influence the result.

If only the holy grail were true, I know now that there is no fool-proof system. They all have positives and negatives, dangers and threats.

I like simple straightforward systems that work – and I can see how they work too. It gives me greater confidence…though without the aid of some decent tech spec like RSB I have little chance of backing up a few theories or dispelling a few myths…You can but try and thats what I seem to doing, pretty much always, I think its best not to complacent in this game though and it always keeps the interest going.

I agree with you Postmark is top notch, top draw stuff and I try to consider their ratings when placing my bets…There good to us those Racing Post guys aren’t they!

[stevedvg]

Peter

Just because a system comes up with value win bets doesn’t mean that the same horses are value each way.

Might be, might not… and, frankly, I don’t bet EW and haven’t thought enough about it to have anything worthwhile to say on the matter.

However, I would say that, if you want to make systemic ew bets, you should test the system that way (ie look at the ew return, not the win return).

Otherwise, you’re just guessing (or presuming) which is the opposite of systems thinking.

(but nevertheless legal in most countries … )

As for the criteria I use to throw out systems candidates: I’m looking for things that are under-reflected in the odds.

For example, a horse’s distance and going preferences are used (along with other factors) by most punters while determining a horse’s chances.

So, I reckon these things are included in the odds of the horse.

However, horses have many more requirements (or patterns in their results) which mean that factors such as:

field size<br>class barriers (such as the Repertory example I gave) <br>whether track is straight or turning

are important.

However, they aren’t paid nearly enough attention by the average punter and only weakly reflected in the odds.  

Plenty of horses start at short prices when these (and other) factors suggest they have little chance.

As for the 3 elements you mentioned (fitness, form, weight), I only consider them as follows:

Either

(1) Within the system’s rules (eg "tipped by Postmark" , "carrying more than 9st")

or

(2) As part of a horse’s individual performance patterns. eg there are horses who need a recent race and those who perform best while fresh.

This individual analysis of horses isn’t part of my systems thinking.

It’s just something additional which takes little time and increases strike rates and profits from systems by cutting out a few losers.

It’s a bit like trying to get better than SP prices for the systems bets.

My systems are tested v SP. If I can get 5% better than SP, then my profits (and margin for error) increase.

Every little bit helps.  

I’m curious. How would you define a "simple straightforward system" and how do you "see how they work"?

I’ll explain the second question as it probably isn’t clear.

If I read about a system with a high strike rate, it’s usually easy to see why the strike rate is high.

However, without testing it myself, I can never tell whether a system will produce profits to go with the high strike rate.  

So, I’d have to test everything myself before I could have confidence in it.

Now, there’s ideas I’ve had for systems that can’t be easily tested with RSB (or other databases I have access to ).  

As you don’t have RSB to test systems, I’m curious to know what process you go through which enables you to have confidence in a system without doing your own testing.

Steve

[peter]

Steve I have tested many systems over the last few years. Many have been bought in good faith and really have shown to pointless and unsuccessful, but others have been shown to work and are still able to get a decent profit from the game.

I have tested quite a lot of my own systems in areas like, runners that are top rated by the Daily Mail Formcast and runners within seven days, plus beaten favourites of course.

Perhaps naively I get confidence from my systems as most have results shown with LSP and VSP profit. I have no way of personally verifying the claims at present, or further researching past results, however I have tested the systems at various stages and have whittled down a lot of systems into 6 which I have a distinct amount of confidence in their ability to attain profits in future.

One of my favourite systems is The Triple Crown, you may have heard of it, this focuses on runners returning quickly to action – within 7 days and it continues to do well, with many adaptations now available on the market.

I value the topspeed ratings in the post, especially over the jumps, one of my systems called Class Of The Field focuses on hurdle races to good effect.

I like simple straight forward systems, because they are easy to remember, gosh its so easy to forget the rules at the bookies, plus I prefer to gain an understanding of why they encourage a certain strike rate, e.g they many be dropping in class, they may be clear top rated by Postmark and so forth.

But at the end of my day at least I am taking a chance and probably a greater one than you Steve, as I don’t have the ability to test a mass of results from years gone by…

[Author]

Posts