Back Fitting Of Horse Racing Systems

The term back fitting when applied to horse racing systems
can be summarised as over engineering selection criteria
putting all logic and common sense aside in order to produce
what on paper looks to be a profitable historic approach.

The big danger in doing so is that non sensible criteria are likely to
lead to future losses even if researched history shows positive results.

Who Can Fall Prey To Back Fitting?

You perhaps

If you research your own systems it is important to try and think logically.
Are there good reasons why your selection criteria are likely to produce good results?
Be careful not to use silly criteria just because by random fluke
the monkeys with type writers of historical results scatter happen to show some positive results.

If you buy systems from a system vendor tread carefully too.
Remember the seller is often primarily concerned with making a quick buck from you.
Often what he cares about is putting up a big sexy past profits pound sign in front of your nose.
They want your cash today and care little about whether what they have sold you with a pile of hype
has significant chance of doing well in the future.

For more reading around this general concept of back fitting see
the article below.

==> Four Simple Horse Racing Systems

Long Distance Handicap Chases by David Renham

In this article I am going to look at long distance handicap chases to see if we can improve our chances of a) picking the winner, but more importantly b) gaining enough value on the winners to make a profit. I have decided that any race 3m2f or longer counts as a long distance chase.

However, what I am not going to do is back track through data and manipulate the stats to some degree. Indeed far too many people do this – they back-fit the results to create a set of rules or a system, that ends up producing a profit. Unfortunately 99 times out of 100 these ‘positive’ findings will not be reflected in future results. My plan to look at a variety of areas in an attempt to find potential starting points – basically these are raw ideas that have the potential to produce winning systems or winning methods. So here goes. The data has been taken from 2007 to October 4^th 2011. All profit/loss figures are calculated to £1 level stakes. SR% stands for strike rate; ROI% stands for return on investment.

Age

When thinking about long distance chasers I think about older horses rather than younger ones. However as the table shows, it is horses aged 6 to 8 that tend to do best in these events:

Combining horses aged 6 to 8 we get 257 wins from 2180 runners (SR 11.8%) for a loss of £232.19 (ROI -10.7%). OK it still makes a loss, but less than 11p in the £, so these horses seem a good group to concentrate on. If we compare their record with horses aged 9 and older we can see it compares favourably – the older group combine for 229 wins from 2796 runners (SR 8.2%) for a loss of £754.48 (ROI -27.0%). So we have losses of 27p in the £, and a lower strike rate to boot.

Position LTO

Recent form is something most punters latch on to, and hence there often is limited value around following a good run last time out. However, let us see the figures for LTO performance in terms of finishing position for our long distance chasers:

LTO winners have actually got a decent record with losses of only 4.5p in the £. Interestingly, horses that finished 6^th or worse LTO have lost only 6.6p in the £ – I would imagine these runners tend to start at better prices than they should due to punters being indifference to their chances. Horses that failed to finish LTO look very poor value and are horses to avoid.

LTO Market Position

Market factors are often used when looking for ideas, angles or systems, but usually it is the actual starting price that is used. The problem with using this as a ‘filter’ is that we don’t know the SP until after the event. We may have a pretty good idea close to the ‘off’ but not many people are able to wait until the very last minute to place their bet. I often look at market factors from previous races. Here, I am looking at their market position LTO. The results are as follows:

The only semi-positive here is the performance of LTO favourites – a decent enough strike rate coupled with losses of around 6p in the £.

Race type LTO

Most long distance handicap chasers race primarily over fences, but there are a few horses that switch back and forth. I thought it was worth looking at whether race type LTO makes any difference:

It looks as if horses that ran over hurdles last time are worth closer scrutiny. They have a better strike rate than those running over fences LTO, and backing all of them actually has made a profit. Now, it is unlikely this profit would be repeated over a subsequent 5-year period, but clearly these runners perform above expectations and probably start at prices that are higher than they should be.

Jockeys

The perception of many is that although claiming/conditional jockeys are able to take weight off the back of a horse, their inexperience means that the claim becomes effectively worthless. The stats for jockeys in these long distance chases makes interesting reading …………

Conditional jockeys claiming 5lb or more have a strike rate virtually identical to professional jockeys, but their losses are significantly less. This is almost certainly down to punters or bookmakers pushing out the prices of horses ridden by conditional jockeys claiming 5lb or more.

Conclusions

I have looked at 5 different angles and the stats have pinpointed some areas that are better than others. So how best can we utilize the stats? Well, the lazy system punter would almost certainly combine the ‘best’ bits of the five areas – so the system would look something like this:

1. won LTO

2. favourite LTO

3. Aged 6 to 8

4. Ridden by conditional jockey claiming 5lb or more

5. Raced in hurdle race LTO

For the record since 2007 this ‘system’ would have produced a strike rate of 66.7% and a return on investment of over 90%. However, before you all start planning your retirement due to this gilt-edged winning system, I should point out that there would have only been 3 selections in the 5 years of study!!! Essentially, this is an example of why back-fitting is not good practice. Too often people use too many rules when creating their systems – this either cuts down the number of bets to ridiculously small amounts, as in this case, or it simply gives a false set of figures even with a decent final sample size.

At the beginning of the article I mentioned the fact I would be looking to find potential starting points – this is rather than producing rigid ‘systems’ that have too many rules. It is my belief that a sensible starting point is when you combine two factors only. So that is what I am going to do. I am going to use the same 5 rules in the lazy system example, but will combine them in twos, not in one block of five. Here are the results:

As we can see, only one combination has produced poor returns – ‘the won LTO and ran in hurdle LTO’ one. Of the remaining 9 starting points, 4 showed a profit, the other 5 showed small losses. Now I am not advocating backing all horses in the future that match these starting points, but clearly any runner that does ‘match’ should be given further consideration. Others factors could then be taken into account such as going, trainer, class, fitness, etc.

Author Bio

Dave Renham works at the Racing and Football Outlook newspaper as their Race Trends expert; has worked on The Racing Post as a writer of the spotlight column and has published several books about horse racing research.

On the Web Dave is heavily involved with the respected horse racing websites www.PunterProfits.com and www.RacingTrends.co.uk

First Time Blinkers

With the flat season gathering pace, I have decided to look at horses that are blinkered for the first time. There are numerous ideas why horses are given blinkers to wear.

These include:

-         To try and get the horse to concentrate as some horses get distracted by the other runners in a race;

-         To help the horse break quickly from the stalls. Blinkers often have this effect when worn for the first time;

-         As a last resort to try and improve a horses’ performance.

What one should realise however, is that generally fitting first time blinkers is a negative rather than a positive. The data for this article is taken from the last 6 complete seasons – flat/all weather racing only. All profits are calculated to £1 level stakes at SP. It should be noted that I have included only runners that are wearing blinkers only (for the first time) – I have excluded runners who were wearing tongue ties also. For the record the combination of blinkered first time + tongue tie produce virtually identical strike rates and returns (from a much smaller sample).

Firstly let us look at the results of all runners wearing blinkers for the first time on the flat over the period of study:

Essentially therefore these runners win around once in every fourteen starts (roughly) for losses of just under 28 pence in the £. Not a great starting point from a betting perspective. However, let us break these stats down into different categories to see if we can either find better betting propositions, or ‘gilt-edged’ laying opportunities.

Age

The perception is that blinkers improve younger horses best, especially 2yos. Let us look at the results breakdown when split by age:

As we can see 2yos actually perform below the ‘norm’. Looking at the ages as a whole, there is no discernable pattern, although much older horses (8yo+) have done relatively well from a very small sample.

Digging deeper into the 2yo stats, I have broken down the data by number of career runs:

Two things that seem to stand out here are firstly that 2yos that have raced several times (7 or more), react to first time blinkers relatively well; secondly 2yos that are assigned blinkers on their 2^nd or 3^rd career start do extremely poorly in terms of strike rate. My guess is that there is an over-reaction to a poor debut run and the addition of blinkers actually makes things worse.

Market position

Generally the market is an excellent guide to the chances of each horse. Let us look at first time blinkered horses coupled with their market position:

At first glance the performance of favourites looks OK. However, when we take ALL flat favourites as a whole, the strike rate is around 30-31% with losses of only 6-7%. Hence, horses that start favourite when blinkered for the first time are not good betting propositions.

Race types

Let us break the data down now by specific race types:

Although the data is limited a combination of first time blinkers and an amateur rider looks one to avoid. Indeed of the 55 losers, only 3 got placed. I decided to look back further just to gather a bigger data set. I looked at first time blinkered runners in amateur contests from 1997 to 2004 – they fared poorly once again with just 3 wins from 104.

Maiden races have seen losses of around 42 pence in the £ which above the base figure of 27.6p. Auction maidens have provided the poorest results for first time blinkered runners with just 12 wins from 279 (SR 4.3%) for a loss of £151.06 (ROI -54.1%). Indeed, 2yo maiden Auction races are even worse with just 3 winners from 160 runners (SR 1.9%) for a hefty loss of £129.50 (ROI -80.9%).

Claiming races have seen close to a break-even situation despite a low strike rate of around 9%. Indeed if you exclude maiden claimers the strike rate rises to 9.3% and profits are made; albeit 9 pence profit for every £ wagered. However, it should be noted that these profit figures are essentially down to a few big priced winners and hence it is not an area where the backer can be confident to make a profit in the future.

Race distance

Let us break the data down now by race distance:

In general there does not seem to be any pattern here. I had expected longer races to produce slightly poorer results but this is not the case.

Turf v all weather

Horses wearing blinkers for the first time perform better on the all weather compared with turf. This could be due to the fact all weather racing is less competitive; or generally of lower grade. Whatever the reason, the stats are worth taking note of.

Jockeys

I decided to see if the experience of the jockey made a difference. The table below compares professional jockeys with claiming jockeys:

The figures seem to suggest the less experienced jockeys struggle when horses are blinkered for the first time. Losses close to 44 pence in the £ combined with a strike rate of under 5% means that one should swerve these jockeys under these circumstances.

Trainers

Some trainers have a better understanding of their animals than others so one would expect a real mix of results for first time blinkered runners. I have included all trainers that have saddled at least 30 horses with first time blinkers:

Sir Mark Prescott has excellent figures considering how poor these runners do in general – a better than 1 in 5 strike rate with profits of over 50 pence for every £ wagered. At the other end of the scale, Milton Bradley and Alan Berry have combined to produce 0 winners from 116 runners.

——————–

To conclude, horses blinkered for the first time are essentially poor investments. However, some are much worse than others as this article has hopefully highlighted. If nothing else, I suspect this article may save you from backing certain horses that have a very poor chance of winning. This should help your betting bank balance.

Dave Renham

=============================

Dave Renahm is a regular feature writer in the Racing And Football Outlook newspaper.  If you are more of an investor rather than gambler at your racing, more of his educational research work may be found online at www.PunterProfits.com and at www.racingtrends.co.uk

HINT: Both sites above will on occasion offer a free trial to sports betting blog subscribers. Why not Register Here for Free so you know when such a deal is available.

The “One Bet Per Month” thread is running over in the full member area of www.PunterProfits.com

The rough idea is that several key contributers contribute one strong or good value bet each month each.

( with several contributers and some somethimes stretching beyond just one each month there is more than one bet per month on the thread )

This concentration and focus is paying dividends so far with 15.5 points profit at SP from 40 selections.

That’s a healthy return on investment of over 38 % !

There is a selection for today.

4.40 Chester Act Of Kalanisi best priced at 5/4 Ladbrokes and VC.

Roughly the same on Betfair.

To visit PunterProfits click here ===> Horse Racing Systems