MISLEADING STATS

One area that is worth analysing closely is the value you attach to form guide statistics. It’s easy to fall into the habit of subconsciously attaching value to certain statistics in a horse’s form and making judgements that seem logical, but are in fact greatly limiting your profit potential. For example, most punters instinctively like horses that:

• Aren’t first up
• Come off a good last start run
• Have drawn a good (inside) barrier
• Have a good win strike rate
• Have proven they can win at the distance
• Have proven they can win at the track
• Have proven they can win in the track condition

… and the list goes on. The more of these traits a horse has, the better it must be as a betting prospect, right? The less it has, then our logic says the more of a risk it becomes, even to the point that it must be a poor bet. There’s no doubt that many of these factors are important. Horses in good form win most races. Horse’s that can’t run the distance or lack race fitness or can’t handle the track condition won’t win.

However, the tendency to make these judgements by simply referring to numbers in the form guide can be very misleading and will greatly limit your potential as a punter!

In this article we will explore a few key form guide statistics and show how the facts are often quite different to the belief we have in them. All of the data shown below is based genuine chances up to $10 SP in the betting market (excluding first starters), in races held on Metropolitan tracks across Australia from 1 Jan 2013 to 10 Feb 2018 (unless otherwise noted.)

It’s natural to take some confidence out of a horse’s proven past ability to win at the distance. The table below breaks down our sample group of horses based on their previous record at the distance of the upcoming race.

Distance Form	Runs	Wins	SR%	POT%
Never won at Distance
0 Starts – 0 Wins	23395	4270	18.3$	-7.0%
1 Start – 0 Wins	11970	2107	17.6%	-5.8%
2 Starts – 0 Wins	6130	1016	16.6%	-5.9%
3+ Starts – 0 Wins	7212	1144	15.9%	-3.3%
Previous Wins at Distance
1 Start – 1 Win	4798	1016	21.2%	-5.5%
2 Starts – 1 or 2 Wins	5288	1010	19.1%	-6.6%
3+ Starts 1 or 2 Wins	18668	3061	16.4%	-6.4%
3+ Starts – 3+ Wins	6414	1055	16.4%	-11.6%

As you can see, there is very little difference in the betting profitability of horses across the categories shown. In fact, the group with the best return are those horses that have had 3+ previous starts at the distance for zero wins. They’ve returned just a -3.3% loss to the punter against the overall markets average of -6.6% The least profitable group are those that have had 3+ starts at the distance and won 3 more races. Their return of -11.4% profit on turnover is a clear sign that the market has overvalued the important of that statistic. It might seem very appealing that a horse has had 6 past runs at the distance for 3 wins, but history shows that statistic isn’t a sign of a more profitable betting prospect at all, one average it’s actually the opposite. If anything, it means you need to be more certain about other traits of the horse helping to present it as a value prospect. It’s great record at the distance is so obvious to the market that it tends to overestimate the chance of those horses compared to the average.

If we look at previous distance form in the simplest terms:

Distance Form	Runs	Wins	SR%	POT%
Never won at distance	48707	8537	17.5%	-6.1%
Has won at distance	35168	6142	17.5%	-7.2%

Remember that these stats exclude first starters. What the data shows is that the strike rate of genuine chances that have never won at the distance is exactly the same as horses that have, with a slightly superior betting return. The ability to run the distance of a race is important, but you shouldn’t just that by form guide statistics. If you have reason to doubt a horse’s ability to run the trip then by all means you should back that judgement. If there is no solid evidence to say the horse is a risk, then you shouldn’t attach any negative value at all to that uncertainty. If you like everything else about the horse as a betting prospect then embrace that lack of distance form in the horse’s history, it’s likely to be helping you to gain a little bit of extra betting value.

A previous good record at the track for a hose can be a sign that it has a liking for the layout (i.e. tight turns, wide turns, short straight, long straight etc.), the surface characteristics or something else intangible. It could also be complete coincidence and not reflective of anything special at all.

There is no doubt that over history there have been many track specialists. Chief De Beers is always one that stands out in my mind from the mid to late 1990’s. He had 20 career wins and they all came at Doomben, including two G1 Doomben 10,000 wins (from a total of 38 starts at the track). He raced across the road at Eagle Farm nine times and never won. In fact, he never won a race at any other track.

Dandy Kid won 15 races at Moonee Valley (more than any other horse) from 52 starts and just 4 races elsewhere from 34 starts.

However, genuine track specialists are a rarity (at least on turf) and the value of taking notice of previous wins at a track as expressed in form guide statistics is minimal at best.

Track Form	Runs	Wins	SR%	POT%
Never won at the track
0 Starts – 0 Wins	24643	4325	17.6%	-7.6%
1 Start – 0 Wins	13368	2214	16.6%	-9.5%
2 Starts – 0 Wins	7192	1289	17.9%	-0.3%
3+ Starts – 0 Wins	10764	1668	15.5%	-5.4%
Previous wins at the track
1 Start – 1 Win	3138	697	22.2%	-8.3%
2 Starts – 1 or 2 Wins	3528	732	20.7%	-5.7%
3+ Starts 1 or 2 Wins	16271	2835	17.4%	-5.2%
3+ Starts – 3+ Wins	4971	919	18.5%	-7.6%

You can see there is very little difference between the categories shown in this table. What is notable is that horses with 2 starts or more for zero wins provide better than the average return, while horses with 3+ starts for 3 or more wins have provided just below average returns.

On the whole I wouldn’t say there is much in the stats about the horses record at a track, one way or another.

It can certainly be different if you are analysing a horse’s past ratings and see a consistent pattern of higher ratings at a track (even if not winning) or if there is something about a horse’s way of racing that you think is suited to a particular track, but don’t lets the raw statistics influence your betting decisions at all.

Win strike rate is simply calculated by taking the number of races a horse has won in its career and dividing it by the number of starts it has had. For example, a horse with two wins from ten starts has a winning strike rate of 2 / 10 = 20%.

“The late Don Scott was a big believer in the proven winning ability of a horse being a key factor to consider from a betting perspective. He wrote in his popular book Winning More, “the best horses to back are horses whose wins are frequent and predictable. The worst horses to back are horses whose wins are infrequent and unpredictable.”

His theory seemed entirely logical and to this day punters still largely consider a high win strike rate a positive for betting and low win strike rate a risk.

However, research of the facts around horses win strike rates paints a different picture. It shows that a bias towards high win strike rates compared to lower win strike rates is unlikely to be adding any value at all to your betting profits and is more likely to be limiting your ability to find good value bets.

The table below shows horses up to $10 SP in non-maiden metropolitan races with at least 10+ starts in their career.

We exclude maiden races as by nature every horse in them has a 0% win strike rate.

We isolate those with 10+ starts so that we can draw conclusions on an exposed group of horses without the statistics being incorrectly influenced either positively or negatively by lightly raced horses.

For example, a horse with 12 starts and a 50% win strike rate (6 wins) is entirely different to a horse with a 50% win strike rate from 2 starts (1 win.)

Equally, a horse with zero wins from 3 starts shouldn’t be viewed as statistically equal to a horse that has zero wins from 15 starts.

Win SR%	Runs	Wins	SR%	POT%
0% – no wins	150	26	17.3%	+48.8%
1% to 10%	5363	818	15.3%	-2.5%
10.1% to 15%	8089	1213	15.0%	-3.1%
15.1% to 20%	12658	1953	15.4%	-7.%
20.1% to 30%	13955	2290	16.4%	-7.6%
> 30%	7907	1523	19.3%	-7.5%

There are very few horses with a 0% strike rate because the vast majority of horses racing in a non-maiden metropolitan race have already won at least one race. That extreme positive result with a sample of just 150 runners is relatively meaningless.

What is very meaningful though is the record of low strike rate horses in the 1% to 15% range when you consider it is metropolitan racing and the horses have all had at least 10 starts.

Their strike rate is only marginally less than horses that win more often, but their betting returns are 2-3 times better.

The table clearly shows that the best betting returns actually come from horses with a lower strike rate. The betting market perceives a low win strike rate on well exposed horses as a distinct negative and offers a slightly better price about them. The truth is that it’s barely a negative at all as far as win strike rate goes and the net effect is a much better return to punters.

There’s no doubt that consistency in a horse is an asset, but the key message here is don’t use simple win% or even place % statistics to judge that. Look at the quality of a horses past runs and competitiveness, especially in races of similar grade to the upcoming one and judge for yourself how consistent they’ve been.

If a horse has the right form and other credentials for a race, then a high win strike rate does not make it more appealing as a betting prospect.

Even more importantly, don’t be biased against a horse that otherwise looks a good bet because of a low win strike rate. Embrace those betting opportunities, because history very clearly shows that you are likely to be getting an extra few percentage points of value in your favour.

First up horses always present some uncertainty and the first point of reference for a punter is to check the horse’s previous first up record.

This is however one of the more misleading statistics in the form guide. Let’s break down the stats for first up horses up to $10 SP in metropolitan races.

First Up Form	Runs	Wins	SR%	POT%
Never won first up before
1 prior run – 0 wins	2349	428	18.2%	-7.7%
2 prior runs – 0 wins	1439	248	17.2%	-1.3%
3+ starts – 0 wins	1215	206	17.0%	+13.8%
Has won first up before
1 prior run – 1 win	1242	287	23.1%	-1.6%
2 starts – 1 or 2 wins	1721	312	18.1%	-10.3%
3+ starts 1 or 2 wins	3174	467	14.7%	-12.6%
3+ starts – 3+ wins75	417	75	18.0%	-13.5%

There is a general pattern here that the more impressive a horse’s first up record gets, the worse the betting returns are for all of those horses as a group. On the flip side, the best returns come from horses with 2+ first up runs in the past for zero wins.

Let’s summarise that table to make the patter clearer and also look at a horse’s previous place record when first up:

First up horses that have never won or placed before have an identical strike rate to horses with a previous first up win, but the betting returns are significantly better.

Remember that we are looking at first up horses that are considered a genuine chance (up to $10) in the betting market. They are all likely well enough credentialed in terms of class, distance of the race and other factors… but there is a clear and incorrect bias by the market that overvalues the important of previous first up wins.

The problem with using first up statistics is that they ignore the class of race that record was earned it and that’s especially relevant when analysing metropolitan racing.

A horse could have a first up record of 2 from 2, but if they were earned in a Country maiden race and then Provincial restricted race, how relevant is that to a first up city class race?

The message is clear. Don’t get carried away with first up statistics. Look beyond them to the class and quality of a horse’s previous runs when first up. Review trials and other information (e.g. jockey booking, support in early markets) to make your judgements about how ready a horse might be to run well.

If you like a horse that is first up, then the last thing you should do is let its lack of previous first up wins or placings put you off. If anything, you should bet with some confidence knowing that the market will be incorrectly biased against that and likely giving you better value.

The universal theory that an inside draw is an advantage in racing and an outside draw is a disadvantage is incredibly limiting when it comes to punting.

The influence of barriers are very much a track / distance specific thing, but a general analysis of inside versus outside barriers at Metropolitan tracks helps to demonstrate the point.

Races up to 1200m

Barrier	Runs	Wins	SR%	POT%
1 to 3	10587	1993	18.8%	-6.2%
10 or wider	3846	616	16.0%	-2.9%

While the strike rate of wider drawn horses is marginally lower, the betting return is more than 2 x better. This shows that the market is biased against wide barriers, but it overcompensates too much in its pricing and provides a much better than average return.

The picture is even more significant is races from 1400m to 1600m.

Races from 1400m to 1600m

Barrier	Runs	Wins	SR%	POT%
1 to 3	8045	1454	18.1%	-4.6%
10 or wider	3699	598	16.2%	-0.4%

The return from horses drawn wide up to $10 in the market is significantly better than those drawn inside.

Races 2000m and beyond

Quite interestingly, we often hear that barriers in staying races matter far less than Sprint races, but the data would seem to pain a different picture:

Barrier	Runs	Wins	SR%	POT%
1 to 3	3405	640	18.8%	-6.3%
10 or wider	1363	199	14.6%	-12.8%

The samples here are small, but a check back to 2008 (the last 10 years) show the results are consistent. Neither inside or outside draws provide better than average returns, but it’s clear that wide draws do return much worse than the average.

These are broad generalisations across all metropolitan tracks and it’s important to recognise that the most insightful view on the influence of barriers will come from understanding the specific track / distance and more importantly the racing style of a particular horse and structure of the speed map. A wide drawn horse that is the only horse with decent early speed in the race should be viewed entirely differently to a horse with moderate early speed that has a number of faster horses drawn inside it.

An inside drawn horse that lacks early speed and could end up buried on the fence in a large field is entirely different to a horse that has the speed to box seat in a moderately run race.

The key message is that you should let go of any universal beliefs you hold about the importance of inside versus outside barriers.

You should especially avoid being biased against a horse simply because it has a wider barrier. Historically they provide much better than average returns (except in staying races) and in the right circumstances can help create terrific betting value about a horse you like.