Introduction

Welcome to Placegetters. We are confident that our association will be a happy one, and mutually profitable.

Since 2004, we have been developing our unique philosophy toward racing. Like many philosophers, we have found it useful to jettison some pre-conceived notions, thereby trying to avoid making common errors. In doing this, we have found that the most important element in racing, as in other forms of wagering or gambling, is the concept of Probability. Indeed, that the consideration of Probability is one of the most quintessential elements of life itself, affecting almost everything we do each day. Of course, we don’t consider it consciously most of the time; that would use up too much of a most precious resource, our time.

However, every time we cross a road, we consider the probability, based on our observations of prevailing conditions, coupled with our past experience, of completing this task in safety. We may not assign a percentage or a decimal point to our summation, but once we perceive enough threats to our safety, we very firmly rule a line against making the attempt at that precise moment. Thus, we move through life mainly with the question of Probability somewhere in the back of our mind, approaching and receding from our conscious thought according to our individually-derived notions of risk or reward, when contemplating various decisions and actions.

So, let’s return to the question of Probability in gambling. As distinct from everyday life, where we may not pay much attention to Probability, in gambling or betting it is absolutely essential to factor in this element. Look at this brief example:

When playing Roulette, we can bet on a great variety of probable outcomes. The normal bets range from 1-1, 2-1, 5-1, 8-1, 11-1, 17-1, up to 35-1. Of course, you can also bet at odds-on, by betting on more than half the numbers (e.g.  an even-money bet plus one or two columns or dozens). A straight-up or single number pays 35-1, and according to the Laws of Probability, is the best value bet on the table. There is mathematical proof of this, but we won’t take the time to go into it here. Suffice it to say that the house advantage is the lowest when a single number is bet.  Furthermore, no matter what combination a punter may bet on the Roulette table, the winning number always pays 35-1. Thus, whenever a player, having bet on 2, 3, 4, 6 or even more numbers gets a hit, they would be well advised to consider that they could have received 35-1 for selecting the single number. Of course, the Probability is lower, but when successful the payout can cover a relatively large number of losing bets, giving a good possibility, if you can find these single numbers, of leaving the table while showing a profit.

Because of the house advantage, anyone who plays for a long time is virtually guaranteed to lose. Therefore, if you wish to win rather than just play for fun, you need to do two things at Roulette.

  1. Minimize the house advantage
  2. Get into profit as soon as possible

And you should add a third condition; have a plan for quitting while ahead, (and use it)!

So, how can we increase our chances of hitting a single number? We know that each number pays 35-1, and all, including the Zeroes, have an equal chance of appearing on any given spin. Nevertheless, if we did nothing for the rest of our lives but spin the wheel in sets of 37 or 38 times (one for each possible outcome, depending on the number of Zeroes on the wheel), we will not live long enough to see all 37 or 38 different numbers come up consecutively, without any numbers repeating. This is because of the randomness of distribution of large numbers. The odds of hitting all 37 or 38 numbers in succession, (in any order at all), is very many billions to one.  In practice, in any series of 37-38 spins, there is an almost 90% chance that there will only be 21-26 different numbers represented. Almost half of that, 40%, will be taken up by series where there are either 23 or 24 different numbers. Obviously, in all these sequences, between 12 and 17 numbers will be repeated at least once (with 13-14 numbers predominating).  Hence, when selecting single numbers to bet, we should focus on those that have already appeared recently. Statistically, in those 40% of cases mentioned before, between 34-37% of numbers will repeat, with higher and lower percentages occurring at either end of the scale covering the aforementioned 90%. If we can get a hit anywhere between 15-35% of the time on an outcome paying 35-1, the Probability is very great that we will eventually show a profit. Especially if you consider the possibility of staking in such a way as to take advantage of the occasions when these hits do occur consecutively, multiplying profits markedly.

In fact, we have personally demonstrated this to our own satisfaction (and profit) over several hundred sessions in a real-life casino, and can state confidently that because of the staking method we use, there is no possibility that the casino will ever recoup the advantage that we have acquired, which is in the many hundreds of chips. The Probability of any number occurring in 1 spin (on a single-zero wheel) is 2.7%: thus cumulatively 54% over 20 spins. Over the life of our tests, the Percentage achieved is virtually exactly that, which is substantially better than the hit-rate mentioned in the last paragraph. And, most importantly, the method involves walking away with the winnings when they occur, avoiding the delusion that it’s possible to just sit there and win endlessly in the face of the House advantage (as if the Casino would simply let anyone do that without barring them anyway)!

Our only regret is that we don’t live close enough to a casino to take advantage of this phenomenon more frequently (there are other things in life besides money).

Ultimately, when we are playing any game where the odds are stacked against us, we are in the same position as rebels resisting a better-armed oppressive regime: the only realistic option is to adopt guerilla tactics. That means picking our moment, making a quick hit and getting out with our gains, to live and fight at another time of our own choosing.

So, our interest in horseracing was piqued. Was it possible to achieve a positive Probability in the so-called Sport of Kings?

Question:  How many horses win races, while paying 35-1?

Answer: Very few. We did some research back in late 2011. We found that from November 1st 2011 and February 2nd 2012 in Australia, 1002 horses raced with an average Starting Price (SP) of $37-and-change, of which only 12 (Twelve!) won.  A less than inspiring 1.2% strike rate!

However, the most interesting observation from that somewhat dismal statistic is that if only 12 won, paying an average of $37, the return from betting a dollar on each one would have been only $444. Now if we get paid $35 on a single number at Roulette, and we know the probability of getting a hit is 2.7% on any spin, we would still lose slowly (if we tried to play continuously), because the house will not pay us our true odds. It would be possible to sometimes win doing this however, if we got one or two early hits, and decided to leave with our winnings. However, in racing, for us to lose anything like as slowly as that backing 35-1 shots, we could argue that they should be paying considerably more than double their actual prices. Why is that?

Firstly, the actual hit rate is less than half what the odds indicate we should expect. That tells us that over a sample of this size, to lose at the same slow rate as we would playing a single number at Roulette, the prices are nowhere near reflective of the horses’ true chances of winning!

What this indicates is that it is difficult to get real or true value from our bets at horseracing. This is partly because of the high take-out by the Totes, but it also indicates that bookmakers are trained carefully to avoid taking a beating when a long shot gets up. The medium-high end of the market is grossly undervalued, and because of the “Favourites Phenomenon”, so is the lower end. Many punters have a haphazard approach, and it’s easy to see that at the closing end of the betting, the favourites are over-represented by last-minute bettors. Simply watching the Tote screens or websites before close and a couple of minutes afterwards almost always displays a lot of last-minute shortening of favourites and lengthening of the prices of the extreme long shots ($40-50 and upwards).

Well, if it’s nearly impossible to get value from the extremes of the market, can it be found elsewhere? If so, what should we be looking for? Theoretically speaking, we need to find horses with a reasonable Probability of winning, but which are overlooked by most of the punters. Let’s consider a few facts. We know that favourites win only about 31% of the time, which immediately tells us that the market makes the wrong choice virtually 7 times out of 10. We already discovered that once the prices get out to over $30 (where the assessed probability of the horse winning is about 3%); they only achieve that result about 1% of the time, or 1/3rd as often as expected. Similar success rate to the favourites, isn’t it?

It starts to look as though the best chance of finding the winners of 70-odd % of races lies somewhere between the average $3 price of the favourites, and the generally hopeless $30-plus roughies. Although these horses win 70% of events, the populace continues to overwhelmingly support the favourites. Why this happens is not our immediate concern here. Clearly though, many of the factors most punters (and pundits) consider to be vital in selecting potential winners may not be as important as generally believed. Or, perhaps other factors need to be granted greater weight of consideration.

Can these contrary factors be identified, and used by people who may not be expert horsepersons, to derive satisfactory punting results in the long term?

We believe that we have identified some of them (there may of course be others) and have compiled a series of results which have shown a consistent profit. Whether these factors can be fully justified by logic remains to be seen, although there is good reason to suspect that anything that happens with a significant degree of Probability will turn out to have a logical explanation.

A further consideration, often overlooked by punters, is that whereas bookmakers must create a market for every race, we as punters can be discerning, and select only races where the information available is likely to be helpful to us. As an extreme example, if there was a race coming up where every single runner was making its racing debut, how would we determine which was the most likely winner? We’d be thrown back purely onto assessments of their pedigrees, and performances in barrier trials, and the qualities of the trainers and jockeys, and precious little else. We could probably get a better feeling of wellbeing by simply donating to our local racing industry body, than by wagering such an event!

To return to our earlier statement, that the Probability of an outcome is the single most important factor in determining a bet.  Just as in Roulette; a player may have their favourite numbers, but would be better advised betting the numbers which are actually going to turn up, insofar as these can be discerned (remembering that a significant proportion of the numbers are going to repeat in the near future). A prime example of this rule will be described in a future article entitled “What Is Probable?” (in our blog section).  In these articles is a description of the Pick-A-Box Paradox, whereby numerous contestants in a TV game lost their bets solely because they clung to preconceived notions and ignored the true Probabilities inherent in the situation.

Let’s diverge for a moment again and consider a poker hand. A player is holding a hand which he knows from his study of percentages, gives him a 30% chance of being a winner. He is one of the last three players in the pot. A bet is made, and he must call or fold. How is he to decide? He must consider the payout potential of the pot, against his odds of winning. If the payout is 4 or 5 times what it will cost him to call, he has an overlay, because his nominal chance of winning is only 1 in 3. If it will only pay him 2 to 1, he is betting an underlay.  In the long run, betting overlays will give him a profit, because although he will lose some of these bets, sometimes he will win (the other players were overplaying their hands or bluffing), but when he wins, he is getting more than his hand is really worth. Conversely, betting under lays will also result in him losing and winning some. However, his losses will represent occasions when he overplayed his own hand, but his wins will be worth less than they should have been, and will not fully compensate him for his losses, resulting in a long-term deficit.

Returning to our racing once more; it is significant if all the form and traditional factors indicate a favourite, if other factors not considered so important by experts indicate an alternative winner, and the odds being paid for that horse represent an overlay. A certain proportion of these less fancied mounts will win, they will pay more than they should (because the favourites are over bet), and will more than compensate for those that lose, just as in the poker game.  It’s all about the Probabilities. How often things happen, balanced against the number of times we attempt them, and whether others are aware of these factors.

Mathematically, if one were to place a series of 1000 bets, with a Win Probability of 30% it is a virtual certainty of getting at least one losing streak of 19, with all the naturally accompanying chances of losing streaks of all numbers up to 19 … not to mention the smaller (but not discountable) chance of getting streaks longer than that. We also need to remember that these losing streaks can occur separated by isolated successes and that if we are to play for an indefinite period, such sequences are mathematically certain. Over any extended period or sequence of bets, virtually every combination of winning and losing streaks can be expected to occur. An expectation of a winning streak unlikely to be longer than 6, ought to bring some sobriety into our thinking. Countering this gloom, however, is the realization that even a losing streak of 19 can be entirely recouped with one or two decent collects, something which is just not possible by betting on favourites.

While a losing streak of 19 out of 1000 bets may appear manageable, betting 5% of a bank on such a series of events will see the bank decline by more than 60%. Acceptable perhaps if playing with $100; not so palatable if our bank is $50,000! There are various ways of coping with this. One is to bet a much smaller amount of the bank, say 1% or 2%. The chance then of serious damage becomes far less likely. A second is to withdraw regularly from the bank, so that a bad streak won’t eat up our starting or existing capital so severely. Another, which we prefer, is to concentrate on the Place strike rate, in this case well over 50%.

The real significance of this however is that with a losing strike rate of about 40%, from 1000 bets one is virtually guaranteed at least one losing run of 9.  However, that is countered by an expectation of at least one winning streak of 11 or 12. The maximum damage to the bank from a run of 5 losses (at 5%) is less than 25%. The Place result is perfectly adequate, especially when considering the possibility of betting 3 or 4 times the amount for the place, as the win. This would return virtually the same profit to the bottom line as just wagering for the win alone. Also, to be considered is the fact that avoiding long losing streaks and enjoying long winning ones is far easier to accept, psychologically and emotionally, and far less likely to cause the crises of confidence which so easily beset most punters, causing constant changes to selection methods and staking.

Finally, it needs to be stressed that the prices we focus upon are Betfair SP. The reason for this is that Betfair consistently outperforms the Totes and bookmakers for win and place dividends, especially once the higher priced horses are factored in. The place dividends, especially once the win dividend gets above $5, also tend to be quite generous compared with Tote or corporate bookmaker prices.