Now that race clubs are no longer tolerating broadcasting of on-course bookmaker fluctuations, any system we can find to provide an idea of what's happening on-course has to be of advantage to the off-course TAB punter.

The idea I have come up with is based on the fact that professional punters oncourse, who bet overs on the TAB, do not start betting until 20 or 15 minutes before a race since it is not until then that bookie markets and betting trends become established.

TAB fluctuations then start reflecting the onslaught of such money. The aim of this article is to show you how to get the maximum amount of information out of this fluctuation data.

Irrespective of the money wagered on the TAB beforehand, I suggest it is the money bet in the 20 minutes prior to the race that is important. It is during this period that on-course investments are going to be following the money in the betting ring.

Furthermore, the on-course TAB investments during this period are big enough for reliable trends to be noticed. Not only are the individual bets larger but there are more of them, since they tend to be placed during the 20 minutes prior to a race, unlike off-course TAB investments, which are spread much more uniformly throughout the day.

In this year's Golden Slipper I looked at the pool and pre-race approximate dividends 20 minutes before and 10 minutes before the race (10 minutes leaves enough time for calculations and bets to be made). During this period I calculated that $54,000 was wagered on the eventual winner Courtza and $49,000 on TAB favourite Mercury. This was a significant change as in the two hours prior to this period, $120,000 was bet on Mercury and $80,000 on Courtza.

Despite the fact that Mercury opened TAB favourite and was landing more bets up until 20 minutes before the start of the race, more money was now going on Courtza-and this coincided with the time- when bookie markets were being established. This, one would argue, suggested that Courtza was shorter in the ring and perhaps even favourite. This was indeed the case as Courtza opened in the betting ring at 5/2 compared to Mercury's 5/1. So the money pointed to Courtza as the bet and, as we all know, Courtza won.

I'll explain now how a few simple calculations can be made to determine exactly how much money is being wagered on the TAB on any horse. All you need to follow are the pool size changes and dividend changes-figures which in most States can be obtained from the Teletext pre-race approximate dividend service (either on Teletext W's or in TAB agencies).

To convert the dividend and pool size data into meaningful information, proceed as follows.

First multiply the pool sizes by 0.86 to get an idea of how much money remains to be shared out in the form of dividends. (The 0.86 factor reflects the fact that 14 percent or thereabouts is taken out as commission by the TAB).

Following our example we find there was around $929,473 available for dividends 20 minutes before the start (.86 times $1,080,782) and $1,172,718 available 10 minutes before (.86 times $1,363,625).

The next step is to divide these figures by the pre-race approximate dividends to calculate how much money is being held on each horse. The difference in these hold figures over time indicates how much money is being wagered on a horse.

For example, 20 minutes before the Slipper there was $211,243 being held on Mercury ($929,473/$4.40) and $138,727 on Courtza ($929,473/$6.70).

With 10 minutes to go, there was $260,604 being held on Mercury ($1,172,718/$4.50) and $192,248 on Courtza ($1,172,718/$6.10).

By subtraction, we find that $49,361 was bet on Mercury during this 10 minute period ($260,604-$211,243) and $53,521 on Courtza ($192,248-$138,727). Hence my conclusion that while Mercury was still TAB favourite, Courtza was the one more money was going for and the one that should have been paying less-and therefore the overlay selection.

We can actually take this analysis one step further and calculate what a horse's true price should be, based on the amount of money wagered over a given period.

In the above example it is clear that Courtza was the value selection. This was because the horse on which the most money was being placed was not the favourite. This will not always be the case and we will find occasions when the horse attracting the most money is favourite yet still at value odds.

To determine whether this is so we divide the amount invested on a particular horse over a given period into the total amount by which the pool increased over the same period. The result of such calculations gives us the dividend a horse should be paying based on the recent investments only.

For example, the Slipper pool increased by $282,843 (from $1,080,782 to $1,363,625). So Courtza should have been showing a dividend of $5.30 ($53,521 divided into $282,843) and Mercury $5.70 ($49,361 divided into $282,843). As already mentioned, Courtza was showing $6.10 and Mercury $4.50 with 10 minutes to go.

Based upon these calculations Courtza was definitely the value selection with 10 minutes to go. Although Courtza's dividend firmed to $5 at close of selling due to a massive investment of $77,000 on the filly during the last four minutes of betting, this still represented value as the filly had been backed from 4/1 into 7/2 in the ring.

Let's look at another example where the horse being backed the most is indeed the TAB favourite. The race I will look at is this year's AJC Derby.

Research was the clear cut favourite on the N.S.W. TAB showing $2.40 with 20 minutes to go. Next in the market was Royal Pardon at $5.50. There was $483,998 in the win pool at this stage.

With 10 minutes to go, Research had firmed to $2.30 and Royal Pardon had eased to $5.80. The pool had increased by $134,158 to $618,156.

Applying the calculations as described earlier (see later summary for further explanation) we find that $57,704 had been placed on Research (.86*618,156/2.30 less .86*483,998/2.40) and $15,978 on Royal Pardon (.86*618,156/5.80 less .86*483,998/5.50).

Dividing these into the amount by which the pool increased indicates that Research should have been showing a dividend of $2.30 (which it was) and Royal Pardon a dividend of $8.40 (it was only showing $5.80). So Research was showing a fair price but not so Royal Pardon.

What you have to be careful about in these situations is by how much a horse is likely to further firm. Research firmed to pay $2 for example. You might have expected this to happen and considered a bet on the horse not on.

On the other hand you might argue that further investments on the horse meant that its true price was in fact actually less than the $2.30 originally calculated.

Of course we should not lose sight of what we are trying to do in the first place-that is predict what is going on in the betting ring. So what was going on? Research was being backed from 11/8 against into 11/8 on. So the evens finally paid by the tote was good value.

I am sure you will agree that this method, although involving a few calculations, certainly seems to be giving us the good oil as to what is happening in the betting ring. To simplify your application of this method, the following step by step guide should be useful. Readers who own computers will be pleased to know that I have written a program to do all this for them. I published it in a recent issue of my newsletter, Punter's PC Digest, copies of which are available from me for just $5. Write to me at GPO Box 2559, Canberra A.C.T. 2601 for a copy or a free brochure on Punter's PC Digest.

SUMMARY
To keep things as simple as possible I have used letters to represent the amounts required in the calculations.

  1. Pick a point around 20 minutes before the start of the race and note the total pool size (A1) and approximate dividends showing for the more favoured horses in the market (B1, Cl, Dl, etc.). Try to pick a time after which the pool size has just been updated since sometimes the pool size is not updated for five minutes or more. you should note pool sizes and dividends as soon as you see the pool size update even if 22 minutes before the start of the race.
  2. Calculate the amounts being held on the horses for which you have just recorded dividends for. For example, the horse showing a dividend of Bl will have investments of .86*Al/B1, the horse showing a dividend of Cl investments of .86*Al /Cl, etc.
  3. Pick a point around 10 minutes before the start of the race and note the total. pool size (A2) and approximate dividends showing for the more favoured horses in the market (B2, C2, D2, etc.). As before, try to pick a time after which the pool size has just been updated--even if as much as 12 minutes before the start of the race.
  4. Calculate the amounts being held on the horses for which you have just recorded dividends for. For example, the horse showing a dividend of B2 will have investments of .86*A2/B2, .86*A2/C2 for the horse showing a dividend of C2, etc.
  5. Subtract the amount being held on a horse at 20 minutes to start time from that being held on it at 10 minutes to start time. This gives how much was bet on it during the period-for example .86*A2/B2-.86*Al/Bl for the horse showing dividends Bl with 20 minutes to go and B2 with 10 minutes to go. Let's call this amount X.
  6. Finally, calculate the amount by which the pool changed, A2-AI, and divide it by the amount bet on the horse you are considering. This figure, (A2-A1)/X, gives the dividend that the horse should be showing as based on the investments placed during the period of your two observations.

You will need a calculator to carry out these calculations and preferably a piece of paper on which you can draw up five columns in which you can write TAB numbers, dividends at first observation, calculated holds on horses at time of first observation, dividends at second observation, calculated holds on horses at time of second observation.

Although this method involves a few calculations, I am sure those of you interested in it enough to use it will find it a powerful tool in your betting activities.

For those of you preferring a more general guide on how to follow TAB fluctuations, keep an eye out for an article on Teletext and Viatel TAB services I am planning for P.P.M. in a month or two's time.

By Neale Yardley

PRACTICAL PUNTING - JUNE 1989