In this article I am going to continue the analysis of last-start winners that formed the basis of last month's article. In particular we will look at last-start winners strike rates for coming second and third in subsequent races.

For those of you who may have missed last months article, here is a brief summary of our findings:

RANDOMLY PICKED HORSES should win around 10% of races.
LAST-START PLACEGETTERS should ain around 12% of races.
LAST-START WINNERS should win around 14% of races.
LAST-START WINNERS AT SATURDAY MEETINGS should win round 15% of races.
LAST-START CITY WINNERS should win around 18% of races.
LAST-START CITY WINNERS THAT LAST RACED WITHIN 0–14 DAYS would win around 19% of races.

These statistics were based on the study of nearly 50,000 race results from the racing season August 1, 1989 to July 31, 1990. As we concluded last month you can see how the application of some fundamental criteria regarding a horse's last start almost doubles the strike rate from the average of 10% to 19%.

Rather than rush on and consider another form factor at this early stage in this series, I have decided to continue with the examination of last-start winners and placegetters and look at place strike rates. As far as I am aware, the comparison of place strike rates with win strike rates has never seriously been carried out in racing literature before.

Place strike rates are important for two reasons.

First, a system or set of rules that produces more wins than seconds and more seconds than thirds is definitely headed in the right direction.

Second, place strike rates are of interest to quinella and trifecta I plot bettors *to wish to profit from the value offered by exotic betting.

When considering the statistics that follow you should keep in mind the fact mentioned above that randomly picked horses should win around 10% of races. This statistic was arrived at by noticing that on average there were around ten horses in every race. Importantly you should note that this means the chance of a randomly picked horse coming second is also 10% as is the of a randomly picked horse coming third.

From the above summary of last months article we look that last-start placegetters (that is horses that finished first, second or third at their last start) win 12% of races. In fact the figure for the racing season August 1" to July 31, 1990 was 12-4%.

So how many last-start placegetters came second and third last racing season? The strike rate, for last-start placegetters coming second was 11.4% and for coming third was 10.5%.

There are two important things to notice from these statistics.

First, the strike rate drops off as we move our consideration from firsts to seconds and then to thirds. The fact that last-start placegetters are more likely to win than come second or third is significant.

Second, while the strike, rate for last-start placegetters coming first next start is a reasonable improvement upon our average 10% figure, the strike rate for them coming second is only marginally better than our 10% average and the strike rate for them coming third is no real improvement at all.

We can conclude therefore that consideration of last-start placegetters is important for picking winners but, by itself, is not as important when trying to pick the placings.

If we confine our selections further to last-start winners as we did last month we find that these horses win 14% of races. The exact figure for the racing season on August 1,1989 to July 31, 1990 was 14.1%.

During the same period, the strike rate for last-start winners coming second was 12.3% and for coming third was 10.4%.

As with last-start placegetters, the percentage of last-start winners that come third in races is still not much better than average. However the strike rate for last-start winners coming first or second is better than average.

It would be fair to conclude, that last-start winners are good candidates for quinella betting. Clearly they are more consistent than last-start placegetters.

If we confine our selections further to last-start city winners we find that these horses win 18% of races. The exact figure for the racing season August 1, 1989 to July 31,1990 was 17.8%.

During the same period, the strike rate for last-start city winners coming second was 13-5% and for coming third was 10.4%.

As with last-start placegetters and last-start winners, the percentage of last-start city winners that come third m races is still not much better than average. The strike rate for them coming first and second however is further improved.

Clearly last-start city winners are also good candidates for quinella betting. They are slightly better propositions than last-start winners since a win on a city track suggests a better class of horse than one that only won its last start on a provincial or country track.

In conclusion we have shown that last-start city winners are not just better win betting propositions than last-start winners or last-start placegetters; they are also better quinella betting propositions.

It might come as some surprise that last-start city winners win races nearly twice as often as randomly picked horses yet do not get placed third any more frequently. Can we produce a method for obtaining a strike rate significantly better than 10% for horses coming third at their next start?

Intuition should tell you that it is very unlikely that a method could be found that picked horses that came third significantly more frequently than they finished in other positions (including first and second). One of the reasons for this is that as we improve a selection method the strike rate for third will increase but so will the strike rates for second and first. If it is a good selection method then the win strike rate will increase more quickly than the second or third strike rates.

Even taking horses that came third at their last start doesn't produce more thirds next time round than other placings. In fact for the racing season August 1, 1989 to July 31,1990 last-start third-placed horses won 10.4% of races, came second in 10.8% of races and came third in 9.9% of races.

The only way we might achieve a good third place strike rate, say around 20%, is to achieve even higher second and first strike rates. We will try and find a way to achieve this next month.