Welcome Guest ( Log In  Register ) 
Articles >> horseracing >> Investigating The Tricast  Part 3
In my last two articles I have been trying to see if it is possible from one of the “exotic” bets – namely the tricast. . To recap I decided to confine myself to 12 runner handicaps on the flat, and split the runners into “thirds” based on their market position. By splitting the market positions into “thirds” this started give me a decent overview of which type of horses reached the first three (in terms of market position) without getting too complicated.
By the end of last week’s article, I had studied four months of data for 2007 and was coming to the conclusion that there were three key combinations “TTT”, “MTT” and “MMT”. (NB. “T” = top third of the betting; “M” = middle third of the betting and “B” = bottom third of the market). These combinations had been responsible for 60 of the 96 tricasts (62.5%). As combinations they had produced the following returns:
These three combinations of the ten possible ones are clearly the ones to concentrate on. I demonstrated last time that perming all horses from the top and middle “thirds” of the betting would produce significant losses due to too many bets per race. With the “MMT” returns so good, the value certainly seemed to lie with making sure that the middle third of the betting was well represented in the perm. My decision now was whether to go for 5 or 6 horses in what I hoped would turn out to be the “optimum” perm.
Looking at the average payouts, would a 6 horse perm of 120 bets be too many? Before pressing on, I decided to include the first four bets from May so that the 96 races became a “round hundred”. They produced 2 wins for “MTT”, 1 win for “MMT” and 1 win for “BTT”. Hence after 100 races the top three permutations gave the following results:
Hence 63 of the 100 tricast bets came from one of these three main combinations. The payout totals for all three combined came to £17,902.15. It should be remembered that the tricast payout includes the initial stake so using a 6 horse perm (120 bets), the outlay over 100 bets would be £12,000. Hence, if we were using three horses from the top “third” of the betting and three horses from the middle “third” to make up our 120 bet perm to make a profit we would have needed to be successful around 7 times in 10 (70%) on the occasions that one of the three favoured combinations come up. Now, using three horses from the top “third” of the betting and three from the middle “third” the chances of hitting that target seems unlikely. The mathematics in working out the true probability is complicated, but I estimate that 3 wins in 10 (30%) is the most likely scenario. Of course one would hope to improve on this performance assuming one’s ability to select horses is better than “random selection”. However, it looks like the 6 bet perm is too risky.
Therefore, how about a 5 horse perm? The outlay would now be £6,000, but for success you would have to hit around 30% success rate in terms of the three favoured combinations. With 30% the likely success rate for 6 horses then with one less horse again our chances look slim of making a long term profit. However, I believe the key maybe to risk a lower success rate, but gain more reward in terms of returns. Hence the value would lie in perming three horses from the middle third of the betting, and two from the top. Again the Maths is complicated but you are looking at a 10% tricast success rate in terms of true probability. Of course the “TTT” combination now is out of the question, but we are concentrating now on the combinations that offer better tricast payouts. Hence 10 wins from 100 will still leave us short of breaking even, let alone making a profit. However, we are still talking about “random selection” here and hence one would hope with some decent form study you could improve the situation somewhat. Whether you could improve it enough is debatable.
What next? A positive outcome was starting to look unlikely and I was in danger of giving up on this whole tricast idea. However, ever the optimist I wondered whether a staking plan would offer a solution. I decided to use the data I already had and to see what would have happened assuming I had 10 winning tricasts out of the 100 races. Looking at the most likely scenario, it seemed I could expect around 5 winning tricasts of “MMT” and 5 wins of “MTT”. My next port of call was to decide on the staking plan – I decided that I would increase the bet after 10 consecutive losers by the initial stake. If I had 20 losers in a row, that would increase again by the same amount. Hence the bet increase would look something like this:
I hoped that a losing run of 41 or more was extremely unlikely but if I happened to have such a poor run, then bets 41 onwards should remain at £24.00 a bet. However, if I had a losing run of 40 that would incur losses of £600 ……. OUCH! Of course the ideal scenario would be to hit a winner soon after a bet increase to maximize the returns. I decided to see what would have happened with this staking plan, and a fairly even spread of 10 winners. The first series of 10 winning bets occurred thus – on bet 3, bet 15, bet 26, bet 35, bet 43, bet 52, bet 60, bet 70, bet 81, bet 91. The 10 winning payouts to a £6 stake were thus:
Now with the increasing staking plan, bets 15, 26 and 81 in this sequence would have seen an increased stale of £12.00 and hence the returns would have been doubled. Indeed for this particular sequence the increased staking plan improved matters:
So in this particular case the staking plan improved matters by £110.44 – quite a difference. A loss was still made, but in the grand scheme of things the loss equated to under 10%. Of course, it was fortunate that the longest losing run was 11 and also there were three winning tricasts after 10, 10 and 11 losers respectively. A losing run of say 19 followed by a small payout is something that could occur and potentially crucify your bank. However, I did take some heart in the results, especially as I hoped that in the real world I would be able to increase the 10% random success rate to nearer 15% or maybe even a little higher.
The problem that many of you will have come across in your own betting is that unfortunately that even spread of results is rare. Results for some bizarre reason seem to come in clusters – you are more likely to say have 2 winning bets in 3 and then 17 consecutive losers than a win every 10 bets. Why it happens like that is hard to say, but betting is never a smooth ride is it? Hence, I decided to take a more random approach to success and placed the winning “MMT” tricasts in one hat and the winning “MTT” tricasts in another. I then pulled out 5 from each hat to simulate a random series of 10 winning bets out of 100. What would the staking plan achieve this time??!
The winning bets on this simulation came on bets 3, 10, 17, 20, 35, 47, 53, 56, 77 and 86 – a much more random and indeed more life like set of results. The staking plan made a significant loss this time, but not as bad as a level stakes approach. The results were:
Again the staking plan proved much more successful than level stakes – this time to the tune of £126.05. However, these are more serious losses. This was partly due to the returns generally being below the respective averages for “MMT” and “MTT”.
At this juncture I feel that the only way forward is to test my ability to improve upon the likely 10% strike rate. I plan to run a trial of 100 bets over the next few months to very small stakes!! When completed I report back on my findings



Copyright PunterProfits.com 