|
You are stating the obvious that one needs to pick winners (as opposed to place or show horses) in horizontal bet sequences such as DD, P3, P4, etc. As I have frequently stated in other posts on this board, there are some races that will be part of such a sequence that are quite predictable, and other races that are more open or less predictable. In fact, in such predictable races, they are likely to be won by the first or second public choice, and it would be folly to leave them off of your horizontal bet ticket.
But betting a P3 or P4 sequence with ALL "predictable" type races will be -EV, even without knowing the payoff. At an extreme, if you bet the top 2 ML choices in each race, you will cash a lot of P3 or P4 tickets, but the payoff will not cover your loses. Lots of obvious winners, but the takeout and overbetting of these combinations will kill you.
As you know, the goal of a horizontal bet is to improve on your payout versus betting a parlay and having the track take deducted from each race. Since you like to test things, take a look at daily doubles where the sequence is a predictable race (likely to be won by the top 1 or 2 betting choices) followed by an unpredictable race (unlikely to be won by the top 1 or 2 betting choices). Since I have done the research, I will save you the time: you will be hard pressed to find a DD payout that exceeds the parlay payout.
Conversely, if you bet the sequence in a DD where the first race is unpredictable and then second race is predictable, your DD payoff will almost always beat that of a parlay bet. Why? Because most will cover the first race with the vulnerable favorite or second favorite in the DD, so you get the advantage in that pool when they do not win. And there is no real need to see the odds of the horses in the second race, since you figure that only 1 or 2 has a real chance to win anyway.
Sure, it is hard to "know" the closing odds these days with the simulcast dump. But just because it is hard doesn't mean you should not try to make an estimate. I am sure, with your experience, you can estimate final odds of most horses at tracks you frequently play. And I am positive you can tell when the ML is way off based on your own knowledge. We play a game of decision making under uncertainty. Estimation of variables is part of the game. Precision is nice, but not always required.
Finally, I don't really know you except from this board (but hope to meet up in August at SAR), but I do know and have been a part of successful betting syndicates in both horse racing and sports. They are ALL price sensitive and deal in probabilities: if the race was run 100 times, how often will each horse win? Pari-mutuel betting with simulcast dumps makes it difficult, but not impossible to estimate whether they are getting paid to take the risk.
Don't make the mistake of thinking any winning bet is a "good" bet. That is ex post thinking. If you are not getting paid to take the risk (i.e., there is a probability of the horse losing = 1 minus it's win probability) you pass.
Finally, all of the professional bettors use a form of Kelly Criterion betting. They do NOT look for each individual horse to be "value", as you put it. They look +EV of the group of horses they are betting. So in your P3 example, they will certainly include "obvious" horses in the group, as long as there is a healthy mix (via the math) of underbet horses as well. That is the beauty of Kelly: by including some slight "underlays" you will increase your win % enough to grow your bankroll faster since you can bet bigger (fewer losing streaks), even though your ROI is reduced. Real value betting is all about estimating win probabilities, NOT JUST betting so-called "overlays."
Richard
|