JCapper Message Board

General Discussion
-- March 2015 HANA Monthly

Home	Register Log In

March 2015 HANA Monthly

jeff
3/12/2015
11:50:05 AM

The new HANA Monthly is out.

Link here:
http://www.horseplayersassociation.org/hanamonthly.html

--quote:

"Is The Rainbow 6 For Suckers?

This post originally appeared on the Thoro-Graph "Ask the Experts" thread and is reprinted with permission.

The general consensus is indeed that the Rainbow 6 is a sucker bet. Andy Beyer has said as much, as has Steve Crist, who advised not playing the bet unless there's a seven-figure carryover or it's the last day of the meet.

Their reasoning? The fact that 30% of the pool, after the initial 20% takeout, goes into the jackpot pool unless there's a single winner, resulting in what they see as an exorbitant 44% effective takeout. Critics like Beyer argue this makes the Rainbow 6 a worse investment than the lottery and that it's pointless to fight such a takeout.

But are they right? Or is it (Jim's point of view on the Thorograph board) that "the 6/6 payouts have been pretty decent" more accurate?

The best way to answer this question is by comparing the actual Rainbow 6 payouts to those predicted based on the equivalent win parlays. As some may recall from my posts in a thread last fall (Robo Betting), I've tracked actual vs. predicted Pick 6 payouts for a number of meets over the years. I've also been tracking the Rainbow 6 payouts at GP since they began.

Here are the results for the 10 meets (from Aqu/Bel/Sar and Hol/SA/Dmr) that I've tracked over the years, for both noncarryover and carryover days, along with the results for the Rainbow 6 (based on an equivalent $2 bet). To be conservative, for the Rainbow 6 I've excluded all lone winner carryover days and mandatory payout days, which are almost always massive overlay payouts relative to the equivalent win parlays.

NYRA/SoCal Noncarryover days (10 meets, 290 days): Median Payout: 27% higher than equivalent win parlay - $14,092

NYRA/SoCal Carryover days (10 meets, 102 days): Median Payout: 78% higher than equivalent win parlay - $31,082

GP Rainbow 6 (12/2/12-3/15/14, 158 days): Median Payout: 98% higher than equivalent win parlay - $30,448

That 98% figure is not a typo..."

--end quote

Read that (and more) at the link:
http://www.horseplayersassociation.org/hanamonthly.html

-jp

.

Acorn54
3/25/2015
4:06:03 AM

hi jeff
i hope it is okay for me to chime in here on the rainbow pick six
although the winning tickets are a great value i ask myself, how often will i get a winning ticket? i have a 2-1/2% hit rate on exactas so can only imagine what, if any tickets i would play on the rainbow pick six would be cashed.
it seems like there would be long time between drinks, so to speak, with an outflow of betting dollars, before getting a winning ticket.

jeff
3/25/2015
2:35:57 PM

You should absolutely expect to go a long time between drinks when you go after high end multi race exotics (pick4, pick5, pick6, etc.)

In a pick6...

IF:
P(1) is the probability of winning leg 1
P(2) is the probability of winning leg 2
P(3) is the probability of winning leg 3
P(4) is the probability of winning leg 4
P(5) is the probability of winning leg 5
P(6) is the probability of winning leg 6

AND IF:
P(ALL) is the probability of winning all 6 legs

THEN:
P(ALL) = P(1) x P(2) x P(3) x P(4) x P(5) x P(6)

The player does have the ability to control the values for P(1), P(2), P(3), etc. by deciding how many (or how few) horses to include in each leg.

If you've seen enough Data Window output with the data broken out by rank for factors like UPR, HDW PSR, QRating, PRating, SilverRating, etc. you can get a pretty solid understanding about win probabilities for each rank.

A win prob by rank table for most of these factors looks (approximately) something like this:

RANK  WIN PROB
----  --------
   1      0.30
   2      0.20
   3      0.15
   4      0.10

Note: Keep in mind that the above values are approximations only and that they will vary according to the influence of other factors such as the odds, field size, etc. Also keep in mind that you can use the Data Window as you see fit to generate your own win prob table(s) for various unique situations.

That said, in the interest of simplicity, I will use the rank-win prob values from the above table in the example sequence presented below.

Example Pick6 Sequence:

Let's assume that you have a strong opinion about the value of the rank=1 horse in the very first leg. Based on that you decide to single that horse in leg 1. The (P1) calculation is easy. P(1) is approximately equal to the rank=1 win prob value from your table:

P(1) = 0.30

In leg 2 let's assume you are looking at a (smallish) 5 horse field and you that you further think the rank 3-4-5 horses are likely to be overbet in the pool. Because you think covering the rank 3-4-5 horses destroys the value of your ticket you decide to limit coverage in leg 2 to the rank=1 and rank=2 horses only. The (P2) calculation is easy. P(2) is approximately equal to the sum of the rank=1 and rank=2 values in your table:

P(2) = (0.30 + 0.20)

or

P(2) = 0.50

For the sake of argument let's assume the following about the remaining 4 legs of the sequence:

a. You don't have any real opinions in the remaining legs like you did in the first two.

b. Even though you could force the prob values for P(3), P(4), P(5), and P(6) to be equal to 1 for each of the remaining legs by using the ALL button - you decide not to do that because you have a limit as to how much money you're willing to risk on today's sequence.

c. Based on the above, you descide to cover the rank=1, rank=2, rank=3, and rank=4 horses for each of the remaining 4 legs of the sequence.

The (P3) calculation is easy. P(3) is approximately equal to the sum of the rank=1, rank=2, rank=3, and rank=4 values from your table:

P(3) = (0.30 + 0.20 + 0.15 + 0.10)

or

P(3) = 0.75

And of course, in this case, your P(4), P(5), and P(6) values are the same as P(3):

P(4) = 0.75
P(5) = 0.75
P(6) = 0.75

Plugging those numbers into the formula, the P(ALL) calculation for getting all six legs right looks like this:

P(ALL) = P(1) x P(2) x P(3) x P(4) x P(5) x P(6)

or

P(ALL) = (0.30) x (0.50) x (0.75) x (0.75) x (0.75) x (0.75)

or

P(ALL) = 0.0474609375

Ticket Cost:

Assuming the example sequence presented above is for a Rainbow 6 at Gulfstream and assuming a base wager amount of $0.20 your ticket cost is: $102.40.

Assuming the example sequence presented above is for a standard pick 6 and assuming a base wager amount of $2.00 your ticket cost is: $1,024.00.

OddsLine or StrikePrice Calc:

Even though you put a little thought into ticket structure and even though you might have a 75% chance (individually) of winning legs 3 through 6...

Your chances of winning the example sequence presented above are only about 1 in 21.

The "OddsLine" or "StrikePrice" calculation for winning all six legs looks like this:

StrikePrice = (1 / P(ALL)) - 1

or

StrikePrice = (1 / 0.0474609375) -1

or

StrikePrice = 20.07

That means (without considering the effect of any consolation payoff) if (in the example sequence presented above) you are putting $102.40 into a Rainbox 6 at Gulfstream, you should have a very strong belief beforehand that if you have all six winners on your ticket - but your ticket turns out to not be the lone winning ticket - the parimutuel payoff will be at least $2,056.00 or 20.07 times the $102.40 cost of your ticket.

If you don't think the parimutuel payoff under those circumstances is going to be at least that amount: I say you have no business betting the sequence in the first place.

That also means if (in the example sequence presented above) you are putting $1024.00 into a standard pick 6, you should have a very strong belief beforehand that if you have all six winners on your ticket - but your ticket is not the lone winning ticket and you have to split the pool with other players - the parimutuel payoff will be at least $20,552.00 or 20.07 times the $1,024.00 cost of your ticket.

Again, if you don't think the parimutuel payoff is going to be at least that amount: You have no business betting the sequence.

Final thought same as the first thought:

You should absolutely expect to go a long time between drinks if you go after high end multi race exotics (pick4, pick5, pick6, etc.)

Before deciding to even get involved with high end multiple race exotics you need to be mentally prepared to handle the bankroll draw downs caused by the lowish hit rates like that of the example sequence presented above.

-jp

.

Acorn54
3/25/2015
5:04:38 PM

thanks,jeff, for the elaborate explanation. I appreciate your time in explaining thoroughly what I would be getting into if I decide in the future to take on the rainbow pick 6.
bottom line, if I understand right, is one has to layout more than is my "comfort zone". personally my track record is have drawdowns typically of $200 or so betting exactas, and that is betting only $1 units. (my records indicate I have a 2-1/2% hit rate with them). I use win bets as my "savers.

JCapper Message Board

General Discussion -- March 2015 HANA Monthly

General Discussion
-- March 2015 HANA Monthly