Del Mar saturday

[shanahan]

late pik 4:
Race 6 - 2YO:
I have been beaten in the pik 4 three straight times by 2YO races.  I did not use the TG sheet for this race.  Mike Joyce is singling the 4 due to blinker addition.  Unfortunately since mid-September, this horse has  been to the track exactly twice to work (both in last 2 weeks).  Tossing.  Top 2YO sires for the 5 6 8 9.  The 3 has a very high 1st time ROI.  Using all 5, hoping one may scr.

Race 7:

Miller stock is running new tops and that puts the #1 in the mix.  4 is third out and improved 4 pts on turf.  In.  6,3,9 logical.  11 has paired 6\'s and could go forward again under Desormeaux.  1,3,4,6,9,11 - another spread.

Race 8:

This is where it gets interesting.  In the last week most of the winners had a TG synth track figure below 14.  At least for the pik 4s I\'m playing.  The two in this race that fit that are #1, and #4.  Thoropattern suggests the 1 regresses from his last pair.  First time Hollendorfer at 27% under Mike Smith is enticing, but if he runs back to his initial fig, he\'s no faster than several.  The celebrity horse drifted in per the notes in his initial foray, so not taking that one.  While # 8 has no idea he ran in a MCL last time, the jump forward was too much for me to take...which leads me to #2 for hot Dan Hendricks.  I\'m willing to overlook the lone poly fig - awful - after consecutive tops, and now after a brief freshening, he\'s back.  Expecting new top.  Singling him for those of you who thought I was going \"all\" for the pik 4.  If some scratches along the way, I may add the 1 back in - per Alan\'s suggestion on the show.

Race 9:
Expecting a bounce for # 7.  Synthetics under the 14 fig are the 3,8,6,9,10 - and 6,9,10 are almost as fast as the #4, so I\'ll take 4 in the finale.

$60 bucks total.

[SoCalMan2]

The payouts on this sequence were as follows:

6th Race -- $21.20
7th Race -- $24.60 ($2 DD paid $233.40 -- Equiv. Parlay $260.76)
8th Race -- $36.20 ($1 P3 paid $4543.10 -- Equiv Parlay $4719.76)
9th Race -- $16.60 (50 cent P4 paid $7900.05 -- Equiv Parlay $9793.49)

Assuming I did my math right, The parlay always paid better than the horizontal exotic equivalent and the discrepancy got worse the longer the horizontal was.  Now, maybe I did my math wrong here. I freely admit I could have done it wrong. Or maybe there is some reason the parlay pays better than the horizontal.  I had always thought that the horizontal should pay better because less take out.  I recognize there are different pools and different people playing so there will never be a match...but I thought the law of averages would have dictated a different comparison.

Have I done my math right?  Is this normal or was today\'s payouts unusual? If this is what one should normally expect, it would seem a little odd to be betting the horizontals when the parlays are a better value. Apologies if i am making some rookie errors here.

[Rick B.]

Those that swear by comparisons between pickers
and parlays will point out that you erred by
comparing $1 and 50 cent picker results to $2
win parlays, and in abstract I would agree...
but I am no fan of these comparisons because
they are built on (other) faulty assumptions:

1. The pools are not indexed to one another, so
the relationships between them (if you believe
it exists) will be imperfect to some degree.

2. Who bets parlays? Most any bettor can afford
a $32 pick-4 play that gets lucky and stabs up
a $7900 payoff...but who among us has the gonads
to pull the trigger on each part of the related
parlay, making a successively larger wager in
each leg?

The last leg is $1180 to win on the #12, Regally
Soul...can you make that bet, the full amount, no
shaving or hedging?

If you can\'t then the parlay is simply a myth,
and the Pick-x vs. Parlay comparison is BS.

Therefore, the only error you and other pool
cross-checkers are making is self-torture, when
you discover that a given Pick-X payoff is \"out
of skew\" with the corresponding parlay. It doesn\'t
mean a damn thing.

[BitPlayer]

I think you have the pick-3-equivalent parlay too high by a factor of 2.  ($10.60 after race 6, times 12.3 gives you 130.38 after race 7, times 18.1 gives you 2,359.88 after race 8.)  The other two look right to me.

[SoCalMan2]

Thank you for the math correction.  I see the error of my ways on that one.  And, it becomes interesting....the Pick 3 was a huge reward over the parlay while the other two the parlay paid better....which to me just goes to show the pools are different and anything can happen.

[Boscar Obarra]

In a pick 3 with 10 horses each there are 1000 combinations.

 Say there\'s a 100k pool and no take out for simplification.

 So each combo would have on average $100 on it, given equal odds.

 $100, not much is it? Take\'s very little to skew a payoff , especially lower.  It\'s actually a minor miracle that the payoffs are as close as they are to the \'expected\' results.

[Mathcapper]

You’re right that the horizontals should, on average, pay more than the equivalent win parlay. This is because, even though the takeout on horizontal exotic bets is higher than on win bets, you only get hit with that takeout once. The more races involved in the bet, the greater this effect.

At Delmar, the math implies the bets should pay, on average:

DD: +8.1% more than the win parlay
Pk3: +26.2% more than the win parlay
Pk4: +49.2% more than the win parlay

In the results you posted:

$2DD paid $233.40. This was -10.5% less than the $260.76 win parlay. The expected payout was $260.76 x 1.081 = $282.
$1Pk3 paid $4,543.10. This was +92.5% more than the $2,359.88 win parlay. The expected payout was $2,359.88 x 1.262 = $2,978.
$.50Pk4 paid $7,900.05. This was -19.3% less than the $9,793.49 parlay. The expected payout was $9,793.49 x 1.492 = $14,612.

As Boscar noted, results can vary widely between any individual sequence, but over the long run they do converge on the expected payout based on the math.

A couple of years ago I looked at the results for the winter/spring meet at Santa Anita to confirm the math. Even though my sample size was only 80 days, I found that the results were already converging to the expected averages:

2012 Santa Anita Winter/Spring Meet (Races 1-5 early Pick 5 sequence - 80 days)

Median DD payout: +12.9% more than the win parlay vs. +8.1% expected
Median Pk3 payout: +32.2% more than the win parlay vs. +26.2% expected
Median Pk4 payout: +61.2% more than the win parlay vs. +49.2% expected

Two thirds of the DD payouts fell between -13% and +42% of the win parlay vs. the +8.1% expected average
Two thirds of the Pk3 payouts fell between +7% and +65% of the win parlay vs. the +26.2% expected average
Two thirds of the Pk4 payouts fell between +11% and +110% of the win parlay vs. the +49.2% expected average

This should give you an idea of expected payoff for each type of horizontal wager and the variation you are likely to see around those expected payoffs.

- Rocky

[SoCalMan2]

Mathcapper Wrote:
-------------------------------------------------------
> You’re right that the horizontals should, on
> average, pay more than the equivalent win parlay.
> This is because, even though the takeout on
> horizontal exotic bets is higher than on win bets,
> you only get hit with that takeout once. The more
> races involved in the bet, the greater this
> effect.
>
> At Delmar, the math implies the bets should pay,
> on average:
>
> DD: +8.1% more than the win parlay
> Pk3: +26.2% more than the win parlay
> Pk4: +49.2% more than the win parlay
>
> In the results you posted:
>
> $2DD paid $233.40. This was -10.5% less than the
> $260.76 win parlay. The expected payout was
> $260.76 x 1.081 = $282.
> $1Pk3 paid $4,543.10. This was +92.5% more than
> the $2,359.88 win parlay. The expected payout was
> $2,359.88 x 1.262 = $2,978.
> $.50Pk4 paid $7,900.05. This was -19.3% less than
> the $9,793.49 parlay. The expected payout was
> $9,793.49 x 1.492 = $14,612.
>
> As Boscar noted, results can vary widely between
> any individual sequence, but over the long run
> they do converge on the expected payout based on
> the math.
>
> A couple of years ago I looked at the results for
> the winter/spring meet at Santa Anita to confirm
> the math. Even though my sample size was only 80
> days, I found that the results were already
> converging to the expected averages:
>
> 2012 Santa Anita Winter/Spring Meet (Races 1-5
> early Pick 5 sequence - 80 days)
>
> Median DD payout: +12.9% more than the win parlay
> vs. +8.1% expected
> Median Pk3 payout: +32.2% more than the win parlay
> vs. +26.2% expected
> Median Pk4 payout: +61.2% more than the win parlay
> vs. +49.2% expected
>
> Two thirds of the DD payouts fell between -13% and
> +42% of the win parlay vs. the +8.1% expected
> average
> Two thirds of the Pk3 payouts fell between +7% and
> +65% of the win parlay vs. the +26.2% expected
> average
> Two thirds of the Pk4 payouts fell between +11%
> and +110% of the win parlay vs. the +49.2%
> expected average
>
> This should give you an idea of expected payoff
> for each type of horizontal wager and the
> variation you are likely to see around those
> expected payoffs.
>
> - Rocky

Thanks tons!  This is fascinating stuff.  Based on your numbers, it seems like the Pick 4 on this particular day (Saturday) was a true outlier.  Expected would have been $14,612 and actual was $7,900....compared to what looks like the \"normal\" range, it would look like something like this should only happen very few times in a relatively long period of time. I could see the situation occurring where the pool is not big enough to support what some outcomes \"should\" be, but this was a saturday at Del Mar pool that should not have been size hampered (I would think).  However, taking Boscar\'s point....I bet that it easily could have been something like an extra $10 on the combination is enough to cause the skew.