light late p5 payout at TB

[dodie]

TB had a ~$26K carry over into it\'s late p5 Sunday.  It paid 9206.80 for a .50 cent bet.  I got suspicious when I saw the late p4 come back $11,334.85.  According to my calculations, and mathcapper, please check me, the .50 cent win parlay on the last 5 TB races paid $26,813.16, for a massive underlay in the P5 that had a modest carryover.  Winning prices in the p5 were 3, 44, 22.20, 24.40, and 6. I didn\'t even bother to figure out the p4 parlay, but it looks light too.  To add insult to injury, the p4 has a 18% takeout, the p5, %15.
Even when I win, it feels like I\'m losing.

[FrankD.]

Dodo Dodo dodie is back !
Just practicing my Trump tweet imitations......

It’s officially triple crown season with a severe outbreak of Derby fever a full 2 days before it’s officially Spring.

Red boards & MORONIC posts, stay tuned kids.

If you bothered to do the math? The pick 4 pool was a whooping 55K meaning post 18% takeout $2 worth takes down the whole pool. Don’t bother telling us you had a $4 pick 4....... The 5 pool was 166K with 26K in dead money so 140 @ 15% yields
$8 worth of conspiracy taking out a pool with a couple of Genii wheeling 1/2 and 2/1 shots. We know you singled the $44 horse.......

THEY ARE  BACK !!!

[dodie]

Thanks for your input, Frank!  I will try to be a good TG citizen this year!  I apologize if I have offended you in the past.  Cheers!

[Al Caught Up]

what\'d you do, kick the guy\'s dog? Sheesh.

[Boscar Obarra]

Anyone using \"Genii\"  is automatically enrolled in my personal Hall of Fame.

[Mathcapper]

dodie Wrote:
-------------------------------------------------------
> TB had a ~$26K carry over into it\'s late p5
> Sunday.  It paid 9206.80 for a .50 cent bet.  I
> got suspicious when I saw the late p4 come back
> $11,334.85.  According to my calculations, and
> mathcapper, please check me, the .50 cent win
> parlay on the last 5 TB races paid $26,813.16, for
> a massive underlay in the P5 that had a modest
> carryover.

dodie --

The $.50 pk5 parlay was actually $6,703 not $26,813 (that was the $2 parlay), which explains most of the perceived underlay. Having said that, it still came back a little light indeed. It paid +37% more than the parlay, but with the 15% take it should pay a little more than twice the parlay at TB on average (+115% more to be exact), and a little more than that on the day in question given that $28K carry.

The late pk3 came in even lighter at -19% less than the parlay. The pk4 on the other hand, came back much better than expected. The $11,334 $.50 payout was 2.5x the $4,468 equivalent parlay (on average it should pay around +73% more than the straight parlay after considering takeout).

No real surprises though in any of the payouts, especially considering the small pools. When you get a scenario like the pk4 pool where there was only 1 $2 equivalent winning ticket, when +1 or -1 winning ticket can produce drastically different payouts, the numbers often vary significantly from expectations based on the equivalent win parlays.

[dodie]

Thanks for the cogent and civil reply, math.  Long division, my old nemesis, gets me again.  Great explanation of what fair payoffs should be.

[TGJB]

Here\'s one for you, Rocky, re the Rainbow 6 at GP and Lrl, and the 5 at FG. On days where there\'s more than one winning ticket they don\'t distribute the whole pool, obviously. Is that chop more than offset by there only being one takeout?

[Boscar Obarra]

Theory goes that the fractional pool payout is well offset by the fact that there is so much dead money bet on near impossible combinations.  20c encourages lots of spreading.

   The one takeout is also in play as with any multi race bet.

    I know I\'m not Rocky and don\'t pretend to be.

[TGJB]

I\'m talking straight math here, one takeout vs only paying out part of the pool. And yes, that aside, the payoffs flatten out, especially when there\'s a mandy payout.

[Mathcapper]

TGJB Wrote:
-------------------------------------------------------
> I\'m talking straight math here, one takeout vs
> only paying out part of the pool. And yes, that
> aside, the payoffs flatten out, especially when
> there\'s a mandy payout.


Jerry â€" I’m not quite sure what you mean by “one takeout” so I hope I’m addressing your question below, let me know if I’m misunderstanding.

I’m not too familiar with the jackpot wagers at Lrl or FG (although I suspect the Lrl wager is similar to GP), so I’ll just speak to the Rainbow 6.

To my mind, there’s always two takeouts associated with pick 6 wagers, be it the traditional version or the jackpot variation.

In a traditional pick 6 like at Santa Anita, there’s a 23.68% takeout. Another 30% of the pool is then paid out in the form of consolation tickets, which I view as an additional “takeout” since that money rightfully belongs to pick 6 winners but is instead being shared with the other 99% of bettors who only selected 5 winners. So it can be argued that the effective takeout on pick 6 winners (notwithstanding a winner’s consolation tickets) is really closer to 1 - (1-.2368)*(1-.30) = 46.6%.

In a jackpot pick 6 like at Gulfstream, there’s a 20% initial takeout. Then another 30% of the pool is carried over to the jackpot pool (unless there’s a single winner), resulting in an effective takeout of 1 â€" (1-.20)*(1-.30) = 44%.

So the effective takeout of a traditional pick 6 vs. a jackpot pick 6 is pretty similar. It’s not an apples-to-apples comparison of course, but a case can be made that the end results are similar in the long run. A traditional pick 6 player gets some of that consolation takeout back in the form of his own consolation tickets during the course of his play, while a jackpot player gets some of that jackpot takeout back through the bonanza positive expectation mandatory payout days, assuming he’s lucky enough to grab a piece of those mandatory day payouts once in awhile.

The one time when there isn’t two takeouts is in the jackpot variation when there’s a lone winner or a mandatory payout, in which case the effective takeout (at GP) is just 20%, which is great when you’re the lone winner. On the surface it would seem even better on a mandatory payout day, but somewhat counterintuitively, the value is often as good or better on the days that are not mandatory payout days.

There’s no mathematical reason why this is often the case â€" it’s a phenomenon that can only be observed empirically, by comparing the actual payouts to the expected payout based on a parlay of the win prices, on non-mandatory payout days vs. mandatory payout days.

What seems to be happening, as Boscar alluded to, is that on non-mandatory payout days, the crowd money is skewed towards longer-priced combinations in an attempt to take down the entire pool, which creates tremendous overlays on logical combinations. Even so, the mandatory payout day, with it’s lone 20% takeout and monster free carryover pool, should in theory produce higher overlays, and indeed it’s often a positive expectation bet, but the resultant value is often mitigated by the fact that so much money is bet on that final day (often 2-3 times the amount of the jackpot carryover) that it mutes the positive effect of all that free money. And unlike the days leading up to the mandatory payout, the crowd money is now skewed the completely opposite way, towards favorites and logical combinations rather than underlaid longshots, the thinking being that even a short-priced sequence will produce a bonanza payoff because of all that free jackpot money. So when those logical combinations do come in, but resultant payoff is often rather underwhelming.

If the question posed at the top of this post relates to the fact that you only get hit with the takeout once in a pick 6 vs. 6 times in a win parlay, the answer is that the pick 6 still has a higher expectation than a win parlay for both a traditional pick 6 and the jackpot variation, even after taking into account the additional drag of the consolation “takeout” and the jackpot pool, respectively.

In a traditional pick 6 like at Santa Anita, with a 15.43% win takeout, the expected pick 6 payout is (1-.2368)*(1-30)/(1-.1543)^6-1 = +46.0% more than the equivalent win parlay.

In a jackpot pick 6 like at Gulfstream, with a 17% win takeout, the expected pick 6 payout is (1-.20)x(1-.30)/(1-.17)^6 -1 = +71% more than the equivalent win parlay.

I’m not sure about that FG pick 5 jackpot wager. From what I see, 50% of the pool is carried over to the following day unless there’s a single winner. Since there’s no consolation tickets in a traditional pick 5 wager like there is for pick 6’s, there\'s only one real takeout - the typically very low 15% vs. an effective takeout of 50%(!) for the jackpot version. It might be worth it to play on some large carryover days, but on a non-carryover day, phew â€" NO THANKS.

I haven’t followed the betting associated with that FG jackpot pick 5 wager, but I’d be surprised of it’s attracting much attention vis-à-vis the traditional version, which is a fantastic bet given that it’s such a low takeout on such a superexotic wager (the value gets better the more races that are included) and with no additional rake for consolation tickets.

It’s no wonder the low takeout traditional pick 5 bet has been such a success. Tracks should take a cue from that success and model the Pick 6 after it. Lower the take to 15% (or even 20%), lower the minimum to $.10 or $.20 like in the Rainbow 6, and eliminate any consolations and jackpots.

Because 25-30% of a traditional pick 6 pool is distributed in the form of consolation tickets, winning Pick 6 players are in effect subsidizing the lousy handicappers, much in the same way that good Rainbow 6 players are subsidizing the “lucky” one-time jackpot winners.

This is the main problem with the takeout structure of the Pick 6. It’s not that it isn’t a good bet. Because of the fact that you only get hit with the takeout once, and because your edge is multiplicative, superexotics like this are still very good bets, even with the higher takeouts typically associated with them, and even with the punitive consolation or jackpot gimmicks, both of which act as an additional “tax” on day-to-day winning handicappers.

Stop subsidizing losing and lucky players, and put the money in the hands of the winning players.

As we’ve seen with the 14-15% takeout Pick 5’s, pools will likely escalate substantially, and the payouts for winning combinations will be demonstrably higher.

Now that’s a bet worth making.

(sorry for the rant - JMHO)

[Furious Pete]

Great explanation Mathcapper, I do enjoy your posts.

However your last point seems a bit short-sighted and simplistic imho:

\"Stop subsidizing losing and lucky players, and put the money in the hands of the winning players.\"

Don\'t you see any longterm benefits in keeping those losing players in the game? And that these kind of \"gimmicks\" actually keeps the game alive, also for those with a positive expectation on their bets?

I guess it\'s politics. It\'s complicated.

[Mathcapper]

Furious Pete Wrote:
-------------------------------------------------------
> Great explanation Mathcapper, I do enjoy your
> posts.
>
> However your last point seems a bit short-sighted
> and simplistic imho:
>
> \"Stop subsidizing losing and lucky players, and
> put the money in the hands of the winning
> players.\"
>
> Don\'t you see any longterm benefits in keeping
> those losing players in the game? And that these
> kind of \"gimmicks\" actually keeps the game alive,
> also for those with a positive expectation on
> their bets?
>
> I guess it\'s politics. It\'s complicated.

Pete â€" I’m all for increasing churn, I just don’t think making pick 6 winners subsidize the other players is the way to go about it.

None of us are in this to lose money. It’s hard enough to beat the game when you’re trying to buck a blended 20%+ take. On top of that, you’re asking pick 6 winners to give up another 30% of their profits in the form of consolation tickets or jackpots. So a winner of a gross pool of $100K, after the double whammy of the track take and the consolations/jackpot, ends up taking home a net of  around $55-$60K, give or take depending on the type of pick 6 and its particular takeout. That’s a big whack, one that is very hard to overcome over time unless you’re a really good handicapper.

When you subsidize other players with the winners’ profits, you’re making the bet that much more difficult to beat for everyone. If players don’t feel like a bet is beatable, they’re not going to play. When the effective takeout is that onerous, the concern becomes less about churn and more about revenue/liquidity.

Imagine if you hit a $100 daily double, and then you had to give up $25 to the other 5 guys sitting at your table who played the bet but only had the odds-on winner of the first leg. After awhile, you might find it’s so hard to come out ahead of the game because you’re giving up all that money every time you win that you end up giving up on the bet altogether. Same for the other guys at the table, since they’re not beating the game either. So the end result is that eventually everyone just stops playing.

By the way, an argument can be made that subsidies aren’t the exclusive realm of pick 6 bettors. To my mind, the granddaddy of all subsidies is given to the biggest players in the game in the form of rebates.

Because of the huge rebates they receive, whales like the computer guys and the arb guys can drive down prices on horses and still show a profit after their rebate. Sure, an argument can be made that when they drive down the price on one horse it creates value on the others, but the computer guys are pretty good at identifying value, and they’re sucking the value out of overlays even more so because the rebates allow them to drive prices down to even below fair value and still make money. They can take a horse that’s a decent overlay at 2-1 and bet it down to 8-5 or 3/2, a price at which no longer offers value to the general public but which is just fine for them after rebates.

We’ve seen the deleterious effect of this “whale welfare” more and more over the years. The favorite-longshot bias has now been mostly arbitraged away. Favorites are winning at a higher clip, and they’re paying less. Value, particularly amongst logical contenders, seems to be much harder to come by these days.

It’s a zero sum game (net of takeout, not gross). After the track takes their cut, there’s a finite pool to be distributed amongst the bettors. So when you put money into the hands of some players (be it in the form of rebates, consolation tickets, jackpots, or whatever) that money by definition has to be taken out of the hands of others. The net result of all this is that it makes it much harder for the general public to make money. Bill Benter himself has estimated that the net effect has been to raise the takeout on the general public by approximately 2 percentage points.

In a perfect world, everyone would be on a level playing field. They\'d all get the same rebate (or preferably, rebates would be eliminated and takeout would be lowered across the board), there would be no consolation tickets and no jackpot gimmicks. If you want to increase churn and keep players in the game, watch what happens when horseplayers start getting more money in their pockets every time they win a bet. Of course we’ll probably never see the current rebate structure go away because if that were to happen, the elimination of whale welfare would likely drive much of  the whale population to the point of extinction, thereby causing  a cascading effect of plummeting revenue and lower liquidity that could in turn lead back to higher takeout and on and on.

But in a perfect world...

[johnnym]

Thanks Math,was thinking of taking a shot at the Rainbow 6 Saturday but know maybe I won’t.

[Mathcapper]

johnnym Wrote:
-------------------------------------------------------
> Thanks Math,was thinking of taking a shot at the
> Rainbow 6 Saturday but know maybe I won’t.


johnnym â€" please don’t get me wrong, the point I was trying to make in these recent posts was that consolations and jackpots reduce the amount of money pick 6 winners receive for their winning tickets, and that pick 6 players would be better served by having a lowered takeout like the 14-15% level offered on pick 5’s at many venues, with no consolations or jackpots.

But that doesn’t mean the pick 6 is a bad bet by any means. On the contrary, because of the fact that you only get hit with the takeout once instead of 6 times in a win parlay, and because your edge is multiplicative, it’s actually a very good bet, as long as you’re bankroll can withstand the roller coaster ride associated with low win rates and windfall payouts.

And make no mistake, the mandatory payout day of a jackpot pick 6 is THE best bet in racing. On mandatory payout days at Gulfstream, there’s only a 20% takeout, with no consolation tickets and no rake for the jackpot, plus all of that massive free carryover money. It’s one of the few days in racing when horseplayers are almost guaranteed to be dealing with a positive expectation bet. On the March 27, 2016 mandatory payout day, players were looking at a +21.8% expectation (see: Rainbow 6 3/27/16), and on the recent Jan 28, 2018 mandatory payout day, they had a +5.3% expectation.

Aside from the mandatory payout days, the non-mandatory payout days are also a very good bet, even after considering the additional 30% rake for the jackpot. This is due to an unintended consequence that I posted about a couple years ago (see: Is the Rainbow 6 a Sucker Bet?). Because everybody is chasing the jackpot, significant overlays are being created on a day-to-day basis on logical sequences, resulting in payouts whose value often exceed the amount seen on typical carryover days of traditional pick 6’s.

On both mandatory and non-mandatory payout days, it’s not uncommon to see payouts 2-3 times greater than the equivalent win parlay or more. On the mandatory payout day it’s almost a given due to the fact that you’re almost certainly dealing with a positive-expectation bet going in.