Pick 5 or Empire 6?

[Mathcapper]

As we approach the end of the Spa meet with another mandatory payout day coming up, I thought it’d be worthwhile to take a look at the value of the Empire 6 bet, both independently and vis-à-vis the Pick 5.

Seeing as we’re less than 20 race days into the launch of the new bet, it’s still far too early to make any statistically significant conclusions, but I’m already seeing the expected tendencies in the Empire 6, and to an even greater degree.

Excluding the mandatory payout day earlier in the meet (which paid +400% more than the takeout-adjusted win parlay), the median payout has been +145% more than the takeout-adjusted win parlay.

To put that in perspective, as I noted in a post on the subject of the Rainbow 6 a few years ago, the median payout over 10 meets (290 days) I tracked of traditional non-carryover Pick 6’s on the NYRA and SoCal circuits was +27% higher than the win parlay. On carryover days, it jumped to +78% higher than the win parlay. And in the 158 days I tracked of the Rainbow 6, it was +98% higher.

At almost two and a half times the win parlay, the Empire 6 is currently paying better than all of them. Much of this is likely due to the same phenomenon we see in the Rainbow 6 (players spreading too deep, driven by both the $.20 minimum and their attempt to capture the jackpot, “sharpies” shunning the bet due to the perceived takeout disadvantage, etc.).

During the same timeframe since the Empire 6 began at the Spa, I also tracked the payouts for the Pick 3, Pick 4 and Pick 5 within the Empire 6 sequences. The median Pick 3 payout was +24% more than the equivalent win parlay, the median Pick 4 payout was +60.4% higher, and the Pick 5 was +123% higher.

So which bet is offering the best value?

To answer that question, we first need to take a look at how much each bet SHOULD pay. In general, the more legs involved in the sequence, the better the value is going to be, because you’re only getting hit with the takeout ONCE, and because you edge is multiplicative. But differences in the takeout between the various bets must also be taken into consideration, which can reduce or increase the value of the particular bet.

Let’s look at the expected value in each bet, after taking into account both of these factors.

The Pick 3 has a takeout of 24%, but because you’re only getting hit with that takeout one time in the sequence vs. three times at 16% in the win pool, the Pick 3 should, on average, pay +28.2% more than the equivalent win parlay.

Similarly, the Pick 4, which also has a 24% takeout, should pay on average 52.7% more than the equivalent win parlay, again, because you’re only getting hit with the takeout once.

The Pick 5 is even better, both because it has an industry-low takeout rate of only 15% and because you now have five legs where you’re only getting hit with the takeout once. As a result, the Pick 5 should on average pay +103% (more than double) the equivalent win parlay.

The Empire 6, another one of the often-derided jackpot bets, is subject to a much higher effective takeout than the Pick 5. With its initial 20% takeout, along with the additional 25% takeout for the jackpot pool, players are looking at an effective takeout of (1-0.8*0.75) = 40%. As I wrote about in my Rainbow 6 post, this may seem quite onerous, but when you compare it to the traditional Pick 6, which has a similar “takeout” for 5-of-6 consolation bets that get paid out to mostly non-winners of the Pick 6, an argument can be made that it’s close to a wash, and the evidence has been that’s it’s actually much better, for the reasons stated earlier.

In any event, the Empire 6, because you’re now looking at six legs where you’re only facing the takeout once, should pay on average +70.8% more than the equivalent win parlay.

All in all, just looking at it mathematically, the Pick 5 currently offers the better value than the Empire 6 (+103% vs. +70.8%). Not a surprise, given the 15% takeout vs the effective 40% takeout in the Empire 6, albeit mitigated by the inclusion of the extra leg in the Empire 6.

However, a mathematical interpretation doesn’t give us the full picture. When we look at what the bets are actually paying, we see differences to what the math is predicting. In the jackpot wagers, as we’ve seen with the Rainbow 6 and now with the early results for the Empire 6, the bet is paying much higher than predicted, likely for the reasons stated earlier.

So while the math is saying that the Pick 5 offers the better value (+103% vs. +70.8%), the actual results are saying that the Empire 6 is a slightly better bet (+145% vs. +123%). A couple of caveats though: (1) the sample size is still way too small and (2) the Rainbow 6, a similar bet to the Empire 6, was shown in a longer study to pay “only” +98% more than the parlay, so it’s quite possible, indeed likely, that the +145% initial early average will come down somewhat over time.

Another factor I should briefly touch upon relates to recent discussions about whether or not the computer guys are being kept out of the Pick 5. This may or may not be true, but the supposition is that if they are being kept out, we may see higher-than-expected payouts in the Pick 5 relative to the other horizontal exotics. It’s too early to tell, but based on the results above, the Pick 5 is offering better value than the Pick 4 and Pick 3 relative to expectations (+123% vs. +103% expected for the Pick 5, +60.4% vs. +52.7% expected for the Pick 4, and +24.2% vs. +28.2% for the Pick 3).

Again, the sample size is way too small right now to draw any conclusions. Day-to-day variations still have too much impact on the overall results, like today for instance where the Pick 3 paid +3.1% more than the win parlay and the Pick 4 paid -19.4% less than the win parlay but the Pick 5 paid a whopping +345% more than the win parlay, most likely driven by Remain Anonymous, the 15-1 ML firster in the first leg of the Pick 5 that got pounded to 5-2 in the win pool (6-1 in the Will Pays).

Over time, I’d expect the Pick 3, Pick 4 and Pick 5 payouts to converge on their mathematical expectations, unless the computer guys are indeed being excluded from the Pick 5, in which case it’s possible we see a little extra value in the wager (although not a certainty). The Empire 6 is another matter. Based on the previous studies related to similar wagers, I’d expect the bet to continue to provide more value than the mathematical prediction, albeit probably less than what may be a somewhat inflated +145% figure we’re seeing in the early going.

Until such time that we’ve gathered enough results, the jury’s still out on whether the Pick 5 or the Empire 6 offers the better value. Whichever the case, they both appear to be two of the best bets in racing, aside from the jackpot mandatory payout day and barring any introduction of a traditional pick 6 with a 15% or 20% takeout with NO jackpot and NO 5-of-6 consolation tickets (one can only wish).

Good luck everyone in Monday’s mandatory.

Looking forward also to Penn Derby day down at Parx and to seeing fellow T-Generates like Joe B. and any cameos from other members of the backyard crew.

Rocky R.

[Molesap]

This is awesome - I LOVE it when you post. Easy to follow with data to back up your contentions along with some conjecture and caveat. Thanks.

[johnnym]

Damn Mr.Capper thanks.

[NoCarolinaTony]

The Rock

[Al Caught Up]

quick question (and apologies if this has already been discussed): Why is it that the computer guys would not be involved in the Pick 5? Is it based on the payoffs you\'re seeing/the data? Is it only a NYRAbets wager? Just curious where this comes from (I\'m not doubting it one way or another, just wondering what the thoughts are on this). Thanks and a really interesting post.

[Fairmount1]

Already discussed but I\'ll post again.

On Whitney weekend (Friday I believe), a NYRA employee was at our tables and told us explicitly that the computer teams do not have access via a code (as I recall) to the LATE pick 5.  Obviously, you can\'t keep anyone from putting in a play.  But keeping them out of the pool for purposes of seeing all the other bets that have been made and putting in many, many bets in a single keystroke in the final seconds is what I believed is what was meant.  

Mathcapper and I have discussed at length and still working our way through what we individually and collectively think.  We have some similar thoughts and some different.  The pools are smaller on the late pick 5 than the early p5 is another reason I believe the employee knew exactly what he was saying re: no computer teams allowed in the late p5.  

Yesterday for example, early pick 5, $738k pool.  Late pick 5, $618k pool.

[Al Caught Up]

thanx--

[philywheel]

how can the empire 6 be a better bet then the pick 5,except on mandatory days. the take out I think is 20% and they hold 25% for the carryover, thats 45% of the pool,dont think you will see that wager again in NY

philywheel

[JohnTChance]

I hit Monday’s Pick-6 at Saratoga for 24k - playing the wager only because of the mandatory payout scenario. But then I looked at the winning 50 Pick-5 - my usual play, but not here - and saw a whopping 66k payout, a stunning amount I think. Seems about 200 tickets hit the Pick 6. And maybe 7 tickets won the Pick 5.

[Boscar Obarra]

You\'re suggesting the computer players have special access to multi race betting  pools (p3 and up) and can see all the bets ?

 Would be news to me , and not good news

[mjellish]

Don’t think that was what he was saying.  They can’t see “All the bets.”  What they can do is rum bet optimizations based on calculated win probability and odds for legs 1 and 2 in the sequence and look for overlays based on win odds in leg 1 and daily double probables and then projected win probabilities in the rest of the races in the sequence and then in ONE PUSH send all those optimized bets in right before the gates open in leg 1.  It is a big advantage when combined with rebates.

In theory, if a track cooperated, an adw site could actually display probables for a pick 3 as well if they plotted on an x-y-z axis and displayed as a cube graphic where you could click on the appropriate z axis coordinate.  Pick your x horse in leg 1, y horse in leg 2, and then click on the program number of your leg 3 z coordinate horse.  Could do same for trifectas.  Wonder what that would do for the bot players, or even the ham-n-eggers.

Thanks for the work and for sharing Rocky.

[Fairmount1]

MJ et al,

A year later, this topic worth revisiting.  

https://www.drf.com/news/jackpot-payouts-put-spotlight-computer-assisted-wagering-programs

Of note and already reported here re: exclusion of CAW\'s from NYRA Late pick 5:

\"And then there is the late pick 5 offered by the New York Racing Association, with a 50-cent minimum and 15 percent takeout.

Launched in the summer of 2017, the bet was first made available only to customers of NYRA Bets, which did not take on CAW operators as customers. It has since been expanded to all NYRA customers â€" but the bet still does not allow CAW operators in the pools. A late pick 5 offered on Saturday linking NYRA races with those at other tracks also does not allow CAW activity.

The bets are not advertised as being “CAW-free,” reflecting the caution exhibited by many racetrack operators when dealing with CAW activity. By drawing attention to the exception, racetrack operators are fearful that everyday bettors might begin to look askance at all the other pools in which CAW activity usually hits 30 percent of the betting.

NYRA officials are counting on the late pick 5 to generate data on the win rates of CAW operators and whether those win rates are cutting into non-CAW handle. Unfortunately, this year’s data will be specious: The COVID-19 pandemic has thrown so many new variables into betting models that conclusions based on 2020 data will be suspect.\"

[Fairmount1]

Further enlightenment on the jackpot topic from @CHRBMike from three July 27th tweets which are entirely from his twitter page.

(1) The California Horse Racing Board reviewed the wagering transactions that led to one account holder winning the entire Pick 6 pool Sunday at Del Mar on an $8 ticket. The records show that the account holder placed 8,613 individual wagers totaling $29,652.

(2) All wagers were placed in batches, the last batch approximately 4 minutes before the first leg of the Pick 6. All 8,613 wagers used just one horse in each race, or using the popular wagering terminology, with six singles. However, the denominations of those wagers differed.

(3) Approximately 70 percent were made in the traditional $2 increment. The rest were in various denominations of $4, $6, $8, $10, $12, $14, $16, $24, $26, and $60. The horse that paid $114 and triggered the massive payoff ($173,912) was on only 1 percent of the 8,613 tickets.

[Mathcapper]

Fairmount1 Wrote:
-------------------------------------------------------
> Further enlightenment on the jackpot topic from
> @CHRBMike from three July 27th tweets which are
> entirely from his twitter page.
>
> (1) The California Horse Racing Board reviewed the
> wagering transactions that led to one account
> holder winning the entire Pick 6 pool Sunday at
> Del Mar on an $8 ticket. The records show that the
> account holder placed 8,613 individual wagers
> totaling $29,652.
>
> (2) All wagers were placed in batches, the last
> batch approximately 4 minutes before the first leg
> of the Pick 6. All 8,613 wagers used just one
> horse in each race, or using the popular wagering
> terminology, with six singles. However, the
> denominations of those wagers differed.
>
> (3) Approximately 70 percent were made in the
> traditional $2 increment. The rest were in various
> denominations of $4, $6, $8, $10, $12, $14, $16,
> $24, $26, and $60. The horse that paid $114 and
> triggered the massive payoff ($173,912) was on
> only 1 percent of the 8,613 tickets.

Interesting. Either the rules are different in other venues or I was given bad info last year.

I specifically asked a fellow at one of the NYRA booths at the Spa about whether or not a higher-than-base minimum $.20 ticket like this $8 winning ticket (I used $1 as an example in my question) would count as a single winning ticket or cause the jackpot to carry.

He said he wasn\'t sure and would have to talk to his supervisor, then got back to me that evening with the following text message:

“Hi Rocky, this is Bill from NYRA. Talked to my supervisor about the $1 Empire 6. He said you would lose the “single winner” chance to take the whole pool if you did this.”

[FrankD.]

Gulfstream’s 🌈 as well Rock, it happened a few years back. The player hit repeat bet by accident, it was a small play that hit and would have been the only single ticket. I recall it was a decent or would have been a decent pay day but not life changing.

IMHO: It defeats the entire purpose of the bets structure. In the get more handle regardless of who or how it hurts/helps business mode, it’s just another advantage for the BOTS.... A 10% or 12% rebate on a $2 10k ticket or a $10 50k ticket? Plus they still collect a single ticket payout or an inordinate number of shared winning tickets/consolations.