Thursday, June 21, 2012

16% to 18% Takeout Equals 25% Long Term Losses for the New Would be Racing Fan

Does the title for this article sound like an exaggeration? Or maybe a tad bit harsh?

Well, read on. As you are about to discover, the headline is absolutely true and betting on horse races is a harsh game.

If you work in track management, if you are a horseman, or if you are a politically appointed state regulator – this article was written for your benefit.

As members of track management, elected leadership of horsemen’s groups, and politically appointed state regulators – you have played an active role in setting takeout levels.

And in so doing you have created a comically poor wagering experience for the new would be racing fan compared to what that same new would be racing fan experiences when he or she walks into a casino.

In a recent article, The Paulick Report covered an appellate court ruling in Kentucky that vacated a decision made by a lower court to allow gaming in the form of Instant Racing to operate at race tracks in the state of Kentucky.

Buried in that article is the following quote:
“According to statistics provided by the Kentucky Horse Racing Commission, $15.7 million was wagered on Instant Racing at Kentucky Downs in May, with $14.4 million returned to the public in winning bets. Total commissions were $1,285,991, with Kentucky Downs receiving $1,050,348 in net commissions. Fourteen percent of the net track commission, or $147,048 go into the purse fund, and $10,503 (1% of net) to a breeders incentive fund. “

Those numbers indicate that Instant Racing takeout is approximately 8.28%.

Bullet Point: This is a case where tracks and horsemen are clamoring for approval to offer an electronic version of racing that has a takeout of 8.28%.

Yet the real thing (live thoroughbred racing) has a blended takeout rate more than double that amount.

Q. Why such a disparity between the two rates?

A. It’s really quite simple. Takeout rates for the live racing product aren’t set by the market.

Instead, takeout rates for the live racing product are dictated by track management, leadership of horsemen’s groups, and politically appointed state regulators – people who to date have done everything in their power to deny even the suggestion that market forces apply to the live racing product.

The 8.28% takeout rate for Instant Racing came about because it was suggested by the market. There is a growing body of real world evidence to suggest that Instant Racing’s 8.28% takeout rate is not too far away from the optimal pricing point for most forms of gaming.

Q. Where does that body of evidence come from?

A. It comes from the casino industry.

According to the American Casino Guide, the average takeout for $0.25 and $1.00 slot machines in the Las Vegas area is less than10%.

25¢ Slot Machines
The Strip - 91.78%
Downtown  - 94.83%
Boulder Strip - 97.07%
N. Las Vegas - 96.48%

$1 Slot Machines
The Strip - 93.49%
Downtown  - 95.26%
Boulder Strip - 96.37%
N. Las Vegas - 96.84%

In case you’ve been living under a rock, in recent years, slot machines have become the most popular form of gaming in Las Vegas. Is it possible that the increased prize payouts for present day slot machines compared to the racing-esque slot machine payouts from the 1970’s have something to do with this form reversal?

When you compare present day slot machine takeouts (less than 10%) to thoroughbred horse racing’s blended takeout rates (more than 22%) it doesn’t take a genius to figure out why racing finds itself in long term decline while slot machine play has become more popular than ever.

Don’t blame us. It’s not like we horseplayers haven’t been speaking up for some time now.

In survey after survey, HANA horseplayer members have made it abundantly clear that one of the primary reasons they bet less on horse racing than they otherwise would is that horse racing has onerous takeout rates.

The problem is that our voices are falling on ears that are deaf to the point of being comically ignorant.

Let’s take at look at the betting experience faced by the new would be racing fan making one of his or her initial forays into our pools.

Below is a chart, generated by computer, where I queried a database to arrive at player expectations for straight bets WIN, PLACE, and SHOW. The database contains every thoroughbred starter at every North American track spanning the time period from January 01, 2012 through Tuesday June 19, 2012. There are approximately 167k starters and 21k races in the sample. (Other large samples that I’ve run from previous years produce nearly identical results.)

The report output shows what would have happened had a player bet $2.00 across the board – WPS – on every starter in every race at every track during the time period of the sample:

     Data Summary         Win      Place       Show
     Mutuel Totals  251758.90  251603.30  250909.10
     Bet           -333938.00 -333938.00 -333938.00
     Gain           -82179.10  -82334.70  -83028.90

     Wins               21486      42746      62566
     Plays             166969     166969     166969
     PCT                .1287      .2560      .3747

     ROI               0.7539     0.7534     0.7514
     Avg Mut            11.72       5.89       4.01

As you can see, the player’s expected return for WIN, PLACE, and SHOW bets averages just 75 cents for each $1.00 wagered.

The true return is far worse than what you might intuitively think given an industry average of 16% to 18% takeout rates for WPS bets.

How can this be? Shouldn’t the player’s expectation be 1 minus the takeout rate?


A glance at the bottom part of the chart, posted to below, where the data is further broken out by odds ranges, will shed some light:

By Incremental Odds Range:
>=Min < Max      Gain       Bet    Roi Wins Plays   Pct Impact
 0.00  0.50   -276.50   2164.00  .8722  721  1082 .6664 5.1783
 0.50  1.00  -1472.70   9304.00  .8417 2281  4652 .4903 3.8104
 1.00  1.50  -1852.00  12838.00  .8557 2499  6419 .3893 3.0254
 1.50  2.00  -2176.30  14266.00  .8474 2243  7133 .3145 2.4436
 2.00  2.50  -2530.30  15460.00  .8363 2018  7730 .2611 2.0287
 2.50  3.00  -2304.30  14742.00  .8437 1688  7371 .2290 1.7796
 3.00  3.50  -2216.60  13760.00  .8389 1375  6880 .1999 1.5531
 3.50  4.00  -2539.70  12554.00  .7977 1068  6277 .1701 1.3222
 4.00  4.50  -2314.40  11744.00  .8029  911  5872 .1551 1.2056
 4.50  5.00  -1859.30  10372.00  .8207  750  5186 .1446 1.1239
 5.00  5.50  -1700.90   9502.00  .8210  633  4751 .1332 1.0354
 5.50  6.00  -2507.70   9012.00  .7217  489  4506 .1085 0.8433
 6.00  6.50  -1481.40   8350.00  .8226  480  4175 .1150 0.8934
 6.50  7.00  -1886.60   7598.00  .7517  372  3799 .0979 0.7609
 7.00  7.50  -1109.00   6936.00  .8401  358  3468 .1032 0.8022
 7.50  8.00  -1269.70   6778.00  .8127  317  3389 .0935 0.7269
 8.00  8.50  -1323.80   6252.00  .7883  269  3126 .0861 0.6687
 8.50  9.00  -1264.20   5968.00  .7882  243  2984 .0814 0.6328
 9.00  9.50  -1112.10   5556.00  .7998  218  2778 .0785 0.6098
 9.50  plus -48981.60 150782.00  .6751 2553 75391 .0339 0.2632

The player’s expectation varies, depending on the attributes of the horse being wagered on. I’m talking about things we as players use to measure horses - such as speed and pace figs, class, the abilities of the horse’s human connections, the condition of the horse, etc. Each of these factors, and many more, are reflected in the odds.

In order for the player’s expectation to closely match 1 minus the takeout, the player must dutch every horse in the race in proportion to the money bet on each horse in the pool.

The above chart shows the player’s expectation for bets made into the WIN pool within incremental odds ranges. For example, in the top row, where the odds are less than 1 to 2, the player’s expectation is a long term loss of approximately 13%. At the other end of the spectrum, in the bottom row, where the odds are greater than or equal to 9.50 to one, the player’s expectation is a long term loss of 32.5%. The brand new (unsophisticated) player has no way of knowing this. When the new player makes repeated blind stabs at picking a winner - quick picks if you will -almost 45% of the starters going to the post will produce a massive losing expectation (long term losses in excess of 30%.)

Race outcomes aren’t the result of random chance (like a slot machine.) Horse race betting is a game of skill where the player must make a careful analysis of the horses and exercise judgment not just to gain an edge – but, quite literally, to have a reasonable hope of cashing a bet.

Please understand that I’m not talking about the skill needed to gain a true long term edge. Most of us know the knowledge and requisite skill to do THAT on a consistent basis doesn’t come easy. The learning curve is steep. In the end, only a tiny percentage of all players ever make it to the promised land. Only a select few ever become skillful enough to consistently beat the game on the up. Let’s leave that discussion for another article and limit the scope of what’s being talked about here in this article to the skill set needed by the new would be racing fan to cash a few bets and have a good enough time doing that during those first few visits to a race track – so that he or she feels as though there’s a fighting chance – and as a result, decides to come back.

The newbie horseplayer, not knowing anything about handicapping is reduced to making what often amounts to plays that are literally no better than blind stabs. Unless horseplayers betting on horses to WIN, PLACE, or SHOW possess some skill or knowledge of what they are doing – the realistic long term expectation (while learning) is a paltry 75 cents for each dollar wagered.

Here’s the harsh reality: Long term losses in the neighborhood of 25% are the norm for the new would be racing fan.

Now let’s say for the sake of argument that the same newbie player walks into a Las Vegas casino. When he or she takes that same $2.00 and feeds it into a slot machine - that player’s long term expectation is a return of approximately 95 cents for each dollar bet.

But that’s not all.

Unlike the new would be horse racing fan, the newbie slot machine player is given that higher percentage return on his or her money without ever facing the need to think.

Make no mistake. The machines are programmed to beat the player’s brains out in the long run - and they will do just that provided the player plays long enough. But the machines are also programmed to pay out often enough and at a high enough rate of return to lead the player to think there’s a fighting chance.

Further, the décor, design, ambiance, people, and sights and sounds found in the Las Vegas casino work together, purposely and by design, to create an environment that on a subconscious level is far more conducive to players believing there’s a fighting chance and opening their wallets than the atmosphere found at most race tracks.

The disparity in expected return isn’t advertised to the new would be racing fan. But over the course of time, the state of the new would be racing fan’s wallet whispers over and over again that maybe race tracks should be avoided – and when the itch to gamble presents itself: best to get it scratched, instead, at a casino.

Atmosphere and décor considerations aside, the player’s expected return of 75 cents for each $1.00 wagered on WPS bets is horribly out of line compared to the player’s expected return of 91.72 cents for each $1.00 wagered on those same WPS bets if made through an Instant Racing machine.

The player’s expected return of 75 cents for each $1.00 wagered on WPS bets is horribly out of line compared to the player’s expected return of 95.26 cents for each $1.00 wagered through a $1.00 slot machine in Downtown Las Vegas. And this is only compounded – at a subconscious level – as the player is affected by the ambiance created by the design considerations of the casino.

But that’s not the comically sad part.

The part that’s comically sad is this:  

Takeout rates for the live racing product aren’t set by the market.

Instead, takeout rates for the live racing product are dictated by track management, leadership of horsemen’s groups, and politically appointed state regulators – people who to date have done everything in their power to deny even the suggestion that market forces apply to the live racing product.

Racing has a serious problem here – one that needs to be addressed. Racing is a gambling game – one that is being priced out of existence by its track managers, horsemen, and political appointees who are sadly and comically oblivious to the betting experience of the new would be racing customer compared to the new customer betting experience when that same customer walks into a casino.

The question you as members of track management,  leadership of horsemen’s groups, and politically appointed racing commissioners need to start asking yourselves is this:

How much longer, by your own actions in setting takeout levels, are you going to continue to drive the new would be racing fan away from a day at the track – and into the hands of your competitors in the casino industry?

Jeff Platt
President, HANA


Steve Zorn said...

Great analysis, Jeff. What do ou think the odds are on the "new and improved" NYRA ever getting close to that optimal price point?

Anonymous said...

ZERO. I predict industry handle of $6B within 5 years. There are virtually no new horse players being created while 1-2% of us die off every year.

Anonymous said...

Nice read Jeff

HANA said...

Great question, Steve...

I can't help but think that with slots money pouring in, NYRA is uniquely positioned among major tracks to give something other than the status quo a try.

What chance is there for this to happen?

I'm really hoping that somebody on the new NYRA board comes to the table with something in the way of vision.

I want to see New York's thoroughbred tracks operated in such a way that NYRA leads the rest of the industry forward - by way of example - in everything they do.

There's a blueprint for success that just about every successful public company follows:

1. They identify their target market.

2. They identify the needs and wants of the customers in that target market.

3. Then they make it their MISSION to satisfy those customer needs and wants.

If NYRA (or any other major racing jurisdiction) would do that - instead of insisting on doing the opposite of that - IMHO, the sky's the limit.

(Call me a dreamer.)

What is definitely not needed are political appointees put in power who will treat NYRA tracks as cash cows to be milked. (Better grab a bigger piece of the pie because there may not be pie tomorrow.)



rocky_rosa said...

Good article Jeff.

This study is additional confirmation of the well-known Favorite-Longshot bias in North America(favorites tend to get UNDERbet, while longshots tend to get OVERbet). It's a subject that's been covered in great length by academia over the years.

Several of the papers can be found in Part IV of Don Hausch's "Efficiency of Racetrack Betting Markets":

The most exhaustive study I've seen demonstating the phenomenom is in Steve Klein's "The Power of Early Speed,"(2005, p.8). He looked at 1,774,850 horses and broke down the results similarly to the way you did it.

I'm sure none of the newbies(nor even some seasoned veterans) are aware of the phenomenom. And since there are more longshots than favorites, I agree, new players run the risk of losing a much greater percentage than the track take if they are basically just throwing darts.

Was surprised to see just how much higher the loss was on an average basis though - interesting stuff.

Equinometry said...


You, me and many others have provided stats to backup the claim that high takeout rates are killing the sport but those that have the power to do something about it don't care. They can't see past today.

The facts are if takeout rates are lowered their revenue may decline in the short term but long term their revenue will increase exponentially.

As for NYRA I wouldn't hold my breath waiting for lower takeout rates. When was the last time the Democratic party, who now runs NYRA, lowered taxes? I wouldn't be surprised if they did the opposite and increased takeout rates to the maximum level under the current law.

Even though the outlook is bleak don't stop fighting for us. If no one speaks up the problem will never be addressed.


Anonymous said...


Couldn't agree more that the takeout rate needs to come down and is nowhere near the optimal rate. Study after study has shown that it is considerably higher than optimal. However, let's remember that there is more to the rate of loss than just the rake. A new fan to racing is going to place a bet at a racetrack they visit about once every half an hour, and will thereby lose on average 20% or so once every half hour. A slots player is pulling the handle or pushing the button maybe four or five times per minute and is thus losing 5% on average many, many, many times over the course of that same half hour. The case can actually be made mathematically that the loss rate is significantly higher on slots. This makes it even more unbelievable that people are attracted to this mindless activity when there is a much better, much more challenging and much more stimulating game available, i.e., playing the horses. The industry needs to capitalize on this inherent attractiveness of horse racing -- proper pricing is one biig factor.

Anonymous said...

I e-mailed this story to a couple of days ago to see if he would post It.

I guess not. hahaha....

Anonymous said...

This is nothing but hot air!

When the clowns who ARE the reason FOR the problem, are the only ones who know or care about said problem, then they ('you') come across as the fools that you are.

On what level can you compare a purely 'random' gamble to a game of skill?

(that's like comparing the pick-4 lottery results to the game of bowling)

The lottery ticket costs a buck, and the game of bowling costs a few bucks. The so-called newcomer has as his goal to get better at bowling. The slots experience is as good as its ever gonna get right off the bat.

Don't you get it? YOUUUUUUU are the collective causes as to why the wager of $2 across the board on every horse returns only .75 ROI.

(Furthermore, your study there is grotesquely flawed in that you failed to incorporate the impact on payoffs OF your $6 wagered on every animal - see that infamous 5th race at Thistledown on May 21 for related understanding)

Brain-dead thinking like this is much of the reason why the HANA movement has had to sink to the level of free and wasted 'handicapping' each week.

The Santa Anita takeout remains what it became, the poor tracks with the reduced takeout and the HANA endorsements are dying on the vine, and the HANA fools are back in the big pools where we always knew they'd be.

Again, You can not become a part of any solution until you collectively cease to be any part of the problem.

Anonymous said...

Wow. What an articulate well thought out comment. Are you in track management? -- An owner?

The days of racing enjoying a monopoly on local gambling are forever gone. Amazingly, racing's brain trust still prices the product as if they still have a monopoly.

Jeff's charts are correct.

Racing HAS created an absolutely horrible new fan gambling experience vs. the competition. (I think that was the main point of the article.)

Yet somehow HANA is to blame for this?

Sticking to the status quo no matter what means a lack of new fans - and a lack of new fan dollars in the pools.

Racing has a serious problem landing new fans who bet. This is absolutely borne out by long term handle trends.

If racing dies on the vine - don't blame HANA. They aren't the ones making your decisions. They are simply pointing out the obvious.

I blame track management and horsemen for sticking to the status quo despite overwhelming evidence that the staus quo is what's causing the problem!

Charlie James
San Diego, CA

Anonymous said...

Sorry Charlie, wrong answer.

No one disputes that Jeff's charts are correct.

However, you are a few Fruit Loops short of a full box if you cannot identify Jeff himself as the cause for why racing's take-out rate is as stated in those charts.

When you and others of your ilk continue to be the problem, you simply can not BE any part of the solution.

And Charlie, you are indeed correct when you state Racing has created a horrible new-fan gambling experience vs. the competition.

So lets see if you can drum it through your collective SKULLS that you ARE the problem:

For what do you think those .75 percentages would be if Jeff and Charlie, and Larry, and Curly, and Mo weren't scraping a little profit off the top, or experiencing a slight 3% loss during each typical 12-month period?

With other forms of gambling, those take-out data are constant, and it is clueless fools such as your collective selves who cause horse racing's takeout to be a variable.

Forget about making every single random bet possible, and study the actual median performance (vs. break-even) of each unique soul who walks through the door of a race track or OTB in a given year.

Imagine for yourselves just at what level that median might sit. I don't care - pick the number of your choosing for this experiment.

Now, having done that, consider your own collective effect on that median over a given year. If done correctly, the lot of you simply can not arrive at any fair conclusion other than that you ARE the problem.

Then consider that other forms of gambling (competition) have a median that sits at or extremely near to (1 - the takeout) while horse racing's median is nearer to .50 than it is to (1 - the takeout). It's time you looked in the mirror, Charlie!

It is for that and similar reasons that "HANA" as an organization is the societal equivalent of present day prisoners petitioning to have a say in the writing of the laws!!!

In closing, Charlie, it is YOU and these many others of your ilk who ARE the problem. You're not merely causing the problem - you ARE the problem.

As such, it is time you stop pretending to have or know the first thing about any solution.