10 July 2014

Analyzing The EquiLottery

Almost 5 years ago I wrote a blog piece Could A National Lottery Help Racing Grow?   That post was made before the Ontario Liberal government pulled the plug on the Slots At Racetrack partnership.  It was evident back then that horse racing was not growing and that there was a need to attract new gamblers with new money.  Now, in Ontario, the situation is even more dire.  Now the government has magically changed a true partnership into a true subsidy that could vanish at any time, horse racing needs to attract as many new players as possible.  One promise made by the new Premier is to integrate the OLG and Ontario horse racing, and this paves the way for horse racing lottery bets.

An announcement was made yesterday that a company called EquiLottery has the software to enable horse racing lotteries.  A Google search of OLG EquiLottery makes me believe that Ontario will be introducing this soon.  Check this PDF found in the search out.

Here is how it works:  Lottery player buys a $2 quick pick for a specified wager on that given day.  It could be a triactor, superfecta, Hi-5, Pick 6, etc.  The wager type can change each day according to the new release.  $1 is put directly into the pool.  The other $1 goes towards the lottery (with the lottery/administrators keeping 80 cents of that). So lets use a trifecta for the sake of simplicity.  The player gets a quick pick ticket in a 10 horse race of 5-2-8 at a cost of $2.  A $1 5-2-8 ticket winds up being bought at the track as soon as the lottery ticket is purchased.  This means the track retains a takeout of 25 cents (or 12.5% of the total cost of the ticket) to share between the track and horsemen.  The lottery portion only pays out 10% of the total cost of the ticket, split among the winners.  So in theory, the lottery player is buying a triactor ticket with a 52.5% takeout.  One has to wonder how long it will take for lottery players to clue in that if they bought the $2 straight ticket directly at the track, their payoff would be over 50% higher on average (all things being equal, which they aren't...I'll explain later below).  In the age of the internet, this scheme will not work very well and most likely dry out in short order even if it is initially successful.

Don't get me wrong, I want to see it work, and there is a lot of stupid money out there when it comes to playing the lottery in Ontario.  Pro-Line for example, does around $250 million in handle a year ($700,000 a day) and has a collective takeout of 40%.  You basically have to be an idiot to play it.  But they get away with it because their most popular bet is a three team parlay, where the parlay includes a tie option.  This is something that is unique so payoffs can't readily be compared with what bookies pay.  A normal three team parlay has a takeout of 12.5% if three teams are picked randomly with a bookie.  It is actually a little lower than that because odds are not a perfect science, and the handicapper can and does take advantage of point spread inefficiencies.

If the EquiLottery were to work as planned, it would be great for the host track because bettors from all over will know there is a good percentage of dummy money in the pool.  In worst case scenarios, it will mean a takeout free type or even takeout positive wager that value players will be all over, adding large amounts of handle that normally wouldn't be in that pool.  And generally when handicappers check out one race or a series of races, they normally make other bets on the card as well.

The Math

Lets use the triactor example.  10 horse race (720 combinations).  Lets say 100,000 $2 quick picks are purchased.  Assuming random equal distribution there will be about 139 winners.  Lets also assume that the average $1 winning ticket pays 60% of the total combinations (I believe that is close to the way it really is). So the average payoff for a $1 triactor in a 10 horse race is around $430.  This means that on average, the winning lottery player will get back $430 plus $143.88 ($20,000 from the lottery portion divided by 139 winners).  So instead of getting back $860 had the lottery player bought a $2 parimutuel ticket, the "sucker" only gets back $573 instead.  That is quite a price to pay for the convenience of buying a ticket at a lottery kiosk versus the convenience of betting the same thing from home and actually having some data to help you have a better chance of winning by picking the number yourself.  I realize, they admittedly aren't going after horseplayers with this but they are relying on a whole lot of stupid.

Because the amount won by lottery players will be verified to be much lower than those who win a parimutuel wager, I just don't see it working.  It has to either be a life changing win (not going to happen with a tri, super, etc.) or the 20 cent lottery component has to be unique with possible carryover potential.

The Fix

The 20 cent lottery component cannot be the same exact wager as the parimutuel one.  The fix is simple.
Using the same triactor example as above, when someone buys the 5-2-8 quick pick, the ticket in this case actually has another 3 or 4 numbers (horse that need to finish 4th through 6th or 7th in exact order).  When it comes to a 10 horse race, the odds of picking 6 horses randomly in the right order is one out of 151,200, 7 horses is one out of 604,800.  Depending on how many tickets are purchased each day, this even creates carryover opportunities which create even more demand.  But using the 100,000 example, this portion of the lottery will be won 2 out of 3 days if 6 numbers are used.  If it is won on the first day, there would be only one jackpot winner (there would still be 139 winners who get the $1 triactor payout) who would win the $20,000 lottery portion of the play plus the $1 triactor payout.  This could fly and be successful.  Also, the chance of cashing the lottery portion become significantly more difficult if the triactor has more than 10 horses, or if a superfecta is the wager type.  For Pick Fours or Pick Fives, etc. the extra number could be an additional number each race, where the jackpot is only paid out if the player has all the exactors in right order for the races involved.  For example, a Pick 4 involving 4 nine horse races randomly will be cashed one out of 6,561 times.  To get 4 straight exactors in a row on a random ticket in 9 horse races, the odds are over a whopping 26 million to one (that is enough to satisfy the life changing element of buying the lottery ticket as the winners amount will be $2.6 million on average).  Even a Pick 3 could work.  Chances of winning on a random ticket is one out of 729 (if three 9 horse races are used), chances of hitting three straight exactors on a quick pick: one out of 373,248.