Takeout – The Price Isn’t Right – Part 3

In the previous articles in the series, I established that the pari-mutuel takeout is both too high and almost impossible to reduce given the belief that player behavior does not increase enough to offset even marginal changes in takeout. For horse racing to get its share of the much larger gambling market, pari-mutuel betting needs to be cost competitive with sportsbooks and poker rooms – the two major venues where there is no house edge besides the price of the bet. Yet the price discovery needed to find that optimal takeout is a risky and difficult proposition, putting tens of thousands of track revenue dollars at stake daily.

 My solution is take the idea of volume discounting and applying it in a novel way to pari-mutuel pools. This idea would put no level of current revenues, either handled or taken out, at risk. It would require no changes to the current betting experience, nor increase ongoing costs for tracks or bet takers. It would share the benefits of increased handle among all interests: tracks, purses, bet-takers and horseplayers. It’s an idea simply explained. It will probably never work.

 Handle-Tiered Takeout – A Proposal

 Summary:  Handle-tiered takeout (HTT) would, for a given pari-mutuel pool, have two takeout rates: the current rate, and a low marginal rate to be applied to all handle above a certain level (baseline handle). For example, if a win pool with a $100K baseline handle has a total handle of $150K, the track will take out 20% of $100K and 5% on $50K (incremental handle). The total take out would be $22.5K, for an average takeout of 15%.

 Baseline handle would be indexed to historical pool averages by type of bet and race – say last year’s exacta pools for maiden special weight at Santa Anita averaged $230K, then SA could set its baseline handle target to $230K, or $250K. Should pool sizes with HTT grow above that, then the new money should truly be considered incremental. With proper promotion of the marginal rates to attract higher levels of play, pools should resettle at new higher totals, with a lower average takeout. This new average takeout would be considered optimal for this particular bet. Total handle is up, total revenue is up (for tracks, purses, and bet-takers) average takeout is down, and payouts are up. The market has determined the price is right.

 Execution: The major investment required to make this change would be in the tote system in the way that final payouts are determined. Now, the tote software simply multiplies handle by a single rate, divides by winning bets, and rounds down. The proposed process would have two multiplications instead of one:

(Handle * Takeout)/Winners –> Round Down –>Payout


(Base Handle * Base Takeout + Incremental Handle * Marginal Takeout) –> Round Down –> Payout

Now, I understand tote software is complex and doing a lot on what is certainly an aging infrastructure. I don’t know how easy this change is, as what’s easy in Excel or a SQL Database for me will be much more complicated in tens of thousands of lines of code in the tote software. The effort would require months of time and hundreds of thousands in investment.

 The second obstacle will be to get all the players to agree on the concept, the right marginal takeout, and how to divide the incremental revenue. What HTT will do is diminish the upside from increased handle. Many in the industry retain hope that handle levels will return to those seen in the glory days of their national gambling monopoly, higher takeouts mattering little when it was the only game in town. There needs to be widespread recognition that those days are gone and are not coming back. The industry must make moves to be competitive – this idea has the least risk associated with it.

 Accomplishing the duel tasks of reprogramming the tote calculations and getting agreement amongst the major players in the industry is why I think it will take a major, vertically-integrated firm to push for this change. There are two candidates to do so – Churchill Downs, Inc., and Magna Entertainment Corp – as they respectively own the two major North American tote firms, United Tote and AmTote. They each own an ADW (Twinspires, XpressBet) and are partners in HRTV. More importantly, they own tracks in multiple jurisdictions including Churchill Downs and Santa Anita and Gulfstream. Would either company take the chance to make the change? I don’t know – I only know it’d be easier for them than anyone else.

 I’d especially make the argument for CDI. Make the changes at United Tote in the background, without impacting any current operations. (Base takeout and marginal takeout need only be the same to replicate status quo). Pick a single track, say the Fairgrounds in New Orleans, that has a well-defined meet in a jurisdiction (Louisiana) that is smaller and perhaps more flexible than a New York, Florida, or California. Theoretically, TwinSpires and track management will be on board. Get a horseplayers group like HANA bought in. Promote the low potential takeout as much as possible. On every single Brisnet chart of a race, show what the new average takeouts are. Collect a ton of data.

Obstacles and Objections: No plan is without its flaws. Even a simple one like this has probably been considered dozens of times and shelved, with the thought that it was either too hard or unnecessary at the time. The main objections I can think about are below:
  • The tote companies will not implement the change themselves because it doesn’t fit their business case, i.e. it won’t make them more money. Agree that some players in the industry will have to fund. Best case scenario is CDI or Magna who can self fund the changes.
  • HTT adds complexity where there was simplicity before. True, but takeout rates are by pool today anyway, and there are few horseplayers who could name off the top of their head the takeout in the pools they are betting on. Part of the idea is to increase awareness of takeout by highlighting it in the final charts of races.
  • Horseplayers are a difficult group to get to act collectively. Without a “first dollar” takeout decrease, the plan will not succeed. Valid objection – but given a reasonable unwillingness to experiment with direct price decreases, this at least provides a no downside option.
  • HTT will have to be heavily promoted in order to work, marketing money that could not be spent elsewhere. No disagreement, but there’s a difference between marketing to a one-time increase, or spotlighting a particular race or racehorse, then raising awareness broadly. The impact of promoting a price is that, as long as the price remains low, the benefit should be permanent.

Final Thoughts: What I’ve proposed is only a solution, not the solution. If the central issue for the gambling side of racing is the price discovery effort for the optimal takeout rate, this is a low-risk idea. I believe that takeout needs to be lower to get new and bigger players into our racing pools, but I do not know how low. I think we can find out.

Takeout – The Price Isn’t Right – Part 2

In the previous installment, I argued that the takeout on bets in thoroughbred racing is too high, both for current players and for other bettors who do not now find racing to be a profitable betting opportunity. In this update, I want to examine why tracks and racing associations are so reluctant to lower takeout. In the next installment, I propose a solution that could lead the industry to finding the optimal price for its product.

Tracks (or frequently, state racing associations owning several tracks, like NYRA) do not want to lower takeout because they believe they will lose revenue doing so. Frequently, they are right. Allow me to explain.

Takeout accounts for most of the direct racing-related revenue that a track earns. The track, however, does not capture all that revenue – it is split among multiple parties, but two are the most important: track operators and bet-takers. The track operator usually must split its revenue share according to certain terms: purses get 25% of revenue, the state gets 5% in taxes, the track keeps 70%. Consider then two scenarios (for simplicity sake, assume takeout is 20% on every bet):

  1. A $1 bet is placed at the track – as the bet-taker, the track gets $0.20, dedicates $0.05 to purses, sends a penny to the state, and keeps $0.14.
  2. A $1 bet is placed at an OTB/ADW – the bet taker now gets $0.12, the track gets $0.08, purses get $0.02 of that, the state still needs its penny, and the track gets $0.05 (and these may be generous terms for simulcasting – I do not have the agreements in front of me.)

As an aside, this dichotomy is frequently explored by Fred Pope in what appear to be quarterly opinion pieces. I’ve linked to a Paulick Report index of his articles. His argument boils down to that this structure is what is hurting racing, not the overall price level. I agree that the revenue sharing is genuinely an artifact of the early days of simulcasting, when new bet-takers needed to be compensated for large capital investments that have long since been recouped. I think Pope mistakes a secondary pricing issue (structure), however, for a primary one (level), since vertically-integrated racing firms like CDI, Magna, and (to some extent) NYRA have largely obviated the structure issue by owning their own ADWs.

Arithmetic explains the rest. In the last installment, I laid out that handle necessarily goes up when takeout goes down because incremental winnings get re-bet. But how much so – in this case, my 20% handle world gets a takeout decrease to 19% – yay. Handle on all pools is $100,000/race, on-track bets only.

  1. After first race, winners collect $1000 more than they normally would have and bet all incremental winnings in race 2.
  2. Handle race 2 is $101K – winners win $101 more than usual….and so on
  3. At the end of the day, handle is up 1.01%
  4. Racing-related revenue is down 4.04%, as the 1% takeout decrease

In this world, horseplayers did not bet more into more-profitable pools nor did new money come in. This is essentially the tracks’ stance: if horseplayers do not change their behavior (i.e. bet more), then the math-driven increase in churn and handle is simply not going to cut it.

In other words, incremental handle decreases are not going to improve track revenues. (This is especially true when you take the bet-taker’s share into account; if the track has to take a $0.01/bet decrease on every dollar, that can mean a 20% decrease in operating revenue from simulcasting.) There is, however, a takeout rate at which thoroughbred handicapping becomes cost competitive with sports betting and poker. At this rate, whales – large handle bettors – should become interested in betting into racing’s pari-mutuel pools. At this takeout level, we should see a step-function increase in handle and track revenues.

What is this takeout rate? I don’t know, obviously, but I imagine it is closer to 10% (the sports betting rate) than 20% (approx. the racing rate). It’s not hard to imagine track officials are frightened to cut racing revenues by 40-60% to find out. A wrong guess can put their livelihoods in jeopardy – who wants to be that guy? Plus, it’s hard – the negotiations would be extensive with horsemen, ADWs, and state commissions, each of whom would be difficult to convince that their revenue would not go down. Instead, we get small, low-risk changes like 15% Pick 5s that won’t be taking money out of any pools, or money out of any pockets.

Price discovery is not a risk-free proposition, but it need not be as scary or difficult a proposition as the industry is making it out to be. I’ll propose a solution with my next 750 words.