Singles or Lucky 15s – a historical results exercise

What shall I bet on – singles or multiples?

Recently we posted a “Betting Multiples video” advocating that betting  in multiples is preferable to singles (see video at the bottom of this blog). Betting multiples minimises restrictions whilst compounding value. 

Which kind of multiple should we bet on? The answer can be complex and depends on your personal risk tolerance. Placing 4 horses at 100/1 into a Lucky 15 can be similar to a +EV lottery ticket.

At bookiebashing we record the ROI of singles on horse racing. We almost exclusively bet in multiples but the issue  with recording the ROI of multiples on a daily basis is it is logistically hard to assume which multiples  would be built. Instead we work under the assumption that if Singles have a positive ROI, then Multiples have a higher ROI. 

The video below attracted some commentary that we were talking nonsense. Therefore we have tried to prove the concept through an empirical analysis of real-world data from the bookiebashing horse  racing tracker. We have attempted to make the analysis as “real-life” as possible:

  • We only placed horses that were +EV on the same day in a multiple.
  • We only placed horses that were +EV at the same bookmaker in a multiple
  • We ensured that two horses from the same race did not exist in the same multiple by only considering one horse from every race.

Sample sizes of bets ranged from 46,606 (singles) to 4,524 (4-selections).

Profit/Loss profiles



  1. Bookmakers reviewed were Bet365, Betfred, William Hill, Ladbrokes, Skybet and Paddy Power. The raw data is available here.
  2. The cover bet with the highest p/l was Lucky 15s. We were able to identify 4,524 Lucky 15s. There was no “WWWW” in the dataset.
  3. The fourfolds performed relatively badly compared to other multiples. However the volatility of this strategy is the highest of all. The average odds of a fourfold was 1.2 million / 1. With a sample size of only 4,524 it is unsurprising that there hasn’t been a single WWWW yet.
  4. The cover bet with the highest ROI was trebles. It should be noted that the trebles data history includes 4 WWW wins that would be considered extraordinary for a dataset of that size, including a winner at 3,755/1.
  5. Concessions such as “all winners bonus” and BOG are not included.
  6. Staking is set at £1 EW

  • Singles, Doubles and Trebles stakes = £2.
  • Trixie stake = £8
  • Patent stake = £14
  • Yankee stake = £22
  • Lucky 15s stake = £30