Skip to content

Your Place Or Mine – Part II

Before proposing wagers across this field, it will be useful to examine how to handle operations for a single selection.

Match Betting the Extra Place

A positive EW edge can only be established when the win edge is not too punishing with respect to the number of places and place fraction on offer. Notwithstanding the general situation of a hard betting patterns in relatively small fields … it is always worth keeping an eye out for anomalies.

A method for exploiting one such opportunity is discussed on web page Extra Place Strategy | Profit from Extra Place Offers ( The bet conditions are:

Bookmaker Back Odds 14/1                                         Betfair Best Back Lay Odds 15.0

Bookmaker Place Offer 1/5th 5 places (not 4)              Betfair Best Place Lay Odds (4 places) 3.85

On Course Profits free Horse Racing magazine

Betfair Commission 5%

In other words, the back and best lay odds are identical with the place and place lay odds tight too … and the bonus of a ‘free’ 5th place.

The pleasing presentation results in the following hedging recommendation:

Hedging both the win and place parts of the original EW bookmaker wager effectively turns it into a very nice ‘bet-to-nothing’ on the extra 5th place (or shall we say a ‘bet to next-to-nothing’).

Yet this discussion is incomplete, lacking two key figures:

  • what stake would a 14/1 shot normally attract?
  • what chance do we attribute to the extra place?

 For the sake of argument, the sequel uses:

  • £10, as originally staked with the bookmaker
  • 1/5th of the chance calculated for the 4 places

The worksheet Extra Place continues calculations, assuming the ‘true odds’ are mid-way between the best back and best lay offers:

Refinements include …

  • if the ‘true’ win price is 14.75 ((14.5+15.0)/2), then the bookmaker back has a small positive edge of 1.7%
  • if the ‘true’ four-places price is 3.825 ((3.8+3.85)/2), then the four bookmaker places have a small negative edge of 0.7%
  • if the four-places chance is 0.2615 … and the extra place 1/5th of that figure … then the extra edge is (0.2615/5) x (2.8+1), or 19.9%, since that extra place returns both winnings and the place stake, which would normally have been lost

… showing that the original EW bet has an edge of roughly 21%. Smashing!

But look at the edge after hedging – only 10.2%. Those two lays on Betfair are very expensive at 5% each in commission.

Would you rather have an edge of 21% with an unhedged risk of losing both £10 stakes … or a 10% edge with a maximum liability of only £1?

My answer: neither!

Were I prepared to lose £20 on a 14/1 shot, then I would be looking to bet £200 EW and make the Betfair hedges (perhaps exercising some prudence by operating in stages). One can stand a halving of edge … if accompanied by a 20-fold increase in turnover!

[Discussion of the embedded solver programming and the Sharpe Ratio deserves its own article. Later.]

Back to the 1.40 at Kelso

Instead of using the Harville formulae, the Sharpe Ratio Calculator uses the Betfair prices as the best indicator of ‘true’ odds. Looking at the favourite first …

… marked to market, the win component has a negative edge of 6.5% and the place component a positive edge of 6.2%, leaving an each-way bet more or less break-even.

Considered on its own, there is no point in hedging either the win or place component of this selection, that would just be giving up the Betfair market spread and commission.

The second favourite is a completely different kettle of fish …

… the win component is again loss-making at negative 6.7% but the place has a lovely edge of 15.4%, leaving a healthy net unhedged each-way edge of 8.7%. Well, that is what the calculator would have shown, had I not introduced the Betfair win lay of 0.93 stakes and place lay of 1.19 stakes: now the net edge is down to only 1.8%. What gives? Can you see?

The mysterious fractions of 0.93 and 1.19 are precisely those required to produce a net return (return, not edge) of 0.02 stakes (actually 0.018 rounded to two decimal places) regardless of the race outcome. In other words, the original bookmaker bet and the current state of the Betfair markets has created an arbitrage position.

The reader is faced with a similar situation to the matched bet discussed earlier: accept the unhedged edge of 8.7% … or make a larger initial wager anticipating hedging for a certain gain.

Did you notice the aptly-named, rank outsider Special Rate? At 50/1 he is the equivalent of a senior in the silly race used to illustrate how a place edge arises (last issue). Here, even if his true price was more like 70, consideration is warranted (notwithstanding the benefit of hindsight).

The Exception Proves …

If there is one race which has the softest of all betting patterns, the largest of all fields and the most room to blow up, it is The Grand National. Not to be touched with a barge pole for place betting, says he. I included it in the racing spreadsheet just for fun … but would you Adam and Eve it, not only did the second and sixth favourites (at my time of data collection) place in the generally available top 6 … but the bloody 7th favourite brought home the sole SKYBET 7 places offer!

Practical Considerations

This article has presented a largely market view of place betting opportunities and some money management ideas. A reader with his own selection method – be it judgemental, statistical or via tipsters – can still use the spreadsheets by inserting his own tissue, overwriting the ‘true’ values.

Further, no consideration of the usual practical difficulties in betting operations has been considered: obtaining a good price; ‘getting on’ (and, if hedging, ‘getting off’); keeping up with market moves; … here too the reader must accommodate the limitations of his chosen scenarios and manoeuvre appropriately.

Temperament might decide whether one prefers to retain a risky 21% edge unhedged … or attempt to bet at larger stakes, anticipating hedging out at a lower risk level.

Inflexibility is the weakness in the place fraction.

John Jackson

For the first part of this article and the links to the spreadsheets used please refer to Issue 103