The Favourite-Longshot Bias
Value betting is the most mainstream concept in the modern betting landscape, writes Tony Keenan. It has passed into the lexicon of gambling and there can hardly be a punter around that hasn’t heard of the idea. We’ve reached a point where Tom Segal, the public face of value betting, is on The Morning Line more often than Tony McCoy and it has even spawned parody comments like ‘value loser’ and ‘you can’t eat value.’
For most, value betting means one thing: big prices. While a value bet can theoretically be a 1/3 shot that should be 1/10, in the main such examples aren’t referenced and the typical comment about a favourite in much racing writing and programming can be summed up as ‘he’ll probably win but he’s a bit short.’ Rarely do we hear mention of the odds-on shot that should be even shorter.
But perhaps the pendulum has swung too far the other way and the runners that appear short, far from being priced lower than their real chance of winning, actually aren’t short enough. This favourite-longshot bias certainly seemed at play over the recent Cheltenham Festival with some of the Mullins hotpots but it’s not just that; the place markets on the exchanges for the Grand National on Saturday were wildly out of sync with the same place odds available in an each-way bet with the bookmakers, the favourites seemingly bigger than they should be and the outsiders representing the flip-side.
I think this applies far more in day-to-day racing now than it used to and we need to be aware that markets are fluid and don’t remain the same: it is important to zig when most people are zagging. And that can be a problem for an old-style value punter who is used to looking on shorties with scorn. In fact, many of these punters – and I include myself here – would have personal betting rules like ‘never bet odds-on’, tenets that might have served you well in the past but have become outdated. As we’ll see later, taking that approach to the current Irish national hunt season would have meant there were 311 odds-on shots (judging on BSP) that you refused to back; 195 of them have won so you were taking out a large chunk of possible betting opportunities.
In the tables that follow, I will look at the fortunes of odds-on shots, defined as having a Betfair Starting Price of 2.0 or shorter, in recent Irish jumps campaigns. Using Horse Race Base, I’ve put together the numbers to see if there are any patterns emerging. When I refer to actual over expected in the final column of the table, it is done to official starting price rather than the Betfair equivalent as this is the method used on Horse Race Base.
Let’s first look at the number of odds-on shots to run over jumps in Ireland in each season since 2008/9:
NH Season | Bets | Wins | Strikerate | Level Stakes | Actual/Expected |
2008/09 | 164 | 104 | 63.4% | +3.16 | 1.02 |
2009/10 | 217 | 125 | 57.6% | -16.23 | 0.93 |
2010/11 | 180 | 109 | 60.6% | -3.89 | 0.97 |
2011/12 | 227 | 130 | 57.3% | -21.69 | 0.91 |
2012/13 | 272 | 162 | 59.6% | -19.37 | 0.92 |
2013/14 | 250 | 159 | 63.6% | +1.36 | 1.00 |
2014/15 | 303 | 184 | 60.7% | -18.06 | 0.95 |
2015/16 | 311 | 195 | 62.7% | -10.00 | 0.97 |
The first thing that jumps out is the gradual rise in the number of odds-on shots; again, punters who are excluding themselves from betting in races with money-on pokes are ruling out a hell of a lot of races. We see that prior to last season, there was not a campaign with more than 272 odds-on shots but there were 303 last year and already we are at 311 in 2015/16 and there are three weeks of the season left. The other notable feature is that backing all these odds-on shots, though not profitable, has come close to breakeven point in each of the last three seasons judging by the actual over expected. Given that those figures are worked out from official SP suggests there is an edge here; Betfair SPs are typically bigger even allowing for commission while many of these winners will have been available at bigger prices before the off.
I suspect at least some of this edge is down to the attitude punters approach the market with, i.e. they want to oppose the fancied horses as they are perceived as too short. There are other reasons though. The concentration of talent in the Mullins yard has played a big part as has that trainer’s approach in that he tries to keep his horses apart as much as possible to maximise the number of races won, rather than run them in what might be seen as the most logical races. Racing has facilitated much of this with a greatly expanded programme book and I’d be in the camp that competitiveness has suffered because there is too much racing. Competitive racing, of course, produces much fewer odds-on shots.
Next, we have the record of odds-on favourites by month and instead of going back as far as 2008/9 here, I’ve worked from 2010/11 to the present day in the hope this will provide a better picture of the current state of the markets.
Month | Bets | Wins | Strikerate | Level Stakes | Actual/Expected |
January | 174 | 107 | 61.5% | -11.31 | 0.94 |
February | 144 | 75 | 52.1% | -30.36 | 0.80 |
March | 126 | 71 | 56.4% | -11.97 | 0.89 |
April | 79 | 46 | 58.2% | -5.97 | 0.92 |
May | 115 | 72 | 62.6% | +2.63 | 1.02 |
June | 50 | 25 | 50.0% | -8.06 | 0.82 |
July | 83 | 55 | 66.3% | +1.72 | 1.04 |
August | 128 | 73 | 57.0% | -9.21 | 0.92 |
September | 67 | 50 | 74.6% | +14.74 | 1.22 |
October | 155 | 104 | 67.1% | +6.37 | 1.05 |
November | 206 | 129 | 62.6% | -8.81 | 0.96 |
December | 216 | 132 | 61.1% | -11.42 | 0.95 |
The findings here are surprising. One might expect odds-on shots to do best in the traditional jumps season from November to April when Willie Mullins has his best horses running. In fact, the opposite is true as betting the odds-on over summer jumps is [much] more profitable with the exception of June. It may just be a quirk of sample size – there are fewer jumps races at this time of the year – but the pattern is still quite marked and it offers food for thought at least.
Different types of Irish national hunt race are also worth considering:
Race Type | Bets | Wins | Strikerate | Level Stakes | Actual/Expected |
Chase | 413 | 261 | 63.2% | -0.63 | 1.00 |
Hurdle | 849 | 513 | 60.4% | -54.64 | 0.94 |
Bumper | 281 | 165 | 58.7% | -16.38 | 0.95 |
Again, the results here are unusual. One might expect chases to be the least predictable of the three disciplines as fences bring in a great possibility of falling but the opposite is true as odds-on shots here do best and come very close to making a small profit off level-stakes.
Finally, the record of trainers with odds-on runners since 2010/11 and for the purposes of the table I’ve looked at those with at least 20 such qualifiers. The table is in alphabetical order.
Trainer | Bets | Wins | Strikerate | Level Stakes | Actual/Expected |
H. De Bromhead | 56 | 39 | 69.6% | +7.80 | 1.14 |
C. Byrnes | 29 | 16 | 55.2% | -2.90 | 0.89 |
G. Elliott | 128 | 74 | 57.8% | -10.76 | 0.91 |
J. Harrington | 51 | 29 | 56.9% | -3.97 | 0.92 |
T. Martin | 38 | 20 | 52.6% | -6.96 | 0.83 |
N. Meade | 71 | 49 | 69.0% | +5.31 | 1.06 |
M. Morris | 20 | 10 | 50.0% | -4.03 | 0.82 |
W. Mullins | 640 | 403 | 63.0% | -32.68 | 0.95 |
E. O’Grady | 31 | 20 | 64.5% | +0.83 | 1.04 |
D. Weld | 36 | 23 | 63.9% | -0.16 | 0.99 |
Willie Mullins has had just a staggering number of odds-on shots; in fact, in the period covered he has sent out 41.4% of the total odds-on favourites. Finding an edge with Mullins is difficult – though betting his shorties at the Cheltenham Festival makes sense – and I prefer to look at other trainers. Henry De Bromhead is one that stands out. He comes out best both on overall strikerate and actual over expected and the best spots to back his short-priced horses are in graded races and over fences and ideally a combination of both. Noel Meade also does well; his 69% return his hard on the heels of De Bromhead and, along with Edward O’Grady, they are the three trainers with actual over expecteds of greater than 1.00.
From a brief glance at the figures for the flat in Ireland, the rise in odds-on shots seems nothing like as marked but it is something I may return to over the summer. For now, be wary of opposing short-priced runners in Irish jumps races for the sake of it and, indeed, don’t be afraid to back some of them; though starting the project at Punchestown might be a little too brave as the meeting can produce some surprising end-of-season results.
- Tony Keenan
Follow Tony on twitter at @RacingTrends