Poor Value Favourites in NH Racing
I wonder, how many readers back favourites from time to time? Is there anyone reading this who only bets on favourites? It certainly would be interesting to know, writes Dave Renham. Maybe you'll share your answer in the comments below this article. I have two reasons for starting with those questions: firstly, I am genuinely interested to know; and, secondly, this article is based around favourites.
I think it is fair to say that favourites have had a bit of a 'bad rap' in the last ten to twenty years, mainly due to punters being bombarded, correctly in the main, with the concept of value. Thirty or forty years ago I am guessing that most punters had less of an understanding of how to assess value prices. Their mindset would probably be more attuned to winners – if they picked enough winners, they would make money. Of course, we know that this is not the case. You can have 50% winners and still lose money; conversely, you can have 15% winners and make a good profit. I still smile to myself when any of my non-racing friends go to a race meeting and ring me up with the question, “Dave, can you pick me some winners please?”
My reply is always the same, “if you want to back winners then back the favourite – this will give you more winners in the long term”. I do clarify that by telling them that is not the right way to bet, but if they want ‘winners’ then my answer gives them the best chance of achieving that on the day in question. And for once a year punters, this is probably the smartest way to bet - trust to luck. But for those of us more regularly engaged with the puzzle, we need to show more discernment.
With the nights drawing in, many of us in the racing world are turning our attention away from the turf flat season and starting to think National Hunt, and the great meetings and races that will take place over the next six months. In this article, then, the focus will be on National Hunt racing and favourites. More specifically, I will be looking to highlight market leaders that the data suggest are showing as offering poor value. Data have been taken from the last six full calendar years of UK NH racing spanning 2017 to 2022, and I will be sharing the results for clear favourites only (excluding races with joint- or co-favourites).
So why look for a race with a poor value favourite? Well, it is not because I am suggesting laying such a horse, although that may be an option in some cases. Rather, it is to do with the fact that if the favourite offers poor value, then there must be value elsewhere in the race on another runner, or runners. Before I start looking for examples of poor value favourites, let me share the breakdown of win strike rate for favourites by year:
As the chart shows, the strike rates do not fluctuate too much, with a difference of only 2.7% between 2022 (the year with the highest SR%) and 2020 (the year with the lowest).
Here is the overall record of NH clear favourites over these six years:
Essentially this is not a bad starting point in terms of returns, with losses of just over 2½ pence in the £. Hence taking National Hunt favourites as a whole, in terms of betting to Betfair SP, they are not bad value. To give some context, the results for backing all fifth favourites would have lost you 7½ pence in the £, sixth favourites 12p in the £. Suddenly 2½p losses look quite good!
So, we have our benchmark figures here for favourites – strike rate of around 37.5%, losses around 2½p in the £, and an A/E index of 0.94. Let's now break this overall group into subsets.
Performance in NH Races by Age of favourite
The first area I want to look at is the age of the favourite in question. Below is a bar chart showing the win percentages for different ages of favourites. There is quite a clear pattern as you will see:
It should be noted that three- and four-year-old favourites make up only 10.6% of all NH favourites and hence I am more interested in the correlation between favourites aged five and older. As the graph indicates, favourites aged five have the best strike rate and, thereafter, that percentage gradually drops for subsequent age groups. Favourites aged ten and older are clearly the worst performing animals in terms of strike rate.
So, have we potentially found a group of favourites that are poor value? We need to see the overall figures for favourites aged ten or older to see if the drop in strike rate effects the bottom line:
As shown, being of that vintage does affect the bottom line, with losses of over 11 pence in the £. Also, the A/E index has dropped to 0.87 which is a secondary indicator of poorer value.
Interestingly, these older favourites have performed worse in chases than hurdles. Chases make up most of the races with older favourites and the chase stats read 247 wins from 807 runners (SR 30.6%) for losses of £146.28 (ROI -18.3%); A/E index 0.82.
While I was trawling through some of the data, I noticed a horse called Midnight Moss who was a ten-year-old last year (2022). He started favourite in his last three races of 2022 (all chases) with the following results:
Clearly punters were happy to keep giving this fella another chance, but in hindsight these races would have given us a good opportunity to look for value elsewhere. For the record, Midnight Moss has run once as an 11yo in 2023... and yes, you’ve guessed it, he started favourite and was beaten into second again.
Performance in NH Races of Favourites that were narrowly beaten last time out
The next group of favourites I want to look at is those runners which were just touched off last time out. I'm including horses that were either beaten by a nose, a short head, a head or a neck in their last race. Here are the figures for that subset of runners who started clear favourite next time out:
One would expect the strike rate to be decent, and it is, but the returns are very poor for these favourites. Losses of over 15p in the £ is extremely hefty considering the overall stats shared earlier. This looks to me a classic case of horses being overbet, the theory being that they ran so well last time that they have a very good chance of getting their head in front this time. And so, despite a decent enough win rate, their actual starting prices have averaged out to be much shorter than their true odds of winning.
Before moving on, if we focus on handicap races only then the results for these narrowly beaten last time out runners who start favourite get significantly worse. Horses that were beaten a neck or less in a handicap last time out, and who are racing in a handicap again as the clear favourite have produced the following results:
Losses of nearly 28p in the £ are quite shocking for this group of NH favourites.
Performance in NH Races of Favourites having their second career start
Horses having just their second run are still relative unknown quantities. Plenty of horses run well on their first start but fail to back it up, whereas others run poorly first time out and then improve out of all recognition next time. It makes sense therefore that horses that start favourite on their second career run may not be the best betting proposition. That was my theory at least; below is the evidence, the table showing favourites who have previously run just once:
The strike rate here is high, a fair bit above the average for all favourites; but returns are relatively poor – losses approaching 10p in the £. The 0.87 A/E figure is low also. As with horses that were narrowly beaten LTO, this again looks a case of a group of horses being overbet driving the prices below their true odds. It is also worth sharing that horses which won on debut lose a little bit more again when favourite next time (losses of just over 10p in the £).
Performance in NH Races of Heavy ground Favourites
To begin this section, here are the win strike rates for NH favourites in terms of going:
The figures are relatively even – the best strike rate has occurred when the going has been the firmer side of good. Having said that, this firmer going is relatively rare, making up just 3% of all NH races. In terms of races on heavy ground, clear favourites have also done well, winning slightly above the norm and losing backers just 2p for every £1 staked.
I am guessing most people will be thinking that previous heavy ground winners are a positive if racing again in such conditions especially if starting favourite (me included). However, this has not been the case as the stats show:
Losses are close to 10p in the £ for these past heavy ground winners.
On the other hand, clear favourites have fared far better on heavy ground if they are yet to have won a race on this going, as these figures clearly show:
Hence, the data for these heavy ground favourites seem clear-cut: be apprehensive of a previous heavy ground winner whereas don’t immediately rule out if not a previous heavy ground winner. What is most interesting, perhaps, is that the win rate of heavy ground win 'virgins' is also higher than those to have won in the deep mud previously.
Performance of Favourites in Handicap chases
Maiden Chasers
Finally, in terms of hard data, I want to explore certain favourites in handicap chases. Firstly, let us consider the performance of handicap chase favourites who had never previously won a race over these bigger obstacles. To qualify they must have raced at least once over fences in their career:
These losses are 7p in the £ above the average figure for all NH favourites, which means once again there should be value by ignoring the favourite and looking at the other runners in the race.
Let's further consider the subset of these handicap chase favourites who had not only failed to win a chase but in all previous chase runs had not been placed either:
Logic suggested to me that these runners might perform relatively poorly as regards favourites and the figures bear that out. Sometimes results turn out like you would expect them to!
*
This article has highlighted several cohorts of potentially poor value NH favourites based on the last six calendar years of UK jump racing. Now, as I always say, articles like this are reporting on the past; there are no guarantees that the figures shared will be replicated in the future. However, most of the sample sizes are decent and the angles are underpinned by credible logic, which gives much more credence to figures. If I was sharing favourite results with only 50 qualifying runners, or where I was unable to explain the results, then it would be right to be sceptical.
Before wrapping up there are a couple of other ideas I have had in regard to finding potentially poor value favourites and both involve using Geegeez tools. The first is looking for NH favourites who display a negative run style. Imagining a two mile handicap chase at Hexham as an example, if we look at the run style figures going back to 2016 (8+ runners) in the Pace Analyser tool, we see the following:
Clearly this course and distance favours front runners/prominent racers. If the favourite happened to be a habitual hold up horse, then this may be an opportunity where the value lies elsewhere. Using Query Tool, here are the results for this group of favourites, by run style (4 is led, 3 prominent, 2 mid-div, 1 held up). There was just one winner from the twelve clear favourites to race in the latter part of the field in these races, whereas those favourites on the lead or prominent won ten from 22 for a better than 22 point profit (ROI 100%) at SP.
The second idea involves using the Instant Expert tab. The idea behind this one is to look for favourites that do not have many ‘greens’ within the traffic light ranking system. Green data is positive, and here is a recent example where the traffic light system seemed to highlight a poor value favourite. The race in question was the 3:33 at Wetherby on 18th October 2023. Here is a screenshot of the Instant Expert screen for that contest:
The favourite was Deyrann De Carjac, but looking along his row, we can see no greens, three reds and two ambers. Now, in this case the figures are looking at win percentages across each horses’ entire career. Whether this is the optimum setting, I'm not sure. You could look at placed percentages instead, or cut the data to the last two years, race code to chase, select handicaps only etc. However, based on the long-term win percentages for the runners in this race, Deyrann De Carjac looked a poor value favourite. Of the remaining runners Mackenberg priced at 15/2 with four greens was arguably offering some value.
Even looking at handicap chase form in the last two years only, Deyrann De Carjac was unappealing, and Mackenberg well suited to conditions:
The result of race was:
As we can see, Deyrann De Carjac was last of the five finishers. Also, for eagle-eyed readers you may have noticed that he was a ten-year-old racing in a chase – a poor value favourite stat I shared earlier. Mackenberg didn’t win to give the ‘dream’ result but ran well to finish second beaten a short head.
I appreciate this is just one example and this is a very tricky one for which to collect historical data. However, as I said before, there is plenty of logic to suggest it ought pay off in the long term: these horses are showing themselves to be either unsuited or unproven against today's conditions and they're being sent off favourite. That doesn't appeal to me!
I hope you have found this article interesting and also illuminating. If you have any ideas to test for poor value favourites, please drop them in the comments.
- DR
Other Recent Posts by This Author:
- More on Price Movement from Opening Show
- A Study of Owners in UK Flat Racing, Part 2
- A Study of Owners in UK Flat Racing, Part 1
- Seeking Value with Female Flat Jockeys
- Early Odds to Opening Odds to Starting Price
Very interesting article this for me Dave as I play the Favourites market on the spreads.From two years of records as you see only NH racing produced a profit…
NH +219 pts
Flat -74pts
AW -168 pts
I work out my own prices so only bet when my figure is above the buy price and this calendar year on NH racing only its showing a 73.5 pt profit.
Thanks Suzanne – good NH figs. I am guessing you are tempted to just stick with NH favs from now on? Good luck with it.
Dave
Hi Dave,
Great article, lots of food for thought. I feel there is a lot of sentiment in NH racing as the horses are with us for many years sometimes. How about looking for horses that won a good quality race in their younger days that are still racing as a 9 or 10 year old?
That’s a very logical thought – I’ll see if I can test out this idea in some way.
A lot to think on there – I do back favourites (even Odds On Fav’s) from time to time when they are highlighted by one of my micro systems and it looks as if conditions will suit them. This is based on the fact that there can be no value picking a horse that you feel cannot win as opposed to one you feel has every chance of winning (The Wisdom of Crowds approach)
Great article Dave giving me food for thought because i go for value when backing Favs so my bets are always at SP but occasionally i find the fav as changed by race time so i’m left with 2nd or 3rd fav
Hi Dave,
Well done on another entertaining and informative article. This is a particularly stimulating area due to the automatic value available where the favourite can be beaten. Personally, I take a different approach to achieve the same objective. Let me explain further.
Firstly because horse racing in the UK is made up of constantly revolving sideshows, I look at what has happened historically in similarly framed races at today’s track to tell me where poor value resides. I apply filters relating to things like race type (to match today’s), age restrictions, time of year, etc. I might also filter on class, distance etc if I feel it’s important.
An example might be a nursery where in some instances a higher exposure is a negative whereas in differently framed 2yo handicaps it becomes an absolute positive. Such variance may be down to the physical demands of the track or it may be the class of the race.
Another example, especially in some of the geographical outreaches like Exeter, you may detect in handicap chases that locally trained horses with recent positive form dominate the market. However research in that niche can often reveal these are dreadful value. The reason for such outcomes could be trainers running horses out of convenience (or desperation!) rather than it being a long-term target. Unexposed younger chasers making a longer journey can be exponentially better value depending on the class of the race.
In some novice hurdles, it’s best to oppose lightly raced market leaders at some tracks at certain times of year where a certain age restriction applies. In some 2yo novice races, you want to be against horses who ran well on their only appearance. The list goes on and on. Maintaining an agile approach is what I personally would recommend to stay in line with the relevance of today’s race.
Some sites do occasionally do race profiling but they are often lazy efforts and only on the big races. It’s best to do it yourself and focus on the minor meetings to get the real value.
Tools like HorseRaceBase are invaluable in this respect but Geegeez Query Tool can assist if people are prepared to do some manual legwork to get into the finer details.
All the best,
Russ
Thanks Russ for detailed and enlightening reply. I agree that race profiling can really help in less fashionable races / meetings. Regardless of ‘angle’ I personally prefer betting at a lower level be it meeting or race (or both). The big races are analysed in so much depth by so many people I find it hard to obtain what I perceive to be value. My only concern with certain ‘niche’ angles or patterns is sample size – is that something that concerns you at all? Thanks again. Dave
Hi Dave,
Yes I definitely agree that sample size can be a problem especially if a really tight set of restrictions are applied. For this reason, I can often look at a day’s racing and find there’s no race that I can confidently single out.
A little bit of relaxation can be applied in terms of trip especially over jumps. I think that time of year needs to be applied more rigidly at certain times (like at present) but as seasons get into full swing you can ease up and increase the sample size.
What I like to do as a start point is break down the data by odds ranges and if there’s an ever increasing better place return as I get to the really big odds ranges (eg, 20/1 to 40/1) then it’s usually enough to reassure me the market is overlooking something. This can help where the sample sizes are limited.
The trick then is to look at today’s contestants for inspiration as to what the golden nugget might be. If there’s a variance between experience, recent finishing positions, origin of sire or dam (these can be really important where contestants are lightly raced) or number of placed efforts (can be a gem), I’ll look to see if these factors have historically been (a) significant and (b) overlooked by the market.
I haven’t been doing this a long time I should say and it’s not my primary approach but it does seem significant in terms of what I’ve seen to date. Patience is key but persistence has led me to some interesting market oversights. You are still often left with 2 or 3 contenders though and if these are lightly raced (or untraced) it can be a challenge to get right. Laying an unsuitable favourite may be the better ploy if the case against it is stronger than the collective positives of the bigger priced rivals.
Finally I would say that Irish racing shouldn’t be overlooked. The increasing infiltration of German sired hurdlers in novices and maiden events is something trainers seem to be latching onto. These horses seem to need about 3 runs under their belt and then they start to fire provided the tracks are not too sharp. There you go, a golden nugget to do some data mining on!
Cheers,
Russ
Thanks Russ – when I get time I’ll do some digging. Cheers, Dave