Category Archives: Cheltenham

Willie Mullins – An Off Season?

Willie Mullins has led the Irish charge to Cheltenham for a number of years. He was the leading trainer every year from 2013 to 2016 and is largely responsible for the recent Irish dominance at the festival. However, this year there his Cheltenham raiders aren’t surrounded by the usual excitement and hype. There is a general feeling that his yard never quite hit form this season, and as a result he sits third in the betting for leading trainer at the festival.

Previous Festivals – Where do the Winners Come From?

Mullins has trained 32 winners at the last 5 festivals, so it’s worthwhile to break them down in order to identify his areas of strength. I’ve divided the festival races into 7 categories:

  • Non-Handicap Novices Hurdles – the Supreme, Ballymore, Albert Bartlett, Mares Novices Hurdle and Triumph Hurdle
  • Non-Handicap Novices Chases – the Arkle, JLT, RSA and National Hunt Chase
  • Open Grade 1 Hurdles – the Champion Hurdle, Mares Hurdle and Stayers Hurdle
  • Open Grade 1 Chases – the Champion Chase, Ryanair Chase and Gold Cup
  • Handicap Chases – the Ultima, Close Brothers, Festival Plate, Kim Muir, Grand Annual
  • Handicap Hurdles – Coral Cup, Fred Winter, Pertemps, County Hurdle, Martin Pipe
  • Other – Bumper, Cross Country, Foxhunters
Non-Hcp Novices Hurdle Non-Hcp Novices Chase Grade 1 Open Hurdle Grade 1 Open Chase Handicap Hurdle Handicap Chase Other
2018 wins 1 2 2 0 1 0 1
2017 wins 2 1 1 1 1 0 0
2016 wins 3 1 2 1 0 0 0
2015 wins 1 3 2 0 2 0 0
2014 wins 2 0 1 0 1 0 0
Total wins 9 7 8 2 5 0 1

So, a few points worth noting:

  • Mullins has had 16 wins in the non-handicap novice races over the last 5 years – in other words, he has trained 16 of the 43 winners of these races in the last 5 years.
  • He has had just 5 wins in handicaps, and no wins in handicap chases.
  • He has had 10 wins in open grade one hurdles and chases, and his other win came in last year’s Bumper.

Novices

It would seem that Mullins’ key strong point is, unsurprisingly, the level-weight novice contests. In the Supreme, Arkle, National Hunt Chase, Ballymore, RSA, JLT, Mares Novices Hurdle, Triumph and Albert Bartlett, he has trained 18.4% of the total field in the last 5 years. His horses accounted for 37.2% of the total winners in that period of time. Mullins trained 27.6% of the horses to make the frame in the last 5 years (including winners).

However, the general consensus seems to be that these wins followed seasons of domination in Ireland, and that the horses were well fancied and in good form when Cheltenham came around. I want to weigh up how Mullins’ novices are performing this season compared to past seasons, so I’m going to look at how his novice hurdlers and chasers performed in graded level-weight contests in past seasons before the month of March. Firstly, his novice hurdlers (note that this doesn’t include juvenile hurdles):

Mullins Novice Hurdlers (Graded Non-Handicaps) Before March

Season Runners Wins W% W/P W/P%
2013-14 25 9 36 14 56
2014-15 31 10 32 16 52
2015-16 27 14 52 18 67
2016-17 30 8 27 14 47
2017-18 31 8 26 11 35
2018-19 31 8 26 12 39

The 2015-16 season was a standout one in terms of novice hurdlers – Mullins won 14 graded events in Ireland before heading to Cheltenham, with over half of his runners in these races winning. However, the following 2 seasons were slightly lower with 8 winners each year before Cheltenham. This season has actually seen the same number of winners to date.

Looking at novice chasers next:

Mullins Novice Chasers (Graded Non-Handicaps) Before March

Season Runners Wins W% W/P W/P%
2013-14 16 5 31 8 50
2014-15 26 10 38 17 65
2015-16 16 8 50 8 50
2016-17 22 6 27 9 41
2017-18 25 6 24 11 44
2018-19 20 6 30 10 50

It’s the 2014-15 season that was the standout this time, but again Mullins has trained the same number of winners in these races as he did in the previous two years. In fact, he has done so with slightly fewer horses, so his win rates (and win/place rates) are actually slightly superior this year to anything else we’ve seen since the 2015-16 season.

These numbers alone tell us that Mullins isn’t having that much of an “off season”. However, a closer look at this year’s winners to date might. I’ve listed the winners in the 2018-19 season below in all cases where they appear towards the top of the festival markets:

  • Cadmium – 20/1 Grand Annual
  • Voix Du Reve – 25/1 Arkle
  • Camelia De Cotte – 33/1 Arkle, 25/1 JLT
  • Ballyward – 6/1 National Hunt Chase
  • Sancta Simona – 12/1 Mares Novices Hurdle
  • Aramon – 12/1 Supreme, 25/1 Ballymore
  • Klassical Dream – 8/1 Supreme

Some of these horses accounted for more than one of the wins in the tables, and some other winners are out for the season and won’t be heading to Cheltenham. This list tells a fairly different story to the tables – Mullins’ horses may be winning races, but they’re certainly not well fancied for festival races. The fact that Ballyward (second favourite for the four-miler at 6/1) is Mullins’ leading hope in the novice contests according to current prices is fairly damning considering the fact that he trained the Supreme favourite or joint favourite in each of the last 4 renewals, and that all of them were shorter than 4/1. He also trained the Arkle favourite in 4 of the last 5 renewals, and 3 of these were odds-on. This year, his leading Supreme contenders sit 3rd and 5th in the market at 8/1 and 12/1, and his leading contender in the Arkle is fourth in the betting at 13/2.

These are just two examples of races in which Mullins doesn’t seem to have the same strength as in past years. Another is the Mares Novices Hurdle – he has trained all 3 winners of the race, each of them the favourite at odds of 4/7, 11/8 and 8/11. This year, his top contender is probably Sancta Simona who can be backed at 16/1, putting her around 10th in the market (although she is as short as 8/1 with one firm).

In short, I’m not sure that it’s fair to say that Mullins is having an “off season” as a whole – we’ve seen that his success in big novice races has been on a par with the last few seasons, and his win strike rate in all Irish races so far this season (25.08%) is more or less on a par with where he finished up last season (26.6%), even if they are below the 3 seasons before last (33.75, 33.21 and 31.52). However, it seems unlikely that his success at the Cheltenham festival (with novices at least) will match that of previous seasons.

The Positives – Handicap Hurdles

So, are there any positives to be taken from this? Well, there is one area which isn’t necessarily associated with Mullins, but in which he has excelled over the years – handicap hurdles. He hasn’t always aimed a large number of runners at these races (from 2007 to 2012 he had an average of just over 3 runners per year, including none in 2008 and 1 in 2009). However, he was nevertheless successful, with 3 winners and another 2 places from his 20 runners in that period. In recent years he has aimed more runners at the handicap hurdles – he has had 55 runners in these races in the last 5 years. His win rate has been relatively low by his own high Cheltenham standards, with 9 winners from these 55 runners (a 9% strike rate with 1.27 A/E) and a further 7 places (a 22% W/P strike rate). The 1.27 A/E is the more interesting point – despite Mullins’ high profile, his horses in these races don’t tend to be as heavily backed as in grade one contests – Bleu Berry was 20/1 when winning the Coral Cup last year, and Arctic Fire was the same price when he won the County Hurdle the previous year off top weight.

An interesting angle is that all of these winners and placed horses ran in a graded race last time out – if we focus just on horses which ran in graded hurdles last time out, it makes the record even more impressive:

Mullins Hcp Hurdlers (Ran in Graded Hurdle LTO)

Runners Wins W% W/P W/P% P/L(BF) A/E
36 5 14 12 33 71.53 1.81

I’ve broken this down a bit more with a few other angles, which can be seen below:

Location of Prep Run

Track Runners Wins W% W/P W/P% P/L(BF) A/E
Leopardstown 15 3 20 5 33 50.87 2.56
Clonmel 3 1 33 2 67 11.59 1.96
Other 18 1 6 5 28 9.07 0.93

As ever, a Leopardstown prep run is proven to be valuable experience heading towards the festival. Interestingly, Clonmel has also produced a winner and a runner-up from 3 runners. Both of these (Don Poli when he won the Martin Pipe in 2014 and Roi Des Francs when finishing 3rd in the same race the following year) ran in the Surehaul Mercedes Benz Novices Hurdle, a grade 3 3m novices hurdle run at Clonmel in February. Battleford took the same route to the Martin Pipe in 2017 but only managed 8th.

Graded or Listed Winner?

Won at Listed Level or Above?

Runners Wins W% W/P W/P% P/L(BF) A/E
Yes 19 4 21 7 37 79.36 2.9
No 15 1 7 5 33 -5.83 0.82

4 of Mullins’ 5 handicap hurdle winners had previously won a graded or listed contest, as had another 3 placed horses.

Which Race?

Race Runners Wins W% W/P W/P% P/L(BF) A/E
Coral Cup 18 1 6 3 17 13.29 0.89
Fred Winter 5 0 0 0 0 -5 0
Pertemps 1 0 0 0 0 -1 0
County Hurdle 18 2 11 6 33 34.48 1.71
Martin Pipe 13 2 15 3 23 10.76 1.57

An obvious question to ask is which handicap hurdles Mullins’ horses are being aimed at and are winning – it’s clear that he doesn’t tend to run many horses in the Fred Winter or Pertemps, but does run them in the Coral Cup, Martin Pipe and County Hurdle.

More Festival Content can be found in the Cheltenham 2019 Section.

The Cheltenham Prep Run – Leopardstown

Cheltenham prep runs are scrutinised annually in order to determine which horses have had the “perfect preparation” for what is generally the main target of the season. Whilst there are a number of different angles, many of them race-specific, I wanted to begin by taking a look at which tracks tend to produce Cheltenham winners.

I’m going to begin by focusing on the tracks with the best win strike rates in the last 10 years. Other important information is included, because a good win rate could be a result of luck (especially from a smaller sample) if not backed up by a good W/P strike rate, or if the track also has a low A/E suggesting that although a few runners won, they actually disappointed based on the market as a whole.

The best 10 in terms of win strike rates are as follows:

Tracks Runners Wins W% W/P W/P% P/L(BF) A/E
Leopardstown 477 57 11.95 155 32.49 175.82 1.16
Navan 99 10 10.1 22 22.22 68.32 1.52
Punchestown 173 17 9.83 42 24.28 -50.2 0.88
Thurles 54 5 9.26 11 20.37 26.74 1.43
Fairyhouse 109 10 9.17 28 25.69 13.86 1.13
Naas 92 7 7.61 25 27.17 88.89 0.84
Exeter 79 6 7.59 15 18.99 22.28 1.27
Kempton 342 25 7.31 69 20.18 54.41 0.99
Cheltenham 535 34 6.36 106 19.81 -188.93 0.82
Warwick 133 8 6.02 27 20.3 10.27 1.04

There’s just one other with a positive A/E:

Track Runners Wins W% W/P W/P% P/L(BF) A/E
Ffos Las 39 2 5.13 6 15.38 14.23 1.21

And one more with a high W/P%:

Tracks Runners Wins W% W/P W/P% P/L(BF) A/E
Clonmel 31 1 3.23 9 29.03 -16.41 0.54

(Remember, A/E doesn’t take places into account).

All of this analysis is interesting in its own right, and there are probably angles which could be exploited for every single track – however, we’re inevitably brought back to one particular track which tends to churn out festival winners on an annual basis – Leopardstown.

Leopardstown

Leopardstown is the second most popular prep track after Cheltenham itself. This is hardly surprising, with the Dublin track hosting the bulk of the quality racing in Ireland between Christmas and the Festival. The Slaney Novice Hurdle in Naas in January is the only Irish grade one outside of Leopardstown from January to March. Since 2018, the quality National Hunt racing between Cheltenham and Christmas has almost all been squeezed into one weekend at Leopardstown’s Dublin Racing Festival, leaving few other options for Irish trainers looking for a prep-run for a top-level horse before the festival.

Handicaps?

First, I want to break this down to look at handicaps and non-handicaps at the festival:

Festival Race Runners Wins W% W/P W/P% P/L(BF) A/E
Handicap 160 11 6.88 44 27.5 63.21 0.96
Non Handicap 317 46 14.51 111 35.02 112.61 1.22

So, runners in non-handicap races (of which there have been about twice as many as in handicap races) performed notably better in terms of wins and places, and have a positive A/E.

Trainers

Predictably, Mullins and Elliott are among the trainers with the best strike rates coming from Leopardstown. Unsurprisingly Jessica Harrington is another top performer, but her win rate (admittedly from a small sample of just 15) is superior theirs, and the market seems to underestimate her runners more:

Trainer Runners Wins W% W/P W/P% P/L(BF) A/E
Jessica Harrington 15 3 20 6 40 25.18 1.95
Gordon Elliott 42 7 16.67 21 50 39.2 1.28
Willie Mullins 149 22 14.77 50 33.56 41.73 1.1

More notable is the lack of success from another top Irish trainer – Noel Meade hasn’t had a single festival winner coming from a Leopardstown prep run in the last 10 years, although his good W/P strike rate makes up for this, and the form of those with SP’s of 10/1 or less is fairly consistent (43343023):

Trainer Runners Wins W% W/P W/P% P/L(BF) A/E
Noel Meade 17 0 0 6 35.29 -17 0
Chases v Hurdles

Looking at type of race, chasers outperform the others in terms of win strike rates, place strike rates and A/E:

Race Type Runners Wins W% W/P W/P% P/L(BF) A/E
Chase 210 30 14.29 76 36.19 46.72 1.27
Hurdle 236 25 10.59 73 30.93 107.75 1.08
NH Flat 31 2 6.45 6 19.35 21.35 0.83
Leopardstown Profits

I haven’t paid much attention to P/L at BFSP so far as it’s a metric that can be skewed a lot by one or two shock winners. However, it is incredible that you can make a profit blindly backing horses which ran at Leopardstown last time out in chases, hurdles and the bumper, regardless of whether the races in question are handicaps and non-handicaps. This has been true in 7 of the last 10 years:

Year P/L(BF) ROI(BF)
2018 29.29 41.25%
2017 6.15 13.67%
2016 -14.61 -30.43%
2015 -6.45 -11.72%
2014 43.7 94.99%
2013 61.26 127.63%
2012 53.43 121.42%
2011 18.16 45.41%
2010 3.26 8.15%
2009 -18.37 -45.93%
Winners

Horses which won their prep race at Leopardstown have a particularly strong record, with almost 20% of them winning at the festival and almost half of them making the frame.

Leopardstown Winner Wins Win% W/P W/P% P/L(BF) A/E
146 29 19.86 68 46.58 93.4 1.25
The Dublin Racing Festival

Last year was the first year that horses came to Cheltenham from the Dublin Racing Festival, and the form of those horses which won at Leopardstown on their last start before the festival last year reads 1312251102025F84. That’s 4 wins and 4 2nd’s from 16 horses (note that not all of these came from the Dublin Racing Festival – Road To Respect, for example, won at Christmas and didn’t run again before Cheltenham). The full record from last year is below:

Arkle Chase Footpad 1st
Mares Hurdle Apples Jade 3rd
Ballymore Samcro 1st
RSA Chase Monalee 2nd
Champion Chase Min 2nd
Bumper Blackbow 5th
Bumper Relegate 1st
JLT Novices Chase Shattered Love 1st
Pertemps Mine Now 15th
Stayers Hurdle Supasundae 2nd
Festival Plate Last Goodbye 10th
Triumph Hurdle Mr Adjudicator 2nd
Albert Bartlett Tower Bridge 5th
Gold Cup Total Recall Fell
Gold Cup Edwulf 8th
Gold Cup Road To Respect 4th

Looking at the list itself probably makes the statistic even more impressive – Apples Jade was 3rd but turned out to be in season, Min and Monalee lost nothing in defeat to Altior and Presenting Percy respectively, and other than the Gold Cup horses the only ones not to finish in the top 3 were in the Bumper, Pertemps, Festival Plate and Albert Bartlett – races which last year had 23, 23, 22 and 20 runners respectively (the horses in question had starting prices of 5/1, 25/1, 8/1 and 33/1 respectively).

More Festival Content can be found in the Cheltenham 2019 Section.

Willie Mullins’ Chasers at the Cheltenham Festival

One interesting note to come from last year’s festival was that Willie Mullins didn’t school Douvan over fences before he went to the Champion Chase. Not only this, but he didn’t school Rathvinden for the four-miler either. Mullins’ priority seems to be getting them to the race safely as opposed to having them jump flawlessly. However, Mullins’ record with chasers at the festival hasn’t really been magnificent through the years.

His record at the festival as a whole doesn’t need to be explained – he has been the undisputed King of Cheltenham for a number of years now. However, if we compare his runners in hurdles and chases, there’s a stark contrast.

Race Runners Wins W% W/P% P/L(BF) A/E
Hurdle 245 33 13% 30% 25.14 1.1
Chase 118 13 11% 31% -77.94 0.79

The hurdles record is outstanding – you actually could have made a profit by blindly backing his hurdlers, and this has been the case in 6 of the last 10 years. For 8 of the last 10 years, his hurdlers’ A/E has been over 1, suggesting that they’re still performing better than the market expects. His chasers, on the other hand, have only been profitable in 1 of the last 10 years to level stakes, and only had an A/E of over 1 twice in that decade.

This obviously isn’t all due to their jumping. After last year’s festival, however, there was a lot of attention given to the fact that he has had 19 fallers (including horses which unseated and one which was brought down) in chases over the last 10 years at the meeting. 6 of these were last year, making this a more pressing “issue” than ever. 19 fallers from 118 runners is a total of 16%.

Before adding to the narrative that Mullins’ horses aren’t good jumpers, I want to compare his faller rates to those of other trainers. I’ve included the only other trainers to have had over 100 chasers at the festival in the last 10 years (Nicky Henderson, Paul Nicholls and David Pipe) as well as Gordon Elliott, as Mullins’ main current rival and the trainer to whom he is most often compared at the moment.

Trainer Chasers Fallers Fall%
Willie Mullins 118 19 16%
Nicky Henderson 153 9 6%
Paul Nicholls 154 20 13%
David Pipe 101 9 9%
Gordon Elliott 48 6 13%

Firstly, Nicholls and Elliott have very similar rates to Mullins, if slightly lower. If we eliminate last year’s fallers (this is hardly unreasonable as there’s a distinct possibility that last year’s high faller rate was an anomaly), Mullins’ rate drops to 12.87%, which is on a par with Elliott, the trainer currently closest to Mullins in terms of success at the top level, and Paul Nicholls, still often referred to as the best trainer of chasers in the game, despite a lack of recent success at the top level.

Looking away from fallers, though, Mullins’ record with chasers compared to hurdlers should still be noted. We can break it down by looking at the record of different jockeys when riding for Mullins – three jockeys have been seen in chases over 10 times for Mullins at the festival in the last 10 years, and these are Ruby Walsh, Paul Townend and Patrick Mullins:

Jockey Runners Wins W% W/P W/P% P/L(BF) A/E
Ruby Walsh 35 9 25.71 14 40 -11.27 1.06
Paul Townend 28 0 0 5 17.86 -28 0
Patrick Mullins 23 2 8.7 7 30.43 -13.9 0.87

Ruby’s record is impressive. He has won on over a quarter of his rides in chases for Mullins and has (just about) outperformed the expectations of the market. Patrick Mullins’ record isn’t bad and he has been on plenty at massive prices – if we narrow it down to those at 10/1 or less, his record is 2 wins and another 3 places from 12 runners (A/E 1.12).

Paul Townend’s record isn’t as impressive, although one would suspect that he has played second fiddle to Ruby over the years, riding second-strings. For this reason, I want to look at his in the same way as I looked at Patrick Mullins’ – narrow them down to horses priced 10/1 or shorter. This narrows the amount of runners down to 8, confirming the theory that he’s regularly on stable second-strings and relative outsiders. His form reads 56F2P2P – the last four were at last year’s festival when Walsh was injured.

He has managed to place twice – however, the prices on the exchange would have suggested 3.2 places from these 8 runners, as well as 1.5 wins. This is admittedly a small sample size and it’s possible that we’re being harsh on Townend (especially considering the fact that one of his second-place finishes was a very respectable one on Min behind Altior in last year’s Champion Chase) but his inferior chase record to Ruby’s may be worth bearing in mind.

More Festival Content can be found in the Cheltenham 2019 Section.

Can a 4-Year-Old Win the Supreme?

One pressing issue this year ahead of the first race of the festival is likely to be the presence of a four-year-old runner. This occurrence isn’t unheard of – there was one last year – but the increased significance this year is understandable, given the fact that Khudha went off at a price of 200/1 last year and finished down the field, as, with no disrespect to the horse or connections, would have been expected. Before that, the last 4yo to run was the admittedly smart Marsh Warbler who had won a grade one juvenile hurdle but went off at 22/1 and was out of his depth, finishing down the field in a Supreme featuring the likes of Sprinter Sacre, Cue Card and Al Ferof.

This year, however, Joseph O’Brien has dominated the Irish juvenile scene and with Sir Erec proving a worthy Triumph favourite in an impressive win at Leopardstown’s Dublin Racing Festival, it’s looking increasingly likely that stablemate Fakir D’Oudairies, who won the Triumph trial on Cheltenham Trials Day with ease, will be re-routed to the Supreme in order to keep O’Brien’s runners apart. So, the question arises again this year, can a four-year-old win the Supreme?

At first glance the record of juveniles in the Supreme looks poor – three horses have tried it in the last 10 years, finishing 13th, 11th and 9th. If we look back over the last 20 years, 15 4-year-olds have gone to the race and only one has won, with another one making the frame. The winner in question was Hors La Loi III in 1999, and the runner-up was Binocular in 2008. They went off at 9/2 and 8/1 respectively, making them two of just three four-year-olds to run in the Supreme at single-figure starting prices in the last two decades.

In short, the form of 4yo’s in the race in the last 20 years looks dire (01F009007002900) but this may be misleading – the form of 4yo’s at single-figure prices is 129.

Admittedly, we’re looking at a very small sample size here, and there may be some logic behind the idea that a 4yo isn’t ideally suited to this test. To my mind, the Supreme is a race for as an experienced horse. This is looked into in detail elsewhere, but to summarise, horses with at least 4 starts over hurdles (in the UK or Ireland) have outperformed those with 3 or less in the last 10 years:

Runs over Hurdles Runners Wins W% W/P W/P% P/L(BF) A/E
1-3 92 2 2.17 14 15.22 -83.76 0.3
4+ 69 8 11.59 16 23.19 50.03 1.52

The two winners with 1-3 hurdles starts before running in the Supreme were Douvan and Vautour, each of whom had run twice in France over hurdles before arriving in Closutton. This seems to be a very strong stat, with the majority of runners in the Supreme in the last 10 years not having had as much hurdling experience as would be desired. To me this is a major negative for this year’s favourite Angels Breath who has run once over hurdles and missed his planned second start due to the influenza. At the time of writing it’s unclear whether he will get a prep run before the festival, but even if he does, he will only have run in two races over hurdles.

Compare this to horses in the field such as Aramon (6 races over hurdles including a win and a second place in a grade one hurdle), Klassical Dream (7 starts over hurdles including a grade one win) and Elixir De Nutz (6 starts over hurdles including a grade one win) and 4/1 seems a very short price for the favourite.

In short, I believe heavily in the idea that experience is a big advantage in a Supreme. With this in mind, I’d be probably be against a lot of four-year-olds running in the race. However, Fakir D’Oudairies has not only run in four races over hurdles (two since he arrived in Joseph O’Brien’s yard and two in France) but also in two chases in France. He’s not lacking in experience, and for this reason I wouldn’t necessarily discard him in a Supreme based solely on his age.

More Festival Content can be found in the Cheltenham 2019 Section.

Cheltenham Handicaps – Spread of the Weights

With the popularity of the festival increasing, there’s extremely high demand to run horses not just in the graded races but also in the handicaps. This means that a lot of handicaps are contested by horses who are good enough to win at graded level, and the ratings can be fairly compressed, ultimately making the actual weights less significant than they are at other meetings. The market possibly hasn’t caught on to this yet, so horses in the top half of the weights can be under-bet as punters prefer something which is theoretically better treated. It’s worth looking at each handicap individually and seeing the spread of the weights. This allows you to determine the relative significance of the weight being carried by horses.

Although every year will be different, so we’ll only know for certain about the spread of the weights in this year’s handicaps when they’re released, it’s worth looking back on previous years to look at the patterns:

The Festival as a Whole

The method of looking at the spread of the weights is simple – just take the top OR in a race, compare it to the bottom OR and find the difference between the two. The average difference between top and bottom OR in last year’s 10 handicaps was 16.5, and this has been fairly steady over the last few years:

Average Difference by Year
Year Avg Dif
2018 16.5
2017 16.1
2016 14.7
2015 15.2
2014 16.7
Overall Average 15.84

It’s worth trying to put this into context, and to do so I’ve looked at the average difference between top OR and bottom OR in two top meetings this year either side of the Irish Sea, each of which hosted 5 handicaps – Leopardstown’s Dublin Racing Festival and Kempton’s Christmas meeting.

Leopardstown DRF 2019
Race Bottom OR Top OR Difference
0-150 2m Hcp Hdl 120 148 28
0-150 2m1f Hcp Ch 113 141 28
2m2f Mares Hcp Hdl 112 133 21
0-150 3m Hcp Hdl 119 139 20
2m5f Hcp Ch 130 150 20
Average Difference 23.4
Kempton Christmas 2018
Race Bottom OR Top OR Difference
Nvc Hcp Ch 121 137 16
2m5f Hcp Hdl 122 142 20
3m Mares Hcp Hdl 120 142 22
3m Hcp Ch 120 144 24
2m Hcp Hdl 114 142 28
Average Difference 22

Both of these are notably higher, so the theory stands (not that we needed much proof, as it’s an easy observation to make) – as more and more horses are campaigned towards festival handicaps, the ratings tend to be more compressed and therefore so are the weights. This also acts as a good reference point – 3 of these 5 races had ratings spreads of 28, and the average difference of the 10 races was 22.7.

Race by Race

Cheltenham does host 10 handicaps over the week, though, and it’s not much use to evaluate them all as a whole. So, I’m going to look at which handicaps have the highest average differences and which have the lowest. In theory, this should show us which handicaps should be looked at as handicaps in the traditional sense (assessing a horse’s capability and comparing this to the mark given by the handicapper) and which should be analysed similarly to level-weights races.

Festival Plate
Year Bottom OR Top OR Difference
2018 137 155 18
2017 133 158 25
2016 135 157 22
2015 137 155 18
2014 131 157 26
Average Difference 21.8
Ultima
Year Bottom OR Top OR Difference
2018 137 155 18
2017 134 155 21
2016 131 153 22
2015 133 155 22
2014 129 151 22
Average Difference 21
Coral Cup
Year Bottom OR Top OR Difference
2018 135 153 18
2017 136 156 20
2016 139 158 19
2015 138 158 20
2014 135 154 19
Average Difference 19.2
County Hurdle
Year Bottom OR Top OR Difference
2018 133 154 21
2017 134 158 24
2016 138 152 14
2015 134 146 12
2014 132 154 22
Average Difference 18.6
Grand Annual
Year Bottom OR Top OR Difference
2018 139 154 15
2017 135 154 19
2016 137 152 15
2015 130 153 23
2014 136 154 18
Average Difference 18
Pertemps
Year Bottom OR Top OR Difference
2018 135 155 20
2017 137 147 10
2016 135 154 19
2015 135 152 17
2014 135 148 13
Average Difference 15.8
Kim Muir (Max 145)
Year Bottom OR Top OR Difference
2018 119 145 26
2017 133 145 12
2016 134 145 11
2015 130 145 15
2014 131 144 13
Average Difference 15.4
Fred Winter
Year Bottom OR Top OR Difference
2018 126 139 13
2017 124 139 15
2016 128 142 14
2015 129 139 10
2014 127 139 12
Average Difference 12.8
Martin Pipe (Max 145)
Year Bottom OR Top OR Difference
2018 136 144 8
2017 135 145 10
2016 135 142 7
2015 135 144 9
2014 133 146 13
Average Difference 9.4
Close Bros (Max 145)
Year Bottom OR Top OR Difference
2018 137 145 8
2017 137 142 5
2016 136 140 4
2015 134 140 6
2014 131 140 9
Average Difference 6.4

So, the Ultima, the Plate, the Coral Cup, the County Hurdle and the Grand Annual, although below the averages set by the Leopardstown and Kempton meetings, can probably still be described as true handicaps, for now anyway. The Pertemps and the Kim Muir are good examples of how much the ratings spread can change from year to year – last year they were 20 and 26 respectively, placing an emphasis on the ratings given to the horse by the handicapper, whereas in 2017 they were 10 and 12 respectively. The Fred Winter, the Martin Pipe and the Close Brothers Handicap Chase tend to have very low spreads, so really the handicap shouldn’t be the main priority when assessing those races.

Application

So how can this information actually be applied to our betting at the festival? Well, my theory would be that a large number of punters haven’t factored this into their analysis, and therefore the market hasn’t factored it in to the odds. Handicaps are viewed as handicaps, and the traditional method of assessing a horse’s chance of winning a handicap is to treat the weight it’s carrying as a key factor, if not the key factor. Punters love to latch on to the idea of a plot horse who has been overlooked by the handicapper – if the weights are less important in races like the Fred Winter, Close Brothers and Martin Pipe, then these horses should be over-bet and those at the top of the weights should be under-bet, therefore representing value.

More Festival Content can be found in the Cheltenham 2019 Section.

National Hunt Chase – An Overview

Class

It’s obvious that in a 4-mile novice chase, you’re going to need a tough horse with plenty of stamina. My angle on this race is a bit different, but even simpler – it’s all about class. In the last number of years, the class of horse competing in this race has improved significantly. 7 of the last 8 winners (including the last 4) were rated 143 or higher. The 5 winners before that were rated 127/129/130/123/120. Looking at last year’s race, 8 of the 16 runners were rated 143 or higher and their form read 1236FFU. The other 8 were rated 142 or lower and their form read 45PPPPPRU. The previous year, the first 5 home were all rated 143+, and these horses accounted for two thirds of the field.

It’s not really good practice to base this trend around the number 143 just because that was 2016 winner Minella Rocco’s rating. For this reason, I’m going to focus on horses rated 140 or higher. This has accounted for between one and two thirds of the field for the last few years:

Year % of Field Rated 140+
2018 50%
2017 67%
2016 65%
2015 35%
2014 20%
2013 25%
2012 16%
2011 19%
2010 6%
2009 11%

It would seem that the number of higher-rated horses competing has increased in the last 5 years, so I’m going to focus on those renewals now:

OR Runners Wins W% W/P W/P% P/L(BF) A/E
0-139 44 1 2.27 4 9.09 -34.58 0.71
140+ 42 4 9.52 11 26.19 12.56 0.85
145+ 27 3 11.11 10 37.04 15.53 0.92
150+ 6 2 33.33 3 50 25.82 2.38

It would appear that it’s a case of the higher the rating, the better. Rathvinden (last year’s winner) and Mossback (around 4th position when he fell 7 out) were the only two horses rated 150 or higher last year (both were rated 150). In 2017, it was just Tiger Roll (winner) and Edwulf (looked sure to finish 2nd until something went wrong after the last and he subsequently collapsed). In 2016, it was just Vicente (14/1 5th of 20) and in 2014, runner-up Shotgun Paddy.

Playing with the figures a bit will always change the results (if you include horses rated 149 it only adds two more runners but one was 2016 runner-up Native River) but the point is clear – the best-rated, classiest horses perform well in this race.

Experience

However, it’s hard to ignore the fact that this is a 4-mile chase at one of the most testing tracks in National Hunt racing. The percentage of horses which fail to complete the race each year is testament to this – in 2016, only 8 of the 20 runners completed, in 2017 10 of the 18 runners completed, and in 2018 (a particularly tough year considering the going) just 6 of the 16 runners managed to complete the race.

While we’ve already seen the importance of class, experience should in theory be vital too. This is backed up by the numbers – 5 of the last 10 winners had at least 8 starts over fences, despite these horses accounting for just 17% of the total field. They notably exceeded the expectations of the market and could have been backed blindly:

Chase Starts Runners Wins W% W/P W/P% P/L(BF) A/E
7 or less 144 5 3.47 22 15.28 -100.23 0.49
8 or more 30 5 16.67 8 26.67 36.69 2.82

It’s worth looking at this for just the last 5 years too, as we’ve seen already that the composition of the race and the type of horse that it attracts have changed in recent years. 3 of the last 5 winners had run at least 8 times over fences, despite the fact that just 13 such horses ran in that period:

Chase Starts Runners Wins W% W/P W/P% P/L(BF) A/E
7 or less 73 2 2.74 10 13.7 -51.55 0.37
8 or more 13 3 23.08 5 38.46 29.53 4.23

The W/P rates are fairly impressive too (bear in mind that although there are technically only 3 places in this race as it’s not a handicap, many bookies are likely to offer 4 on the day if there’s a big field).

So, we’ve got two clear angles here – we’re going to be looking for a horse who has proven class (we tested this by looking at those with higher official ratings) and plenty of chasing experience (we tested this by looking at those with 8 or more previous starts over fences).

The Ideal Candidate

When there’s a particularly strong angle into a race, as ever, I’m making out a “profile” for the ideal candidate for the race. As ever, it’s rare to find a horse that ticks all of the boxes, and strong past results aren’t too reliable as we’re basing the profile itself on the success of these horses. However, if we went with the following criteria:

  • OR 145+
  • 8+ chase starts

There would have been just 4 horses in the last 5 renewals which fit the bill – 2015 winner Cause Of Causes (8/1), 2017 winner Tiger Roll (16/1), 2018 winner Rathvinden (9/2) and 2018 runner-up Sizing Tennessee (8/1). The reason that I’ve focused on just 5 renewals is the fact that there have been more runners with higher ratings in recent years, and the race seems to be changing into more of a top-class affair. However, if we were to extend it back to the last 10 renewals, there would be two more qualifiers: 2011 winner Chicago Grey (5/1) and 2011 4th Alfa Beat (11/2).

Looking at this year’s market, ante post favourite Ok Corral is rated 153 but has only run twice over fences. Delta Work is likely to run in the RSA (and only has 3 chase starts anyway), Cubomania has run 10 times over fences but is rated just 142, Mortal is rated 148 but has only run twice over fences, Ballyward has run just twice over fences, and Santini is likely to run in the RSA (and has only run twice over fences). If Cubomania was to put in a good run at the Dublin Racing Festival, he would likely be of interest. However, we could end up with no “ideal candidate” this year (there were none in 2012, 2013, 2014 or 2016) or with something at a big price.

Mares Hurdle – An Overview

The Quevega Factor

When looking at the Mares’ Hurdle, I’m not going to use the same stats as I did when looking at other races. The reason for this is that Quevega won 6 of the last 10 races, so any win figures would be completely skewed by whether Quevega fit the criteria, as in turn would any win%, P/L(BF), and A/E figures. However, I still want to try to find an angle into the race, so I’m going to focus not only on the winners but also on the placed horses. So, my main way of taking on this race is by comparing the percentage of the total field which fit a stat to the percentage of the total places filled by these horses.

Stamina

This race is (interestingly) run over 2m 4f. There’s no grade one hurdle run over a middle distance at the festival open to all horses (there is this race, for mares, and the Ballymore for novices) and the argument is that if they were to introduce one, it would dilute the quality of both the Champion Hurdle and the Stayers Hurdle (the latter doesn’t tend to be a particularly exciting division anyway and doesn’t need its quality diluted any more). A few weeks later, there’s a grade one over this trip at Aintree’s Grand National meeting so any horse which isn’t quite suited to 2 miles or 3 miles gets their chance there.

The distance is often a forgotten factor in this race, though. People forget that at 2m 4f, it is a test of stamina, not just pure speed like the 2m Champion Hurdle that takes place before it. All of the last 10 winners had won over this exact distance. These horses accounted for 18 of the total places – that’s 60% of the total places from 38% of the total field. If we were to count Quevega as just one winner for a moment, forgetting about her other 5 wins, this would be 52% of the total places from 37% of the total field – still an over-performance. Looking at horses which had previously won over 2m 4f or further, these horses took up 28 of the 30 total places on offer.

Britain v Ireland

All of the last 10 winners were Irish-trained (in fact, 9 of them were trained by Willie Mullins and the other, Apples Jade, moved to Willie Mullins’ yard from Michael O’Leary’s following the Gigginstown split). Irish-trained horses accounted for 18 of the 30 horses which made the frame – that’s 60% of the total places on offer from 27% of the total field. It would appear that the Irish have taken notice of the lucrative rewards that are now on offer for having top-class mares in training a bit quicker than the British. The form of the top-placed English finisher in the last 10 renewals reads 2224232342 – there’s often a British horse there or thereabouts, but there has always been a better Irish contender.

Class

8 of the last 10 winners won last time out and these horses over-performed significantly, filling 60% of the total places from just 28% of the total field. This points back to some recurring themes in these top festival races – firstly, form, and secondly, class. The latter might be of even more interest. 9 of the last 10 winners were officially rated 150 or higher. These horses accounted for just 11% of the total field, but they managed to fill 50% of the total places. Just 19 horses have run in the race in the last 10 years were an official rating of 150 or higher, and their form reads 7213171111F21132173. Last year was actually the exception – the two contenders were Apples Jade (3rd) and La Bague Au Roi (7th). However, it is probably worth pointing out that Apples Jade clearly wasn’t at her best last year, and we were subsequently given the explanation that she was in season.