UPDATED (weight) Cheltenham 2015: Handicap Hurdles (inc 2 micro angles)

Cheltenham Handicaps

Handicap Hurdles

Well it was about time I started thinking about the Festival wasn’t it. Today I want to have a look at the handicap hurdles. This is an overview and looks at all the races combined, rather than individual races/trends etc. That will come in time.

For now I want to see what type of horse generally does well and what type of profile we need to keep an eye on.

As well as being interesting ( I hope) I also want to highlight a couple of mini-angles to keep onside. I will highlight the selections on each day of the festival and while I may not back them blind, they will be essential in helping me analyse the races.

Of course there will always be issues with sample sizes, but that is life. This is a festival and I don’t think that means we shouldn’t dig into the stats as a way of trying to find the type of profile that normally does well. I also don’t think there is much point looking pre 2010. 5 years is plenty, especially given that, for example, training methods change and improve, and so does the quality of a horse that a trainer can attract. I would think Paul Nicholls would say he is a better trainer now than 6 years ago? Breeding and buying patterns may change as well and all told I am happy with looking over the last 5 years. I know some users of trends only look at the last 8 years or so of a race.

Where I end up discussing micro systems I have obviously tried to ensure there is logic in every rule. It can be easy to unintentionally back-fit (or intentionally if you are looking to flog crap) but I don’t think I have done that. Normally if there is a logical reason for a rule then you are fine. And of course sample sizes will be quite small. While I won’t necessarily back the selections blind, I think they will be useful in providing another level of information for race analysis.

The sole aim, as with everything I write, is to try and help us all make some money…

So, let’s get to it.

This is a very long post. (10 hours +, nearly 4000 words)  If you want to get excited you can jump straight to the micro-systems at the bottom (which are explained in detail below also) and then come back and read the rest 🙂 

***

Headline stats 2010-

625 Bets / 25 wins / average of 25 runners per race

(when you see ‘places’ below, that figure means all wins and places combined)

 

Odds

It is best not to get too ‘creative’ here. There has been no winner priced 50/1 or higher, from 111 to have tried. Only 3 have placed.

In fact the stats for horses ISP over 20/1 read: 329 bets / 5 winners

21/1 to 49/1: 218 bets / 5 wins / 30 places / AE 0.73 (perform 27% below market expectation)

Compared to horses priced 20/1 or under: 296 bets / 20 wins / 67 places / AE 0.81

So, clearly bigger priced horses do win, but they are hard to find. And don’t bother with anything 50/1 or over. Of course there will be a winner at those odds at some point in the future, but I doubt I will be on. 20/1 or under appears to be the place to focus on.

Interestingly 2013 was responsible for 3 winners priced over 20/1. There were none in 2010 and 2011 and 1 in each of the other two years.

Odds should never be used as a strict rule for things such as the festival. (some would argue they should never be used, I don’t agree) The hope of course is that when you back a 25/1 or 33/1 or 50/1 shot in the morning that it shortens. But, it is something to keep in mind. You have to back quite a few big priced horses before your pin lands on a winner.

For the following stats I will focus on those priced 20/1 or under… (that does mean I have given 5 winners away already, but I get 3 of those back later)

 

Breeding

USA and German Breds are 0 from 20, and only 4 places to show. Indeed I don’t think there has been a German bred winner at the Festival for a very long time but haven’t checked that. I read it somewhere, Ben Aitken’s excellent NarrowingTheField blog I think…a must read for any NH fan.   

GB: 62 bets / 4 wins / 12 places / AE 0.82

IRE: 141 bets / 10 wins / 32 places / AE 0.9

FR: 72 bets / 6 wins / 19 places / AE 0.84

Not too much to go on at this stage, a fairly even spread although Irish and French breds do better from a win % and Win/Place%

 

Age

Again not too much at this high level.

The only things of note is the poor record of 4 year olds: 59 bets / 2 wins / 13 places / AE 0.39.

They really underperform against market expectations and have an abysmal win %. Especially when you consider that 5 year olds, from the same number of runners are: 59 bets / 6 winners / 18 places / AE 1.16  performing 16% above market expectations, and are profitable to 24.5 points at BFSP

11 year olds are 0/5 with 1 place.

All other ages are fine. 7 year olds are 0/52 but for the life of me I can think of a logical reason why this is the case. 8 year olds are: 26 bets / 4 wins / 7 places / AE 1.83 and would have made you nearly 50 point at BFSP

Not too many strong conclusions bar the fact that 4yos struggle. Excluding 7 year olds from any analysis would be back-fitting in my opinion. No logical reason, given spread either side, why they should struggle simply due to age. 4 year olds are in all likelihood still maturing and may not have the experience to handle these competitive big field races.

 

Career Runs

Some interesting observations…

Horses with fewer than 10 career runs running in handicaps: 137 bets / 6 wins / 30 places / AE 0.47

4 of those wins come from horses with between 1-3 career runs, who are clearly still open to massive improvement and may have enough hidden from the handicapper: 30 bets / 4 wins / 11 places / AE 1.28

That means horses that have had between 4 and 9 career runs are: 107 bets / 2 wins / 19 places / 0.21

Horses with 10+ career runs: 155 bets / 14 wins / 31 places / AE 1.2

I am trying to think of the logic here, as I write…maybe on the one hand you have horses that are open to bags of improvement and are clearly well handicapped (1-3 runs). Those that have had 10 runs +  are clearly more battle hardened and up for the fight. They either still have something up their sleeve or are back to a mark to be competitive from. And those in between with 4-9 career runs are exactly that, in between – they are neither unexposed enough to cause a shock, and at the same time don’t have the experience needed to get competitive. Your thoughts on this phenomenon would be interesting…I can’t think it is just an anomaly in the stats and I think there is some logic there! Or maybe it is a statistical anomaly..  

 

Career Wins

The only thing worth noting is that horses yet to win at all are: 16 bets / 0 wins / 3 places.

Winning experience is clearly important. There hasn’t been a winner with more than 8 career wins yet, but only 6 such exposed horses have tried.

 

Career Wins in a Handicap

FASCINATING

Those horses that have 3 or more wins in handicap company are: 34 bets / 0 wins / 4! places .

I find this surprising. Clearly being ‘exposed’ is no good thing here. Now, it is only 34 selections priced 20/1 or under, and something will surely win at some point. But it is the 4 places that I find surprising.

Those with 0 wins in handicaps (but, as above, clearly with a win in novice/maiden company against their name) are: 146 bets: 6 wins / 32 places / AE 0.49

Those with 1 or 2 handicap wins to their name are: 116 bets / 14 wins / 31 places / +142 BFSP / AE1.41

There must be another mini system here with those stats…I will come back to that later if I can find anything.

For now…

 

***

MINI SYSTEM 1

The unexposed angle interested me…

Rules

Handicap Hurdles

Horse-Run Career: 1-3

Horse- Win Career: 1 or 2 (0 wins are 0/10)

Odds: 14/1 or under (Over 14/1 are 1/14, including a 40/1 winner..there is something interesting about bigger priced horses, give me a moment…)

Results: 14 bets / 4 winners / 9 places / +31 BFSP / AE 2.13…

BUT..here is the exciting thing… the TRAINERS…

If I were to tell you that all runners come from 6 trainers in total, but that ALL 4 winners come from just 3 trainers…I think you could guess who they are…yes Messrs  Henderson, Nicholls and Mullins..

Henderson: 3 bets / 1 win / 2 win, place total +16

Nicholls: 2 bets / 1 win / 1 win, place total +3.5

Mullins: 3 bets / 2 winners / 2 win, place total + 18

As impressively these three are also responsible for the majority of the big priced placed horses and the one big priced winner… 

Henderson = 40/1 winner ; Nicholls – his two other runners came 4^th and 3^rd at 25s; Mullins a 4^th 25/1 and 6^th 20/1

David Pipe is the other powerhouse who has had 4 runners with the rules above. He has yet to have a winner but 2 have placed. This could be his year?

Just focusing on those 3 powerhouses, with no odds cap…

13 bets / 5 winners / 9 w,place / +85 BFSP / AE 4.03

These types of handicap hurdlers, from these powerful yards, always do well, regardless of price. Yes it is a small data set…but come on, it is the festival. Do you want me to wait until they have had 100 runners each 🙂

Of course none of the above may have a winner this year but the information should be useful nonetheless when looking to dissect the handicap hurdles.

***

 

Right, back to the general analysis…

 

Horses Runs in 90 Days

Handicap Hurdlers at the Festival that have had no runs in last 90 days are: 32 bets / 0 wins / 6 places

You want a horse that has proved their fitness…

Horses with just 1 run have the best overall stats:  103 bets / 11 wins / 27 places / +62 BFSP / AE 1.17

More than 1 run inside last 90 days: 161 bets / 9 wins / 34 places / AE 0.69

So, 1 run looks optimum but don’t be put off by more. Do be put off by those that haven’t run at all in the past 90 days. Although if you really fancy it, and the price is right, you should probably have something on :)C

 

Days Since Run

ALL 20 winners in the sample (those priced 20/1 and under) ran between 21 and 75 days ago. This makes perfect sense. You need a horse that has had enough of a break since their last run, but at the same time they still need to be sharp.

Handicap Hurdlers returning within 6-20 days are: 27 bets / 0 wins / 4 places

Those returning after 75 days off are: 40 bets/ 0 wins/ 7 places

This appears to be material and should be a standard rule in any other mini system I think.

 

Season Runs + Wins

Nothing much on seasonal runs, however, and unsurprisingly, those having their first run are 0/7.

Wins during the season… well the only thing of note is those that have won 3 or more times already. They are: 36 bets / 1 win / 7 places / AE 0.3.

This has clear logic. It would suggest that they do not have anything in hand of the ‘capper..which you would expect having won a few times already.

There is a pretty even spread of stats and figures for those that have 0, 1 or 2 wins this season.

 

Cheltenham Experience

There isn’t really anything of note here and maybe too much is made of this. Those that have never ran at the track are: 134 bets / 8 wins / 27 places / AE 0.65

1 or 2 runs at track: 118 bets / 6 wins / 26 places / AE 0.66

3 or 4 runs is more solid: 31 bets / 4 wins / 11 places/ AE 1.7

I wouldn’t be put off by no track experience although clearly if your selection has some, and they have proved they handle the track, that is one less thing you have to worry about. Horses that have never ran at the course can clearly win. Maybe experience of larger fields (something I can’t test) is more important.

 

Class Move

Very interesting…

Those horses running in the same class or dropping in class (1 class through to 4) are: 96 bets / 2 wins / 16 places / AE 0.27

This looks significant. Clearly you need a horse at the top of his game and a drop in class suggests they may be out of form or have been pitched into a level they found difficult. I wouldn’t want to be backing class droppers based on this evidence. I suppose people get sucked in by horses dropping in class, rather than those who are progressing and climbing the ladder. Those running in the same class, included in the stats above are:  44 bets/ 1 win / 10 places.

Those going up in class by 3 or more classes don’t perform much better either: 46 bets / 1 win / 7 places / 0.22

This is a bit easier to explain I think. Clearly most of these are not good enough. Dossing up in C4 races say and going up to graded hurdles is clearly rather challenging! It looks like those taking a big rise in class are to be avoided.

The figures for those running up 1 or 2 classes are: 77 bets / 9 wins / 18 places / +90 BFSP / AE 1.36

Interesting. Yes they are not massive samples but it appears, from the evidence so far, that it is wise to focus on those progressive horses (or well hidden from handicapper) that are going up in class, but not too much.

 

Distance move

Those dropping down any distance from their last run are as follows… 53 bets / 3 wins / 11 places / AE 0.7

Those going up in distance from their last start.. 181 bets / 14 wins / 42 places / AE 0.89

All in all not too much to read into those taking a move in distance I don’t think. Certainly I would slightly favour those running at the same distance or taking a step up in trip. Quite a few stepping up in distance clearly improve for doing so.

 

Finishing Position Last Time Out

Those that won last time out undoubtedly do best: 97 bets / 8 wins / 22 places / AE 0.82

The rest are: 195 bets / 12 winners / 44 places / AE 0.82

Nothing in it in terms of market expectations. Last time out winners are a good place to start but I wouldn’t let last time out finishing position put you off a bet. (unless strong trends for a certain race)

 

Track Last Time Out

Maybe not one to dwell on too much but there are some interesting observations…

Those tracks that have a negative record… Ascot 0/19, Sandown 0/16. The others that have yet to produce a winner haven’t had 10 runners as yet. (Cheltenham is 1/23)

Those tracks that have produced more than 1 winner are: Fairyhouse 2/7; Haydock 2/15; Kempton 2/22; Leopardstown 4/39; Newbury 2/47 (similar to Cheltenham given number of runners)

These figures are likely to do with the races contested and their strength, rather than the nature of the track. Who knows, but interesting nonetheless.

 

Trainers

I have removed the odds cap for this just so we can see their overall record.

Winners…

Those with more than 1 winner: Gordon Elliot 2/21 7 placed; Henderson 3/52 8 placed; Malcolm Jefferson 2/2; Willie Mullins 4/35 11 placed; Nicholls 3/38 12 placed; David Pipe 2/36 5 placed;

Single winners include: Tim Easterby 1/2; Hobbs 1/33; King 1/27; Lavelle 1/7; Martin 1/6 ; McCain 1/21; ALT Moore 1/1; T Mullins 1/; Jonjo O’Neill 1/26;

Those with notable poor records over the past 5 years include: Venetia Williams 0/22 2 placed; Evan Williams 0/10 0 places; NTD 0/14; Longsdon 0/13 0 places; J Ferguson 0/9 0 places;

***

So, that is it. I should probably try and pull some of it together into another profitable angle…  

I have shown you one micro system already. Caution should be advised with any approach that relies on small samples but it is all we have. (do I sound like a broken record)

The idea of what follows is to try and home in on the type of profile that has done well in recent years. Whether their success is repeated we shall see.

I hope you will be excited by what I am about to reveal…

I have built this mini system numerous different ways to ensure it is robust. Every rule has been looked at individually as a starting point. I am content that it has sound logic and there is no reason not to follow it this year…( at least as a ‘way in’)

So pulling together all of the above…

Rules:

(starting with no odds cap for now)

1. Cheltenham Handicap Hurdles
2. GB/Irish/French Bred (USA/German breds yet to have a winner in hurdles, that could change)
3. Days since run: between 21 – 75 days ago. Material…only 1/169 outside of this timeframe)

In a way that was the starting set of data. With no odds cap this left 413 bets / 24 winners…

4. Class Move UP 1 or 2 Classes ONLY
   • Those running in same class, or taking a drop: 136 bets / 2 winners / 19 places
   • Those up 3 or more: 70 bets / 1 win / 6 places
5. Distance Move: NOT a drop in trip at all from last run. Those taking a drop: 33 bets / 1 win / 7 places
   • This is not as robust as some of the above…I may remove this rule to see all selections but without doubt those running at same trip or going up in trip perform better.

Up to this point we get 93 bets / 11 winners / 19 places / +185 BFSP / AE 2.51

This is a solid starting point for any form of further analysis come race day but there will be quite a few selections per meeting/day/race

I know from earlier analysis that those that had already won 1 or 2 handicaps faired best from the larger data set – would that be the same here (yes!)

6. Horse WINS in a handicap: 1 or 2 ONLY
   • From this smaller data set above those with 0 wins in a handicap are: 35 bets /1 win / 4 places. Clearly it is hard to win a handicap having just won a novice race say.
   • Those with 3 or more handicap wins are: 15 bets / 0 wins. Now that is obviously not a large data set, but none have even placed. Clearly not the place for more exposed horses (in general, one will probably win this year at some point)

That leaves: 43 bets / 10 winners / 15 places / +150 ISP / +214 BFSP/ AE 5.00

However, no odds cap as yet and we know big priced horses don’t do that well…

7. Odds 33/1 or below.
   • This angle finds 2 winners from 8 at 33s, with 5 winning or placing!
   • By getting rid of the no hopers, those priced over 33/1 we are left with a decent micro-angle. No winners priced over 33/1 from the 17 to have tried, only 1 place

 

RESULT

26 bets / 10 winners / 14 wins or places / +167 ISP / +231 BFSP / AE 5.78

Incredibly, these stats do not include the two Willie Mullins winners from the micro angle earlier, nor the Henderson winner from that micro angle. Adding those two would make it 32 bets / 13 winners

That is 13 of the 25 winners since 2010 from just 32 bets.

Full results…

	Bets	Wins	Win%	P/L(SP)	Places	Place%	Races	Race%	P/L(BF)	P/L(Plc)	A/E
ALL	26	10	38.46	167.5	14	53.85	14	71.43	231.85	60.55	5.78
2014	10	4	40	56.5	6	60	4	100	97.69	29.02	5.41
2013	6	2	33.33	45	2	33.33	2	100	53.1	6.16	3.92
2012	4	2	50	34	4	100	4	50	41.66	22.2	11.11
2011	3	1	33.33	18	1	33.33	2	50	21.27	1.99	8.33
2010	3	1	33.33	14	1	33.33	2	50	18.12	1.18	5.56

 

By Trainer…

Trainer	Bets	Wins	Win%	P/L(SP)	Places	Place%	Races	Race%	P/L(BF)	P/L(Plc)	A/E
Easterby, T D	2	1	50	32	2	100	2	50	71.64	18.43	16.67
Greatrex, W J	1	0	0	-1	0	0	1	0	-1	-1	0
Henderson, N J	3	1	33.33	12	1	33.33	3	33.33	12.23	1.52	6.25
Hobbs, P J	3	1	33.33	2.5	1	33.33	2	50	3.21	-0.99	2.94
Jefferson, J M	1	1	100	20	1	100	1	100	25.65	4.66	20
King, A	3	1	33.33	31	1	33.33	3	33.33	35	5.03	7.14
McCain Jnr, D	2	1	50	15	1	50	2	50	17.01	2.87	11.11
Newland, Dr R D P	1	0	0	-1	1	100	1	0	-1	7.22	0
Nicholls, P F	3	2	66.67	26	3	100	3	66.67	30.72	12.25	11.11
ONeill, Jonjo	1	0	0	-1	0	0	1	0	-1	-1	0
Pipe, D E	4	2	50	34	2	50	4	50	41.39	5.17	4.17
Tizzard, C L	1	0	0	-1	0	0	1	0	-1	-1	0
Williams, Miss Venetia	1	0	0	-1	1	100	1	0	-1	7.41	0

 

I think this is exciting.

Yes we are playing with small data sets. I have looked to build this starting with every different ‘rule’ first. I have turned it inside out and upside down. I don’t believe this is back-fitted. There is logic for every single rule, and when combined. Feel free to disagree. 

Stats from 2003-2007 are 0/34. 08 and 09 had 1 winner each year and were profitable. The rise of certain trainers, change in methods and how trainers target horses are responsible for this I believe. Maybe it is just a statistical anomaly and we will have 3 years where this approach doesn’t pick a winner. However, it is not many bets per year and while it may be foolish to back the selections blind, they should form a decent starting point for analysis. Although, even if we did back them all blind this year it wouldnt cost much. 

As an Each Way angle it is exciting. Time will tell if it works this year. There is only one way to find out.

‘Selections’ will be posted on my daily Cheltenham Blog Posts.

****

To recap, the 2 micro systems…

1. Unexposed Hurdlers
   • Handicap Hurdles
   • H-Run Career: 1-3
   • H- Win Career: 1 or 2
   • Odds: 14/1
     2. Results: 14 bets / 4 winners / 9 places / +31 BFSP / AE 2.13…
   • No odds cap and focussing on Nicholls, Henderson and Mullins
     4. 13 bets / 5 winners / 9 w,place / +85 BFSP / AE 4.03

 

2. Handicap Hurdlers
   • 33/1 or shorter Industry Starting Price
   • GB/Irish/French Bred
   • Ran between 21-75 days ago
   • Handicap Wins : 1 or 2 only
   • Up in class by 1 or 2
   • NOT dropped in trip from last run
     5. 26 bets/ 10 winners / 14 wins or places / +167 ISP / +232 BFSP / AE 5.78

 

Right, that has taken 10 hours +…I had better move onto handicap chases. Once my brain has cleared I will go back through the handicap hurdles to see if there is anything else of interest/that i have missed. Handicap Chases will be posted next week at some point. There is some decent racing tomorrow and Saturday to look at in the meantime. And I have more flat profiles to find…that bank is building up nicely.

 

As ever without the excellent Horse Race Base (and the V4 system builder in particular) this analysis would not be possible. So, thanks to Chris B and the team. If you join Horse Race Base, or have recently on back of my video, you can say that i sent you. when you donate (josh wright, username: ynwajim ) I think for every person that does that I get 1 month free, and every little helps 🙂 Of course, you don’t have too! 

***

UPDATE 

As one helpful reader pointed out to me I didn’t mention weight! That goes to show you just how little attention I pay to such things in my punting…but given it is clear certain types of horses win time and time again at The Festival it was somewhat of an omission for a piece on handicaps….

So, using the top set of data with no odds cap (ie all handicap hurdles since 2010)…

Th only thing of note to me really is that there have been no winners carrying 10 stone or less: 20 bets / 0 wins / 0 places . The 20 bets by itself is maybe not so standout but the fact that not a single horse has placed is significant. I would suggest these horses are simply not good enough. 

When you then get into the higher weights you can dice it up how you like to fit your own ends really. 

There is no real reason why I have chosen these cut offs…

10-1 to 11-5 : 504 bets / 21 wins / AE 0.8 

11-6 + : 104 bets / 4 wins / AE 0.66 

Everyone always likes to focus on top weights…. 

11-10 : 8 bets / 2 wins / 2 wins,places

11-12 : 16 bets / 1 win / 4 wins, places 

I couldn’t safely say ignore those carrying 11-12 based on that data set. 

 

If we add the 20/1 cap which gets rid of most of the no-hopers…

10-1 to 11-5: 233 bets / 17 wins / 53 wins,places / AE 0.87 

11-6 + : 53 bets / 3 wins / 14 wins,places AE 0.63

 – this is  how you dice it up. If I say those carrying 11-7 + the stats are then: 39 bets / 2 wins / 11 places / AE 0.63

11-10 : 5 bets / 1 win 

11-11 : 3 bets / 0 win 

11-12 : 8 bets / 1 win / 2 wins,places 

If  I look back at my main micro- angle those carrying 11-10 + are 2 from 3. 

All in all, bar those at the bottom of the weights, I cant read too much into them. I suspect for certain races there will be clear weight trends, maybe influenced by the distance of the race for example, we shall see. 

You could say that 10-1 to 11-5 (or 10-1 to 11-6) is optimum based on market expectations but if the price is right I wont be put off by the weight, bar those caveats above (low weight/clear trends)