Data won’t tell you everything about why horses fall, writes Tony Keenan. There are too many intangible factors at play, variables that can’t be number-crunched. Race flow plays a big part; horses can be drawn into mistakes by how the race unfolds, be it the pace it is being run at or simply by being distracted by another runner jumping alongside it.

Physical issues can affect jumping; an underlying injury can be found out in the heat of battle while the effort of in-race exertion can cause a fresh problem. That’s not to mention mental issues: some horses seem unable to concentrate on jumping consistently or lack the self-preservation instinct to get from one side of a fence to the other.

But numbers can still tell us plenty, not least because of a large sample size of chases and fallers/unseats each season. Unless otherwise stated, I have looked at all Irish chases since the start of the 2003/04 season until the end of the 2014/15 campaign (in hurdle races, jumping just isn’t as important with fewer and smaller obstacles).

In that period, there were 4,932 chases with 57,626 runners; 6,107 horses fell or unseated, a faller rate of 10.6% with an average of 1.24 falls/unseats per race. When I refer to ‘faller rate’ I mean the combined number of falls and unseats.

 

Irish and UK Chase Faller Rate by Season

Season	UK	Ireland
2014/15	7.7%	9.9%
2013/14	7.1%	9.3%
2012/13	7.1%	8.7%
2011/12	8.3%	9.6%
2010/11	8.1%	10.5%
2009/10	9.3%	11.3%
2008/09	8.7%	11.9%
2007/08	9.1%	11.1%
2006/07	8.8%	10.3%
2005/06	9.6%	10.5%
2004/05	9.4%	11.7%
2003/04	10.4%	11.8%

 

The first thing that stands out is general downward trend of faller rate in the UK and Ireland over the period covered; there are some blips along the way as the table above shows but the broad picture is clear. On the whole, fences in the UK appear to be easier which isn’t the greatest surprise; the animal rights lobby, regardless of what you may think of them, are certainly stronger there than in Ireland.

The Grand National fences are the most high-profile example of this but another interesting case, albeit with a small sample size, was made by Matt Tombs in his recent book ‘How to Bet and Win at the Festival’. Tombs points out that while there were 22 fallers at the 2014 Cheltenham Festival there were just eight in the most recent iteration, a marked decline. Irish faller rates are dropping too, though they remain higher than the UK which may help Irish horses on their raids as they are more tried and tested jumpers.

 

Irish Chase Faller Rate by Track since 2003/4

Track	Faller Rate
Sligo	6.0%
Naas	6.5%
Roscommon	6.6%
Tramore	7.5%
Killarney	7.7%
Ballinrobe	8.0%
Wexford	8.2%
Kilbeggan	9.1%
Navan	10.2%
Galway	10.3%
Fairyhouse	10.9%
Punchestown	11.0%
Tipperary	11.0%
Limerick	11.2%
Leopardstown	11.3%
Gowran Park	11.4%
Downpatrick	12.0%
Clonmel	12.0%
Thurles	12.0%
Cork	12.6%
Listowel	12.7%
Down Royal	15.2%

 

Faller rates at the various Irish chase tracks produced surprising results, not least the bizarre mix of tracks that comprised the bottom five. Down Royal is a big galloping track with one of the widest circumferences in the country and a few tricky downhill fences that are met at speed, Listowel is flat and tight but they tend to stick close to the rail over fences and racing room is at a premium, Cork is flat and galloping, both Thurles and Clonmel have downhill fences but beyond that have few similarities.

Perhaps the most encouraging finding is where the main winter jumping tracks fall in the table. The big four of Leopardstown (eighth), Punchestown (eleventh), Fairyhouse (twelfth) and Navan (fourteenth) bunch around the middle which is a good sign as they host the majority of our graded races. These tracks should be fair, a test without being an ordeal, and the numbers suggest this is the case.

Down Royal is the outlier here. Not only is their chase track top in terms of faller rate, it is also 2.5% higher than any other track in the country and there is no bigger discrepancy between one track and the next anywhere else in the survey as there is between the highest and second highest. The fences at the Ulster course have long seemed ultra-stiff and it’s probably not the place to run an iffy jumper or even to start one off over fences.

 

Irish Chase Faller Rate by Trainer since 2009/10

Trainer	Faller Rate
R. Tyner	5.5%
C. Roche	6.9%
C. Swan	7.9%
H. De Bromhead	7.9%
N. Meade	8.0%
F. Flood	8.2%
D. Hughes	8.4%
E. Bolger	8.7%
G. Elliott	8.9%
E. O’Grady	9.0%
P. Nolan	9.4%
A. Moore	9.4%
M. Morris	9.5%
M. Hourigan	9.8%
J. Mangan	9.9%
J. Harrington	10.4%
O. McKiernan	10.5%
T. Martin	10.6%
P. Rothwell	10.9%
J. Lambe	11.0%
W. Mullins	11.4%
C. McBratney	11.6%
C. Byrnes	12.0%
L. Burke	12.6%
J. Hanlon	12.7%
J. Walsh	15.8%
J. Ryan	16.7%
C. Murphy	18.1%

 

There is a host of contributing factors to why a trainer may have a high or low faller rate. The type of horses they typically handle plays a huge part; if they tend to get national hunt types, their faller rates should be lower as such sorts are more physically able to jumps fences while forcing flat types to do the same is a somewhat Sisyphean task. Good schooling facilities have to help too as would access to good jockeys, both for homework and on the track.

This table takes into account the top twenty-five trainers in terms of chase runners since 2009/10. Colm Murphy comes out worst on these numbers and he’s been cursed by some of the worst jumpers around in recent times. Zaarito (five falls/unseats), Big Zeb (four) and lately Empire Of Dirt (four) have all tried his patience though whether this is randomness or something to do with the trainer is impossible to say. The numbers suggest that Willie Mullins horses aren’t the best jumpers and the visuals back this up; a few of his stable stars have had their issues over fences but it hasn’t necessarily stopped them winning.

There isn’t a huge correlation between those with low faller rates and high return in terms of winning chases. I looked at trainers’ success rate in various types of races last month and the bottom five here – Tyner, Roche, Swan, De Bromhead and Meade – have a mixed record. De Bromhead (18.0% over fences since 2010) does very well, Tyner (12.0%) and Meade (11.7%) do ok while Roche (9.3%) is below average.

 

Irish Chase Faller Rate by Jockey since 2009/10

Jockey	Faller Rate
M. Walsh	6.4%
B. Geraghty	6.8%
N. Madden	7.1%
T. Doyle	7.2%
P. Carberry	7.3%
A. Lynch	7.3%
P. Townend	7.5%
B. O’Connell	7.8%
A. McNamara	7.8%
A. Crowe	8.1%
R. Power	8.2%
P. Enright	8.3%
R. Walsh	8.4%
D. Condon	8.7%
D. Russell	8.8%
J. Cullen	9.1%
B. Cooper	9.6%
D. Casey	9.6%
M. Ferris	10.0%
A. Heskin	10.5%
S. Flanagan	10.8%
M. Darcy	11.2%
P. Mangan	11.6%

 

The sample size for fallers with jockeys is bigger than for trainers and their faller rates have nothing like as wide a spread. Like the trainers’ table above it takes into account the top twenty-five jockeys in terms of chase rides since 2009/10. The two JP McManus-retained riders come out very well, particularly the much-improved Mark Walsh who tops the table.

I certainly won’t get into jockey bashing here when you consider that one fifth of the riders listed above aren’t riding any longer; David Casey, John Cullen, Davy Condon, Andrew McNamara and Tom Doyle have all retired recently. One thing that emerges is the overall level of competency across the board; even the worst faller rate is only 1.3% higher than the national average since 2003/04.

- Tony Keenan