A National Nightmare?

Since the terrific but tragic 2012 Grand National I have heard plenty about what should be done and what should not be done to make the race safer, if it needs to be made safer at all. I have always thought that this or that might be the problem, and that this or that might be a solution, but I have not really had an  in depth understanding about exactly what happens, where it happens, why it happens and so on, only general theories. I would not say the Grand National is my favourite race, far from it in fact, but I appreciate both its positive and negative contribution, and its importance to the sport.

I have looked in detail at the 2012, 2011 and 2010 Grand Nationals to try and discover more. Time and resources limit me from looking further and I appreciate that the time period is small, and that the period I have looked at is not necessarily reflective of the overall picture. However, I am sure that useful and interesting conclusions have been reached which have enabled me to understand the race better. I hope they help you do the same.

And, before you begin, I must apologise for the length and detail, and any errors you may find. 

What Happened - 2012

In the 2012 Grand National 15 horses completed the race and, of the 25 that failed to complete, 10 fell, 7 unseated riders, 4 pulled up, 3 were brought down and 1 refused.

Start.

Viking Blond - Fell 1st. Clear view of the fence with no interference.

Junior - Fell 2nd. Clear view of the fence with no interference.

West End Rocker - Fell 2nd. Clear view of the fence with no interference.

State Of Play - Unseated rider 5th. Clear view of the fence with no interference.

Rare Bob - Brought down 5th. Interfered with when Noel Fehily fell into his path.

Chicago Grey - Brought down 5th. Interfered with when Rare Bob fell into his path.

Synchronised - Fell 6th (Becher's Brook). Relatively clear view of the fence with no interference.

Alfa Beat - Fell 7th. Relatively clear view of the fence with no interference.

Killyglen - Unseated rider 8th (Canal Turn). On the inside with a lot of horses around him but no interference.

Black Apalachi - Fell 8th (Canal Turn). Clear view of the fence with no interference.

Organised Confusion - Unseated rider 8th (Canal Turn). Direct result of interference.

Tatenen - Unseated rider 8th (Canal Turn). Direct result of interference.

Becauseicouldntsee - Unseated rider 8th (Canal Turn). Direct result of interference.

Treacle - Fell 9th. Clear view of the fence with no interference.

Arbor Supreme - Unseated rider 9th. Clear view of the fence with no interference.

Giles Cross - Pulled up before 10th.

Always Right - Unseated rider 15th (The Chair). Clear view of the fence with no interference.

Quiscover Fontaine - Fell 17th. Clear view of the fence with no interference.

Deep Purple - Pulled up before 19th. 

Vic Venturi - Refused 19th.

Mon Mome - Pulled up before 22nd (Becher's Brook).

Postmaster - Pulled up before 22nd (Becher's Brook).

On His Own - Fell 22nd (Becher's Brook). Interfered with by loose horse and unsighted.

According To Pete - Brought down 22nd (Becher's Brook). 

Weird Al - Fell 27th. Clear view of the fence with no interference.

Finish.

If the four horses that pulled up and the one refusal are eliminated then the twenty who failed to complete as a result of the fences remain. 

Total - All Fences

• 10 Fallers (10/20 = 50% of total)
• 7 Unseats (7/20 = 35% of total)
• 3 Brought Down (3/20 = 15% of total)
• 20 Total

To get an indication about where the problems are occurring the race can be split into three stages. The result is as follows:

Stage 1 - Fences 1 to 10

• 7 Fallers (7/10 = 70% of fallers)
• 6 Unseats (6/7 = 85.7% of unseats)
• 2 Brought Down (2/3 = 66.6% of those brought down)
• 15 Total (15/20 = 75% of total)

Stage 2 - Fences 11 to 20

• 1 Faller (1/10 = 10% of fallers)
• 1 Unseat (1/7 = 14.2% of unseats)
• 0 Brought Down
• 2 Total (2/20 = 10% of total)

Stage 3 - Fences 21 to 30

• 2 Fallers (2/10 = 20% of fallers)
• 0 Unseats
• 1 Brought Down (1/3 = 33.3% of those brought down)
• 3 Total (3/20 = 15% of total)

What Happened - 2011

In the 2011 Grand National 19 horses completed the race and, of the 21 that failed to complete, 11 fell, 3 unseated riders, 5 pulled up and 2 were brought down.

Start.

That's Rhythm - Fell 1st. Clear view of the fence with no interference.

Becauseicouldntsee - Fell 2nd. Clear view of the fence with no interference.

Vic Venturi - Brought Down 2nd.

Ornais - Fell 4th. Clear view of the fence with no interference.

Calgary Bay - Fell 4th. Clear view of the fence with no interference.

Dooney's Gate - Fell 6th (Becher's Brook). Clear view of the fence with no interference.

The Tother One - Fell 6th (Becher's Brook). Tight for room and interfered with.

Or Noir De Somoza - Fell 6th (Becher's Brook). Dooney's Gate falls in front of him.

West End Rocker - Brought Down 6th (Becher's Brook). Dooney's Gate falls in front of him.

Tidal Bay - Unseated Rider 10th. Relatively clear view of the fence with no interference.

Quolibet - Unseated Rider 11th. Clear view of the fence with no interference.

Grand Slam Hero - Fell 13th. Clear view of the fence with no interference.

Quinz - Pulled Up before 16th.

Can't Buy Time - Fell 18th. Clear view of the fence with no interference.

Majestic Concorde - Unseated Rider 24th (Canal Turn). Clear view of the fence with no interference.

What A Friend - Pulled Up before 27th.

Santa's Son - Pulled Up before 27th.

Killyglen - Fell 27th. Clear view of the fence with no interference.

Arbor Supreme - Fell 28th. Clear view of the fence with no interference.

Hello Bud - Pulled up before 29th.

Comply Or Die - Pulled Up before 29th.

Finish.

If the 5 horses who pulled up are eliminated then the 16 who failed to complete as a result of the fences remain. 

Total - All Fences

• 11 Fallers (11/16 = 68.8% of total)
• 3 Unseats (3/16 = 18.8% of total)
• 2 Brought Down (2/16 = 12.5% of total)
• 16 Total

If the race is split into three stages then the result is as follows:

Stage 1 - Fences 1 to 10

• 7 Fallers (7/11 = 63.4% of fallers)
• 1 Unseats (1/3 = 33.3% of unseats)
• 2 Brought Down (2/2 = 100% of those brought down)
• 10 Total (10/16 = 62.5% of total)

Stage 2 - Fences 11 to 20

• 2 Fallers (2/11 = 18.2% of fallers)
• 1 Unseat (1/3 = 33.3% of unseats)
• 0 Brought Down
• 3 Total (3/16 = 18.8% of total)

Stage 3 - Fences 21 to 30

• 2 Fallers (2/11 = 18.2% of fallers)
• 1 Unseat (1/3 = 33.3% of unseats)
• 0 Brought Down
• 3 Total (3/16 = 18.8% of total)

What Happened - 2010

In the 2010 Grand National 14 horses completed the race and, of the 26 that failed to complete, 10 fell, 8 unseated riders, 7 pulled up and 1 refused to race.

Start.

King Johns Castle - Refused To Race. 

Eric's Charm - Fell 1st. Clear view of the fence with no interference.

Pablo Du Charmil - Fell 2nd. Clear view of the fence with no interference.

My Will - Fell 4th. Clear view of the fence with no interference.

Made In Taipan - Fell 5th. Clear view of the fence with no interference.

Can't Buy Time - Unseated Rider 8th (Canal Turn). Clear view of the fence with no interference.

Irish Raptor - Fell 14th. Clear view of the fence with no interference.

Royal Rosa - Unseated Rider 14th. Clear view of the fence with no interference.

Arbor Supreme - Unseated Rider 15th (The Chair). Clear view of the fence with no interference.

Beat The Boys - Pulled Up before 19th.

Madison Du Berlais - Fell 19th. Clear view of the fence with no interference.

The Package - Unseated Rider 19th. Difficult to see but possibly slightly impeded though not significant.

Backstage - Unseated Rider 20th. Hampered after fence by falling loose horse. 

Vic Venturi - Fell 20th. Clear view of the fence with no interference.

Nozic - Unseated Rider 20th. Hampered by fall of Vic Venturi.

Flintoff - Pulled Up before 21st.

Maljimar - Fell 22nd (Becher's Brook). Clear view of the fence with no interference.

Ballyfitz - Fell 22nd (Becher's Brook). Clear view of the fence with no interference.

Ellerslie George - Unseated Rider 23rd. Clear view of the fence with no interference.

Dream Alliance - Pulled up before 24th (Canal Turn).

Mon Mome - Fell 26th. Clear view of the fence with no interference.

Palypso De Creek - Unseated Rider 27th. Clear view of the fence with no interference.

Ballyholland - Pulled Up before 28th.

Niche Market - Pulled Up before 29th.

Conna Castle - Pulled Up before 29th.

Ollie Magern - Pulled Up before 29th.

Finish.

If the 7 horses who pulled up and the 1 who refused to race are eliminated then 18 remain. 

Total - All Fences

• 10 Fallers (10/18 = 55.6% of total)
• 8 Unseats (8/18 = 44.4% of total)
• 0 Brought Down
• 18 Total

If the race is split into three stages then the result is as follows:

Stage 1 - Fences 1 to 10

• 4 Fallers (4/10 = 40% of fallers)
• 1 Unseat (1/8 = 12.5% of unseats)
• 0 Brought Down
• 5 Total (5/18 = 27.8% of total)

Stage 2 - Fences 11 to 20

• 3 Fallers (3/10 = 33.3% of fallers)
• 5 Unseat (5/8 = 62.5% of unseats)
• 0 Brought Down
• 8 Total (8/18 = 44.4% of total)

Stage 3 - Fences 21 to 30

• 3 Fallers (3/10 = 33.3% of fallers)
• 2 Unseat (2/8 = 25% of unseats)
• 0 Brought Down
• 5 Total (5/18 = 27.8% of total)

What Happened - 2012, 2011 and 2010 Combined

Total - All Fences

• 31 Fallers (31/54 = 57.4% of total)
• 18 Unseats (18/54 = 33.3% of total)
• 5 Brought Down (5/54 = 9.3% of total)
• 54 Total

Stage 1 - Fences 1 to 10

• 18 Fallers (18/31 = 58.1% of fallers)
• 8 Unseats (8/18 = 44.4% of unseats)
• 4 Brought Down (4/5 = 80% of those brought down)
• 30 Total (30/54 = 55.6% of total)

Stage 2 - Fences 11 to 20

• 6 Fallers (6/31 = 19.4% of fallers)
• 7 Unseats (7/18 = 38.9% of unseats)
• 0 Brought Down
• 13 Total (13/54 = 24.1% of total)

Stage 3 - Fence 21 to 30

• 7 Fallers (8/31 = 25.8% of fallers)
• 3 Unseats (2/18 = 11.1% of unseats)
• 1 Brought Down (1/5 = 20% of those brought down)
• 11 Total (11/54 = 20.4% of total)

Size Of The Field

If, on average over the past three renewals, 55.6% of incidents occur in the first third of the race then one plausible explanation would be that the size of the field is the problem. This might be because horses suffer interference or are unsighted at a fence when there are so many other horses around them. 

Using the descriptions of each incident outlined above:

2012

• 10 Fallers and 1 (On His Own) was a direct result of interference. 
• 7 Unseats and 3 (Becauseicouldntsee, Organised Confusion and Tatenen) were as a direct result of interference. 
• 3 Brought Down (Rare Bob, Chicago Grey and According To Pete)
• 7 of the 20 horses (35%) failed to complete as a result of direct interference with another horse.

2011

• 11 Fallers and 2 (The Tother One and Or Noir De Somoza) were a direct result of interference.
• 3 Unseats and 0 were a direct result of interference.
• 2 Brought Down (Vic Venturi and West End Rocker)
• 4 of the 16 horses (25%) failed to complete as a direct result of interference

2010

• 10 Fallers and 0 were a direct result of interference.
• 8 Unseats  and 2 (Backstage and Nozic) were a direct result of interference.
• 0 Brought Down
• 2 of the 18 horses (11.1%) failed to complete as a direct result of interference.

Combined

• 31 Fallers and 3  were a direct result of interference.  
• 18 Unseats and 5 were a direct result of interference.
• 5 Brought Down.
• 13 of the 54 horses (24.1%) failed to complete as a direct result of interference.

Over the last three renewals of the Grand National, nearly a quarter of those who fell, unseated or where brought down suffered interference which led directly to the incident. It is clear that interference is a problem but the argument for a reduced field size would be strengthened if cases were concentrated in the early part of the race when there are more horses around. If interference occurs regardless of the number of runners then there is little to support a reduced field size from the start and it must be other factors causing the problem.

Stage 1 - Fences 1 to 10

• 18 Fallers and 2 were a direct result of interference.
• 8 Unseats and 3 were a direct result of interference.
• 4 Brought Down
• 9 out of 30 (30%) incidents were a direct result of interference.

Stage 2 - Fences 11 to 20

• 6 Fallers and 0 were a direct result of interference.
• 7 Unseats and 2 were a direct result of interference.
• 0 Brought Down
• 2 out of 13 (15.4%) incidents were a direct result of interference.

Stage 3 - Fence 21 to 30

• 7 Fallers and 1 was a direct result of interference.
• 3 Unseats and 0 were a direct result of interference.
• 1 Brought Down
• 2 out of 11 (18.2%) incidents were a direct result of interference.

The above shows that interference is responsible for a higher percentage of incidents in the first third of the race when compared with the final two thirds. This would suggest that the number of runners is an important factor. 

If a field of 40 is too big, then what size would be more appropriate?

Over the past three years 40 runners have lined up on each occasion. Stage 1 has, on average, accounted for 10 horses (15 in 2012, 10 in 2011 and 5 in 2010). Therefore, by the start of stage 2 there were, on average, 30 horses still running. Stage 2 accounted for 4  horses (to the nearest horse), leaving 26 still running. The interference data shows that the number of cases of interference does not fall from stage 2 to 3. If, on average, there are 30 horses running at this stage, then 30 might be a more appropriate number with which to start the race. The above would suggest that cases of interference may be almost halved if the field was reduced to 30.

However, it is not that simple because interference can be caused by horses that are no longer in the race. There were 13 cases of interference and 3 (23.1%) of these were a result of loose horses. These three incidents occurred at Fence 20 (Backstage unseated after hampering by a falling loose horse) and Fence 22 (On His Own fell after interference from a loose horse and brought down According To Pete). 

There were 4 cases of interference in Stages 2 and 3 of the race, and 3 of these are the result of loose horses. Once the initial field is reduced to 30 runner (on average) by fence 10, the cases of interference from horses still in the race falls even more dramatically than the bare result would suggest.

Speed Of The Race

The speed of horses is often quoted as a reason for fallers, and a reason why falls can be fatal. It is difficult to be completely accurate with crude hand timing but it can at least provide an indication as to whether there is a significant pace difference between the first and second circuit. Times are for the leading horse, except where indicated, from take off to take off. MR = Melling Road, F1 = Fence 1 and so on.

Note 1: Where fences have been bypassed, averages are calculated using the figures available. 

Note 2: F1 to F2 on the 2nd circuit in 2012 I have used the time of the 3rd horse. Richard Johnson and Planet of Sound steadied the pace and it is noticeable that the field close up on him indicating that the leader's speed does not reflect that of the race. Shakalakaboomboom was hampered on landing in 2nd. The time for Planet Of Sound is 12.7 seconds.

Note 3: I have ignored Conna Castle in the 2010 renewal who set an erratic pace which was not reflective of the race.

The sample of three races is not ideal but the above table still provides a hugely interesting insight. The noticeable trend is that the pace is fast over the first three fences, particularly into fences 1 and 2, and then settles down. I have often heard the term "the race to Becher's", and, whilst I think the race to Becher's might be overstating the case, there is a pace problem over the first couple of fences. 

We now know that the early pace is fast before steadying towards the 3rd fence. The next question is why? Despite the build-up, I find it hard to believe that experienced jockeys will ride unnecessarily fast just because it is the Grand National. If they are going too quickly there must be a reason. Many riders say "I want to sit handy early", "get a position" and very few seem to want to "drop in" in the Grand National. The need for 'a position' is crucial and this could be the cause of the fast early pace. Why is 'a position' so important? When there are so many horses the best place to be is out the front with a clear view of the fence and some space, out of the way of possible interference. This could generate a vicious circle where jockeys need to travel too fast and risk a fall to get a good position in the race, so that they avoid the problems of being in behind. A catch 22 if you like. 

It is often stated that handy horses run well in the National and that it is difficult to make up ground from the rear. This fact (or myth) could be the problem. So is it fact or myth? In the last 5 Grand National the first 4 home have recorded the following in running comments:

2012

1. Mid division
2. Held up in mid division
3. Tracked leaders
4. Mid division

2011

1. With leaders
2. Tracked leaders
3. Mid division
4. Chased leaders

2010

1. Mid division
2. With leaders
3. Prominent
4. Tracked leaders

2009

1. Towards rear
2. Mid division
3. Mid division
4. Prominent

2008

1. Tracked leaders
2. Held up in mid division
3. Chased leaders
4. Held up in mid division

The above gives the theory that you need to be handy, or at least in mid division, some sort of credibility. By way of comparison, during those years a total of 81 horses were described as being in worse than mid division using Racing Post comments in running. There were 16 in 2012, 18 in 2011, 17 in 2010 (not including King Johns Castle who refused to race), 17 in 2009 and 13 in 2008. That is 40.5% of the 200 horses that took part and yet only one (Mon Mome in 2010) was able to make the first 4. It seems that being handy pays.

In the past three renewals, the pace has been relatively fast approaching the first and second fences. Does this fast early pace increase the chance of fallers?

Fence 1

• 2012 - 1 Faller
• 2011 - 1 Faller
• 2010 - 1 Faller

Fence 2

• 2012 - 2 Fallers
• 2011 - 1 Faller, 1 Brought Down
• 2010 - 1 Faller

The first two fences have claimed at least one casualty in each of the past three years. It is hard to say whether this is a direct result of the fast pace so when the pace begins to steady on the approach to the 3rd and 4th are there fewer fallers?

Fence 3

• 2012 - No Fallers
• 2011 - No Fallers
• 2010 - No Fallers

Fence 4

• 2012 - No Fallers
• 2011 - 2 Fallers
• 2010 - 1 Faller

So, since 2010, fences 1 and 2, when the pace is relatively fast, have claimed 8 horses. In the same period fences 3 and 4, when the pace begins to steady, have claimed 3 horses. Most interesting is the fact that fence 4 was the subject changes implemented after the Review because of its difficulty. The sample size is small, but the difference is noticeable. It seems the speed at the first couple of fences which steadies approaching the 3rd and 4th does have an impact.

If they are going too fast and the reason for this is the need for a position and the increased pace does have an impact, then what can be done? Again, if the field size was reduced then that might give horses more space, reduce the risk of interference and make it easier to move through the field, reducing the need to be out the front. It might not. 

Fences

ST1 = Stage 1 and so on, F = Fell, UR = Unseated Rider, BD = Brought Down, % of Totals.

The above table is a fence by fence account of what happened where, as already outlined previously. Having investigated the impact of field size and the pace of the race, the next factor to consider is whether any particular fences have had a significant impact over the last three years. 

7 fences (3, 13, 16, 21, 25, 29 and 30) have had no incidents in the past three years.

10 fences (7, 10, 11, 12, 17, 18, 23, 24, 26 and 28) have had one incident.

4 fences (9, 14, 15 and 19) have had two incidents.

4 fences (1, 4, 20 and 27) have had three incidents.

2 fences (5 and 22) have had four incidents.

2 fences (2 and 6) have had five incidents.

1 fence (8) has had six incidents. 

There are nine fences that have averaged one incident per year or more over the last three renewals. In numerical order: 1, 2, 4, 5, 6, 8, 20, 22 and 27.

Before looking at why these fences are more influential than others there are a few points that need to be made. The first is that Fence 20's appearance on that list could be (though may not be) an anomaly. All three incidents occurred in 2010 when a loose horse fell hampering and unseating one, and one horse fell, hampering a third horse who also unseated his rider. Only one of the three actually fell at the fence, the two others were as a result of interference. I am, therefore, going to eliminate it from the list of influential fences. The second fence that might appear like an anomaly is fence 27. However, it has claimed one horse for the past three years, each with no other reason, and so it remains.

Fence 1

• 3 Fallers (1 in 2012, 1 in 2011, 1 in 2010) 
• Interference responsible for 0.

Fence 2

• 4 fallers (2 in 2012, 1 in 2011, 1 in 2010)
• 1 Brought Down (2011)
• Interference responsible for 1.

Fence 4

• 3 Fallers (0 in 2012, 2 in 2011, 1 in 2010)
• Interference responsible for 0.

Fence 5

• 1 Faller (2011)
• 1 Unseated Riders (2012)
• 2 Brought Down (2012)
• Interference responsible for 2.

Fence 6

• 4 Fallers (1 in 2012, 3 in 2011, 0 in 2010)
• Interference responsible for 3.

Fence 8

• 1 Faller (2012)
• 5 Unseated Riders (4 in 2012, 1 in 2010)
• Interference responsible for 3.

Fence 22

• 3 Fallers (1 in 2012, 2 in 2010)
• 1 Brought Down (2012)
• Interference responsible for 2.

Fence 27

• 2 Fallers (1 in 2012, 1 in 2011)
• 1 Unseated Rider (1 in 2010)
• Interference responsible for 0.

The problem fences can now be split into two groups depending on whether it is the fence itself which is the issue, or whether interference is the problem. 

Fences with 3 or more incidents not explained by interference - 1, 2, 4, 8, 27.

Fences with less than 3 incidents not explained by interference - 5, 6, 22.

Using the speed analysis above, it would appear that fences 1 and 2 are problematic because of speed. The times from the past three renewals suggest that horse travel relatively fast into the first two fences before slowing. This could explain why fences 1 and 2 feature on this list and fence 3 does not. Fence 3 is the first open ditch so does this have anything to do with it? Perhaps the pace is quick over the first couple of fences because they are plain, 'easier' fences, but, the pace then steadies in preparation for the open ditch at fence 3. It is impossible to say with any certainty whether this is the case but it is a theory which the evidence could support.

Fence 4 was the first full height plain fence on the course and was found to be one of the most influential fences for fallers in the Grand National Review with 12.6% of all falls occurring here since 1990. Furthermore, of the 32 incidents at the fence, 4 were fatal, a much higher proportion than at any other fence. The Review concluded that the fence should be lowered by 2 inches. It is at this point that the omission of fence 20 from the list of problem fences may seem erroneous because, of course, fence 20 is fence 4 on the second circuit. Perhaps the fence has been a problem in the past but, whilst it is far too early to say whether the changes have been effective, there were no incidents at the fence in 2012 (after the changes) and in the previous two years there were 4 fallers and 2 unseated riders. Many of the changes as a result of the 2011 Review have been given short shrift but in this instance perhaps more credit is deserved. Time will tell.

As I am sure you are all aware, fences 6 and 22 are Becher's Brook on the 1st and 2nd circuit respectively. This would suggest that the major problem with Becher's Brook is not the fence itself but the frequency with which interference occurs. There were 3 incidents in 3 years at Becher's Brook (1 at fence 6 and 2 at fence 22) which could not be explained by interference. This means that of the 8 incidents at Becher's Brook, 5 were the result of interference rather than the fence itself. This is not what I expected, and probably not what you expected either. However, the unexpected is exactly what this is about. I want to know what actually happened rather than what people believe happened. Why have there been 5 cases of interference in the last 3 years? After this year's race, ex-jockey Mick Fitzgerald said:

"I spoke to Robbie Power (rider of Killyglen) after the race and he said normally when you got to Becher's there was a lot of room because the way the drop was before nobody, except the really brave men and the guys who were on horses they knew would he able to cope with the drop, went down the inside. It meant the whole field spread out when they got to Becher's whereas now the drop's been levelled off nobody moved off that inside. Because of that you had a bit of a pile-up situation and a knock-on effect almost. That's the danger. Suddenly no-one wants to go to the outside of the fence. They all piled up on the inner, hence the reason you get a faller and one being brought down."

The data from the last three renewals suggests that interference rather than the fence itself is the problem so could it be that Mick's argument is correct. Is Becher's now too easy? I somehow doubt that making Becher's harder, and, therefore, ensuring the horses are spread across the track is a suggestion that will find favour with those opposed to the race. However, I am not trying to appease, I am trying to find out what needs to be changed and the evidence suggests that interference is the major problem at Becher's and not the fence itself. 

Before getting too carried away it must be remembered that the fall of On His Own, who brought down According To Pete, was a result of interference from a loose horse, a horse without a rider choosing which path to take. In this instance, the 'riders choosing to come down the inner' argument does not hold sway. If that case (On His Own. Both On His Own and According To Pete's jockeys did choose where to be) is eliminated then that leaves 7 incidents at Becher's Brook in the past 3 years. Of which, 4 were the result of interference and 3 which were not. It is clear that, even if allowances are made, interference is the bigger problem over the past three renewals.

If interference is a bigger problem than the fence, then what is the solution? The obvious one is to have fewer runners. A smaller field means more space and less interference. Is this the right solutions? If it is, then more cases of interference would be expected at the first Becher's than the second. Over the past three renewals, there have been 3 cases of interference at the first Becher's and 2 at the second. The average field size over the past three years at the first Becher's is 35 (to the nearest horse) and at the second is 23 (to the nearest horse). On average, the field size is 2/3rds as large on the 2nd occasion, and there is only 2/3rds of the interference. This would suggest that fewer horses means less interference and support a smaller field. 

An alternative is the argument suggested by Mick Fitzgerald (and others) outlined above, the drop element of the fence could be reinstated to encourage (or force) jockeys to take a wider course, spreading the field across the track. However, given the evidence that appears to support the case for a smaller field it will be difficult to argue that the fence should be made more difficult in the face of so much criticism to the contrary.

Fence 8 is the first Canal Turn and is the most difficult fence to categorise in that of the 6 incidents that happened there in the past 3 years, 3 were as a result of interference and 3 were not. The incidents at fence 8 predominantly occurred in the 2012 renewal which could skew the results. Of the 6 incidents, 5 were in this year's renewal. Also of note is the 5 of the 6 incidents were unseated riders with only 1 fall. The 2nd Canal Turn (fence 24) has only been responsible for 1 unseated rider in the past three renewals. I think it is fair to say that it is congestion which is the problem at the first Canal Turn. As the runners swing in to angle the fence, some interference is inevitable. This is unlikely to ever be eliminated but it could be argued that it would be reduced with a smaller field.

Fence 27 is the 4th last fence and one horse has come to grief here in each of the past three seasons. Their comments in running are:

• 2012 Weird Al - Behind when hampered 2nd Becher's, fell 4 out.
• 2011 Killyglen - Disputing 3rd and staying on when fell next.
• 2010 Palypso De Creek - Behind well fell 4 out. He actually unseated.

So with the exception of Killyglen, the other two horses were in rear at the time. It could be tiredness that is responsible, but no other fences at this late stage have had such an impact. However, fence 27 is the last open ditch and, this, together with the tiredness factor, could be the explanation. The fence is 5ft high so perhaps it might be lowered should the high incidents trend continue. 

Experience

Another theory put forward is that horses lack experience of the Grand National fences and that this leads to problems. In the past three renewals the first 4 horses home were:

2012

1. Neptune Collonges - No experience of the fences.
2. Sunnyhillboy - No experience of the fences.
3. Seabass - No experience of the fences.
4. Cappa Bleu - No experience of the fences

2011

1. Ballabriggs - No experience of the fences.
2. Oscar Time - No experience of the fences.
3. Don't Push It - Previous experience.
4. State Of Play - Previous experience

2010

1. Don't Push It - No experience of the fences.
2. Black Apalachi - Previous experience.
3. State Of Play - Previous experience.
4. Big Fella Thanks - Previous experience.

This shows that a lack of experience over the unique Aintree fences is no barrier to success. This is all well and good, but, unfortunately, the fallers and the unseats are those under consideration. You can win without experience of the fences but would a 'trial race' over the fences eliminate some of those who are unsuitable.

F = Fall, UR = Unseated Rider, F/UR = Fall or Unseated Rider Combined, BD = Brought Down, PU = Pulled Up, R = Refused, RR = Refused to Race.

None = No previous experience of the Grand National fences

% None = Percentage of those with no experience to fall etc.

Comp = Previously completed a race over the Grand National fences.

DNF = Previously experienced the Grand National fences but not completed a race.

LT10 = Less than 10 chase starts before running in the Grand National. These horses are included into the other categories as appropriate.

All = All starters.

What can be deduced from this? The list of placed horses in the past 3 renewals indicates that a lack of experience of the fences is no barrier to success. However, those lacking experience do fall or unseat more often than those that have previously negotiated the track. However, those without any experience of the track were roughly on a par with the average over the period. It was the inexperienced horses and those that had already failed to complete the track that fared worst of all.

This would suggest that a trial race might have some benefits. There is nothing to stop a horse running well on its first attempt at the fences but nearly 70% of horses that have tried and failed to complete the course either fell or unseated in the past three Grand Nationals. Therefore, a trial race could be used to eliminate those that are not suitable, rather than finding those that are. A general lack of experience was also a negative with 60% of horses with less than 10 chase starts either falling or unseating in the past three renewals. 

The completion percentages tell a similar story. The horses that had no experience of the fences fared slightly worse than average, the horses that had already completed the course fared best, inexperienced horses fared the same as those experienced but without experience of the fences, but, again, it was the horses that had tried and failed in the past that had the lowest completion rates (just 15.4%). 

Loose Horses

Using the figures above, on average over the past three renewals 45% (40.8% fell or unseated + 4.2% brought down) of the field can be expected to part company with their rider at some point in the race.  That means we can expect 18 loose horses per race.

The problem of loose horses is a major concern. 5 of the last 10 fatalities suffered in the Grand National were by horses running loose, as has been shown already, they can impede other runners.

In the past three renewals there have been 3 incidents that could be attributed to interference by a loose horse. 

How to catch loose horses is a difficult question. Often the loose horses follow their herd instinct and run with the field. It is likely that these will always be impossible to 'extract' from the field. These are the horses which cause interference with the rest of the field but I cannot see a way to prevent this.

However, it is noticeable that loose horses running with the field jump much better than those running alone in rear. Without any evidence or figures, I would suggest that the horses injured when running loose are those in rear who end up taking on the fences in a half hearted manner. This is certainly the fate that befell Synchronised. I have heard the idea of US style outriders put forward as a solution to the loose horse problem. It could be that they might be able to have a positive impact in catching loose horses in rear. 

Fatalities

As you can probably tell by the fact that it has taken me this long to address the most pressing question, I am uneasy with the complete focus on just those who died. To my mind there are plenty of others who were in similar situations but live to fight another day. 

By way of explanation, 3 horses were brought down in 2012, but only According To Pete was fatally injured and a total of 20 horses lost their jockeys (10 fallers, 7 unseated riders and 3 brought down) and yet it was only Synchronised that suffered a fatal injury when running loose. The reduction of fatalities is the ultimate aim but focussing on what led to their demise to the exclusion of everything else is short sighted in my opinion. It is more important to focus on the broader picture and reduce incidences which could have led to fatalities, and, obviously, eliminate to the best of our ability, the factors which have. 

Another example is how both fatalities in the 2012 Grand National fell at Becher's Brook. It might, therefore, seem that getting rid of Becher's Brook is the way to go. It might be, but these two horses cannot be used as statistics to support such an argument. As already stated, According To Pete was brought down, like 2 other horses, and yet was the only one to lose his life. It could so easily have been one of the others and the fact that it happened at Becher's is a coincidence. Similarly, Synchronised fell at Becher's but suffered his injury at Fence 11 when running loose. This could have happened to any of the 17 horses that lost their jockeys at other fences (9 other fallers On His Own also fell at Becher's). Whilst their importance is paramount, and their elimination the ultimate objective (though unachievable), they must not cloud our judgement.

Conclusion

In the immediate aftermath of the 2012 Grand National, Gavin Grant, Chief Executive of the RSPCA, said:

"As far as the Grand National is concerned there are lots of factors, Firstly, the scale of the field. Forty horses is a heck of a lot. Secondly, there are unique jumps there that horses aren't experienced in going over and I think we need to look at those jumps again. Becher's Brook has claimed another casualty [According To Pete] and perhaps it's time for that to go. We need to look at the landing areas. Some improvements have been made there, but when you've got a drop on the other side of the fence a horse isn't expecting that. And the going. The ground conditions are very important. Aintree has made a lot of progress making sure the going is softer because when it's hard the horses run faster. There is lots of work to be done to take the risks to horses out of this."

He outlines a number of concerns: the number of horses, the 'unique' fences, Becher's Brook, the going and the speed of the race. It could be argued that the number of horses does contribute to the number of fallers early in the race, the fast pace approaching the first and second fences and the incidences of interference. It can be shown that horses that have previously shown a liking for the unique fences do fare better than either those who have no experience or those with bad experience. However, a lack of experience is no barrier to success in the Grand National. Over the past three seasons, the major problem with Becher's Brook has been the number of cases of interference. These have outweighed the number of incidents not caused by interference. Therefore, it could be argued that the problem is not the fence itself but perhaps the size of the field, leading to crowding and other problems. I do not have the access or time to analyse any further back than the past three years but, if going was the major concern, then you would expect incidents to be spread around the track assuming that the going is similar all over. This has not been the case indicating that it is other factors that are causing the problems.

It is essential that no changes are implemented for the sake of changes. There must be a reason, and a valid explanation for what a particular change is designed to improve, and how. Just looking at the past three years is insufficient to come to such conclusions, but nevertheless, general themes can be discovered. Any changes that are made will most likely be irreversible so it is crucially important that a comprehensive review and understanding of the situation is first carried out.