Phil Smith's Handicapping Of The National

Cesarewitch runners alloted 9-5+

2014 6/33
2013 4/33
2012 2/32
2011 2/33
2010 2/32

got bored then

More compelling stats that the correlation between weight and distance isn't what people perceive it to be. 11% representation has produced a 25% win ratio.
 
I'm wondering if in actual fact the reverse of what he believes is true. Perhaps top weights in the big 2 mile handicap chases and hurdles are more affected, and weight impacts speed more than it does stamina.

Does anyone have anything to hand that may prove or disprove this?
 
To frame it correctly you need to generate a null, convert the winning performances into Z scores, and then calculate some coeffcients of determination in order to establish cause and affect. My experience of doing this on horseracing data very often (though not exclusively) is that you end up with something where the correlation isn't as strong as it first appears and can usually be accounted for within the laws of normal distribution.

The anomoly of the cross country race is one I wa spondering years ago, and never really answered. Does a horse use up more energy having to sustain a gallop, or does it use up more energy if it's required to continually slow and then reaccelerate again? I always felt the answer lay in there. You should be able to work out the average speed per furlong between a cross country race and a conventional race, but it's only going to be loosely indicative at best. If Eider horses are running faster than the Cross Country horses though, you should have a pretty strong clue
 
I'm wondering if in actual fact the reverse of what he believes is true. Perhaps top weights in the big 2 mile handicap chases and hurdles are more affected, and weight impacts speed more than it does stamina.

Does anyone have anything to hand that may prove or disprove this?

What is speed? it's a much abused word

You can have cruising speed, or you can have acceleration. Ultimately speed if expressed as cruising speed, (which 2 mile championship races are normally run at) is really an expression of stamina
 
some extracts from the weight chapter from the excellent book Against The Odds by David Lee Priest where he analyses the results of thousands of races to come to his conclusions.

"An eminent backer has demonstrated that weight only significantly encumbers horses when they are travelling at maximum speed; horses slow considerably to undertake turns and thus weight is a greater issue when horses are racing in a straight line" (There's your X-C explained right there.)

talking about using raceform ratings and whether or not to adjust them for weight carried - "I found that adjusting the ratings for weight carried in the conventional fashion weakens the relationship between ratings and performance"....." I found that the weight correction which yielded the best results was 0.25lb for every pound of additional weight carried"

From handicap races he produces the following table based on how far from top weight the beast is, showing number of runners, strike rate and level stakes profit.


[TABLE="width: 448"]
<tbody>[TR]
[TD="width: 64"][/TD]
[TD="class: xl63, width: 192, colspan: 3"]Flat[/TD]
[TD="class: xl63, width: 192, colspan: 3"]Jumps[/TD]
[/TR]
[TR]
[TD]Rating[/TD]
[TD]No[/TD]
[TD]SR%[/TD]
[TD]Profit[/TD]
[TD]No[/TD]
[TD]SR%[/TD]
[TD]Profit[/TD]
[/TR]
[TR]
[TD]Top Wgt[/TD]
[TD="align: right"]9619[/TD]
[TD="align: right"]11.8[/TD]
[TD="align: right"]-0.23[/TD]
[TD="align: right"]5182[/TD]
[TD="align: right"]15.9[/TD]
[TD="align: right"]-0.19[/TD]
[/TR]
[TR]
[TD]1-3lbs[/TD]
[TD="align: right"]8656[/TD]
[TD="align: right"]10.1[/TD]
[TD="align: right"]-0.26[/TD]
[TD="align: right"]2841[/TD]
[TD="align: right"]13.2[/TD]
[TD="align: right"]-0.2[/TD]
[/TR]
[TR]
[TD]4-6lbs[/TD]
[TD="align: right"]10625[/TD]
[TD="align: right"]9.4[/TD]
[TD="align: right"]-0.25[/TD]
[TD="align: right"]3367[/TD]
[TD="align: right"]11.3[/TD]
[TD="align: right"]-0.29[/TD]
[/TR]
[TR]
[TD]7-9lbs[/TD]
[TD="align: right"]11813[/TD]
[TD="align: right"]7.7[/TD]
[TD="align: right"]-0.34[/TD]
[TD="align: right"]3862[/TD]
[TD="align: right"]11.4[/TD]
[TD="align: right"]-0.28[/TD]
[/TR]
[TR]
[TD]10-12lbs[/TD]
[TD="align: right"]12452[/TD]
[TD="align: right"]7.5[/TD]
[TD="align: right"]-0.35[/TD]
[TD="align: right"]4328[/TD]
[TD="align: right"]10.6[/TD]
[TD="align: right"]-0.29[/TD]
[/TR]
[TR]
[TD]13-15lbs[/TD]
[TD="align: right"]12189[/TD]
[TD="align: right"]7.6[/TD]
[TD="align: right"]-0.28[/TD]
[TD="align: right"]4516[/TD]
[TD="align: right"]10.5[/TD]
[TD="align: right"]-0.19[/TD]
[/TR]
[TR]
[TD]16-18lbs[/TD]
[TD="align: right"]10716[/TD]
[TD="align: right"]7.2[/TD]
[TD="align: right"]-0.29[/TD]
[TD="align: right"]4302[/TD]
[TD="align: right"]9.7[/TD]
[TD="align: right"]-0.23[/TD]
[/TR]
[TR]
[TD]19-21lbs[/TD]
[TD="align: right"]8980[/TD]
[TD="align: right"]6.4[/TD]
[TD="align: right"]-0.33[/TD]
[TD="align: right"]4327[/TD]
[TD="align: right"]9.9[/TD]
[TD="align: right"]-0.05[/TD]
[/TR]
[TR]
[TD]22-24lbs[/TD]
[TD="align: right"]7437[/TD]
[TD="align: right"]6.2[/TD]
[TD="align: right"]-0.3[/TD]
[TD="align: right"]5632[/TD]
[TD="align: right"]8.3[/TD]
[TD="align: right"]-0.17[/TD]
[/TR]
[TR]
[TD]25-28lbs[/TD]
[TD="align: right"]7942[/TD]
[TD="align: right"]5.5[/TD]
[TD="align: right"]-0.37[/TD]
[TD="align: right"]6459[/TD]
[TD="align: right"]6.8[/TD]
[TD="align: right"]-0.2[/TD]
[/TR]
[TR]
[TD]29lbs+[/TD]
[TD="align: right"]5733[/TD]
[TD="align: right"]5.3[/TD]
[TD="align: right"]-0.3[/TD]
[TD="align: right"]168[/TD]
[TD="align: right"]7.1[/TD]
[TD="align: right"]-0.21[/TD]
[/TR]
[TR]
[TD][/TD]
[TD="align: right"]106162[/TD]
[TD="align: right"]7.8[/TD]
[TD][/TD]
[TD="align: right"]44984[/TD]
[TD="align: right"]10.5[/TD]
[TD][/TD]
[/TR]
</tbody>[/TABLE]
 
My first question there Simmo is how the hell have you reproduced a table and maintained the formating?
 
some extracts from the weight chapter from the excellent book Against The Odds by David Lee Priest where he analyses the results of thousands of races to come to his conclusions.

"An eminent backer has demonstrated that weight only significantly encumbers horses when they are travelling at maximum speed; horses slow considerably to undertake turns and thus weight is a greater issue when horses are racing in a straight line" (There's your X-C explained right there.)

talking about using raceform ratings and whether or not to adjust them for weight carried - "I found that adjusting the ratings for weight carried in the conventional fashion weakens the relationship between ratings and performance"....." I found that the weight correction which yielded the best results was 0.25lb for every pound of additional weight carried"

From handicap races he produces the following table based on how far from top weight the beast is, showing number of runners, strike rate and level stakes profit.


[TABLE="width: 448"]
<tbody>[TR]
[TD="width: 64"][/TD]
[TD="class: xl63, width: 192, colspan: 3"]Flat[/TD]
[TD="class: xl63, width: 192, colspan: 3"]Jumps[/TD]
[/TR]
[TR]
[TD]Rating[/TD]
[TD]No[/TD]
[TD]SR%[/TD]
[TD]Profit[/TD]
[TD]No[/TD]
[TD]SR%[/TD]
[TD]Profit[/TD]
[/TR]
[TR]
[TD]Top Wgt[/TD]
[TD="align: right"]9619[/TD]
[TD="align: right"]11.8[/TD]
[TD="align: right"]-0.23[/TD]
[TD="align: right"]5182[/TD]
[TD="align: right"]15.9[/TD]
[TD="align: right"]-0.19[/TD]
[/TR]
[TR]
[TD]1-3lbs[/TD]
[TD="align: right"]8656[/TD]
[TD="align: right"]10.1[/TD]
[TD="align: right"]-0.26[/TD]
[TD="align: right"]2841[/TD]
[TD="align: right"]13.2[/TD]
[TD="align: right"]-0.2[/TD]
[/TR]
[TR]
[TD]4-6lbs[/TD]
[TD="align: right"]10625[/TD]
[TD="align: right"]9.4[/TD]
[TD="align: right"]-0.25[/TD]
[TD="align: right"]3367[/TD]
[TD="align: right"]11.3[/TD]
[TD="align: right"]-0.29[/TD]
[/TR]
[TR]
[TD]7-9lbs[/TD]
[TD="align: right"]11813[/TD]
[TD="align: right"]7.7[/TD]
[TD="align: right"]-0.34[/TD]
[TD="align: right"]3862[/TD]
[TD="align: right"]11.4[/TD]
[TD="align: right"]-0.28[/TD]
[/TR]
[TR]
[TD]10-12lbs[/TD]
[TD="align: right"]12452[/TD]
[TD="align: right"]7.5[/TD]
[TD="align: right"]-0.35[/TD]
[TD="align: right"]4328[/TD]
[TD="align: right"]10.6[/TD]
[TD="align: right"]-0.29[/TD]
[/TR]
[TR]
[TD]13-15lbs[/TD]
[TD="align: right"]12189[/TD]
[TD="align: right"]7.6[/TD]
[TD="align: right"]-0.28[/TD]
[TD="align: right"]4516[/TD]
[TD="align: right"]10.5[/TD]
[TD="align: right"]-0.19[/TD]
[/TR]
[TR]
[TD]16-18lbs[/TD]
[TD="align: right"]10716[/TD]
[TD="align: right"]7.2[/TD]
[TD="align: right"]-0.29[/TD]
[TD="align: right"]4302[/TD]
[TD="align: right"]9.7[/TD]
[TD="align: right"]-0.23[/TD]
[/TR]
[TR]
[TD]19-21lbs[/TD]
[TD="align: right"]8980[/TD]
[TD="align: right"]6.4[/TD]
[TD="align: right"]-0.33[/TD]
[TD="align: right"]4327[/TD]
[TD="align: right"]9.9[/TD]
[TD="align: right"]-0.05[/TD]
[/TR]
[TR]
[TD]22-24lbs[/TD]
[TD="align: right"]7437[/TD]
[TD="align: right"]6.2[/TD]
[TD="align: right"]-0.3[/TD]
[TD="align: right"]5632[/TD]
[TD="align: right"]8.3[/TD]
[TD="align: right"]-0.17[/TD]
[/TR]
[TR]
[TD]25-28lbs[/TD]
[TD="align: right"]7942[/TD]
[TD="align: right"]5.5[/TD]
[TD="align: right"]-0.37[/TD]
[TD="align: right"]6459[/TD]
[TD="align: right"]6.8[/TD]
[TD="align: right"]-0.2[/TD]
[/TR]
[TR]
[TD]29lbs+[/TD]
[TD="align: right"]5733[/TD]
[TD="align: right"]5.3[/TD]
[TD="align: right"]-0.3[/TD]
[TD="align: right"]168[/TD]
[TD="align: right"]7.1[/TD]
[TD="align: right"]-0.21[/TD]
[/TR]
[TR]
[TD][/TD]
[TD="align: right"]106162[/TD]
[TD="align: right"]7.8[/TD]
[TD][/TD]
[TD="align: right"]44984[/TD]
[TD="align: right"]10.5[/TD]
[TD][/TD]
[/TR]
</tbody>[/TABLE]

That makes sense in the context we've been discussing and where we've got to. It's logical that horses over two mile trips will be travelling faster than horses running over four mile trips. Therefore what he asserts fits with the way our discussion has developed, and in fact horses over the minimum trip are likely to be more inconvenienced statistically by carrying more weight.

It would still be worth doing some research on the key 2 mile handicap hurdles and chases (those generally run at a good gallop) to see if the opposite is true of what we've found for the extreme staying races.

It also makes me think that there will be further information to be gained by researching the correlation of weight to results depending on the times and how those races were run against course pars.

Being able to predict the pace of a race (which can usually be done), may give a huge advantage when searching for value.
 
Actually one other thing we haven't considered is the impact the ground has on the weight carried.

I'm kicking myself for missing something so obvious. The stats you produced on the key staying chases simmo, do they change significantly if you split the races run on good to soft or better from soft or worse?
 
i've used flatstats for the Cesarewitch

Results since 2005

wins/runners

9-4+:3/43 = 7.0%
8.08 to 9.03: 4/131 = 3.1%
up to 8.07: 3/153 = 2.0%

worth noting that those carrying 9.04+ were just 13% of the runners

now those that were on ground slower than Good..more stamina sapping

9-4+:2/15 = 13.3%
8.08 to 9.03: 3/68 = 4.4%
up to 8.07: 1/114 = 0.9%

the more a test of stamina it is.... the more it favours top weights..opposite of what Smith believes
 
In the same theme

These are every handicap run on SOFT+HVY since 2005 over 16f or more..only 20 races..but all will be a test of stamina above the norm

9-4+:9/74 = 12.2%
8.08 to 9.03: 5/81 = 6.2%
up to 8.07: 6/55 = 10.9%

again no evidence that carrying weight is a bar to winning
 
Do flatstats also provide the same for placed horses Alan? If the same correlation exists, which logically it should, it's likely the higher weighted horses are underbet because of the perception that carrying weight is a disadvantage.

I guess the logical final step is to look at the distribution of starting prices.
 
yes they do..and they show the AE..which answers your price query too

again..16f+ handicaps..soft/hvy

AE [performance against market expectation]

9.04+: 0.98
8.08-9.03: 0.65
up to 8.07: 1.27

placed

9.04+: 32.4%
8.08-9.03: 21.0%
up to 8.07: 30.9%
 
Last edited:
Going back to the National

A possible theory why top weights appear to be disadvantaged. Many horses near the top may be targetted at the National..as showed by Simmo earlier...but only after they have given their best trying to win Graded events for a year or two previoulsy..basically when a horse is at its peak they keep away from the National..then as a later career move decide the horse isn't good enough to win those..although still has a high rating.

In the examples Simmo gave..Madison De Berlais...Imperial Commander..Tidal Bay...could fit that criterion...and i bet there are more like them.

How many times do we get a new kid on the block like the first 3 in the GC this year..who are then targetted at the National the season after..or same season...very few imo. Many Clouds is pretty unusual in that he breaks that trend...but i think many higher weighted GC horses have their best days behind them. I'll bet if horses that are potential GC winners didn't go that route and went the National route..then more higher weights would win..with no concession needed. Its clear that would never happen as you are clearly going to have a couple of go's at the GC with such a horse..but in using those years trying to win the GC you are losing those best years for that purpose. Hence you get to the National..with a big weight..and your best years behind you.
 
Last edited:
So whichever way we look at it, the higher weights overperform over staying trips as is evidenced by both win and place stats.

Therefore I draw three conclusions:

1. Class can overcome weight, and therefore statistically there is more value at the top of the handicap in staying races in both codes.
2. It's worth doing the work to see if the outcome changes as trips reduce, and what impact pace has over the shorter distances.
3. Phil Smith is a muppet
 
with that thought in mind..equating to best days behind you

using the 16f+ results again..i just ran it for 6yo+ horses

9.04+: 2/37 = 5.4%
8.08-9.03: 2/31 = 6.5%
up to 8.07: 2/22 = 9.1%

Interesting how carrying more weight now seems to be a disadvantage

the AE's also show horses underform against odds when older

9.04+: 0.46
8.08-9.03: 0.74
up to 8.07: 1.28

I'd say if horses are carrying weight..then they need to be carrying it when at the peak of their powers..not after

It seems to me that all Phil Smith is going to do is help old hosses with high ratings win a National ..and he will make sure anyone with a GC horse..who is in its prime..like Many Clouds..win even more often.

It looks to me that if you have a GC horse..get it shoved in the National instead..you get a big edge with a horse in its prime..and a smaller edge if you have already had a few goes at the GC

Imo he has dropped a big clanger with this concession. But for some astute trainer..it could be a regular supply of National winners:)

When you look at Many Clouds..you would have to say if he had not run in a gruelling GC..he would have probably won the GN by 10 lengths if just targetted straight at it
 
Last edited:
Going back to the National

A possible theory why top weights appear to be disadvantaged. Many horses near the top may be targetted at the National..as showed by Simmo earlier...but only after they have given their best trying to win Graded events for a year or two previoulsy..basically when a horse is at its peak they keep away from the National..then as a later career move decide the horse isn't good enough to win those..although still has a high rating.

In the examples Simmo gave..Madison De Berlais...Imperial Commander..Tidal Bay...could fit that criterion...and i bet there are more like them.

How many times do we get a new kid on the block like the first 3 in the GC this year..who are then targetted at the National the season after..or same season...very few imo. Many Clouds is pretty unusual in that he breaks that trend...but i think many higher weighted GC horses have their best days behind them. I'll bet if horses that are potential GC winners didn't go that route and went the National route..then more higher weights would win..with no concession needed. Its clear that would never happen as you are clearly going to have a couple of go's at the GC with such a horse..but in using those years trying to win the GC you are losing those best years for that purpose. Hence you get to the National..with a big weight..and your best years behind you.

The most successful trainers of the modern era have been those that have thought differently to the crowd and found new angles. Given the value of the race it's almost worth reversing the conventional thinking.

The bigger stables, and Nicholls immediately springs to mind given his constant riches in the staying chaser department, could easily target this rather than the Gold Cup.

What if Southfield Theatre for example had his mark protected and they targeted the National. He'd be a classy 8 year old staying chaser getting in off 154, possibly also getting some additional help from Phil Smith.

Typically he would be targeted at the Hennessy where he'd likely go up to the 160's, or if things didn't go to plan he's win a graded race around Cheltenham which would have the same effect. Either way you'd potentially be looking at a mid 160's horse off top weight rather than a 154 horse going off closer to 11st.

Are you reading Mr Nicholls?!!!
 
Last edited:
yes..Nichols could win the National nearly every year if he set his stall out..and he is smart enough..he is already exploiting Cheltenhamitis that most trainers have got..is there a saturday when he hasn't mopped up a big pot?

he could aim two or 3 at it every year. ..with horses that are still in their prime like Many Clouds
 
I've often wondered why some of the top trainers don't start a top staying chaser's campaign in the new year or at Kempton over Christmas.

What's worth more and which would you rather win, The Hennessy or The Betfair Chase, or the Gold Cup and The National?
 
Fair question, but it's also worth looking at the place money for the latter two compared to the first two.

National -
  • £561,300.00, £211,100.00, £105,500.00, £52,700.00, £26,500.00, £13,200.00, £6,800.00, £3,600.00, £2,000.00, £1,000.00
Gold Cup -

  • £313,225.00, £117,535.00, £58,850.00, £29,315.00, £14,740.00, £7,370.00
Hennessy -

  • £99,662.50, £37,397.50, £18,725.00, £9,327.50, £4,690.00, £2,345.00
Betfair -

  • £112,873.35, £42,733.33, £21,553.33, £10,933.33, £5,653.33, £2,993.33

That alone is a no-brianer. It's worth throwing the King George and the Lexus in the mix as well.

King George -

  • £113,900.00, £42,740.00, £21,400.00, £10,660.00, £5,360.00, £2,680.00

Lexus -

  • £75,000.00, £23,750.00, £11,250.00, £3,750.00, £2,500.00, £1,250.00

If you just look at the place money for Aintree and Cheltenham, financially it's a better option to target these with a fresh horse. And if the conclusions we're drawing are correct, the National itself should be an obvious target for a classy 2nd season chaser with it's mark looked after.

Don't forget, if a horse doesn't get round the National fences and it's fresh enough there's still a good option at Punchestown. Al Co lines up in the Scottish version tomorrow after its early fall. There's also Sandown, so 2 or 3 options as consolation prizes if necessary.
 
Last edited:
Actually looking at the National prizemoney Paul Maloney must have picked up over £400,000 in prizemoney for his placed efforts for the Rucker horses Alvarado, Cappa Bleu, and State of Play in the last 8 years. That's more than you get for being a Gold Cup winner!
 
Last edited:
Back
Top