derby fact 64 years

[sbcaris]

In the last 64 years only 4 horses won the roses if they were stone closers. I define a stone closer as being 10 lengths or more off the pacesetter at the 1/2 mile point in their last race before the Derby.

The four that were exceptions to the above rule were Grindstone 1996, Pleasant Colony in 1981, Cannonade in 1974 and Carry Back in 1961.

That's 4 in the last 64 or ONLY 6.25% who were exceptions. This rules out Brodys Cause and perhaps a few others in this years Derby.

Note: It doesn't prevent them from running second or third but it sure is a big disadvantage in the win position of the Derby. 93.75% of the winners were closer than 10 lengths behind at the 1/2 mile call in their last start before the Derby.

[SecretAgentMan]

I'm glad I didn't hear about this angle the year I picked Monarchos. But you're right, deep closers do very bad in the KD.

We've also only had 2 horses wire the field since 1988, that's 2 for 28.

[sbcaris]

Monarchos was not a stone closer by definition. He was only 7 1/4 lengths behind the leader at the 1/2 mile point in the Wood Memorial.

[SecretAgentMan]

Monarchos was not a stone closer by definition. He was only 7 1/4 lengths behind the leader at the 1/2 mile point in the Wood Memorial.




Oh Ok, I've always thought of him.as a deep closer. I rank 3 divisions, pace setters, stalkers & closers. Of course there are also the horses that come from left field 20 lengths back, but rarely win. See SuddenBreakingNews.

SBN isn't as fast as Monarchos was tho. Crazy how Monarchos never won another race after almost breaking Secretariats time in the KD.

[SecretAgentMan]

Who are these stone closers this year?

[Spiderman]

Monarchos was not a stone closer by definition. He was only 7 1/4 lengths behind the leader at the 1/2 mile point in the Wood Memorial.


Monarchos' connections had no intention of winning that Wood. it is was ceded to the front running Congaree.

[ArlJim78]

In the last 64 years only 4 horses won the roses if they were stone closers. I define a stone closer as being 10 lengths or more off the pacesetter at the 1/2 mile point in their last race before the Derby.

The four that were exceptions to the above rule were Grindstone 1996, Pleasant Colony in 1981, Cannonade in 1974 and Carry Back in 1961.

That's 4 in the last 64 or ONLY 6.25% who were exceptions. This rules out Brodys Cause and perhaps a few others in this years Derby.

Note: It doesn't prevent them from running second or third but it sure is a big disadvantage in the win position of the Derby. 93.75% of the winners were closer than 10 lengths behind at the 1/2 mile call in their last start before the Derby.
Derby or not isn't it a big disadvantage for any horse coming up to a 10 furlong race? Unless you know the figure for all 10 furlong races you can't really call it a big derby disadvantage. It's what goes with the territory for that run style.
Also the fact that only 6 or so percent win with this angle, does not by itself rule out anyone. What's to say that Brody's Cause won't turn out to be one of the 6.25%?

[burnsy]

I tend to ignore some of this. Every race and every year is it's own entity. The 6.24% times it did happen.........its a loser if one thought this way. Orb didn't lag like that but he made quick middle moves and in the derby he started out 3rd (17th) from last. I kind of look at how the derby is run more than the prior prep. These kind of horses lose so much ground they become "stone cold" closers when the derby pace is fast. I toss a stat like this if I think the pace is going to be an issue. That's the key to this entire race because a few of the favorites should be in the mix early.........will one or more last? I watch the derby replays, just to see and remember.....you get the right fractions, like it was in 2013. Palace Malice and Verrazano folded fast and "Verrazano" was supposed to be "all world"...which never panned out. Right off the top my head, I've seen Orb, Animal Kingdom, Mine That Bird and Giacomo run them down from the rear, watching the replays. I could care less about how the "race before" unfolded. Basically, to me, that stat means.....not much. Because you put some pace up there and most will come back late.......that's when the closer strikes.

[CincyHorseplayer]

I don't think you can look at charts from the past KD and draw any conclusions going forward. A lot of horses you think you have nailed down will have different running styles in this race and other than vaguely it's hard to say where exactly they will be and what they can do from there. These aren't older horses where the guessing game is more concrete.

That said my criteria is that any horse no matter how far back, if they have shown they can run down a fast pace they are worthy. Fast paces strand bad closers. There are more than a few closers that have shown this ability.

[f2tornado]

The better question is how many stone closers ran in the Derby relative to the number of winners? I don't see many win a 9F prep from the clouds.

[CosmicWon]

Although it's a very fair track overall, in general it's not very advantageous to be a big closer at CD unless your name is Street Sense, Mine That Bird, or Drosselmeyer.

Revolutionary, the Dallas Stewart placers, Ice Box all made big runs to hit the board but they get so far behind even Zenyatta couldn't do it successfully. Keen Ice in 2015 Clark another example of too little, too late in that stretch.

Jon White has his Derby Strikes and one is a horse has to be in position to take the lead at the 8th pole which is where the winner of KYD usually asserts himself. California Chrome, Barbaro, I'll Have Another, Big Brown, American Pharoah, Smarty Jones all had tactical speed and sat the perfect Derby trip about 2w and pressing in maybe 5th or 6th (or closer) before opening up into the stretch.

Mor Spirit, Gun Runner, Destin, Mohaymen, Dazzling Gem (if he gets in), maybe Outwork are some of the animals likely to sit that trip in 3wks imho.

[MONEY]

In the last 64 years only 4 horses won the roses if they were stone closers. I define a stone closer as being 10 lengths or more off the pacesetter at the 1/2 mile point in their last race before the Derby.

This statistic has no value because we don't know how many total horses fit that criteria.
We also don't know how many of the last 64 Kentucky Derbys had one or more horses that fit your criteria.

There are 12 horses entered to run in Oaklawn's Arkansas Derby this Saturday.
None of the horses were 10 or more lengths behind at the half mile call during their last race.
So a horse that was 10 or more lengths behind at the half mile call during their last race can't win the
Arkansas Derby this Saturday.

[depalma113]

Looking over the past 15 years, almost all of the horses that fit this criteria were hopeless longshots. Dialed In was 4-1, Revolutionary was 6-1 and Circular Quay was 8-1. The rest were double digit.

Add in Ice Box at 11-1 and you have probably 4 legitimate contenders that fit this angle over the last 15 years.

[Lemon Drop Husker]

This statistic has no value because we don't know how many total horses fit that criteria.
We also don't know how many of the last 64 Kentucky Derbys had one or more horses that fit your criteria.

There are 12 horses entered to run in Oaklawn's Arkansas Derby this Saturday.
None of the horses were 10 or more lengths behind at the half mile call during their last race.
So a horse that was 10 or more lengths behind at the half mile call during their last race can't win the
Arkansas Derby this Saturday.

Suddenbreakingnews and Creator were both 10 lengths behind at the 2nd call in the Rebel.

[Cratos]

Although it's a very fair track overall, in general it's not very advantageous to be a big closer at CD unless your name is Street Sense, Mine That Bird, or Drosselmeyer.

Revolutionary, the Dallas Stewart placers, Ice Box all made big runs to hit the board but they get so far behind even Zenyatta couldn't do it successfully. Keen Ice in 2015 Clark another example of too little, too late in that stretch.

Jon White has his Derby Strikes and one is a horse has to be in position to take the lead at the 8th pole which is where the winner of KYD usually asserts himself. California Chrome, Barbaro, I'll Have Another, Big Brown, American Pharoah, Smarty Jones all had tactical speed and sat the perfect Derby trip about 2w and pressing in maybe 5th or 6th (or closer) before opening up into the stretch.

Mor Spirit, Gun Runner, Destin, Mohaymen, Dazzling Gem (if he gets in), maybe Outwork are some of the animals likely to sit that trip in 3wks imho.
I agree with you; Churchill Downs at 1-1/4 miles is not very conducive for "come from behind" type horses because about 60% of that race distance is in the straight-away at Churchill and the turns are very tight.

[sbcaris]

The reason I researched this angle was as follows: I was wondering after Brody's Cause won the Blue Grass how many stone closers won the roses after a major closing effort in their last start when they were 10 lengths back at the 1/2 mile call.

I did not realize that only around 1 or 2 horses per year ever enter the Derby with such a slow beginning in their last major prep race. The impact value for this rule is probably around 1.00 ( 6% starters divided by around 6% winners) which means it doesn't have much clout at all.

Interesting tidbit: Brodys Cause with the Field in pool 2 pays $155 for 2 bucks. Whereas the Field over Brodys Cause pays ONLY $92 for 2 dollars. It may only be a parimutuel inequity but the people that wagered on the above exacta combinations might have been thinking that Brody's Cause has much more chance to finish second, third or fourth rather than winning the roses.

[castaway01]

I'm glad I didn't hear about this angle the year I picked Monarchos. But you're right, deep closers do very bad in the KD.

We've also only had 2 horses wire the field since 1988, that's 2 for 28.

Which in a 20-horse field is still more than you would statistically expect, showing how these stats can be misinterpreted or misleading.

[pandy]

In the last 64 years only 4 horses won the roses if they were stone closers. I define a stone closer as being 10 lengths or more off the pacesetter at the 1/2 mile point in their last race before the Derby.

The four that were exceptions to the above rule were Grindstone 1996, Pleasant Colony in 1981, Cannonade in 1974 and Carry Back in 1961.

That's 4 in the last 64 or ONLY 6.25% who were exceptions. This rules out Brodys Cause and perhaps a few others in this years Derby.

Note: It doesn't prevent them from running second or third but it sure is a big disadvantage in the win position of the Derby. 93.75% of the winners were closer than 10 lengths behind at the 1/2 mile call in their last start before the Derby.

This is probably true for all dirt races at all distances and classes. The percentage won't be as bad as the Derby, but most winners aren't that far back. When all is said and done, it's very tough to win if you don't play the percentages. One could make the case that a smart bettor should never bet a horse that figures to be that far back.

[MONEY]

Suddenbreakingnews and Creator were both 10 lengths behind at the 2nd call in the Rebel.

According to Brisnet Creator was 10th by 31⁄2 and Sudden move was 13th by 71⁄2 at the second call in the Rebel.

[f2tornado]

Interesting tidbit: Brodys Cause with the Field in pool 2 pays $155 for 2 bucks. Whereas the Field over Brodys Cause pays ONLY $92 for 2 dollars. It may only be a parimutuel inequity but the people that wagered on the above exacta combinations might have been thinking that Brody's Cause has much more chance to finish second, third or fourth rather than winning the roses.

I suspect you're correct. I know in one of the early Future wagers I took Mohaymen over the field and several other horses including Brody. In hindsight, I should have taken the field over and under several horses instead of focusing on one. In the case of Brody, I have a hard time with that putrid BSF he got in the Bluegrass but it wouldn't be the first putrid final prep BSF to win roses. I still have a hard time overlooking that Storm Cat on top as well. These guys have hit the board but have yet to win. I'll use him mostly underneath just as the Future wager suggests more folks are doing.

[Cratos]

The reason I researched this angle was as follows: I was wondering after Brody's Cause won the Blue Grass how many stone closers won the roses after a major closing effort in their last start when they were 10 lengths back at the 1/2 mile call.

I did not realize that only around 1 or 2 horses per year ever enter the Derby with such a slow beginning in their last major prep race. The impact value for this rule is probably around 1.00 ( 6% starters divided by around 6% winners) which means it doesn't have much clout at all.

Interesting tidbit: Brodys Cause with the Field in pool 2 pays $155 for 2 bucks. Whereas the Field over Brodys Cause pays ONLY $92 for 2 dollars. It may only be a parimutuel inequity but the people that wagered on the above exacta combinations might have been thinking that Brody's Cause has much more chance to finish second, third or fourth rather than winning the roses.
In all due respect, from what you have stated this is a conditional probability problem which is more complex and need more info than what is being given.

[DeltaLover]

This statistic has no value because we don't know how many total horses fit that criteria.
We also don't know how many of the last 64 Kentucky Derbys had one or more horses that fit your criteria.

There are 12 horses entered to run in Oaklawn's Arkansas Derby this Saturday.
None of the horses were 10 or more lengths behind at the half mile call during their last race.
So a horse that was 10 or more lengths behind at the half mile call during their last race can't win the
Arkansas Derby this Saturday.

:ThmbUp:

Good posting. People are easily deceived by statistics like the one presented in this thread. I will repeat: This statistic has no value and should not be considered for handicapping purposes

[sbcaris]

I see two horses in the Arkansas Derby who were 10 lengths behind at the 1/2 mile point in their last race.

Creator according to the Daily Racing Form was in 12th position at the 1/2 mile point call and it is listed as 10 lengths behind. At six furlongs creator was 10th 3/12 lengths behind.

Suddenbreakingnews according to the Daily Racing Form was in 13th position and 10 lengths behind at the 1/2 mile call. At the six furlong call he was in 13th and 7 1/4 lengths behind.

[dilanesp]

:ThmbUp:

Good posting. People are easily deceived by statistics like the one presented in this thread. I will repeat: This statistic has no value and should not be considered for handicapping purposes

In general, people should be very wary of Derby statistics in handicapping. The sample size is so small.

It's similar to what Donald Trump is doing to political scientists' election models. They were quite sure something like Trump could not happen (and that a contested convention, which is now the only alternative to a Trump nomination also could not happen). And now it has.

That's because you can't jump to conclusions based on really small sample sizes.

[Lemon Drop Husker]

In general, people should be very wary of Derby statistics in handicapping. The sample size is so small.

It's similar to what Donald Trump is doing to political scientists' election models. They were quite sure something like Trump could not happen (and that a contested convention, which is now the only alternative to a Trump nomination also could not happen). And now it has.

That's because you can't jump to conclusions based on really small sample sizes.

The biggest thing with Derby statistics, is that since 1999 when Charismatic upset the apple cart, a lot of weird things, trends, never happened before kind of stuff has now happened.

The old tried an true is no longer tried and true.

War Emblem's prep was the Illinois Derby. Chrome was outside of any breeding angle whatsover. Funny Cide was a gelding. Giacomo and Mine That Bird were "no-chancers". Animal Kingdom couldn't run on dirt.

The "never has happened before" list will keep dwindling until there is no more angles to be played.

[dilanesp]

The biggest thing with Derby statistics, is that since 1999 when Charismatic upset the apple cart, a lot of weird things, trends, never happened before kind of stuff has now happened.

The old tried an true is no longer tried and true.

War Emblem's prep was the Illinois Derby. Chrome was outside of any breeding angle whatsover. Funny Cide was a gelding. Giacomo and Mine That Bird were "no-chancers". Animal Kingdom couldn't run on dirt.

The "never has happened before" list will keep dwindling until there is no more angles to be played.

Yep. My Derby prognostication record is no better than anyone else's, but if I were to offer an angle, it would be "approach this thing as much as possible like it is a normal 3 year old stakes race". There's no reason to think that any of the statistical angles are actually positive expected value rather than just being statistical noise over a small sample size.

[sbcaris]

Dilansap: Is 43 years of data too small a sample size to draw conclusions from? My angle for the Derby which consists of a final 3/8 time of 37 4/5 has a high degree of predictableness for the Derby. In fact it has gotten nearly 70% winners from only 30% of the starters over the last 43 years. A Fischer test of significance gives a positive p value for such a statistic.

Do you have any proof that the above is not valid? If so, explain it to me.

[Lemon Drop Husker]

Dilansap: Is 43 years of data too small a sample size to draw conclusions from? My angle for the Derby which consists of a final 3/8 time of 37 4/5 has a high degree of predictableness for the Derby. In fact it has gotten nearly 70% winners from only 30% of the starters over the last 43 years. A Fischer test of significance gives a positive p value for such a statistic.

Do you have any proof that the above is not valid? If so, explain it to me.

I like that angle. I actually don't even consider it an angle as much as it is a qualifier.

These are 3YO horses looking to go the longest distance in their lives in an actual race. Contenders and pretenders are very much sorted out in the 9F preps, and horses that are coming forward late are the ones that have to get your attention.

[dilanesp]

Dilansap: Is 43 years of data too small a sample size to draw conclusions from? My angle for the Derby which consists of a final 3/8 time of 37 4/5 has a high degree of predictableness for the Derby. In fact it has gotten nearly 70% winners from only 30% of the starters over the last 43 years. A Fischer test of significance gives a positive p value for such a statistic.

Do you have any proof that the above is not valid? If so, explain it to me.

43 races is a very small sample size. Take it out of the context of the Derby. Have you ever run bad handicapping over a stretch of 43 races? Gone 1 for 43, or 2 for 43, or even 0 for 43? Would you make conclusions as to your performance as a handicapper from a negative result over 43 races?

I think you are hiding behind significance tests. This is common sense. If you were going to study some characteristic of $10,000 claimers, you'd want hundreds of races. Maybe thousands.

Additionally, the Derby has changed, a LOT, since 1974. For the first half of your sample, the race routinely did not draw 20 starters. There were even a couple of Derbies that didn't even draw 15 and didn't use the auxiliary starting gate. Silver Charm's Derby, in 1997, is NOTHING like the modern Kentucky Derby. If you don't believe me, pull the race up on youtube and watch it.

The sample includes tracks that were dead, like the 1989 Derby, and tracks that were lightning fast, like the 2001 Derby. It includes Derbies with lone front runners, like in 1985 and 2002, and Derbies with logjams up front, like 1981. It also includes Derbies after the eligibility requirements were changed to exclude sprinters, in the last few years.

Your sample also ignores that the average number of starts of a Derby starter has gone way down. Derby horses have far less experience now than they did in 1974. And it ignores that the average length of time between the last start and the Derby has decreased immensely. In the 1970's, the Derby routinely featured horses who had run in races like the Derby Trial (one week before) and the Blue Grass (at the time, less than two weeks prior). It almost always featured the Wood Memorial winner (at the time, exactly two weeks prior). Some of these horses, like Pleasant Colony and Alysheba, won the Derby.

For your invocation of significance to be valid, you would need to be repeating the same, exact trial 43 times. But that's not what the Derby is, not at all. It's a totally different race now.

[sbcaris]

Dilansap: You have not really provided any proof that the above indicator is not a solid one.

If one looks at my data over the last 43 years of Derbies we find that 30 Derby winners in the last 43 years fit my final fraction indicator. That's 69.8% winners. Since the percentage of starters is 29.8% the impact value is a strong 2.39. Horses qualifying on my final fraction indicator are winning the roses more than twice as often as is expected. The ROI is a strong 41% over the last 43 years and the p value of the Fisher test comes to .001 which means the data above is statistically significant and not likely to be due to chance or a statistical aberration.

[dilanesp]

the p value of the Fisher test comes to .001 which means the data above is statistically significant and not likely to be due to chance or a statistical aberration.

http://influentialpoints.com/Training/Fishers_exact_test_use_and_misuse.htm

Pooling of frequencies from different blocks or replicates in experimental designs is also invalid. There is a clear risk of bias when categories are collapsed to create 2 × 2 tables from r × c tables....

The fact that Fisher's exact test is generally used when sample sizes are small leads irrevocably to another problem with its use - namely that, irrespective of the result of a statistical test, one cannot have much confidence in results based on very small sample sizes. Any test on such data will have low power to disprove the null hypothesis, and the chance of such a sample being representative of a population is low. As always, convenience sampling means that inference cannot be extended outside the sample

Cliffs: Fisher's exact test only delivers a valid significance level when the exact same experimental design was used to generate each sample (not true, as I demonstrated, for the last 43 Derbies), and that even when the significance level is valid, the sample size problem remains because you can never be sure of the representativeness of the sample.

[sbcaris]

Dilansap: You said, "one cannot have much confidence in results based on a small sample size."

I say you can and the stat I presented is a solid one.

How much more confidence do you need? 43 years of the Derby and 70% winners from only 30% of the starters. If the indicator was meaningless one would expect 30% winners from 30% starters. Horses qualifying on this indicator are the strong contenders. They are outperforming their opportunities. They finish fast in their prep races at 9 furlongs and are the ones most likely to finish fast when they go another furlong in the Derby.

[dilanesp]

Dilansap: You said, "one cannot have much confidence in results based on a small sample size."

I say you can and the stat I presented is a solid one.

How much more confidence do you need? 43 years of the Derby and 70% winners from only 30% of the starters. If the indicator was meaningless one would expect 30% winners from 30% starters. Horses qualifying on this indicator are the strong contenders. They are outperforming their opportunities. They finish fast in their prep races at 9 furlongs and are the ones most likely to finish fast when they go another furlong in the Derby.

At this point we are talking past each other. I will try to summarize:

You think the discussion begins and ends with the fact that you have a data set of 43 Derbies.

I think that would only be true if the 43 Derbies all constituted independent repeated instances of the same trial, when in fact the nature of the Kentucky Derby has changed so much over time and in so many different ways that you cannot consider the 1974 Derby to be the same event as the 2015 Derby. And that makes your significance statistic invalid.

[sbcaris]

No it doesn't because the following data indicates that my final fraction indicator has done well in all the decades from 1973 to the present:

My data goes from 1973-2015 and the following segments show that in each decade horses qualifying on my indicator outperformed their opportunities:

1973-1979 There were 5 winners in those 7 years that qualified on my final fraction indicator.

1980-1989 There were 6 winners in that 10 year period that qualified

1990-1999 There were 8 winners in that 10 year period that qualified

2000-2009 There were 6 winners that qualified in that 10 year period.

2010-2015 There were 5 winners that qualified in that 6 year period.

My indicator of final fractions performed well above expected in the 70s, 80s, 90s, and 2000nds. Field size to me is irrelevant because my final fraction indicator has been above average in all field sizes since 1973.

[PhantomOnTour]

In the last 64 years only 4 horses won the roses if they were stone closers. I define a stone closer as being 10 lengths or more off the pacesetter at the 1/2 mile point in their last race before the Derby.

The four that were exceptions to the above rule were Grindstone 1996, Pleasant Colony in 1981, Cannonade in 1974 and Carry Back in 1961.

That's 4 in the last 64 or ONLY 6.25% who were exceptions. This rules out Brodys Cause and perhaps a few others in this years Derby.

Note: It doesn't prevent them from running second or third but it sure is a big disadvantage in the win position of the Derby. 93.75% of the winners were closer than 10 lengths behind at the 1/2 mile call in their last start before the Derby.
Forget the Derby...what % of runners win ANY dirt race when 10 back at the half mile pole?

[Lemon Drop Husker]

Forget the Derby...what % of runners win ANY dirt race when 10 back at the half mile pole?

I'd set the Over/Under at about 5%, and I'd lean heavily towards the Under.

[SecretAgentMan]

Forget the Derby...what % of runners win ANY dirt race when 10 back at the half mile pole?




Not many, so when I see horses like Brody's Cause, SBN or Mo Tom, things have to go perfectly right w/o any doubt. Mo Tom always seems to find trouble, & he's an automatic toss for me.

Unless there's a serious pace meltdown & fast fractions in the KD, like Giacomo's year, the deep closers won't be winning the derby.

The percentage is very low as me tioned above.

[Lemon Drop Husker]

Not many, so when I see horses like Brody's Cause, SBN or Mo Tom, things have to go perfectly right w/o any doubt. Mo Tom always seems to find trouble, & he's an automatic toss for me.

Unless there's a serious pace meltdown & fast fractions in the KD, like Giacomo's year, the deep closers won't be winning the derby.

The percentage is very low as me tioned above.

Actually, the Derby percentages are way way higher than most races.

Street Sense. Mine That Bird. Giacomo. Orb. Big Brown. Monarchos. Animal Kingdom.

3 of those essentially came from DFL (Dead effin' Last) to win the Derby.

[SecretAgentMan]

Actually, the Derby percentages are way way higher than most races.

Street Sense. Mine That Bird. Giacomo. Orb. Big Brown. Monarchos. Animal Kingdom.

3 of those essentially came from DFL (Dead effin' Last) to win the Derby.



I don't consider Big Brown or Street Sense deep closers the races before the derby.

[Lemon Drop Husker]

I don't consider Big Brown or Street Sense deep closers the races before the derby.

Most deep closers that have won the Derby aren't considered that way.

Takes some imagination with a 15+ horse field.