Early 20-20 for the Derby

Post Reply
Kennedy
Posts: 1043
Joined: Thu Sep 12, 2013 9:58 pm

Thu Apr 21, 2016 5:14 pm

It's a shame you can't upload spreadsheet's directly to the message board. So hopefully screenshots and a long winded explanation will do the trick :)

For those who have no idea what I'm talking about the 20-20 system as I call it is just a pet project of mine that I've been working on for a little while. The source data for the original 20-20 profile was from the years 1996-2007. Since "going live" with the system in 2008 I have made one major revision in 2015 which resulted in changes to 4 factors. The goal of the changes was to remove arbitrarily selected numbers that had been used as guidelines. I have since been able to see how the system works from 1992 to the present.

The whole intention and vision of the Derby 20-20 system is to identify key statistical factors and trends and compile them into a profile that could be used for more accurate and profitable wagering on the Derby. I do not advocate simply wagering blindly on a system like this. Common sense must always be applied. This is just a tool to help shape your view of the viability of each contender.

#1 The entrant must earn a Beyer Speed Figure around two turns as a three year old that is equal to or superior to the average two turn prep winning figure.
In other words I take the winning Beyer figures from all the two turn stakes prep races represented in the Derby and then average them. Horses who did not qualify with this factor were a cumulative 169-2-7-6 over the last 24 runnings of the Derby. Mine that Bird and Sea Hero were the exceptions.

This years average is 93. That means Brody's Cause, Gun Runner, Majesto, Mo Tom, Oscar Nominated, Shagaf, Tom's Ready and Whitmore are not fast enough. Lani gets a pass because he doesn't have any figures.

#2 Last two Beyer figures achieved must total the average two turn prep winning figures times two minus six.
Seems a bit mathematical but essentially this is an effort to weed out one time big fig horses. It demands consistent speed over the entrants last two races. So the average prep winning figure is multiplied by two. But we also recognize that these horses are improving and it is not necessary that a horse be "fast enough" to win the Derby 2 starts before the Derby itself. So we minus 6 Beyer points from the total to allow for this improvement. 6 is not an arbitrarily selected number. 6 is the average Beyer improvement shown by Derby winners during their prep races. Horses who did not qualify with this factor were a cumulative 212-3-9-10 in the Derby. Mine That Bird, Sea Hero and Giacomo were the Derby winners who did not qualify.

So it means for this year that the entrants last two Beyer figures have to equal 180. Brody's Cause, Majesto, Mo Tom, Mohaymen, Shagaf, Tom's Ready and Trojan Nation were not fast enough over their last two starts. Lani and Oscar Nominated get a pass because of lack of figures or because only main track races are considered.

#3 The entrant must have been first or second at the 1/8th pole in one of their final two prep races.
This is a non-numerical measurement of speed. Plodders rarely win the Derby. Although the race is 10f it is often really just a race to the 1/8th pole. Most of the horses who are in front at the 1/8th pole are in front at the wire. We want a horse to have enough tactical speed to get them into a contending position when it matters. Horses who did not satisfy this criteria were a cumulative 122-1-8-6 in the Derby. Even Giacomo managed to show enough tactical speed in a prep race to qualify. Sea Hero was pretty much a stat buster on the speed front and he was the culprit again here.

That is bad news for Mo Tom, Mor Spirit, Shagaf and Suddenbreakingnews. None of them appear to have the tactical speed to be in contention when it gets serious so they'll have to be chasing late.

#4 The entrants last two prep races must have been run around two turns.
Two turn experience is vital for having a horse ready to go 10f in May. Horses with a single prep race around one turn (or less) are 44-1-1-1 in the Derby. Horses like Bellamy Road, Showing Up and Favorite Trick were all non-qualifiers. Big Brown was the only one to overcome this lack of stamina seasoning.

Nyquist and Lani are the ones trying to defy the odds this year.

#5 The entrants finishing position must be no worse than their position at the 1/8th pole in both final preps.
This is the second fitness factor or maybe a toughness factor. Derby winners rarely get passed in the stretch. They should be at least able to hold their position through the final 1/8th. The non-qualifiers for this factor were 130-3-9-6 in the Derby. Silver Charm, Super Saver and Mine That Bird all managed to overcome this factor and win the Derby

Danzing Candy, Exaggerator, Mohaymen, Shagaf and Tom's Ready all failed to hold their position in the stretch in one of their last two races. Lani gets a pass because of the lack of reported points of call.

#6 The entrant must have earned higher than field average BRIS Late Pace figure in their last race
In order to win the Derby you have to finish well. If you’re gasping for breath in the final stages of a 9f race you won’t handle the Derby distance as well. This is one of the factors that changed because I used to use actual closing fraction averages instead of pace figures but given the fact that I don't use raw times for races, I prefer speed figures, it seemed odd to then use raw final times for races. Pace figures are just a little sharper. This factor is kind of a hybrid though because I don't have any BRIS pace figures for any year prior to 99'. So to evaluate those older Derbies I still use the raw closing fraction data. All in all a Derby record of 189-3-8-11 for horses failing this stat. Silver Charm, Mine that Bird and Go for Gin were the outliers. Mine that Bird is the only one since the raw data was replaced by LP figures.

This year many of the final preps were really weird so Outwork, Trojan Nation, Shagaf, Mor Spirit, Danzing Candy and Mohaymen all failed to close out their last preps in a satisfactory manner. Lani gets a pass. Foreign horses are really hard to calculate.

#7 The entrant must have finished in the money in their last prep.
This one is simple and a pretty common and obvious way of judging form. A horse needs to show that they are in fine form and one of the most obvious ways they do that is to finish in the top 3 in their last race. The horses who failed to qualify for this factor were a cumulative 105-4-4-5 in the Derby. Giacomo, Thunder Gulch, Mine That Bird and of course Sea Hero were the culprits who defied this factor.

#8 The entrant must have finished in the money in at least two races as a three year old.
This value of this factor is two fold. First of all it tries to identify consistency among the entrants, but secondly it also means that a horse with two preps has very little wiggle room. The profile does not have any factor concerning the actual number of prep races. It only makes minimum requirement of what a horse must get from their prep races. In this case it effectively means that any horse with one prep is eliminated and also any horse with two preps who did not run well in both. Of course horses who do poorly over 3 or 4 preps can also be eliminated. The horses who failed to qualify for this factor were a cumulative 70-2-3-1 in the Derby. Sea Hero and Mine that Bird were the outliers and if you've read this far you've likely noticed that they were just about impossible from a statistical perspective but it's racing. Things happen.

That could spell bad news for Brody's Cause and Lani

#9 The entrants last Beyer Speed Figure cannot be worse than his previous two.
This is the last measurement of form. I feel that a horses last three Derby prep races are really the only ones of importance and you don't want the horses worst race to be his most current. I don’t mind slight regression but I want the trend of the speed figures to be up overall. If a horse only had two prep races they can be exempt from this factor. However not all horses with two preps will be exempt. For instance if a horse only has two preps but runs a career high in his last prep race then it’s obvious that his latest figure would not worse than his previous two. The horses who failed to qualify for this factor were a cumulative 62-0-2-1 in the Derby. Prairie Bayou, Bluegrass Cat and Imperialism were the only ones to overcome this and run in the frame.

Mohaymen, Mo Tom, My Man Sam and Danzing Candy all failed to qualify with this factor. Nyquist gets a pass which counts neither fore nor against. Since his last prep was worse than his second last we cannot say with authority that he is not regressing but this pattern is also based on 3 races and two races really tells us nothing. Lani also gets a pass because he has no speed figures

#10 The entrant must be a stakes winner.
It's been a while since Sir Barton. Not only do we want our Derby contenders running in stakes races, we want them winning them as well. Now they do not have to be Graded Stakes races but I do want my Derby horses to have beaten more than maiden company. The cumulative record of horses who failed to qualify for this factor is 95-1-6-3. Giacomo of course was just a maiden winner, but the overwhelming majority of these horses ran poorly because most of them simply are not good enough.

This year it applies to Trojan Nation, My Man Sam, Majesto, Tom's Ready and Whitmore

#11 The entrant must have earned a BRIS LP figure that is equal to or higher than their E1 or E2 pace figures in at least one route race as a 3yo.
This one is kind of geeky and is the newest factor. I tracked this stat for a while but it became an official part of the profile before the 2015 running. What we're trying to measure here is that a horse is capable of finishing a race. Generally you find that if a horses slowest part of their race is the last part they're not going to love a stretch out. Things happen in some races and dynamics can get weird which is why we give horses a wide berth and only require that they do it once in a route race in the current year. Horses who failed to qualify for this factor are 30-0-1-2

Danzing Candy and Shagaf are both in non-compliance.

#12 The entrant must have at least one prep race at nine furlongs.
This factor is about the entrant’s experience. Could a horse stay 10f without having prepped at 9f? Sure they could, they are merely not as tested and generally less likely. Derby entrants need a good solid 9f prep race. Horses who did not prep at 9f went 32-0-1-1 in the Derby.

That is a strike against the highly regarded Destin. All other entrants had 9f prep races this year.

#13 Must have exceeded the winning avg projected 10f time at least once in one of their last 2 preps
It's hard to explain this one briefly. Dr. Steve Roman publishes projected 10f times calculated from the prep races. It's really just math but what it attempts to do is predict what the time of the race would have been had it been 10f. Not a hard and fast measurement by any means but I take these projections and much like the speed figure measurements I get an average and I make sure that the entrants have exceeded the average at least once in their last two preps. The record of the non-qualifiers is 114-3-6-3 which actually makes it the least accurate of all the profile criteria but I still think it provides some value. Animal Kingdom, War Emblem and Mine That Bird failed to qualify.

Gun Runner, Lani, Oscar Nominated, Shagaf and Tom's Ready all failed to comply this year.

#14 Highest Lifetime Beyer must be achieved in a two turn race.
This is a pretty straightforward measurement of suitability to the Derby conditions. We don't want sprinters in the Derby, a quality distance horse will be faster in longer races than they were at short distances. 43 horses entered the Derby with this factor going against them and very few have emerged with credit (43-0-1-1)

This is maybe the biggest factor that stands against Nyquist. It also affects Shagaf.

#15 The entrants sire or broodmare sire must show a progeny average winning distance of 8.00f or more.
Only North American stakes races are included in the measurement of AWD. This is the only aspect of breeding I consider. Dosage is misleading but this factor looks directly at how the sire and broodmare sires best runners have fared in stakes races. If they are predominantly sprinters the average will show less than 8f and a sprinters pedigree is not ideal for winning the Derby. Horses whose sire and broodmare sire have less than 5 stakes runners are not included and therefore those horses are given a pass. The cumulative record of non-qualifiers is 39-0-1-1. Horses like Street Sense and Monarchos were given passes because of insufficient data on their sires at the time.

Mo Tom and Nyquist are both in violation but their sire is a freshman. Danzing Candy and Outwork get a pass because of insufficient data on either the sire or damsire side.
Attachments
Kentucky Derby 20-20 2016.png
Kentucky Derby 20-20 2016.png (54.08 KiB) Viewed 5835 times
Kennedy
Posts: 1043
Joined: Thu Sep 12, 2013 9:58 pm

Thu Apr 21, 2016 5:20 pm

For anyone who survived that first post and is wondering how the 20-20 profile actually worked out over the last few decades the answer is displayed graphically below.

This shows the number of "perfect qualifiers" meaning horses with no strikes against them. The figures show a year by year profitability based on a $2 win wager on each perfect qualifier.
Attachments
Kentucky Derby 20-20 results.png
Kentucky Derby 20-20 results.png (19.15 KiB) Viewed 5831 times
slewobarney
Posts: 202
Joined: Sat Sep 14, 2013 8:45 am

Thu Apr 21, 2016 8:22 pm

Great job, Kennedy......thank you!

If Adventist were to make the field, any idea where he would fit here? Thank you, again.
BaroqueAgain1
Posts: 9879
Joined: Sat Sep 14, 2013 6:16 pm

Thu Apr 21, 2016 9:26 pm

If I have read your fabulous research correctly, Kennedy, it looks like Creator is the only horse who ticks all your boxes. Interesting. 8-)
Somnambulist
Posts: 7382
Joined: Thu Sep 12, 2013 5:59 pm

Thu Apr 21, 2016 9:40 pm

His name reminds me do the awful metal band. I can't do it.

Your research/stats are always really interesting. I respect the time that went into it. You've got to be one of the most objective people I have ever encountered on the Internet, if not the most. Thanks for sharing.
"Life's no piece of cake, mind you, but the recipe's my own to fool with."
User avatar
Sparrow Castle
Posts: 4884
Joined: Wed Sep 18, 2013 6:44 pm

Thu Apr 21, 2016 9:58 pm

Thanks so much for posting this, Kennedy. Lots of good factors backed with data is certainly food for thought. I often don't bet the Derby because it seems to me the racing gods hold all the cards in that chaos (notable exception last year when I was there at the track). But I plan to take a shot this year, as long as I still have money in my account after Friday.
User avatar
Dusty
Posts: 602
Joined: Thu Sep 12, 2013 9:51 pm
Location: Dana Point, Ca
Contact:

Thu Apr 21, 2016 10:43 pm

WOW - very nicely done Kennedy - the time and thought invested in this - has me in awe - FABULOUS job and thanks for it!
May they run with the WIND
BigCountry
Posts: 82
Joined: Tue Dec 17, 2013 9:45 pm

Thu Apr 21, 2016 11:08 pm

Somnambulist wrote:You've got to be one of the most objective people I have ever encountered on the Internet, if not the most. Thanks for sharing.
Seconded!
middleground
Posts: 433
Joined: Fri Sep 13, 2013 1:29 pm

Thu Apr 21, 2016 11:17 pm

This is most impressive, Kennedy. Big thanks!
User avatar
Treve
Posts: 4394
Joined: Fri May 08, 2015 5:12 pm

Fri Apr 22, 2016 1:08 am

I am still very much a 'baby handicapper' and I don't bet but I love statistcs (for knowledge... knowledge is power they say). Thank you so much for compiling this and sharing with us! will be interesting to see how this year's derby fits in!
A filly named Ruffian...

Eine Stute namens Danedream...

Une pouliche se nommant Trêve...

Kincsem nevű kanca...


And a Queen named Beholder
MySaladDays
Posts: 907
Joined: Sun Mar 23, 2014 3:16 am

Fri Apr 22, 2016 3:23 am

Beautiful presentation, Kennedy!

It was truly a labor of love for you to write out all that, let alone create such a neat system.

Best of luck in the derby, you do the work!
peeptoad
Posts: 2713
Joined: Thu Sep 12, 2013 6:53 pm

Fri Apr 22, 2016 6:20 am

Nice job, Kennedy! I look forward to your 20-20 every year... and I'm glad to see my feelings about Creator are somewhat validated here. By no means would I consider him my top pick, but the 20-20 is another tool I am using to try and sift through all these guys. The margin between at least half of them is pretty razor thin this year.

Incidentally, when we were discussing Mohaymen in another thread a couple of weeks ago and I was talking of ground loss on the winner this is the stat I was trying to recall: #5 The entrants finishing position must be no worse than their position at the 1/8th pole in both final preps.
Kennedy
Posts: 1043
Joined: Thu Sep 12, 2013 9:58 pm

Fri Apr 22, 2016 3:01 pm

slewobarney wrote:Great job, Kennedy......thank you!

If Adventist were to make the field, any idea where he would fit here? Thank you, again.
The 20-20 can shift a little based on the entrants because many of the factors are averages and the averages will change if you add or remove horses. But assuming you were to leave the field as is (I'm not sure who to remove :)) and just add Adventist to the mix I think he'd have 5 strikes against him, so that would put him statistically on par with Moyahmen.
Kennedy
Posts: 1043
Joined: Thu Sep 12, 2013 9:58 pm

Fri Apr 22, 2016 3:49 pm

peeptoad wrote:Nice job, Kennedy! I look forward to your 20-20 every year... and I'm glad to see my feelings about Creator are somewhat validated here. By no means would I consider him my top pick, but the 20-20 is another tool I am using to try and sift through all these guys. The margin between at least half of them is pretty razor thin this year.

Incidentally, when we were discussing Mohaymen in another thread a couple of weeks ago and I was talking of ground loss on the winner this is the stat I was trying to recall: #5 The entrants finishing position must be no worse than their position at the 1/8th pole in both final preps.
If you want the full spreadsheet I can pass that on as well. No need to wait for me to write a long winded post :)
Kennedy
Posts: 1043
Joined: Thu Sep 12, 2013 9:58 pm

Fri Apr 22, 2016 5:13 pm

For those who are interested I've also been working on a tool that is aimed at predicting Derby winners from the other side of the coin. So the 20-20 essentially creates rules then throws out horses that do not abide by them. I've been working on a much more simple way to try and predict a winner and that is to simply enter in data points and see who ranks higher.

The thought is that it's simply relying on readily available stats to see which factors actually point you towards winning a Derby.

For example you could just take the last Beyer Speed Figure earned by each entrant and rank them 1-20 in descending order and bet the top 3 every year. In the 17 runnings you'd have 8 winners and a decent +51% return.

Use the top 3 by 3yo average Beyer and you'd still have 8 winners in 17 years but you'd have lost 20 cents.

Try average BRIS LP figure and Last BSF and you've have 9 winners and a +63% ROI. Expand those same stats to the top 5 contenders and you'd have 12 winners but the ROI would slip a little to +45%

The spreadsheet allows you to mix and match stats easily and tells you how it would have worked out in past years.

If you want numbers that aren't in there already you'll have to add them yourself but if anyone wants a copy of this spreadsheet just let me know (send me your email by PM). It's a little geeky but I think it's a fun tool to play with.
peeptoad
Posts: 2713
Joined: Thu Sep 12, 2013 6:53 pm

Fri Apr 22, 2016 6:03 pm

Too late for me on the long-winded post, but I'll take it. :)
Apollo
Posts: 294
Joined: Thu Sep 26, 2013 1:05 pm

Fri Apr 22, 2016 6:54 pm

Very good. I particularly appreciate that you didn't try to catch every winner. Not every outlier is expainable via normal variables that are likely to attach again. I'm seen so many backfit systems in sports and racing that make ridiculous weaves and turns while desperate to include every bizarre result.

I look forward to the other side of the coin. From my experience in sports betting and Excel data back to 1992, "bet with" systems are far more reliable than "bet against" angles. In fact, it's not remotely close. That's partially why I always have to laugh at posters here who announce some horse as a "bet against" target. Well isn't that wonderful? Do you collect if he doesn't win? That seems to be the assertion, with notable examples like Bodemeister. Hilarious. As a matchup bettor I have a clear path to profiting if my horse defeats the bet-against nominee. Otherwise it's majority blowhard noise.
Tessablue
Posts: 3506
Joined: Fri Sep 13, 2013 11:29 am
Location: Boston

Sun Apr 24, 2016 3:36 pm

Fantastic stuff as always, Kennedy! I really appreciate you taking the time to write all this up and share it with us, and it's got me thinking even more about Creator as a legitimate contender. Also felt some inspiration to go looking for other exclusionary angles, these are rough and mostly for fun but just in case you're interested:

1. Must improve on 2yo top Beyer by at least 3 points at 3: horses who did not qualify are 65-3-2-2

2. Final prep must be within 5 points of career-best: 99-2-3-3

3. Did not regress >3 points in first try at 9f: 123-3-3-4
(I excluded horses who had never run at 9f from this category, but just running through the PPs, it does look like a prior 9f start is absolutely crucial which has cooled me on Destin somewhat)

Horses who failed in at least one of these categories are 179-3-6-5 total. The winning exceptions are Street Sense, whose final prep was just impossible figure-wise, Barbaro, who had the turf angle complicating things, and Sea Hero, who as you noted was just a total anomaly. I'm also bringing this up because it echoes what your 20/20 found- Mohaymen and Nyquist both alarmingly fail in two categories, while Creator and Exaggerator are both safe. Food for thought and thanks again for sharing!
EquineAnne
Posts: 685
Joined: Mon Sep 23, 2013 7:22 pm

Sun Apr 24, 2016 4:41 pm

Somnambulist wrote:His name reminds me do the awful metal band. I can't do it.

Your research/stats are always really interesting. I respect the time that went into it. You've got to be one of the most objective people I have ever encountered on the Internet, if not the most. Thanks for sharing.
hear, hear!

Thank you, Kennedy, for taking the time to share. To this inexperienced fan, I find it really, really, interesting.
Post Reply