Have a look at this chart/system from American Turf Monthly

[AutumnLotus]

In searching for the contents within the September 1964 issue of ATF, a Ray Taulbot article regarding his Dutching / hedging system when a handicapper frequently narrows the race down to three likely contenders, I came across a link from a short thread within the deep pages of this very message board, the link and brief description was posted sometime in 2011 by one of our very own frequent pace advantage users.

I am hoping to find the article in its entirety, I emailed American turf monthly yesterday so hopefully they can help further.

In the interim, the gentleman who posted the link provided the meat of the system Taulbot created.

I had to take a screenshot of it, refer to the image, give it a look over, would love to hear your opinions.

The dutching program I use calculates the same thing once a target wager is entered. Handy little chart to have nonetheless.


I have become very interested in dutching/hedging as of late, and have been working relentlessly on something that I do not yet want to share, but so far the results have been incredible, and I look forward to sharing it!
BELOW IS ABOUT MY DIVE INTO DUTCHING/HEDGING BEFORE COMING ACROSS THIS SYSTEM - i.e., screenshot

Some of you may have seen recently I posted a thread about the best way to approach three contender selections. I realize it is a loaded question that can spin off in many different opinions on the matter and for the more well versed of you, I am also aware that to give a proper opinion on the matter solid statistical data is needed to answer such questions such as what is the percentage overall that one of the three contenders win/ as well as what is the percentage that two of the three come in enough times to warrant betting exactas/ and so on.

To add some perspective about my situation prior to laying my questions out I will tell you that at the moment I have Just a couple of days under a full two months of back data/ back checking over two tracks (Penn National and Charles Town), Of which by this evening I will be up to date move / March. My situation is threefold:

1. To get a dependable percentage regarding one of the three for the WIN SPOT.

2. To record the true odds of my three selections which will then be scrutinized, painstakingly I might add, getting a clear and concise picture of the spread between the odds / 3 selections, And Inputting those odds into a three horse dutching / hedging program I have, Of which I will then record and analyze the average spread between the odds in an effort to find and record the “sweet spot” and make a chart relative to the information so that I do not have to continue going back and forth between my printouts and hedging program, etc.
The sweet spot/chart is to reveal the average fluctuations/drop/rise of the odds up to the very last moment that my wager would be accepted.
*So far, I have a standard odds chart divided into sections based on what I am uncovering with the odds changing at the last moment, and for the most part it is working. During live testing of this, the odds have not went below my guestimation when odds are FINALIZED/TRU ODDS. In fact, I am being more than conservative and going up to 4 STEPS (steps not groups) LESS than the real final odds, so that any money wagered in the near future, granted will be less profit, will however be covered.

For example, of my 3 selections there is usually a 1st or 2nd favorite that unless super chalky I immediately assume 3/2. If the odds are 2/1 5 min before Post Time I enter it into the program as 3/2. Unless hedging I do not go lower. Then regarding the other 2 selections, obviously one wants a good spread in odds between them, and high odds for a good payoff, I adjust accordingly.

The groups are:
1/9 to 4/5 Not that I would play within this group, just posting how I separated the odds based on my research.
1/1 to 7/5
3/2 to 2/1
5/2 to 4/1
9/2 to 7/1
9/1 to 12/1
13/1 & UP

3. And the last thing that I intend to do once I have this data up to the most current races Of this month is to go over and above with a fine comb all the info about my three selections in its entirety and try to make a set of rules and guidelines to follow and that will of course be scrutinized perhaps every two weeks to a month to make sure the rules and guidelines continue to work. Some of the guidelines I intend to uncover to make things easier will be for example one of the guidelines that I use already in testing is if one of my selections is lower than 5 to 2 and or Has the lowest odds between the other two in the vicinity of 3/1 and lower as post-time approaches I automatically enter 3/2 odds. It sucks to say this but with the small fields and depressing payoffs (I'm looking at you turf paradise. Ridiculously elementary) The majority of my data shows a large portion of the lower odds when one of the three selections are included it is best to make it a rule and enter it as 3 / 2 odds and depending on the odds of my other two selections I will either hedge the 3/ 2 selection or not.

My overall agenda is to reveal via the data a near concrete way of confidently and safely betting on three selections - one for the win spot, Alongside making the one of three selection win spot wager as concretely positive as possible my parallel agenda as a zed before is to Find the sweet spot within the spectrum of fluctuating odds between my selections and create a ready fast chart / table to not only quickly and accurately calculate whether or not to play a race but instead I want to base that decision upon setting a daily profit target.

[P.Rosa]

From your thumbnail,
"Make sure that your top three contenders win at a clip of 75 to 80%."

I don't know anyone that is that good.

[Poindexter]

The typical way dutching works(and the chart that you posted does the same with some rounding off) is you use the odds (best estimate of what the final odds will be-probably should be using daily double probables at least as a big input into what that estimation is).

So if your 3 top choices are going to go off at

3-1, 4-1 and 6-1, you have to determine how much you need to bet to get back $100. In this instance 25, 20 and 14 (to determine the amount the formula is (1/(odds+1))*100 ) would have a cost of $59. So the math involved is you either lose 59 if all 3 horses lose or win 41 if one wins (there is some round off error-14 to in on the 6-1 only comes back 98 if it pays $14.00 even). Since that dutch would cost you $59, the break even point for dutches with these 3 exact odds is 59%. It your horses are 8/5, 2-1, 4-1, then you would have to bet 38, 33, 20 or $91 to get back 100 and would need a 91% hit rate to break even........

My contention is that you can't say that because you hit 72% of your 3 horse dutches that if you play any dutch that costs you maybe $65 or less or $60 or less, whatever you decide on that you can insure a profit. Logic tells you that horses with higher odds win less often and you will have to figure out at which point you can dutch the 3 horses and make a profit. Thus you have to analyze the data you have. Also just because you might make money with one 2 month sample doesn't mean the next two month sample will be as generous.

To analyze your data you are probably going to have to type in the cost of each 3 horse dutch and the profit or loss. Sort them by cost and figure out if there is any cost that makes you profitable (under $70, under $60, under $55, under $50......). You are probably going to have to go by your odds estimates as the cost and actual profit and loss on each dutch. Obviously your goal is to find that number that seems to be profitable and accumulate more data and verify that is indeed the right number long term as your sample grows larger.

[MJC922]

Finding three overlays with the field sizes we typically see today is going to be a challenge. It can happen but for it to occur there almost has to be a horse in the race (typically if not always it would have to be the actual race favorite) which is extremely overbet in your estimation.

If you wanted to do this a lot more often like as an every race strategy then maybe HK racing with the larger field sizes might be the way to go. One of the three can certainly be a break-even horse that just boosts the win% but still in 6 horse fields let's face it you're typically going to be spinning your wheels or cutting the profits from your largest overlay horse by using more than a single vs the takeout.

I like the idea of dutching and I guess you could say that I do something somewhat similar but with doubles. It has to be a specific scenario to work though, as stated above you need an overbet favorite to lay against. When you've identified that, spread, then be more direct in the other leg.

[Saratoga]

Originally Posted by P.Rosa

From your thumbnail,
"Make sure that your top three contenders win at a clip of 75 to 80%."

I don't know anyone that is that good.

Hit it on the nail....

With MLO and fields of 8+ this year

you can barely hit 60%

[Jake]

Originally Posted by Saratoga

Hit it on the nail....

With MLO and fields of 8+ this year

you can barely hit 60%

Dave Schwartz, do you want to jump into this discussion? I know
you have have systems to do this kind of hedging. I will say that I
can certainly do it, systematically with considerable margins above
60%. However, I don't use a hedging approach. Just play my top 3
in Pk'3 or doubles, which beats hedging in my long experience.

I don't need the consistency of hedging. I do like the ROI of top 3
contenders in a single leg of horizontal betting. Different strokes for
different preference.

Jake

[Saratoga]

This is HTR's famous K Rating for 2023...


I guess you can get close to those numbers ..if your looking at

fields of 7 and less .....which is not mentioned

Pic 1 ....8+

Pic 2.....7 and less

[lefty359]

It's hard to dutch these days because odds fluctuate wildly after horses get in gate and are running.
Lately some of my winners went in gate and odds dropped to almost nothing.
If those odds have been revealed before the race I wouldn't have played them.


But I certainly wish you good luck and suggest you play on paper until you're
sure you can be profitable.

[Jake]

Thank you. I have spent tons of time getting to know Massa's stuff.

I'm going to suggest that anyone can simply look at the relative win probabilities of the top 5 or 6 ML horses, and see what the cumulative joint probabilities tell them. Just using your particular handicapping method to remove the weakest 2 or 3 weakest horses will get you to 60% or higher in many races. Not to steal the hedge focus of this thread, I just find actually playing those remaining horses in horizontal bets pays more.

When the ML favorite is bet down, its win odds move close to 40% and up, and as Lefty bemoans, it's hard to know where the final odds end up. That makes hedging more difficult. I don't mind, because I know roughly 60% of the time that horse will lose, and I have 2 other horses in that betting leg to keep the Pk3 now alive at even higher payoff odds. There are other reasons for my preference for horizontal betting over vertical hedging in a single race, but that's the main one I try to exploit in my handicapping.

I'm impressed with anyone that makes hedging works long term. That's why I suggest Dave Schwartz speak up here. He has that kind of expertise in his HSH software, and no doubt in his new forthcoming AI software. Not a plug for his software, I just know he has spoke about this approach on his site.

Jake

[hracingplyr]

I have a Crammer Research report that talks about significant odds changes in certain situations. Drop me a line at robertj1949@gmail.com for details.

Bob

[Dave Schwartz]

Originally Posted by Jake

Dave Schwartz, do you want to jump into this discussion? I know
you have have systems to do this kind of hedging. I will say that I
can certainly do it, systematically with considerable margins above
60%. However, I don't use a hedging approach. Just play my top 3
in Pk'3 or doubles, which beats hedging in my long experience.

I don't need the consistency of hedging. I do like the ROI of top 3
contenders in a single leg of horizontal betting. Different strokes for
different preference.

Jake

Thank you for the kind words, Jake.

Yes, I've done a lot but the impact of the Whales has taken everything in a different direction.

The biggest challenge is that the GATE ODDS (what I call them) are completely unreliable to bet into --- except for the whales, that is.

(For my shortest possible answer to this, just scroll to the bottom.)

Short version of how people handicap:
The Logical Steps
Step 0: Gather Data

This is the mark up your DRF step when you build factors that are important to you.
(Works with or without colored pens. LOL)

Step 1: Pick Contenders

Most people only do the following steps for their contenders.

Step 2: Estimate Chances

Think "I make this horse 3/1," meaning I think he's got about a 25% chance of winning.

Step 3: Estimate Odds/Consider Tote Board

This is usually done early but then adjusted when horses are in the gate.

Step 4: Determine Value

This is the point when they consider the combination of steps 2 & 3 to come up with who the VALUE PLAYS are.

Step 5: THE BETTING DECISION

Of course, this is where the rubber meets the road.

Step 6: WHALES START HERE!

At this point, we've made our decisions based upon our perception of the hit rates and the odds.

In HSH we expressed the "value proposition" as a letter grade, which went from A+ down to F, where an A+ projected to be at least a $2.40 $net (i.e. +20% edge).

A- was break-even.

What does, "Whales start here" mean?
It means that their work BEGINS where ours ends.

Imagine I have an A+ horse who is also the top hit rate horse.

Do I really believe that I am the only one who can see that this horse is a good bet AND a likely candidate for the winner's circle?

This is when we have a tendency to believe that our way of picking is just soooo unique.

What we conveniently forget is that even if we are the smartest handicappers on the planet, at least a few other people (with their obviously lesser methods) are betting the same horse.

THE POINT: The Whales beat us because they've made a business from adding step #6: CONSIDERING THE 0 THRU 5 STEPS and ADDING HOW THE TOTE WILL REACT TO THOSE STEPS!

Step 7: ANTICIPATE WHO THE WHALES WILL BET AND ADJUST

This is the goal of my AI, but you can mimic this on your own.
(How to do that is a deep topic and outside the scope of this post.)

Think of it like a chess game.
A simple player looks at the current state of the board and asks himself, "What's the best move?"

A master player, asks himself, "What move will my opponent likely make?"

A Grand Master, mentally studies the most likely moves and his opponent's likely responses.

Then he follows the path of each of those "threads" searching for his best course of action.

The shortest possible answer I can think of.

Almost 75% of all winners are going to be bet down from their GATE ODDS.

If you are a two-horse bettor, and you are wagering some simple approach like 60/40% on lower odds vs higher odds, that means that over 2/3s of the time you would be better off betting 40/60.

With 3 horses that gets way more complex.

I have tested 16 ways from Sunday and determined that the Gate Odds are actually WORSE than using a GOOD SET of PROJECTED odds. (HSH has those and I've shared the factors that we use many times.)

The Morning Line is worthless for this, but shockingly, a tiny bit better than the Gate Odds.

The key point of misunderstanding is that any form of ODDS PROJECTION is NOT to be measured by what the final odds WILL be.

Projected odds represent what the final odds SHOULD be.

Gate Odds WILL have value with that.




.

[Jake]

Great reply, Dave. Tough game.

Jake

[ranchwest]

If you like dutching and catch a reasonable number of non-favorites, I suggest you consider making a portion of your wager a flat bet, research 25% flat and 75% dutch.