Dutching

[PaceAdvantage]

With the late moves on the tote after the race is off, is dutching all but dead? And was it ever really alive to begin with?

I had some spare time and wrote a dutching tool into my "AutoBet" program, but after a couple of test bets, I am trying to understand why I went through the limited trouble. If there are any "master-dutchers" out there, drop me a few hints as to when this might come in handy....you know, certain field size, certain odds distributions, etc. etc.

[the little guy]

I've sort of wondered that myself and have had the same feelings as you. I can say it was alive at one time, as I do know one of the all-time " master dutchers ", Speedking could confirm this, but it feels like the late money has to have seriously altered this. I guess it still exists as long as you're getting a substantial rebate.

I'll try an look into it for you. I am curious myself.

[Hosshead]

The simo dump has thrown a big wrench into dutching.
Net Pool Pricing - I may be wrong, but from the way I understand it, the win odds are still accurate, so hopefully that won't affect dutching. However the money (changed to dollars) that comes in from other countries because of Net Pool Pricing, might make the "after the bell" odds change, even more dramatic, albeit a larger total pool.
If that outside U.S. money could be regulated, to be put into the pools before post, thus enlarging the pool before post, it would take more money to change the odds after the bell.
But then I suppose (the computers of) the people who make the large last second bets, would increase their bets accordingly (to the total pool).

Now if you could work out a dutching method using ML odds, that works (in the long run) despite the simo dump..................

[cj]

Can someone post an example of dutching? I've been betting horses quite a while, but don't really know how to do this. I think it is used to bet multiple horses in the same race while attempting to get the same return if any of them win, correct?

[Pace Cap'n]

There is even a program for it...

smartbet.homestead.com/ (http://smartbet.homestead.com/)

Had this link for several years, never got around to trying it.

[timtam]

When I first started fiddling with different types of wagers I bought Earl

Thomas' book called Strategy. It gave various combinations on how to wager

on multiple horses and still bring back a profit. It was due column wagering

only using 3-5 horses. Man did I get my doors blown in. I almost lost my

shirt. It looked so easy in the examples :blush:

[bobbyb]

I use a Hedge app. and have had much better results with it than any other dutching tool/application.

Info Sheet can be found @:

http://www.dmtc.com/handicap/tools/wt_hedge.php

I have it on my primary at home and my PDA for live.
(PDA app. is Windows CE based)

bobbyb

[cj]

I have found a free one here (http://www.ukhorseracing.co.uk/tools/DutchingCalculator.asp).

It appears you have to add one to the odds to make it work.

For example for even money and 3 to 1, enter 2 and 4 respectively.

[sjk]

I look at both the win and exacta pools and have found that it is much more likely that a horse's price will hold up if there is value in both pools.

If a horse is 6-1 or less on the ML and is going off at a higher price with a few minutes to post but provides no value in the exacta pool, the win odds almost always come down (I only pay attention to horses I like; perhaps the ones that are really bad act differently).

On the other side, if the win odds on two contenders are not great value but the exacta is paying well there is a good likelihood that the exacta price comes down, especially if there are no real long prices involved.

In these circumstances I usually wait until 0 MTP before pulling the trigger and even then get burned after the break. I bet a horse yesterday at 7-1 at 1 MTP that crossed the wire at 7-2 (not first) and wondered if I had bet too early.

If a horse shows good value in both the win and exacta pool I can bet at 2-3 MTP and seldom have the odds do something drastically bad. Usually for me most of the value is in the exactas.

[Dave Schwartz]

CJ,

The simple definition is to divide 1 by the odds plus one for each horse, then normalize the amount of th ewager across said horses.

A= 3.00:1
B= 5.00:1
C= 2.50:1


A= 0.250
B= 0.167
C= 0.286

Sum=0.703

A's bet= 0.250 / .703 or 35.56% of the total wager.
B's bet= 0.167 / .703 or 23.75% of the total wager
C's bet= 0.286 / .703 or 40.68% of the total wager.

(You get to deal with the small rounding.)



A couple more hints:

Want to know what your "net odds" are? (Has nothing to do with "net pricing.")

NetOdds= (1.000 - Sum) / Sum
NetOdds= (1.000 - .703) / .703
NetOdds= ( 0.297 ) / .703
NetOdds= 0.42:1 or about 2/5.


Another valuable one:

Suppose you want to know how much of the pool is wagered on each horse:

Simply do the "booking percentages" of all the horses and sum them, then normalize the field to 100% to determine the pool pct wagered on each horse. It is not perfect but works very well. (3/1 = 25% is commonly referred to as the "booking pct.")

[toetoe]

This is a good lesson on value even if you're only betting one horse. Ex.: If you demand 100% juice, you can't take 2-to-1 on your 50% shot.

2-to-1/1-to-1 is faulty money-management math.

[(even + 1) * 2] - 1 = 3. 3-to-1 is what you would require. At big prices, the add-one gets almost inconsequentially small, but the point should be kept in mind.

[Hosshead]

........................DUTCHING TABLE.................
ODDS....%......... ODDS.....%........ODDS.......%

1-1...-..50...........5-2...-..29.........7-1...-...13
6-5...-..46...........3-1...-..25.........8-1...-...11
7-5...-..42...........7-2...-..22.........9-1...-...10
3-2...-..40...........4-1...-..20........10-1..-.....9
8-5...-..39...........9-2...-..18........11-1..-.....8
9-5...-..36...........5-1...-..16........12-1..-.....8
2-1...-..33...........6-1...-..14........13-1..-.....7

Add the % of each horse you play, deduct from 100% and (if you win), you have the (approx.) amount won.

[Handiman]

I don't know what to say..or how to say it. But since PA started this thread I will forwarn all readers...This is a mini advertisement. I will be glad to pay PA some money if anyone is interested in buying the program.


I have written a dutching program that allows the user to dutch up to 4 horses in a race. It gives the user opportunity to set his own level of risk and then creates bets based on his risk tolerance and bankroll.

It also displays optimum bet percentage based on odds ...independant of user risk level. Program only creates bets if their is group overlay.

Again....I realize this if self promotion, but it is in response to PA's thread. I have not included a link to site as I don't want to get into trouble with PA. I will respond accordingly to whatever direction I'm given by PA and level of interest by members.

Dutching is still a formidable way of betting, since the late money action will at times affect your bets adversly and at times affect your bets positively. In the end, it should balance out.

That's my take.

Handi

[Dave Schwartz]

Toetoe,

2-to-1/1-to-1 is faulty money-management math.

What does that mean?


Dave

[sjk]

Dave,

I don't see Toetoe around but I have always practiced what he suggests so I will reply. If you require a 100% overlay percent you need an $8.00 payoff to give you the needed value for a horse that you give a 50% chance to win. (I am currently using 80% so I require $7.20.)

I think this is in keeping with Kelly where bets are proportional to edge x (1/(1+[real odds of winning]).

I have been doing this for a long time so if this is wrong thinking please straighten me out.

[Dave Schwartz]

SJK,

Now I get it.

You're saying that you need a large overlay to make sure you are profitable.

I thought he was saying something about bad math.


As for my own personal comment on what you said, I see that "overlay margin" as a function of one's line accuracy. If your line is more accurate, you don't need so much margin.

Of course, verifying line accuracy is a science all unto itself.


Dave

[sjk]

Easy enough to verify a line's accuracy if you have the data and computer horsepower to do significant backtests.

Choosing the overlay percentage to use is a trade-off between ROI and total dollars won. You might win a few extra dollars by lowering you ROI requirement and making a lot of marginal plays to chase a few more dollars. I find it more comfortable to use a somewhat higher requirement to keep up the ROI and reduce some of the streakiness.

[classhandicapper]

This is an interesting thread because I bet 2 horses to win all the time, but I suspect I do it for different reasons.

Occasionally, I hate the favorite and want to get involved in a race. Sometimes, because of the complexities of the race, I feel I have almost no ability to make an accurate odds line on the INDIVIDUAL contenders I might be interested in. In those situations, even though I can't make an odds line on the individuals, I can usually quantify the group's chances of winning vs the return I wouldl get if I bet them all. If the group (2-3 horses) will produce profits, I then simply dutch them knowing full well I have some combination of underlays and overlays that nets out slightly positive.

However, if I can guage the chances of the individual horses, I NEVER bet more than one horse unless they are both overlays.

Why include an underlay just to increase your win percentage?