Dutching Win Bets

[Dave Schwartz]

About Kelly...
For the Kelly Criterion to be correctly applied one must know two things:

1. Hit rate
2. Payoff

In horse racing we do not actually have either of those.

Sure, assuming we have a large sample (which none of us have) we can know the hit rate on (say) all horses we have bet at 7/1. But the truth is that the hit rate on a given horse is a pure guess relative to the entire sample of 7/1 horses.

The payoff is absolutely a moving target.

Now, there is a way to improve that but it still isn't as perfect as proponents of Kelly would have you believe.

That would be to play "betting sessions" and track the session win percentage and payoffs.

Example: I play a $300 session. The session is won when I am up $300 or more and the session is lost when I hit a downturn of $300.

My session hit rate is down a little to 69%. When I win a session, my average return is $681 (profit of $381, or 1.27/1).

When I lose, my average loss is $281.

Therefore, my average risk is $281 (as opposed to $300) and my actual odds are $381/281 = 1.356/1

This is equivalent to hitting 69% with a $4.71 payoff.
(A 62.5% advantage)

Since Opt Bet = Advantage / Odds my optimum per session would be:

Opt%=62.5 / 1.356 = 46% of bankroll.
A laughable percentage.

But it is why I am hard to beat.

[BarchCapper]

Originally Posted by BCOURTNEY

This is an excellent point, time horizons are important as well. For example a player might not want to play a 30000-1 pick 6 ticket with a positive EV due to the fact the combination likelihood of being physically dead before the wager realizes itself. In fact one can argue if the player is reinvesting winnings, there is a time horizon versus value to compounding returns which is not evident nor accounted for in most strategies.

[AltonKelsey]

Originally Posted by BCOURTNEY

This is an excellent point, time horizons are important as well. F....

Glad someone appreciates the nuance.

[ubercapper]

Originally Posted by InsideThePylons-MW

You gave both horses the exact same 25% chance of winning.

If you think that betting $11 to win on the 7-1 and $9 to win on the 9-1 is optimal then just quit now.

Anybody with one ounce of wagering common sense knows that $20 to win on the 9-1 will outperform the above in the long run based on your opinion.

Go ahead and keep betting the most on the horse with the least edge.

The point, from the OP on, was to double check the discussion you and I had on Twitter and your opinion on dutching.

That Twitter discussion prompted me to start a topic here on PA which I thought would involve a number of people with thoughtful opinions, and it did.

I was open to being corrected for recommending dutching two horses to win. I was not, and it further appears the opinions on the topic in this thread put your opinion in the minority.

You are entitled to your opinion, but insulting someone because you don't agree with that opinion is not appropriate. I'd respectfully ask you to refrain from statements like "just quit now" and "anyone with one ounce of wagering common sense" as they do not add to an intelligent discussion of this, or any, topic.

Thankfully, some pretty intelligent people like Tony@HSH, Blake, Thaskalos, Tom, Nitro, CJ and others I left out, believe proportional betting is a totally acceptable approach. If you want to insult all of them as well be my guest.

[Robert Fischer]

Originally Posted by ubercapper

The point, from the OP on, was to double check the discussion you and I had on Twitter and your opinion on dutching.

That Twitter discussion prompted me to start a topic here on PA which I thought would involve a number of people with thoughtful opinions, and it did.

I was open to being corrected for recommending dutching two horses to win. I was not, and it further appears the opinions on the topic in this thread put your opinion in the minority.

You are entitled to your opinion, but insulting someone because you don't agree with that opinion is not appropriate. I'd respectfully ask you to refrain from statements like "just quit now" and "anyone with one ounce of wagering common sense" as they do not add to an intelligent discussion of this, or any, topic.

Thankfully, some pretty intelligent people like Tony@HSH, Blake, Thaskalos, Tom, Nitro, CJ and others I left out, believe proportional betting is a totally acceptable approach. If you want to insult all of them as well be my guest.

Enjoy reading both of your opinions.



say I have a False Favorite 'C', and two Alternatives 'A&B', and both A and B have a nice margin of safety, and..., I can't definitively say A >(, or <) B in terms of value, then I will project the odds of A,B and use that relative proportion to 'fudge' a 'probability estimate' while giving myself a moderate edge (say 5% for each horse to leave room for margin-of-safety).
That's how I want to come up with my probability in this case.

I am not smart enough to know the exact percentages.

I'm not going to bother deviating from the projected tote odds at that point.

If I had a special insight(not 'priced-in') on A or B, I wouldn't be dutching them both, (I'd only be betting the special insight). All my value in this case, is coming from the fact that the favorite is false, and that I've narrowed it down to two contenders.

[ubercapper]

Originally Posted by Robert Fischer

Enjoy reading both of your opinions.



say I have a False Favorite 'C', and two Alternatives 'A&B', and both A and B have a nice margin of safety, and..., I can't definitively say A >(, or <) B in terms of value, then I will project the odds of A,B and use that relative proportion to 'fudge' a 'probability estimate' while giving myself a moderate edge (say 5% for each horse to leave room for margin-of-safety).
That's how I want to come up with my probability in this case.

I am not smart enough to know the exact percentages.

I'm not going to bother deviating from the projected tote odds at that point.

If I had a special insight(not 'priced-in') on A or B, I wouldn't be dutching them both, (I'd only be betting the special insight). All my value in this case, is coming from the fact that the favorite is false, and that I've narrowed it down to two contenders.

I appreciate the way you put that, particularly the last sentence about the value coming from the favorite considered false.

It happened again in the Pegasus Turf. On one of the blogs I wrote Magic Wand (who I presumed to go favored) was vulnerable as she rarely seemed to get there evidenced by having a lot of second place finishes vs wins. I also did not mention Without Parole or Mo Forza. All three went off at 3/1 to 9/2 as the three top public choices.

I recommended dutching Instilled Regard and Zulu Alpha at 3/1 or more. They went off at 11/1 & 10/1. Zulu Alpha paid $24.

There are those who would argue betting one of the two was (and is) the only correct way to bet. There are those who recognize proportion betting may not have the mathematical edge of the former approach but it offers other positives such as psychological as it cuts down on variance, which, especially for many of the people who read my opinions and who only wager occasionally, is far more important to them.

[Robert Fischer]

Great call, in the Pegasus Turf.


With a couple of 10/1 or so shots, preferred over a block of false favorites, you can't go wrong.

[cutchemist42]

I used to read the brisnet library slot back when I was starting and they had a few articles related to 2 win betting when identifying false favorites.

As a Canadian I used to have Betfair access and it was definitely easier to manage the 2 win bet on an exchange with known odds.

It seemed to have merit to me but not truly sure what the math truly says.

[ultracapper]

I use the place pool, but it's the same idea as betting 2 horses to win. I use the place pool to keep cash flow disruption at a lower rate. I've looked into backing up my choice by also betting my 2nd choice, but I've had better results with the place back up bet. I know I'm looking at betting 2 horses to win from a different perspective than the OP, but it is another way to look at it.

[castaway01]

Originally Posted by ultracapper

I use the place pool, but it's the same idea as betting 2 horses to win. I use the place pool to keep cash flow disruption at a lower rate. I've looked into backing up my choice by also betting my 2nd choice, but I've had better results with the place back up bet. I know I'm looking at betting 2 horses to win from a different perspective than the OP, but it is another way to look at it.

Are the place pools big enough these days to get any leverage? I would think in order to make any real money you'd have to bet so much that you'd kill your own price, but I'm far from an expert on betting to place. If it's working for you then it's working for you.

[ultracapper]

Originally Posted by castaway01

Are the place pools big enough these days to get any leverage? I would think in order to make any real money you'd have to bet so much that you'd kill your own price, but I'm far from an expert on betting to place. If it's working for you then it's working for you.

I'm not using the place pool to win money. It's to get my investment back if I get beat. Since this thread was started, I've been looking at the possibility of taking the money that I would put in the place pool, and using it to cover secondary opinions in the win pool. It's still "back up" money, but handled differently.

The OP is betting 2 horses in a weighted manner to make money. I'd be betting one horse in the win pool to make money, and one, or maybe even 2 others, to get my investment back if my choice failed. IDK, betting 2 horses in the win pool has always been counter to my nature. I probably should just let it go.

[ubercapper]

Originally Posted by ultracapper

I use the place pool, but it's the same idea as betting 2 horses to win. I use the place pool to keep cash flow disruption at a lower rate. I've looked into backing up my choice by also betting my 2nd choice, but I've had better results with the place back up bet. I know I'm looking at betting 2 horses to win from a different perspective than the OP, but it is another way to look at it.

"Keeping cash flow disruption at a lower rate" is very descriptive when used in the context of wagering, whether it be by dutching win bets, betting to win and place, boxing exactas or utilizing other approaches which attempt to make a profit without as much concern for maximizing profit.

We are all, at times, and some people more than others, risk tolerant, risk averse or somewhere in between.

I would not argue with someone who is entirely risk tolerant and who states win betting only is the best long term approach to maximize profit. Personally, I feel those bettors are in the minority although I would not try to estimate what percentage of the total they represent.

IMO the majority of fans & bettors fall into the camp of being a bit risk averse and as long as they are aware some wagers cut into overall profit in order to make a profit that's fine.

[vegasone]

“In horse racing we do not actually have either of those.
Sure, assuming we have a large sample (which none of us have) we can know the hit rate on (say) all horses we have bet at 7/1. But the truth is that the hit rate on a given horse is a pure guess relative to the entire sample of 7/1 horses.”


Having a large sample is not as much an issue as the fact that a 7/1 now is not the same as a 7/1 ten or twenty years ago. Just makes things harder to analyze all the time.

[Nitro]

Originally Posted by Nitro

The use of these odds generated by the Win pool in each race could certainly be applied differently by those viewing them. However, the bottom line especially when Dutching is the odds value being applied to the selected entries. I generally view the odds only as a value barometer to determine playability no matter which type of bet I want to make.

You might have misunderstood my reference to statistical “probability”. I have no use for these sorts of things.
I prefer to rely on my personal on-going betting records as far as determining the probability of how well the analysis does with specific race conditions. My confidence level is based on how well I’ve done in the past (hit frequency %) with certain races. It's NOT on how well the odds on a tote board have proved to be right or wrong in terms of determining the probability of winning based on just a specific odds value. As far as I'm concerned there are huge holes in that way of thinking and its not a very successful approach to this game.

Originally Posted by vegasone

“In horse racing we do not actually have either of those.
Sure, assuming we have a large sample (which none of us have) we can know the hit rate on (say) all horses we have bet at 7/1. But the truth is that the hit rate on a given horse is a pure guess relative to the entire sample of 7/1 horses.”

That’s the point! Some of us do!
Having a decent sized sample of what you’re doing with a solid Dutching scheme provides you with the confidence to continue making a decent profit.

These are the typical results (in terms of the O.A. Profit margins) that I’ve been experiencing over the last 5 years at Hong Kong when playing a 3-entry Dutch Win bet:
http://www.paceadvantage.com/forum/s...G+KONG&page=53
.
.

[castaway01]

Originally Posted by ultracapper

I'm not using the place pool to win money. It's to get my investment back if I get beat. Since this thread was started, I've been looking at the possibility of taking the money that I would put in the place pool, and using it to cover secondary opinions in the win pool. It's still "back up" money, but handled differently.

The OP is betting 2 horses in a weighted manner to make money. I'd be betting one horse in the win pool to make money, and one, or maybe even 2 others, to get my investment back if my choice failed. IDK, betting 2 horses in the win pool has always been counter to my nature. I probably should just let it go.

Got it.