trifectaMikes pure oddsline formula

[douglasw32]

This may have nothing to do with this since I did not read the other post, but I rank my horses, then take the morning line and from lowest to highest give the odds the ML has to my horses. so if the fav on the ml is 9/5 my top ranked gets the 9/5 and on and on. maybe that is a dumb way to do it ? the only thing I have trouble with is how high the public makes a horse where I say...okay I have something wrong here and it is not an overlay but a dumb bet... like I have one at 5-1 and it is 30-1

[MPRanger]

Originally Posted by TrifectaMike

Can you suggest an alternate formulation, so that we may test it against empirical data.

I'll award you a $500 check for such a formulation, Of course a formulation that has to be universal.

Mike (Dr Beav)

Hi Trifectamike,

Certainly. Now, this isn't the way I usually do it but you asked for a
universal formulation. This way is much better than your tables.

1.) Start with the morning line.

2.) convert the fractional odds into decimal values.
1-1=50%, 2-1=33.33%, 3-1=25% etc.
Take the track win pool takeout and multiply by .01
ex; GP takeout = 17% 17*.01=.17

3.) Subtract .17 from 1 1-.17=.83

4.) Multiply .83 times the decimal values ofthe ML odds
1-1=50 50*.83=41.5, 2-1=33.33 33.33*.83=27.69

Now we have stripped the takeout from the odds.

5.) Use the decimal values to make adjustments between the
horses where you disagree with the ML oddsmaker.

As you know the oddsmaker is supposed to be predicting the
publics behavior so he's just keying the obvious. Theoretically,
he could have an opinion different from his line. But you are
a handicapper so you are looking for a place where he and the
public will be wrong. You make your adjustments on these points.

6.) Convert your new numbers back into fractional odds.

Now you can watch the toteboard for overlay opportunities in the
win pool.

It's handy to use a booking table to convert between fractional
and decimal values.

There is a universal way to make an accurate oddsline.

[TrifectaMike]

Originally Posted by MPRanger

Hi Trifectamike,

Certainly. Now, this isn't the way I usually do it but you asked for a
universal formulation. This way is much better than your tables.

1.) Start with the morning line.

2.) convert the fractional odds into decimal values.
1-1=50%, 2-1=33.33%, 3-1=25% etc.
Take the track win pool takeout and multiply by .01
ex; GP takeout = 17% 17*.01=.17

3.) Subtract .17 from 1 1-.17=.83

4.) Multiply .83 times the decimal values ofthe ML odds
1-1=50 50*.83=41.5, 2-1=33.33 33.33*.83=27.69

Now we have stripped the takeout from the odds.

5.) Use the decimal values to make adjustments between the
horses where you disagree with the ML oddsmaker.

As you know the oddsmaker is supposed to be predicting the
publics behavior so he's just keying the obvious. Theoretically,
he could have an opinion different from his line. But you are
a handicapper so you are looking for a place where he and the
public will be wrong. You make your adjustments on these points.

6.) Convert your new numbers back into fractional odds.

Now you can watch the toteboard for overlay opportunities in the
win pool.

It's handy to use a booking table to convert between fractional
and decimal values.

There is a universal way to make an accurate oddsline.

I'm going to get some poor soul to apply for food stamps and send them to you. You are going to need them based on what you have described.

Mike (Dr Beav)

[MPRanger]

Originally Posted by TrifectaMike

The crap you offer REALLY stinks!

Mike (Dr Beav)

Doe this mean I'm not gonna get the $500.00?

[maddog42]

Originally Posted by MPRanger

This is absolute meaningless nonsense.

Is anyone actually going, "Ok, there's 8 horses
in the field . So, after I rank them in order, I'm
going to assign these pre determined values
to them"? Hmmmm...? Is that what this is
supposed to be? Hahahaha HaHaHahahah

I posted the results of a very clever formula, that trifectamike came up with. This is a mathematical representation of an oddsline based on horses rank. No one, no way, no how said it was to be used to bet. If you think that it is easy to come up with such a formula, my hat is off to you, but I think you are mistaken. I posted these results to get people INTERESTED in making an oddsline. It is a good starting point.
I personally like the betting line templates that Steve Fierro devised in The Four Quarters of Horse Investing. I HATE horse racing books with Investing,
or Wall Street or similar titles, that make this game seem like investing in CD's.
So I was prepared to hate this one too, but it is very good.

[stu]

When I did the morning for a couple of tracks in the Mountain time zone I used a set of templates for each field size. I constrained myself with a couple of rules to limit the number of templates.

My rules:

• The First Three Favorites Have Different Odds
• Favorite Is No Lower Than 9:5 and No Higher Than 3:1
• The Longest Shot Is No Lower Than 12:1 and No Higher than 20:1
• All Percentages Total to +/- 1% of 100 plus the field size plus the take out

I know that the rules are arbitrary but they were meant to reduce the options for the purpose of saving time.

For 7-horse fields I had 14 templates from which I would select what seemed to match the race. First, I would guess what horse that the crowd would make the highest price and the corresponding price. Then I guess how low the favorite would be.

e.g. If I thought that the bomb would be 15-1 and the favorite 9-5 I would then chose from the following templates:


7 HORSE FIELDS - (9:5 5:2 7:2 x x x 15:1)

9:5 5:2 7:2 9:2 10:1 15:1 15:1

9:5 5:2 7:2 9:2 12:1 12:1 15:1

9:5 5:2 7:2 5:1 10:1 12:1 15:1

9:5 5:2 7:2 6:1 8:1 12:1 15:1

Each race was different so I would try to best match the race to one of the four templates. The templates were designed to keep me from perpetually redoing the math everyday to add up the percentages.

I had 15 pages of templates within which I would work (never just one per field size). If the race had an outlier that deserved heavy favoritism or serious shunning I would abandon the templates and add up the percentages to accommodate a 4:5 or 50:1.

[proximity]

Originally Posted by MPRanger

Doe this mean I'm not gonna get the $500.00?

at this point i'd just be happy with the food stamps and the equation......

[MPRanger]

Originally Posted by proximity

at this point i'd just be happy with the food stamps and the equation......

Why? Do you think your unhinged friend has made some point?
Do bluster and name calling pass for premises and conclusions
here? I'm not impressed yet.

Actually, since you and him are so happy with that method.
Then I am happy for you both.

[MPRanger]

Originally Posted by maddog42

I posted the results of a very clever formula, that trifectamike came up with. This is a mathematical representation of an oddsline based on horses rank. No one, no way, no how said it was to be used to bet. If you think that it is easy to come up with such a formula, my hat is off to you, but I think you are mistaken. I posted these results to get people INTERESTED in making an oddsline. It is a good starting point.
I personally like the betting line templates that Steve Fierro devised in The Four Quarters of Horse Investing. I HATE horse racing books with Investing,
or Wall Street or similar titles, that make this game seem like investing in CD's.
So I was prepared to hate this one too, but it is very good.

Thanks for the explanation.
I can see its usefullness.

[thaskalos]

Originally Posted by MPRanger

Thanks for the explanation.
I can see its usefullness.

Be careful, my friend...

I was an infrequent poster once too, and then I got into an argument similar to the one you have gotten into here...and I have been addicted to posting here ever since...

I would hate to see this happen to you too...

[maddog42]

Originally Posted by thaskalos

Be careful, my friend...

I was an infrequent poster once too, and then I got into an argument similar to the one you have gotten into here...and I have been addicted to posting here ever since...

I would hate to see this happen to you too...

He's right mpranger, you just fell into our evil trap. Brouhahaha!!! Brouhahaha!!!Brouhahaha!!!Brouhahaha!!!Brouhahaha! !!Brouhahaha!!!Brouhahaha!!!Brouhahaha!!!Brouhahah a!!!Brouhahaha!!! Now you must post 5,000 posts and we MIGHT let you go. This place is like Heroin and internet porn. The first ones are always free. Brouhahaha!!!Brouhahaha!!!Brouhahaha!!!Brouhahaha! !!

[Dave Schwartz]

There is a universal way to make an accurate oddsline.

I will venture a small comment:

IMHO, if one were to wager using Ranger's line, the underlays would return more than the overlays.

That is what happens when you use "How the Public Usually Bets" to make the line and bet into "How the Public Actually Bet." I know - I tested it.

The opposite is actually a pretty good approach: "Take how the public DID bet and use THAT to bet (backwards) into what the odds should have been.

Let that sink in and you will see a methodology that actually has promise.

Regards,
Dave Schwartz

[proximity]

do you mean ranger or douglasw32 ?? i think ranger was making race specific adjustments in step 5 of his process.....

[TrifectaMike]

Originally Posted by maddog42

I posted the results of a very clever formula, that trifectamike came up with. This is a mathematical representation of an oddsline based on horses rank. No one, no way, no how said it was to be used to bet. If you think that it is easy to come up with such a formula, my hat is off to you, but I think you are mistaken.

Why use mathematical equations (formulas)?

Mathematical equations (formulas) contain information in densely packed form. That is the single most important reason why data is subjected to the process of formulation. Just imagine having to describe processes and experimental data by pages and pages of derived and raw data. Not only would would careful study of the data be tedious and unlikely, but data trends would be most difficult to discern.

Mathematical equations can tell us at a glance the efect of one variable on another. How much of an increase or decrease of a variable is impotant or not? What is the shape of the increase or decrease?

Mathematical equations remove undesired variations from data ( noise or artifacts ). which can range from strictly random events to systematic effects.

Mathematical equations can be used to derive theoretical implications concerning the underlying principles relating variables to one another. Is one variable related to another in a linear, exponential, periodic or some combination of forms? Each of these forms carries with it a fundamental notion concerning the connection between one variable to another. There is confidence that the relationship means MORE than just a blind attempt at describing data, and that interpolation or extrapolation can be obtained without the expectaion of too much error.

Mike (Dr Beav)

[classhandicapper]

I think everyone understands that these probabilities are not meant to be automatic formulas for creating an odds line in various size fields. However, if they are reasonably accurate on "AVERAGE" in various field sizes, IMO they are a pretty good guideline for starting the process of making an odds line. You would then simply tweak them to reflect the actual horses in the race. That's better than a totally arbitrary process.