HOW MUCH OVERLAY IS ENOUGH?

[captvancouver]

Thanks to the direction of some of the more experienced members here, I am starting to recognize the importance of overlays. I am at the point that I do assess the chances of each horse in the program, and compare that to the tote.I think I am getting a lot better at it . How much overlay should there be before I bet? Are we talking 4:1 here. I realize the answer may vary according to the accuracy of the odds that the bettor establishes.Any pointers here? I suppose I could just keep records and get some sort of an average.

[cj]

I use 20% for my top pick, 40% for my second pick, 60% for my third pick, and so on.

[46zilzal]

Mitchell (a champion of the overlay) used to always say that betting on John Henry to place and show was his BREAD and BUTTER since given his probability of placing they were ALWAYS overlays..

I have a system called "two lines better than the field" that is hitting over 80% winners and over 95% at least place. Odds are low (8/5 to 3/1) and these represent only about 5 bets per week since these are consistent, good horses but that is a probabilty of 0.8 so MOST are overlays SO ONE CAN SAFELY unload ON THEM so they are still overlays

[MichaelNunamaker]

Hi 46zilzal,

You wrote "I have a system called "two lines better than the field" that is hitting over 80% winners and over 95% at least place."

That's amazing. My odds line only picks about five horses a year to have an 80% chance of winning, and they generally go off at very very low odds. Any hints on what you are doing?

Mike Nunamaker

[xfile]

Thanks to the direction of some of the more experienced members here, I am starting to recognize the importance of overlays. I am at the point that I do assess the chances of each horse in the program, and compare that to the tote.I think I am getting a lot better at it . How much overlay should there be before I bet? Are we talking 4:1 here. I realize the answer may vary according to the accuracy of the odds that the bettor establishes.Any pointers here? I suppose I could just keep records and get some sort of an average.

What works well for me over the last few years is playing overlays with my oddsline up to 6-1 but are going off at 8-1 or higher. I won't touch anything less than 8-1. My win% is low but who cares. My roi is there. In certain situations I will dutch but mostly one overlay to win. At times I also wheel the overlay in the exacta if I see tremendous payouts with my horse. 90% of my bets are simply a win bet on one overlay in about 10 races a day (2-3 tracks) usually at Kentucky, New York circuits and Gulfstream. A bit at Oaklawn too. Obviously there won't be good overlays in every race but if I select from 2-3 tracks I come up with enough action.:cool:

[JustRalph]

Hi 46zilzal, Any hints on what you are doing?

Mike Nunamaker

Tap dancing ..................

[classhandicapper]

I want my ROI to be + 20% if my odds line is correct. That gives me enough of a margin of safety to be comfortable of case my odds line is off a little. If I am fishing with a live longshot, I'll insist on a slightly higher ROI because smaller errors in my estimate of the win percentage are more meaningful.

[46zilzal]

You wrote "I have a system called "two lines better than the field" that is hitting over 80% winners and over 95% at least place."

That's amazing. My odds line only picks about five horses a year to have an 80% chance of winning, and they generally go off at very very low odds. Any hints on what you are doing?

Mike Nunamaker

with Sartin software, anytime TWO of the last few lines rank ONE TWO over the ENTIRE field they have such dominance, that they still win (low odds) at that rate now ofr over two years

[twindouble]

I want my ROI to be + 20% if my odds line is correct. That gives me enough of a margin of safety to be comfortable of case my odds line is off a little. If I am fishing with a live longshot, I'll insist on a slightly higher ROI because smaller errors in my estimate of the win percentage are more meaningful.

I still don't get it, why bother to make your own line or put weight on the published lines. First off, you have to handicap the race and figure out who you like to win the race or get it down to who the competitors are. Only two things can happen when the race goes off, either the majority of the public agreed with you or they don't. What's new? If you are saying your lines are more accurate than the published lines, that just says your a better handicapper. What's new?

If your just structuring your wager around stats like the win percentages of any given odds or published lines, your not handicapping. If anyone can show me how to make money wagering that way, I'm all ears.

I've had many situations where in my mind the horse looked like a 2-1 shot and went off long so there shouldn't be any limit on your overlays when it comes to your opinion.

T.D.

[classhandicapper]

This is what I do.

I handicap the race, identify the contenders, and rank them in order of probability of winning. I don't go so far as to assign exact percentages (I used to), but I'm very aware of how those contenders stack up against each other and how many non-contenders there are.

Are the contenders evenly matched?

Is one much better than the others?

Are two close but clearly better than the other two? etc...

Then I compare my opinion to the odds board.

If my opinion differs with the odds board, I still don't get involved unless it differs by enough to give me a margin of safety in case I am misunderstanding something or my information is not correct/incomplete. I used 20% as a guideline because when I used to assign exact percentages, that's what I wanted. Now it's more of a gut feeling. I look at the board and somtimes I say "How the hell can that horse be 4-1. He should be the favorite!".

That's when I bet.

Much of the time I pretty much agree with the odds board. So I pass.

Other times, it's a struggle to be sure if it's an overlay or not. So I pass.

Sometimes it leaps off the pages at me like above. (not very often and that's why I pass so many races)

[kenwoodallpromos]

An average overlay has to be at least 25% better than its tote odds. That insures a profit.
Odds line- You can lose a lot of money picking winners if the payoff is too low. People who sell picks are good at picking winners but their ROI is mostly terrible (just pick even odds or less favorites!).

[twindouble]

This is what I do.

I handicap the race, identify the contenders, and rank them in order of probability of winning. I don't go so far as to assign exact percentages (I used to), but I'm very aware of how those contenders stack up against each other and how many non-contenders there are.

Are the contenders evenly matched?

Is one much better than the others?

Are two close but clearly better than the other two? etc...

Then I compare my opinion to the odds board.

If my opinion differs with the odds board, I still don't get involved unless it differs by enough to give me a margin of safety in case I am misunderstanding something or my information is not correct/incomplete. I used 20% as a guideline because when I used to assign exact percentages, that's what I wanted. Now it's more of a gut feeling. I look at the board and somtimes I say "How the hell can that horse be 4-1. He should be the favorite!".

That's when I bet.

Much of the time I pretty much agree with the odds board. So I pass.

Other times, it's a struggle to be sure if it's an overlay or not. So I pass.

Sometimes it leaps off the pages at me like above. (not very often and that's why I pass so many races)

Nothing wrong with that! Spot players can do just fine but most players find it boring but that don't mean you can't play more tracks for action. In most cases in a 9 or 10 race card I can find a couple races or more to key in on, ESP in the gimmicks or picks.

Good luck,

T.D.

[Overlay]

If you base your fair-odds determination on hard statistics rather than subjective opinion, and if you also size your wagers according to the disparity between your fair odds and the horse’s actual odds, you don’t need to demand as much of a cushion, since you can have greater confidence in your assigned probabilities to begin with, and you’ll also automatically be betting less on occasions where the actual odds are not that much higher than your fair odds.

[captvancouver]

I use 20% for my top pick, 40% for my second pick, 60% for my third pick, and so on.

Thanks for the input. Just to make sure I have it right.If you have horses A,B, and C, with A being top pick etc,and you have no overlay on Horse A, you will bet B to win if it is 40% overlay, even with the strong chance that A will finish ahead of B, or did you mean you would bet B to place if B was 40% greater than the second horse on the tote board. I don t know if I like the idea of betting against my own horse-"A"..but I ll give it some thought.

Thanks. And thanks to everyone for their input. I am amazed at the food for thought on this forum.

All the Best

Captain Vancouver

[cj]

I make a line for races where the horses have raced often. I have a tool I use to make an odds line. So, for example, let's say I have the 3 at 5/2 fair odds, the 5 at 3/1 fair odds, the 6 at 5/1 fair odds, and the 8 at 8/1 fair odds.

I would start with the 3. I'd take 5/2* 1.2 = 3, so I'd bet at 3-1 or more. If not an overlay, I move to the next horse. I take the 3 * 1.4, to get 4.2, so I'd bet at 9/2 or more. If not an overlay, I move to the next and multiply 5 * 1.6, and get 8, so I'd bet at 8-1 or more. If still no overlay, I'd look at the 8 horse and take 8 * 1.8 = 14.4, so 15-1 or more is needed.

I want more of a premium to bet my non-top selection.

I also change bet size according to the odds and the difference in my line, but that is another thread :)

[JimG]

I use 20% for my top pick, 40% for my second pick, 60% for my third pick, and so on.

Excellent rule of thumb Craig! That is pretty much what I do as well.

[BillW]

I realize the answer may vary according to the accuracy of the odds that the bettor establishes.Any pointers here? I suppose I could just keep records and get some sort of an average.

You hit the nail on the head. All suggestions here are a good starting point but the final answer lies in your own records. You mention not liking the idea of betting against your top horse, well once you start keeping records, you'll really hate betting against what you are staring at in black and white. :) Best of luck,

Bill

[captvancouver]

Sounds good

All the best

Captain Vancouver

[cj]

Here is an example from today, Thistledown, 5th race:

I use my numbers, and run a 5,000 race simulation to get these win percentages:

1 0%
2 7%
3 8%
4 9%
5 19%
6 3%
7 14%
8 40%

I eliminate any horses that don't have at least a "natural odds" chance of winning, so in this case, a field of 8, it has to be 12.5% or more to be a contender.

I'm left with 5, 7, and 8. The percentages convert to fair odds of 5 = 4.2, 7 = 6.0, and the 8 = 1.5.

I then do the math for overlays:

8 = 1.5 * 1.2 = 1.8, or 9/5.
5 = 4.2 * 1.4 = 5.9, or 6/1.
7 = 6.0 * 1.6 = 9.6, or 10/1.

That is it, now I just sit and wait. I have this mostly automated, so it doesn't take very long anymore and I can look at many, many cards.

[traveler]

Yes, Mitchell was a big proponent of bettin overlays. He used to claim there were many better handicappers than he, but he understood value. Interestingly, I have an audio tape of him and barry Meadow discussing making an odds-line, and Mitchell admitted that his fun bets had a higher R.O.I. than his serious bets where he made a line and bet the value. I think he looked for 20% overlay. Allways made me question the point of painstakingly making a line. In looking for value has anyone ever tried his "Group Overlay Method"?(if I,m straying from topic, sorry)

[captvancouver]

Craig, you sure have it down to a science. But still, there has be some sort of handicapping from the person who invented the simulation hardware. Are these 5000 races based on random factors? And wouldn t it be better just to find one or two whopper races, and bet like hell. Otherwise I would be in front of my computer all day.And I need to get out..I live only a few blocks from some of the most magnificent natural forests in the world..right in the middle of the city. I need to be with the trees sometimes.

Regards

The good captain

[cj]

Craig, you sure have it down to a science. But still, there has be some sort of handicapping from the person who invented the simulation hardware. Are these 5000 races based on random factors? And wouldn t it be better just to find one or two whopper races, and bet like hell. Otherwise I would be in front of my computer all day.And I need to get out..I live only a few blocks from some of the most magnificent natural forests in the world..right in the middle of the city. I need to be with the trees sometimes.

Regards

The good captain

LOL, I get out plenty, it really doesn't take that long. But I do spend around 20 hours per week in front of a computer betting.

As for the betting a few whoppers, I don't know which ones I'm going to win! :) I don't like losing streaks or betting too much on any one event, so I play this way.

By the way, the 8 at 5-1 was beaten by the 7 at 9-2, with the 5 third as the favorite at 2-1 maybe, don't remember exactly.

[kitts]

Mitchell/Overlays
Maybe a tad off thread. The "Group Overlay" has been tested a whole bunch by Cynthia Publishing. It is built into All-In-One V6 software as a component of the Platinum Strategies. It works, but is quite time consuming. Unless you have the software. Find out more at:

http://www.cynthiapublishing.com

[headhawg]

Not to discourage anyone from AIO, but here is a link at Cynthia Publishing to a post that has a link to a website that uses a modified version of the Group Overlay method. (Eeeew. Awkward sentence.)

http://www.cynthiapublishing.com/cgi-bin/yabb/YaBB.pl?board=AV6;action=display;num=1095377493

And you can probably use this article to create a spreadsheet to do it. Here is the link:

http://cynthiapublishing.com/MitchellArticles/dm20001104.htm

Good luck.

[classhandicapper]

CJ,

Do you ever bet 2 horses to win in the same race if there are 2 overlays?

[46zilzal]

You wrote "I have a system called "two lines better than the field" that is hitting over 80% winners and over 95% at least place."

That's amazing. My odds line only picks about five horses a year to have an 80% chance of winning, and they generally go off at very very low odds. Any hints on what you are doing?


went one for one today with it in the opener at philly...could not reproduce the entire screen but the dominance is apparent and this is what I look for. A caveat: NO maiden races, and the style of the horse has to be the same as the majority style of the distance at that track

[captvancouver]

I ve given the subject of overlay serious consideration as a newbie. I m impressed with the idea, that no matter how good or bad my handicapping skills become, if I know what the overlay factor is in my betting, then I can bet always withan advantage..if I keep records. But until I get to that stage, I have to choose.
For now I am considering the major pychological personal trauma that would result if my own horse beat me.I could have nightmares.And it wouldn t do much for the household harmony either.
She:"Did you win dear?" .
Me "Well yes but my own horse beat me".
She:"It figures".(LOL)
Actually, I looked at your figures, Craig, and I see that you would still win,because you have taken the "redundant" overlay factor into account, and besides it works.(And if my own horse does beat me,I ll tell her…"well sortta" and quickly change the subject.
LOL)

All the best

Keith

[BillW]

Keith,

You need to differentiate the horse that gives you the best chance to cash a ticket v. the horse that gives you the best chance to make a profit (long term).

Hope this helps your dilema :)

Bill

[cj]

CJ,

Do you ever bet 2 horses to win in the same race if there are 2 overlays?

Yes, I do this if the odds are high enough.

[captvancouver]

I think I get it.

Regards

[Binder]

with Sartin software, anytime TWO of the last few lines rank ONE TWO over the ENTIRE field they have such dominance, that they still win (low odds) at that rate now ofr over two years


The problem with this is people. If you don't already know
The Sartin programs do nothing until you enter a pace line
So 46's method has many build "yes buts"

Because its all depended on each users pace line and contenders
selections with Sartin Programs

unless I'm wrong and He is basing this on the Preanalysis
screen that is availible on the program he is showing
which is a post Sartin homegrown version of Synthesis

[46zilzal]

me and my shadow....

[46zilzal]

I get notes all the time from people who have found that this angle is universal..

Here is another response today........

Date: Thu Nov 17, 2005 2:27 pm

Also, thanks to 2 horse better than the rest of
the field the 8th at Laurel gave me a very nice priced winner paying $12.40. Thats a great price as most in my analysis are usuall favs.