Make Your Own Line

[socalsportsbook]

This may be a fun discussion so I hope I can make my question clear.

I'm making my line for today's races (hypothetical).
I like the 2 horse and the ML is 2-1 (33%). I think he'll get off at 3-1 (25%). Remember this is hypothetical.

I make him 4-1 (*40%). If all goes according to plan I'll have an overlay on the 2 horse. I'll play at 6-1.
*That means that if they ran this race 100 times my horse would win 40% of the time. Here is my problem; everything I have read or read simply glosses over this estimation of what I value the horse at. It seems to me this is a little difficult to do accurately.

So here I stand at the window with my $2 in hand and watching the odds until the last minute hoping I get my price and all this energy is spent assuming I'm right on my "guess" that this horse is valued properly at 4-1.

My question or observation is: This step in the process may be the most crucial, time consuming and damn near impossible part of the process. Yet I still base my decisions on this educated guess. Did I miss something.

Well its off to Los Alamitos for today's races. I like Lady Essex (5) in the 7th at Bay Meadows. I don't know if this horse can ever win this race but the price should be good.

[socalsportsbook]

I knew I'd screw this up.

Let me try again:

I make the horse 2-1 (33%)
ML is 4-1 (20%) not 40%.
I believe that if they run this race 100x my guy wins 33%. Based on these numbers I have an overlay if the horse gets off at 3-1.

NOW, my question: Is deciding that my horse will win 33% of the time a difficult if not impossible task?

Sorry for the confusion

[anotherdave]

Originally posted by socalsportsbook

NOW, my question: Is deciding that my horse will win 33% of the time a difficult if not impossible task?


Being that precise is close to impossible. If it is 28% then 5-2 is fair. 25% now it is 3-1. I am a statistician, and I usually do set a line, but only as a guide.

I find that when the time comes to make a bet and I have the race clear in my mind, then I know whether it is an overlay or not. When I look at my horse, most of the time I can sense whether the bet is good or not(overlay or not), whether I have made a formal line or not.

A few years ago when I was just getting this concept clear in my head, I made up an odds line and stuck with it pretty closely. But as the concept became clear, I found that I could "feel" the bet. What is it Michael Pizzola says : "Let the bet make you" This may sound a bit shaky, but I think that most winners, whether they actually set a line or not, have an intuitive line in their head. Only comes from experience, I guess.

AD

[andicap]

I'm feeling lazy at work today so I'm posting a lot:D

I agree 100% with Dave about odds lines. How the hell -- unless you keep unbelievably good records -- can you know whether the horse will win 28% or 33%. And what about the lower odds? 9-2 is 18 or so and 5-1 is 16%!!! 6-1 is 14 and 7-1 is 12.5%.

I can see the thinking process. "Hmmm the 2 horse has a 16% chance the win, but the 4 horse has a 20% chance to win."
No one you can be that precise. I think these odds lines are a joke.
NOW, I think they are good to do as a rough guide that's all. I'll do one like this:

8 horses in the race. 5 contenders. My top 5 win 85% of the time. All being equal, they have a 17% chance to win = 9-2 or 5-1. Decide which of the 5 I like best and create a continuum that adds up to 85%. Do I like the horse a lot or a little. If a lot I'll create a cushion between the first and secondhorse -- and so on down the line. If I like the 2nd horse a little more than the 3rd horse I'll keep that margin think then I'll keep fudging the numbers until they look OK as a GROSS ESTIMATE.
So if i end up at 2-1 (33%) 4-1 (20%) 5-1 (16%) 6-1 (14%) ..that's 4 horses and 83%..oopps...gotta give the 5th horse a decent odds, say 7-1 (13%), so I'll reduce the others accordingly so it fits at 85%.
But I won't take those odds ultra seriously. It means if the horse I had at 5-1 is at 8-1 I've got some thinking to do...is it an overlay for me? These are the judgement calls that kill me. But at 10-1 I'll bet it confindently knowing that even if the horse was actually 6-1 it's an overlay.

What I'm working on to make my life easier is an estimate as to how my 2nd choice finishes. Does my 2nd ranked horse win 20% of the time across a wide representative range of odds? (If all my winners are 3-1 and below and my losers are 4-1 and above then why would I bet my 2nd ranked horse? But if my 2nd ranked horses win at higher odds as well, I hope to get away with betting it when the odds are 6 or 8-1. )

[Rick]

The worst part about computing an accurate odds line is that the actual odds themselves are a pretty accurate estimate of the real probabilities in the race. If your method is a good one and is not too similar to what the public is doing, then you can have overlays but they will always be less than you think they will be. For example, you think the horse should be 2-1, but the toteboard says 6-1. Most likely, the horse won't win 33% of the time but more like maybe 20%. So, you still have an overlay, just not as big a one as you thought.

If you combine the two, then you can come up with a very accurate odds line, but of course you can't use final odds since those won't be available to use. You could probably do it at 2 minutes to post with some automation and be pretty accurate though. It probably would be even better to use late money and discard early money for the "public probabilities".

Another way to look at it is to use the morning line odds or ranks (adjusted for scratches) since they account for about 80% of the closing odds anyway and largely represent "dumb money". Or, bet against the 10 minute odds, which are considerably less accurate than the closing odds. If the odds drop, it is more often due to well-informed money, so your chances of winning will increase enough to compensate for your loss in odds.

Another way to look at it is how much do you hate the favorite? If your method rates the favorite way down the list, say 4th, you probably have an overlay on your top horse no matter what the odds are because you've eliminated so much in terms of probabilities. If you give a horse with 33% of the money on it only a 12% chance to win, you can bet on almost any other logical contender (or all of them) and win money. I'd say that's the most logical basis for multiple horse betting.

[cj]

Originally posted by Rick
... You could probably do it at 2 minutes to post with some automation and be pretty accurate though.

That used to be true, but I doubt it unless the pools are pretty big. On Monday and Tuesday, I played LS and Crc. Both tracks, I saw horses pounded in the last minute. One horse dropped from 5-2 to 7-5 as they headed into the turn at Crc, while a LS horse was 3-1 at 1 mtp and paid that same whopping $4.80. Tough to build that into a betting line! Good news was at least they both won.

CJ

[Rick]

I just don't believe that money coming in from off-track sites is all that "smart" on the average. Now, you can probably come up with examples to the contrary but I think that you don't need to worry about simulcast money coming in after the race is off because on the average it is "dumb" money. That's just my opinion. I could be wrong.

[anotherdave]

Originally posted by Rick
I just don't believe that money coming in from off-track sites is all that "smart" on the average. Now, you can probably come up with examples to the contrary but I think that you don't need to worry about simulcast money coming in after the race is off because on the average it is "dumb" money. That's just my opinion. I could be wrong.

I don't think it is smart, either. But in my experience that money seems to drop the odds on favourites and increase the odds on longer horses. So it looks "smart" more often because lower odds horses win more often.

AD

[Dick Schmidt]

Thoughts on line making.

In one of his books, Dick Mitchell says that a good handicapper has to be able to differentiate between a horse that should be 5-2 and one that should be 3-1. That's when I stopped even considering making a line. I can't tell the difference between a horse with a 28% chance of winning and one with a 25% chance of winning. Once more, I don't think anyone else can, either. I've been to the races a lot with Dick, and I sure don't believe he can do it.

People who make lines like to wrap themselves in numbers and absolutes, saying such things as "if this race was run 100 times" such and such would happen. Like they really know. The truth is that no race is ever run twice. Even if the same horses ran again, they would be in different places in their form cycles, have different post positions, different jocks. Not gonna happen.

The true basis on all handicapping lines is a guess. I think (guess) this horse is worth 2-1. If I make an 80% line, then my next pick is (I guess) 3-1 and so on. They are doing the same thing I do; based on the facts I can see, I guess this is the best horse, that one is second best and so on. I just don't assign numbers to my guesses and thus lose a chance to appear to be scientific and a "math guy."

The thing that all line makers need to acknowledge is that their line is a guess, and has no validity in the real world, no connection to the race in question. Now, that's not to say it can't be a damn good guess, and that lines aren't a useful tool to many people, just don't imbue them with magic or the aura of precision. They are just a guess, based on incomplete information, on how the horses rank in a race. A shorthand way of saying, Horse A is twice as good as Horse B (I think).

So, if you can't tell the difference between a 9-5 horse and a 2-1 shot, don't fret. Neither can anyone else, no matter how sexy their line may look, no matter how much math they use to make it. It is still a WAG.

Dick

[Rick]

Count me in the WAG philosophy group as well. Although it is possible to come up with a more accurate line than the public's, it's really not necessary anyway. If a horse is ranked first using my method but third by the public, I'm positive that I have an overlay and I don't really care what the odds are. If I've selected the favorite, it's probably not an overlay. Although there are exceptions, you won't go wrong very often using that approach.

[Bob Harris]

There is no such thing as an "accurate" betting line. You can't take gaming theory as it applies to roulette and transfer it to a game with living animals with different form cycles who are bumping and cutting each other off all the way around the track.

Making an oddsline is simply a way for the handicapper to quantify his *perception* of the race. Handicappers using different methods are very likely to have different perceptions and therefore produce different oddslines.

Without having a general sense of an investment's worth or potential, how can you possibly know if you're getting a good deal ? Would you buy stock or real estate without going thru this process?

All winning players, even those who scoff at "value", go thru a risk/reward procedure...some put it down on paper and some just do it in their heads.

[Rick]

Bob,

I have to agree with you. When I said "more accurate" perhaps it would have been more correct to say "less fuzzy". The inaccuracies are especially evident when you get down to the smaller probabilities. I don't think anyone can distinguish between a 40-1 shot and a 50-1 shot, for example because there's just too much room for error. But you should be able to distinguish between a 2-1 shot and a 4-1 shot because the difference in assumed probabilities is so large.

[Dick Schmidt]

Bob,

You're right, we all come to some conclusion in each race about the value of a bet. I use a dutching program, and insist on even money or better from the combination of all the horses in the bet. Others find a line is comforting. As the wise man said to the aged monkey, "Whatever floats your boat."


Dick

[freeneasy]

Are we allowed to say sex
yeah well my esx lines are better then your esx line
hehaw

[rrbauer]

Here’s an exercise that might help with regard to building a betting line. Although the product from the exercise is not a betting line, per se; it will give you some indication whether you’re on the road to positive-expectation plays.

Think about each race as an opportunity to build a virtual stable with however many horses you think that you need, in order to take an even-up stand against the rest of the field. Sometimes it will be one horse. Most often it will be two or more horses.

Based on the current betting, demand composite odds of even money for your horses. (Composite odds implies “dutching” for multiple horses.)

At crunch time, if you get your even money, or better: Play.
If you don’t: Pass.
(I say make the play at even money because you have probably “stacked” the deck in your favor in assembling your “stable”. As a practical matter you’re probably odds-on!)

At first when you use this approach you will have too many horses in your stable to get a play. So what you need to do is refine your selections. You can do this by repeating the “us” versus “them” process for just your stable. Get rid of the horses that do not contribute much margin to the rest of your list.

As you produce refinements you will come to appreciate the value to your stable of a horse with some chance of winning at high odds versus the value of the same horse at low odds. That’s as it should be. Value (in horse-race betting) has two components: Chance of winning and the Payment when you win.

I agree that trying to assign finite probability expectations is a hair-splitting exercise. As someone pointed out, when experienced players have an overlay they just “know”. It’s intuitive at that point. The process that I’ve described can be done initially on paper as a simple adjunct to whatever handicapping process you currently use. If a horse from your stable is usually the winner, that’s testimony to your handicapping ability. If the other stable is beating you frequently, then you have work to do. (Or maybe you should just use the other stable!)

The benefits here are twofold. First, it helps to hone your skills in contender selection. And, it forces you to consider your play against a financial benchmark of profitability.

[Dave Schwartz]

Richard,

What an awesome approach! Thanks. (Hope you don't mind if I apprpriate it.) <G>

Regards,
Dave Schwartz

[anotherdave]

Thanks for the posting, Richard. That was great. Just the kind of thing I come here for.

AD

[Dick Schmidt]

RRBauer,

Very articulate explanation of a real-world friendly approach. I've been doing almost exactly that for some time now, using a dutching program, but had never conceptualized it in this manner. Thanks for the insight, I'm sure it will help at that most difficult moment in handicapping.

Dick

[Tom]

That was very well put. In fact, it was the most profound
thing I heard all day (I was in a staff meeting all day!) ~G~

[rrbauer]

OK...now that I'm on a roll I will go over to the "Selections" section and do a thread on my stables for tonight's HOL card.

You are all very kind.

Thanks!