Handicapping with 2 or 3 factors only

[jackad]

A message on another forum refers to studies that showed two factors with positive ROIs nationally, later confirmed by a study of California races. Unfortunately, the message does not identify these factors.
What do you think about handicapping taking only 2 or 3 factors into consideration? (Most software programs seem to take 15 or more factors into consideration to make their selections.)
About how many factors do you tend to consider?
Jack

[andicap]

One of my favorite topics!!!
I am about to begin (if time permits -- damn work and family!) to post selections based primarily (but not entirely) on my BRIS pace formulas.
My theory, which worked (win and place betting) during my trip to Saratoga last month, is to take my top three contenders (through a point system I developed) and bet them at certain odds. Research still continuing, but could become:

top choice: 3-1
2nd choice: 6-1
3rd choice 8-1

something like that, more reserach needed.

A problem I see, however with the top choice is if its odds are too high, that could be a red flag.

I am experimenting with two methods

-- Just the points (two sets of formulas)
-- Points plus some form factors used to adjust the odds (e.g. distance, trainer, jockey, horse for course, running style, etc.)

I hope to post my top 3 picks as much as possible when Belmont opens.

[ranchwest]

I use five, but that includes computed figures, aggregates, rankings, etc. that are done through the computer because of the complexity of the calculations. The bottom line of what I look at is only five data items.

[Rick]

I use four factors now but have used as few as two successfully in the past. The reason I favor fewer factors is that it's much easier to find a few relatively independent significant factors than it is to find many. If they're too similar, your top ranked horse will have a worse ROI than the 2nd ranked. A simple example would be a method based on combining speed ratings with average earnings. If you check all of the horses that rank 1st in both categories, you'll get a good win % but the ROI will be horrendous, probably something like a 30% loss.

So it's best to start out with a few factors that are not too similar to one another and experiment with adding new things. Anything that lowers the ROI of your top ranked horse should be tossed out. It's not impossible to find independent factors though. The Hong Kong syndicate has done that over a period of many years. Each additional factor contributes less and less to the overall effectiveness of the model though and it may not be worth the additional effort.

Also, some data is difficult to use directly. Post position is a good example of this. It's certainly an independent factor, and that's good, but statistics are not normally readily available in the proper format. You have to adjust for number of starters to get the significance right. You should look at the exact distance, and you may need to look at multiple years in order to get a large enough sample. It's quite common to see a percieved bias reverse from one year to the next since track management notices those things too. It may also be desirable to smooth the data some in order to get rid of the noise.

If you're just looking for something to improve your win %, you can use today's odds with your other factors. That will improve it more than any other factor you can find, but I think I'll pass on that one.

[JustMissed]

Some of you guys always talk about linking ROI with odds. I read on another sight that true odds can only be determined after the race is finished which on the surface makes sense to me, but aren't final odds simply a way of expressing how much money was bet on a certain horse compared to all the money bet on all the horses. My question is: What do odds have to do with the probablity of a horse winning the races? After all, track odds are the track handicappers opinion of the likelyhood of a certain horse winning, the same with tote odds-the public's opinion, and the same with your own odds line-its just your opinion of how well a horse will run.

After all, the winner's probablity of winning is 100% and his ROI is the payoff divided by the amount bet. The second place horse's probability of winning is 0 and it's ROI is 0.

It seems to me that the only odds that matter in the area of handicapping is your own odds line. To me the tote odds only matter in relationship to the betting decision, i.e. Will I risk $2 if I can only make $2, as opposed to-If I can win $8 I am willing to risk $2 on a horse that I think has a chance to win.

I realize that your opinion of the horses chances to finish in the money has a great bearing on your decision as to the amount and types of bets you place or whether you even bet at all, but I don't see how you can link the handicapping probabilites with the betting probabilities in a mechanical fashion and get winning results.

JustMissed
:)

[ranchwest]

JustMissed,

The thoughts you've shared are in line with what I and a few others have been saying here for awhile. The winner wins 100% of the time.

However, I do think that to keep your handicapping in line over time, not based on a single race, you need to begin with the premise that the public is basically always an honest gauge. Horses win in proportion to their odds.

Also, at all odds ranges the public loses about the amount of the take.

Now, where does that take us? Well, it takes me to the fact that we can win at any odds range if we can beat the take plus get the profit we require. However, we cannot focus on a narrow odds range but rather are better off playing horses of various odds ranges.

That brings me right back to the winner winning 100% of the time. You have to have a metamorphosis of thought to realize that the best horse always wins. You simply have to arrive at a new definition of "best horse".

So, I just play the horse that I think is the best horse for the race, whether it is 1/5 or 40/1. Well, I guess I should add that I'm very selective about which races I play. Also, I rarely get a chance to see the tote board anyway.

[JustMissed]

I agree with 100 percent of what you say. The longer I study this business the more I am convinced it is like learning golf. Both games can and should be simple but we write thousands of books and articles to explain things that have always been and are now quite simple.

One other thing that I have noticed is that by placing so much emphasis on betting the winning horse it seems we may be missing the boat by passing bets on losing horses. For example, some shrewd players play high odds KEY(K) horses and construct their bets this way: K to W/P; KA/KAB or K/AB & AB/K, etc. When I first saw this I thought it did not make sense to predict that just because K had high odds it could win til I realized that this bet also predicits that K will lose, BUT by placing second the bet was constructed so that even if K loses you still collect on a Place ticket as well as on an exacta. Once I realized that the better players are picking horses that COULD POSSIBLY win or place or show and constructed their bets so as to benefit from the various outcomes, it really changed my thinking.

My approach now is to select the horses that I think have a chance to be in the money, arrange them in slots as to where I think they might finish and STEP 2-either construct a bet that will give me a likelyhood to profit or pass the race.

Also, in the little example above, when I first saw this I did not realize that the author was also picking the two best horses, A & B, and playing them with the K horse. The same senario but a different way to look at it.

JustMissed
:)

[Tom]

I think the more factors you "factor in" the more good prices you are going to "factor out."
Try to mix a positice and a negative factor and see what you get - say,
1. Top three Beyer from last race
2. Finished out of the money

The good Beyer may suggest good speed, form and/or class in today's field, the poor finish might scare off the public money.

Years ago, I used to use every factor in Ainsle's Complete Guide.
I literally spent and hour a race getting the final contenders, then ran them through each of the 77 plus factors in the back of the book! Good thing there was no simulcasting back then.
At the end, I usually had fit, well meant horses that ran good.
That won't work anymore, or at lest when it does, you get a 3-5 shot that everyone else in place has.

[hdcper]

Hi Everyone,

I can't remember where I got this link(it may have been on this site) and whether the URL I have is correct, but it talks about a study of handicapping factors and what number works best.

Here is hoping it works, I believe you will find it interesting reading:

www.odci.gov/csi/books/19104/art8.html


Hdcper

[baravot]

Hi hdcper,

Thanks for the link to that article- it is loaded with good info.

[JustMissed]

Thanx for the info-very good stuff.

JustMissed
:)

[andicap]

can someone put this into English for me?

[JustMissed]

I read the article a couple of times and thought I understood it. What do you need help with?


JustMissed
:)

[Triple Trio]

Originally posted by hdcper
Hi Everyone,

I can't remember where I got this link(it may have been on this site) and whether the URL I have is correct, but it talks about a study of handicapping factors and what number works best.

Here is hoping it works, I believe you will find it interesting reading:

www.odci.gov/csi/books/19104/art8.html


Hdcper

I think the study should be interpreted this way: human beings are not much good in handling information once the amount of information exceeds a certan point. What it doesn't say is that five items of information is good enough for handicapping a race. If you use a computer to help you handicap a race, you can use as many factors as you want to analyze the race.

[JustMissed]

Great reply to Andicap's query. I would also like to ask you about your comment about the computer analyzing as many factors as you like. I think I now what you meant except the article dealt with decision making and the computer cannot make decisions.

The computer can sort, query(answer yes or no), list and retrieve files, and most importantly for us, rank in order of programmed criteria, but the computer does not make decisions.

The only reason I mention this is because I learned many years ago that a computer cannot do any that I can do with a pencil and a piece of paper, assuming I had enough time.

Not to say that computer programs don't do a great job of ranking horses according to winning criteria-because they do do that very well every day of the week and some of them do very well- but it is the mind of the programmer that makes the decision-not the computer.

JustMissed
;)

[ranchwest]

I don't want to get into an argument over the semantics involved with the word "decision", but while a computer only directly interprets "1"s and "0"s, the data the computer works with can be used to provide artificial intelligence, making the computer "smarter" and "smarter" in a way that a human cannot.

[JustMissed]

Your contributions to this board are great and I would never consider anything you said to be argumentative. I posted my response just to caution Andicap to be careful when comparing studies about the human mind and computer application. To be honest I don't think anyone really knows how the human mind works anyway and we can only study processes & results.

Your post actually proves my point. If I did a search of this website site looking to identify troublemakers, I could type in certain words like adverse words including the word 'argument'. Your name would pop up as having used argument but that doesn't mean you are a troublemaker. The search engine only did what I told it to do, but I'm the one that made the decision that you are not a trouble maker because I have read alot of your post and believe you to be a good guy. I also certainly believe in AI but I think a program has to be able to rewrite itself(of parts of itself) as it accumulates data and I don't think we are quite there yet. Some of the programs are so good at filtering and sorting that they give the illusion of decision making but I don't believe that is in anyway comparable with "A Beautiful Mind'.

Just My Opinion

JustMissed
;)

[ranchwest]

Hmmm, my daughter just got that for her birthday, guess I should borrow it. :)

Yeah, I've seen some people who seem to make a lot of right decisions and I'm not really sure how or why.

[hdcper]

Hi Just Missed,

Sort of confused by the your post below:

Great reply to Andicap's query. I would also like to ask you about your comment about the computer analyzing as many factors as you like. I think I now what you meant except the article dealt with decision making and the computer cannot make decisions.

If I said anything to this effect I don't remember it. Maybe you have me mixed up with another's post!

Just wondering,

Hdcper

[JustMissed]

Sorry for the confusion. The original post that started the thread was about computer software as follows:

ORIGINAL THREAD WAS STARTED WITH:
"What do you think about handicapping taking only 2 or 3 factors into consideration? (Most software programs seem to take 15 or more factors into consideration to make their selections.)"

When you referred us to the article which was about human beings handling information I did not make the crossover and was still thinking about software, not mental capacity in decision making.

Your response listed below was great but I was only making the point that I had been confused by associating the article(which only dealt with mental capacity) with the original post(which dealt with software programing). Thats all I meant.

YOUR RESPONSE WAS:
"I think the study should be interpreted this way: human beings are not much good in handling information once the amount of information exceeds a certan point. What it doesn't say is that five items of information is good enough for handicapping a race. If you use a computer to help you handicap a race, you can use as many factors as you want to analyze the race."

JustMissed

[hdcper]

Just Missed,

Triple Trio said that, not me! Understand you thinking it was me since my handle is just above the comments.

Make sense to me,

Hdcper

[Triple Trio]

As far as I know, some of the most successful teams in the world use a completely mechanical approach and let the computer make the betting decisions. The software typically includes many many factors, both subjective and objective. Where a subjective factor is used, it is often given a numerical value so that the information can be blended with the objective factors.

Now most of us don't bet like that. Depend on our style, we may depend more or less on the computer. We mostly use the computer to rank, sort, etc. the data for us so that we can make the betting decisions more easily. Since the final decision is a still a subjective one, bearing in mind the CIA study and the limitation of we human beings, regardless of how many factors our software uses in its computations, it makes sense the reduce the final output to maybe 5 or 6 measures so that we don't get overwhelmed in the decision making process.

[JustMissed]

Triple Trio,

Are you talking about a black box program where the raw data is entered(presumeably from equibase or whatever) and the program selects the probable winners and tells you how to bet them?

I'd like to hear more about this if it is for real, and especially the results.

Thanks,

JustMissed

[Dave Schwartz]

JustMissed,

I believe Triple is referring to this:

http://www.wired.com/wired/archive/10.03/betting_pr.html


Regards,
Dave Schwartz