How much information is enough?

[garyoz]

From today's Wall Street Journal's Market Mover Column

"Dresdner Kleinwort Wasserstein strategist James Montier says that overconfidence seems to be such an ingrained human response that even professionals can't escape it. Studies have shown that accuracy of bookmakers in setting the odds for a horse race, the ability of psychologists to reach a diagnosis and the ability of candidates for master's degree in business administration to predict earnings improve only marginally when they are given more than basic information. But their confidence in their abilities increases more."

Thought the reference to the bookmakers was interesting.

[Overlay]

I definitely think that there's a point of diminishing returns in handicapping, after which additional information only confuses the picture rather than clarifying it. Because I love the game, I could spend hours dissecting peripheral or redundant minutiae, but there would come a time when I would have to ask myself if the knowledge gained was justified by the value of the time expended. I've reached a balance between oversimplification and "death by analysis" that I'm comfortable with, and which (most importantly for me) gives me insight into the availability of betting value for whole fields, rather than focusing solely on picking the one horse most likely to win in absolute terms.

[andicap]

I absolutely feel that is the key to winning (well, one of the major keys).
Analysis by paralysis will kill you, but so will leaving out a critical factor like a horse you think is in great shape and has a top trainer and speed figures, is one of 4 speed horses, parked in the 11 hole and could be 4 wide

The problem is you have to know where to stop. How do you know you have left out a critical piece of information? When does the Law of Diminishing Returns come into play? Have you underhandicapped or overhandicapped?

Well, you know by experience and lots of practice of course. But no one's perfect and we all slip. I've seen people handicap races effectively in a couple of minutes and for others -- even experienced people -- it takes 20 minutes. None are right or wrong -- it just what you feel comfortable with.

If you're winning you're on the right track, If you're losing you're not.
And by "you," I mean "I" since I've been there and back.

A big problem is that in some races you will overhandicap -- as I did with War Emblem in the Derby -- and some races you will underhandicap. That leads to second guessing ("why didn't I see Billy Bob had a 47.1 gate work?"), hesitancy, lack of confidence and disaster at the windows because you will not have a clear plan of attack in your handicapping.

Maybe the answer is keeping a checklist of information you feel is critical and making sure you go through those paces. If you have a consistent list you will win some and lose some but if you're good it should all balance out in the end.

[Overlay]

Maybe the answer is keeping a checklist of information you feel is critical and making sure you go through those paces. If you have a consistent list you will win some and lose some but if you're good it should all balance out in the end.

You hit the nail on the head there. Standardizing the handicapping process around key factors, and having sufficient confidence in the results based on past success, will ensure that all the main points get covered in a systematic fashion, while avoiding the line-by-line, race-by-race agonizing that it's all too possible to get drawn into.

[john spencer]

could not agree more people . i think maybe we live in an era where sometimes we tend to over complicate matters and on occassions miss the obvious . in part where possible the old keep it simple method will put one in good stead . for yeasr i was looking for a way to have some input into doing the form but keep things short and sweet given my lack of time to do full time research on the horses yet enough reseacrh to enjoy my past time and all the systems and programms i found all had one common trait . over complicated . out of all teh sytems and programms i bought they all had the same results though just varying degrees of time in inputting the data . for want of refernce though have bought one taht i can cut and past my own information into and have a bit of fun with as well from racehorsedata.com (http://www.racehorsedata.com/) . it keeps things clean and simple and gets back to basics. just personal opinion i know, but when used this with the free data from risa.com.au (http://www.risa.com.au/) for australian races and even some of beyers speed figures when playing races in the usa it works for me . anyway . happy punting all . could not agree with you guys moreso.


John

[Dave Schwartz]

IMHO, the issue for many players is that they have lots of good information but very little idea about how to best use that information.

Database players are very susceptible to this malady because we have a natural tendancy to think the answers all lie right there in the numbers themselves.

It is not the numbers!

It is extracting value from the numbers.

One of the best lessons I learned about handicapping came from Dick Schmidt when he told me how to "work on" an approach.

He said something like, "You need to work through a few hundred races one by one. Start by looking at the results first, and then look at how the approach works."

He went on to explain that he knew looking at the results first "flies in the face of conventional wisdom" because that prevents it from being a legitimate test. That is because it is not really a test at all. It is about learning how to win down this particular path.

That was very profound! More profound than I believe Dickles realized. I mulled this over for several weeks, whenever I thought about winning. Understand that I do not have an opportunity to concentrate on my own winning ways very often because about 97% of my time is spent writing software, doing tech support, etc.

Why was it profound? Because it is not the way I do things. I test, or I look at the "big picture." I spend my time testing systems by trying something and measuring the performance of said system. Whether it is research through the database or click-click and record results, I am simply trying to verify that one scheme or another works.

Sure, I twist the dials and try this and that, but I NEVER look at races one at a time and question the individual race.

The above is from chapter 2 of this article:
http://www.horsestreet.com/freestuff/articles/LoseToWin/index.html

The entire article is here:
http://www.horsestreet.com/freestuff/articles/LoseToWin/Chapter2.html

FWIW.


Regards,
Dave Schwartz

[Tom]

I've heard enough!

[classhandicapper]

I've always thought that each race is more or less decided by a big edge in one factor and that handicappers can do well by mastering one or more of them.

For example:

In a race where all the contenders have been running similarly at the same class level, if one has a big edge in speed figures it is usually decisive.

In a race where all the contenders have been running similar pace/speed/performance figures, if one of them has been performing successfully at a much higher class level it is usually decisive.

In a race where all the contenders are similar from a numeric performance figure/class perspective, if one of them is the only speed or one will benefit from a likely duel, that is usually decisive.

In a race where all the contenders are similar from a numeric performance figure/class perspective, if one of them will benefit from a bias or all but one will be hurt by a bais, that is usually decisive.

In a race where all the above are similar or not a factor, if one of them is trained by a "super trainer" or has some other trainer pattern going for it, that is usually decisive.

If a horse has 2 or more things going for it, it is very decsisive.

It's when the information is conflicting or you have some pluses and some minuses that you run into problems. Then you are dealing with accuracy issues and relative benefits/costs. That's when you take headache pills. :lol:

[kenwoodallpromos]

No such thing as information overload- But there is application overload. No one is forcing you to use all the information available. Just decide which info suits the race you are handicapping and forget the rest.

[Valuist]

All I know is that I keep adding more and more things each year as the public gets smarter, not doing less and less. Edges are not getting easier; they are there but you have to work to get them.

[toetoe]

Hate to use the 'A'- word, but if you have a handful of angles, you have practically automatic bets, provided the odds are high enough.

If you love the Canani claim, you can't worry that FMar may get outridden by Matt Garcia (as happened).
If you love the Adam Kitchingman debuter, you can't pass because I. Enriquez is riding.
Sometimes, you get it all. Rosemary Trela privately buys (apparently) a horse previously trained by Mario Olivares and waits for Fairplex. She enters him in a $5,000 cl.nw2l., against a 1-to-2 M. Mitchell dropper you hate. You also get Agapito Delgadillo, a good rider who's an expert on the bullrings. No excuses if you refrain from banging him.

Hey, I came up with three examples from a weekday card at Fairplex. Wow.

1) barn change ... claim
2) tr. spec. ... debuter
3) barn change ... private deal

Mix those in with what Quinn calls comprehensive handicapping, and you will have action galore. You might even make some dough.

[oddswizard]

Pace wins sprint races up to 7 furlongs. Class horses win distance races. So determine which pace and class formulas win for you. It really is that simple. Remember the old saying "Keep it simple, Stupid?" I am a believer.

[garyoz]

I think what is interesting about the topic is that more information doesn't seem to change the decisionmaking much, but that it gives the decisionmaker more confidence to go with the decision. This is regardless of whether the decision is right or wrong.

[ezpace]

turf checklist

Claimers checklist


all other classes dirt races check list


then it's

apply proprietary 1948 speed figures

apply CJ performance numbers

apply/spend MOST time setting up pace running styles

proprietary** FORM #'s **/Current class (review)

decision model

gamble!!
********

keeping in mind all races are like a figerprint = different

confidence in model for each will provide good % and "real" ROI... fwiw

[ezpace]

turf checklist

Claimers checklist


all other classes dirt races check list


then it's

apply proprietary 1948 speed figures

apply CJ performance numbers

apply/spend MOST time setting up pace running styles

proprietary** FORM #'s **/Current class (review)

decision model

gamble!!
********

keeping in mind all races are like a figerprint = different

confidence in model for each will provide good % and "real" ROI... fwiw

[46zilzal]

i think maybe we live in an era where sometimes we tend to over complicate matters and on occassions miss the obvious .



Could not agree more

[ezpace]

dbl post please..

[toetoe]

EZ,
Could you repeat that request please?

[Vegas711]

The fly in the ointment concerning horse racing is that never will every horse come together in its same exact form cycle and race against all the same horses ever again. Every race is a unique, never to be repeated occurance.

This is the reason why you can never set an exact probability on the chances of any horse winning. An oddsline like anything else is just a guesse.

This is what protects the game and makes it so damn fun.:bang: :confused: :lol:

[Vegas711]

Pace wins sprint races up to 7 furlongs. Class horses win distance races. So determine which pace and class formulas win for you. It really is that simple. Remember the old saying "Keep it simple, Stupid?" I am a believer.

Sorry dude ,but this is flat out wrong. Pace wins races up to 1 1/8 miles beyond that not. My highest level of success using pace is route races.My betting records show this. I never look at man made class levels.

K.i.s.s...... this is right on.:cool:

[Overlay]

As Vegas711 says, no one can dispute that each race is unique, and that it's never possible to say with 100% certainty that any one horse will definitely win any single race before the race is run. However, there are patterns of performance that recur with enough predictability, and that can be interrelated with a sufficient degree of confidence in the results, to permit meaningful calculations of what each horse's winning chances are (similar to projecting how often each horse in the field would win if it were in fact possible to run the same race a sufficiently large number of times). This permits judgment of whether the odds on each horse represent a favorable or unfavorable betting proposition, which in turn allows for systematic, long-term profit, even though each race will never be run under the identical set of circumstances again. (As several handicapping authors have pointed out, the old saying, "You can beat a race, but you can't beat the races" should be reversed to read, "You can beat the races, but you can't beat any one race.")

[dav4463]

There may be something to pace wins sprint races and class wins distance races. I don't use pace figures at all. I show a profit on route races, using primarily class moves, past class, and trainer intent, along with qualifying Beyers. I was showing a loss on all races under one mile. I have to use a different approach to sprint races. I quit using my same approach to sprints, so tried something else..... ( I use Overlay's book ! )

[Art P]

Just bet the speed and collect!

[nobeyerspls]

The form provides a great deal of data and information now. For me, some is helpful and some is not. I save a lot of time by ignoring data I don't need such as jockeys, weights, and speed figures. Most races cannot be handicapped with a useful degree of certainty so these can be skipped. When you find a way to identify these quickly, you'll have more time to spend on races that can be profitable.

I know my home track well so it gets the most play. The other tracks might produce one play per day. Confidence and concentartion of capital are a good combination.

[parlay]

I purchased Betting Thoroughbreds in 1979 and i love rereading it.
The Forward by Andy Beyer and Davidowitz Introduction are worth
the price of the book. IF YOU HAVEN'T READ IT GET IT NOW!!!

[Figman]

How about digesting this?
http://tinyurl.com/8guwg

[oddswizard]

Pace in distance races may be right for you but it is wrong for me. I pick 3 horses per race . One of these 3 horses will win 70% of the races using Pace in sprints & Class in routes. However, this is not about you or me being right. It is about making money. If your formula for Pace wins over 70% in route races please let me know the formula you use. All I want to do is win. Thanks.