Database Delusions

[Handiman]

I have a question for everyone, that cares anyway. Are we deluding ourselves with the use of databases?

I ask this question for a couple of reasons. I thought that during our two weeks of daily play with the Black Box approach and with the Wall Street approach that a database guy would step up and knock our socks off. As far as I know that didn't happen.

The main reason I am looking for input on this subject is cause I am making plans and researching the best database to use with Handifast. But if there are guys much more sophisticated than I am with database access that aren't just slaying the tracks, what chance do I or anyone else have using a database with Handifast.

If for instance the top pick with Handifast can give me a break even or small profit, which so far, from what I have seen or heard, is just as good as the database gurus. If I am wrong, please straighten me out and show me the error of my ways.

I'm asking this here instead of the Handifast sub forum because this is a database question and not program specific. I need all the help I can get, before I go to all the work of attaching database capabilities to the program.

Handi :)

[Partsnut]

Hi Handi,

In my personal opinion, maintaining a database for each track is a waste of time. I've modelled many tracks over the last couple of years and found
that what wins today or has won in the past will or may not win today.

I believe that the Brohammer approach using FX and FW is the only way to travel. The Brohammer model, in my opinion is by far the most credible for my puposes because it gives you an idea of how specific tracks will run and will give you a fair idea of an individual tracks bias. This in combination with running styles and pace shape has produced some good results for me.

The fact that some of the software providers that declined to participate in your contest with an excuse of being busy or whatever, tells me that their touted products are less then credible. I would believe the best way to promote ones product is to show all how well it does in real time.
I applaud those that did participate because it tells me that they believe in what they are selling.

[sjk]

I have enjoyed good success with database derived handicapping over many years and recommend it.

As I mentioned in the thread leading up to the Black Box comparison I think you really need to test a process over a much larger number of races to draw any meaningful conclusions. It would take a significant time commitment to post that many races in advance which would be hard to get anyone to do. Meanwhile no one is willing to believe what someone else says about their historical returns.

[traveler]

The main reason I am looking for input on this subject is cause I am making plans and researching the best database to use with Handifast.

Handi :)
Hi Handi - Forgive what may be a really dumb question but isn't Handifast a piece of software? Are you saying you want to use another program with database capabilities, in conjunction with Handifast? If thats what you want I can tell you for a fact Schwartz's HSH will do Brohamer type modeling dynamically, in other words, if you open the 1st at AQU and its a 6f clm you can easily see whats winning a 6f at AQU for Fx Fw and other factors. The next race you look at it 1 1/16 at HOL and you can get the same info. Dave has some Pace Handicapping videos he did that I don't think are on his "public" site. If you haven't you may want to call him. He's got that seminar tonight with a special pricing going on. As a disclaimer I do use his HSH product. I don't believe Dave uses this approach personally but it can be done. If I've totally missed the point it wont be the last time but maybe this helps somebody else with questions. Good Luck

[douglasw32]

My 2cents..I was a databse guy before i learned the little I did about programming to allow the code to be advanced.

I can do more, more quickly with Microsoft Access, or sql than I could ever pretend to do with VB.

As for coming up with something more than handifast, the main thing a database can be used for other than checking results of a large sample of charts quickly would be in the power of combining other works.

For example one could export the DRF files, pull them into a database set up to incorporate the BEYER figure into the FAST number to get a whole new number.

The possibilities are endless...once it can be kicked out into a csv.

Actually, I regularly use an access database of the program (Before it became a program) that gives me a very nicely formatted summary.

Of course it simply matches the programs results but it is an old habit I have a hard time breaking.

My point is, the endeavor of writing the code to give the program database possibilities is well worth any effort in my book.

[Handiman]

Maybe I didn't make my self clear. I do not want to use another program in conjunction with Handifast.

What I plan to do is add two modules to Handifast. So there will be 3 modules. The first module will be Handifast as it sits now as a program for live play. In addition, when a race card is completed, a player can produce a csv file.

The second module will be a batch processor. It will basically allow the user to take downloaded files all in a single folder, and run them. The module will handicap every race card and then kick out a csv file.

The third module will allow a user to take those csv files and load them into the Handifast database. Then that database will allow the user to run queries.

The three databases I am looking at using are sqLite, Tsunami or Cheetah. They all can be managed with libety basic. So it can all be under one umbrella.

I hope that makes more sense. Right now the first track will kick out a csv file. I have 5 more tracks to code. Plus tonight, I finished up writing a scorekeeper for live contests. We play with about 10 to 17 people at a time and trying to keep count of their winnings by hand is crazy. But with this scorekeeper, it does it all with minimal effort.

And there are several different ways to keep score based on whether you play win only, WPS, WP, PS, Exacta boxes....whatever.

So now I plan on finishing the csv coding for the other 5 tracks. And then move on to the database coding.

Thanks,
Handi

[raybo]

As an Excel user, I have incorporated databases, within Excel that allowed me to test ideas against past races and results. It also allowed me to determine my personal strengths and weaknesses, as well as ticket structures/wagering strategies. I have never used one for modeling as I feel the game is too dynamic. However, modeling, if used dynamically with smaller samples, could be useful in predicting future, dynamically changing, trend forecasting, allowing the user to stay ahead of the game.

Having never created or used a "full blown" database, I'm sure there are those, who have, that have discovered other important uses for modeling, possibly with a dramatic improvement in their ROI. As a "data/math/logic" guy, I am all for any platform that allows detailed, flexible, analysis of play history vs result history. As a matter of fact, I've been thinking, for quite some time, of incorporating this into AllData PPs.xlsm (2007 version, with it's increased limits and power).

Looks like Handi is a little ahead on this one. I salute you Handi, and think it will be well worth the effort you're going to have to put into it.

By the way, Handifast and it's csv output will be included in the AllData PPs output to the database. Hey, it's free and open source, too, right? :ThmbUp:

[raybo]

As a side note, in the interest of furthering the goal of increasing the number of winning players, and thus the number of total players, in our game, I believe that offering free, full featured applications such as Handi intends, and myself (with much help, from another PA member to be named later), will help drive the state of racing software, at all levels, due to the natural, expected competition for users.

Our game must produce more successful players (we can do that part) and must publicize those successful players (that has to come from savy industry promoters), in order to turn the sport around and propel it in the right direction.

I don't know about you, but, I'm planning on the extra income continuing during my "golden years". :cool:

[acorn54]

i can't see doing anything but using past results in horseracing to predict what is going to happen in the future. isn't that what all of us are doing in some way? modeling based on sartin of brohamer is using the recent past to forcast the future. same with database research. also the validity of speed figures is based on past history.
i mean what else can you use?

[raybo]

i can't see doing anything but using past results in horseracing to predict what is going to happen in the future. isn't that what all of us are doing in some way? modeling based on sartin of brohamer is using the recent past to forcast the future. same with database research. also the validity of speed figures is based on past history.
i mean what else can you use?

There are other ways to be successful, however, the "big boys" that make millions per year, are all using databases and it's analysis as an integral cog in it's drivetrain. That should be sufficient to validate the value of DBs.

[Handiman]

Ray you are on the right track for sure. Because I have discovered why they call them "The Golden Years." It's because to get through them you better have a shitload of Gold...:lol:

And what a better way to line your pockets than with horse racing. Or I should say more fun than racing. A better way would be hitting the power ball or super lottery.

I will press on!

Handi :)

[raybo]

Ray you are on the right track for sure. Because I have discovered why they call them "The Golden Years." It's because to get through them you better have a shitload of Gold...:lol:

And what a better way to line your pockets than with horse racing. Or I should say more fun than racing. A better way would be hitting the power ball or super lottery.

I will press on!

Handi :)

"Keep on truckin'", dude!!!! I'm right behind ya.

[46zilzal]

I have a question for everyone, that cares anyway. Are we deluding ourselves with the use of databases?


YES, each race is an independent happening, never happening again. PERIOD

[Handiman]

While you are technically right 46, in a larger view you are also a bit mistaken I believe.

I've watched thousands of races and Jim Bradshaw was pretty much right. The race starts and one horse will get the lead. And then it either goes wire to wire, or another horse passes it up and wins.

Every race is the same. The spot where the early horse is alone on the lead may differ from race to race, but the mechanical action still happens. It's our job to figure out which horse is which. There in lies the problem or challenge, depending on how you look at it.

Handi :)

[raybo]

YES, each race is an independent happening, never happening again. PERIOD

If you're using knowledge, based on your previous experiences when you handicap, then you're using a database.

[hcap]

If for instance the top pick with Handifast can give me a break even or small profit, which so far, from what I have seen or heard, is just as good as the database gurus. If I am wrong, please straighten me out and show me the error of my ways.My 250 race test playing the top FAIR ODDS horse in non maiden races supports this.
Throw in only if one of the top 5 ranked jocks is riding, and a boost to positive ROI territory.
Found it out with a database set up in Excel. Also did exceptionally well in >8 furlong Turf races

[jonnielu]

While you are technically right 46, in a larger view you are also a bit mistaken I believe.

I've watched thousands of races and Jim Bradshaw was pretty much right. The race starts and one horse will get the lead. And then it either goes wire to wire, or another horse passes it up and wins.

Every race is the same. The spot where the early horse is alone on the lead may differ from race to race, but the mechanical action still happens. It's our job to figure out which horse is which. There in lies the problem or challenge, depending on how you look at it.

Handi :)

Databases will probably be a great tool one day, once someone figures out some meaningful data to enter.

jdl

[Vinnie]

Handi:

It sounds like all that you have mentioned in the body of your post is a major undertaking, but, man oh man would it make Handifast a completely devastating tool (heck, it already is) for the handicapper.

Handi, I hope that you have been feeling much better lately than you have in months past.

Wishing you nothing but continued health and success. :)

[acorn54]

There are other ways to be successful, however, the "big boys" that make millions per year, are all using databases and it's analysis as an integral cog in it's drivetrain. That should be sufficient to validate the value of DBs.


i hear alot on this board about the whales betting and making millions at u.s. tracks
can anybody document who these people are, like provide names and documentation like irs returns that validate, that in fact people are making millions on betting the u.s. racetracks?

[Houndog]

i hear alot on this board about the whales betting and making millions at u.s. tracks
can anybody document who these people are, like provide names and documentation like irs returns that validate, that in fact people are making millions on betting the u.s. racetracks?

Dave Schwartz mentioned Dana Parham in his webinar yesterday. Following is a link to an article on the web.

http://leftatthegate.blogspot.com/2007/12/whaling-stories.html

[Red Knave]

As well as storing past performance information, the true value of a database is storing the results of the calculations and rankings you generate for each race. For example, what were the FX and FW results and ranks in a race similar to this last February at AQU? Those are some of the things I find useful.

[Dave Schwartz]

One thing I find amazing about that article is that the author speaks of the annual return as 5-6% and says (basically) "What's the big deal?"

Of course he apparently isn't sharp enough to realize that an ROI of 5% when you turn the bankroll 125 times per year does not amount to 5% annually. It is actually more like 625% per year!

[ranchwest]

Independently give 1,000 people the same database (without preset software solutions) and every one will find something different. There is no single way to compartmentalize database users or the use of databases.

[Handiman]

Ranch....I believe you are right. It really boils down to each individual's perception and how that perception is focused.

Thanks to all who have contributed to this conversation.

I could barely sleep last night as I contemplated getting to work with a database that I create. And here is part of those thoughts that like Sugar Plums danced about my head.

I create a database. It looked to me to be a two pronged approach. Not so much the database itself, but what I put into it. So tell me what you think of this......an internal approach and an external approach.

Batch processor runs race cards. After running one iteration, the csv output is the summary based on default settings. But the settings don't really matter because what is used in database research here is the rankings of each factor's raw numbers. For example, how does the top ranked class factor horse do in whatever parameters you set.? Then maybe test the Fast column. Top ranked number in that column. And so on and so on.

Second prong, would be running the race cards with different settings. And then researching the fair odds horses and how they fare depending on what the setting's numbers are and which factor mix you use.

Now this is where I need a little help from you math gurus. The program has 20 factors I believe. So if I ran race cards through using a 5 point interval in each factor, the following would be true as to the number of combos.....
total points available for each factor....50........20 factors = 11 raised to the power of 20? I think this is right, but not sure. My algebra is a bit rusty. When you tell me I am full of you know what, don't forget the 0 is an interval also, thus the 11 intervals in 50 points.

Handi :)

[PaceAdvantage]

Databases will probably be a great tool one day, once someone figures out some meaningful data to enter.

jdlAmazing. Jonni apparently knows everyone on the planet running a database for handicapping, AND he knows what data they enter into their databases.

You are awesome!

[Roger]

Hi Handi,

In my personal opinion, maintaining a database for each track is a waste of time. I've modelled many tracks over the last couple of years and found
that what wins today or has won in the past will or may not win today.

I believe that the Brohammer approach using FX and FW is the only way to travel. The Brohammer model, in my opinion is by far the most credible for my puposes because it gives you an idea of how specific tracks will run and will give you a fair idea of an individual tracks bias. This in combination with running styles and pace shape has produced some good results for me.

The fact that some of the software providers that declined to participate in your contest with an excuse of being busy or whatever, tells me that their touted products are less then credible. I would believe the best way to promote ones product is to show all how well it does in real time.
I applaud those that did participate because it tells me that they believe in what they are selling.

Partsnut...I'm familiar with Brohamer's work..I have MPH'capping but find no reference to FW that you mention....Maybe I'm having a senior moment...could you elaborate.

thank's
Roger :confused:

[lsosa54]

Partsnut...I'm familiar with Brohamer's work..I have MPH'capping but find no reference to FW that you mention....Maybe I'm having a senior moment...could you elaborate.

thank's
Roger :confused:

Factor W is the average of Early and Sustained Pace. In some of Doc's readouts, it was labeled as AP or Average Pace.

[Roger]

Now, that rings a bell! Thank's.....BRAINFOG. :bang:
Roger,

[Partsnut]

Roger: Partsnut...I'm familiar with Brohamer's work..I have MPH'capping but find no reference to FW that you mention....Maybe I'm having a senior moment...could you elaborate.

It looks like Lou gave you the right answer.

[raybo]

i hear alot on this board about the whales betting and making millions at u.s. tracks
can anybody document who these people are, like provide names and documentation like irs returns that validate, that in fact people are making millions on betting the u.s. racetracks?

I would recommend asking Dave about this, I'm sure he has some pretty good sources for what you ask, as I believe, some of them are using software he created.

[Dave Schwartz]

Not my software. All of the teams of significance use software they developed themselves.

[hcap]

Ranch....I believe you are right. It really boils down to each individual's perception and how that perception is focused.

Thanks to all who have contributed to this conversation.

I could barely sleep last night as I contemplated getting to work with a database that I create. And here is part of those thoughts that like Sugar Plums danced about my head.

I create a database. It looked to me to be a two pronged approach. Not so much the database itself, but what I put into it. So tell me what you think of this......an internal approach and an external approach.

Batch processor runs race cards. After running one iteration, the csv output is the summary based on default settings. But the settings don't really matter because what is used in database research here is the rankings of each factor's raw numbers. For example, how does the top ranked class factor horse do in whatever parameters you set.? Then maybe test the Fast column. Top ranked number in that column. And so on and so on.

Second prong, would be running the race cards with different settings. And then researching the fair odds horses and how they fare depending on what the setting's numbers are and which factor mix you use.

Now this is where I need a little help from you math gurus. The program has 20 factors I believe. So if I ran race cards through using a 5 point interval in each factor, the following would be true as to the number of combos.....
total points available for each factor....50........20 factors = 11 raised to the power of 20? I think this is right, but not sure. My algebra is a bit rusty. When you tell me I am full of you know what, don't forget the 0 is an interval also, thus the 11 intervals in 50 points.

Handi :)

You can adjust weights in your database. Feed it the default settings and use a multiplier for each factor.
So you really only have to run a card thru Handifast once. Then use your database to test different configurations. Much faster.
I started to do this in Excel, (2007 can now query 1 million records. Pre-2007 65,000) but am waiting for the completion of batch processing.
Useful weights and configurations found should be easily be transferable back to Handifast

[ranchwest]

Ranch....I believe you are right. It really boils down to each individual's perception and how that perception is focused.

Thanks to all who have contributed to this conversation.

I could barely sleep last night as I contemplated getting to work with a database that I create. And here is part of those thoughts that like Sugar Plums danced about my head.

I create a database. It looked to me to be a two pronged approach. Not so much the database itself, but what I put into it. So tell me what you think of this......an internal approach and an external approach.

Batch processor runs race cards. After running one iteration, the csv output is the summary based on default settings. But the settings don't really matter because what is used in database research here is the rankings of each factor's raw numbers. For example, how does the top ranked class factor horse do in whatever parameters you set.? Then maybe test the Fast column. Top ranked number in that column. And so on and so on.

Second prong, would be running the race cards with different settings. And then researching the fair odds horses and how they fare depending on what the setting's numbers are and which factor mix you use.

Now this is where I need a little help from you math gurus. The program has 20 factors I believe. So if I ran race cards through using a 5 point interval in each factor, the following would be true as to the number of combos.....
total points available for each factor....50........20 factors = 11 raised to the power of 20? I think this is right, but not sure. My algebra is a bit rusty. When you tell me I am full of you know what, don't forget the 0 is an interval also, thus the 11 intervals in 50 points.

Handi :)

I hope I'm understanding.

My suggestion is to have several sets of tables:

1) Raw data, such as what you download from a data provider.

2) Results

3) Computations (including rankings)

4) Multipliers for the computations

5) Totals after the multipliers are applied.

If you leave the raw data unaltered, you can always re-process that data to alter (adjust) the computations and multipliers.

[atlasaxis]

One thing I find amazing about that article is that the author speaks of the annual return as 5-6% and says (basically) "What's the big deal?"

Of course he apparently isn't sharp enough to realize that an ROI of 5% when you turn the bankroll 125 times per year does not amount to 5% annually. It is actually more like 625% per year!

Excellent point!

[46zilzal]

Correlation does not now, or ever has meant, cause and effect.

If one studies the history of science you will discover this oft quoted and oft found erroneous belief that with JUST THE RIGHT rules, all you have to do is pour in the variables and EVERYTHING will be understood (Newtonian reductionism).

With what we have learned from chaos, sensitive dependence on initial conditions and RANDOMNESS, these conclusions of databases garner less and less ability to do anything other than review what HAPPENED and less and less ability to predict what will.

The charts comparing the stock market by the leading predictors for two separate 5 year periods in the book FOOLED by Randomness wake one up to these realities. Same people, same prediction factors, WILDLY different results.

EACH race is mutually exclusive to the next bringing together a field in varying stages of form cycles over tracks that vary in their speed characteristics etc etc

[Tom]

The only thing that matters is the results.
People are using and making money using dbs as tools.
There is nothing you can post that changes that.
You keep harping on the EL graphs - how can they be predictive if every race is unique unto itself.

46, you are using a db and have been for quite some time.

[Dave Schwartz]

Not my software. All of the teams of significance use software they developed themselves.

I got an email questioning this post from one of the "mini-whales."


Permit me to clarify -

None of the "top 6" use my software. That is, nobody wagering $100 million or more. Several of the "mini-whales" - those wagering (say) $10m or more do use HSH and several others I have consulted with.


Regards,
Dave Schwartz

[harntrox]

From a custom database, 2 win bets *every race* at Santa Anita Meet only
Santa Anita ROI by Race Type (http://harntrox.com/sanroi.jpg)

By factoring 50 common-sense 'angles' into each of the 5 handicapping methods. Only the final blended Handicap analysis is shown here; each of the other 4 methods require a seperate report.

If this is a delusion, it sure is interesting that 5 of 7 tracks in the database show similar positive ROI (+5-17 %). Win % ranges from 40-51 % on the Socal tracks i been playing since 2008.

re: Handicapping

So many times i have heard the old story ' Works on paper, but not at the track.' I havent experienced that for some time now, and its refreshing to not feel at odds with the percentages, and know that im on the positive side of things, even if only slightly with the 2 win bet thing.

re: Betting discipline
How i wish i could make my computer do something other than pattern bet ...

But How do you design and code this: ?
The chaos that arises with all the possibilities of parrallel vs. series bets, conditional betting, money management - and avoiding greedy hunch plays -
this is an entirely different subject than either handicapping or analysis.

Instead i bet the entire database in a simulation using a constant pattern, and then run a filter to try to improve the ROI by using sprint vs. route or surface etc. I usually get the daily double or exacta going 2 deep in the positive range of +20-80% ROI.

If you argue that using the past to predict the future is akin to rolling lopsided dice, consider the http://en.wikipedia.org/wiki/Law_of_large_numbers numbers is not the Law of Averages/gamler's fallacy/Correlation doesnt prove causation. Indeed: With magnitudes of oversampling a proven mean average can be achieved for each factor; comparisons to that mean indicate trends over time that can be quantified as having velocity in positive or negative direction.

Constant regression testing shows changes in some of the factors are needed over the course of each meet to correct trends that were 'noise' or cross-cancelling a higher order variable. This is what the velocity is used for.

[ranchwest]

Correlation does not now, or ever has meant, cause and effect.



Screw cause and effect. Winning is the objective. We don't need to know the cause. We only need to correlate factors to winning results.

If nobody (including you) ever found any correlations, then nobody (including you) could ever be a consistent winner.

[46zilzal]

The Drunkard's Walk: How Randomness Rules Our Lives by Leonard Mlodinow

Let this fellow tell you first hand why the over-dependence upon these sources are full of holes and bogus conclusions

[ranchwest]

The Drunkard's Walk: How Randomness Rules Our Lives by Leonard Mlodinow

Let this fellow tell you first hand why the over-dependence upon these sources are full of holes and bogus conclusions

So, randomness trumps pace and your handicapping sucks, right?

[46zilzal]

Once you comprehend its significant limitations as a predicting tool you don't rely on it as heavily ever again.

I was taken aback with what I discovered here within.

Couple that with that the studies exposed in the book Smart Choices by Hammond, Keeney and Raiffa (psychological aspects in decision making) and you might recognize the mass of erroneous influences go into many decisions

[castaway01]

Screw cause and effect. Winning is the objective. We don't need to know the cause. We only need to correlate factors to winning results.

If nobody (including you) ever found any correlations, then nobody (including you) could ever be a consistent winner.

I know zillyzally occasionally makes good points, but there's so much "nothing works for anyone except my specific method that only I know" that it's a waste of time debating him. He's got blinkers on.

[46zilzal]

I know zillyzally occasionally makes good points, but there's so much "nothing works for anyone except my specific method that only I know" that it's a waste of time debating him. He's got blinkers on.
I'd ask you to quote the highlighted part but then you would never find it. OVER-reliance on the subjectiveness of database conclusions are well debunked in the literature, READ IT FOR YOURSELF

[Handiman]

46,

Based on what you say, can you tell me what approach you honestly take when handicapping then. I am serious in asking, as I don't understand how in this game you can handicap without the use of previous run race information.

And if you don't rely on it as heavily as you used to, then what do you rely on...Tarot Cards, Dice, Names, Colors? Again I am serious in asking this, because by not relying on PP's as heavily insinuates that you have replaced that portion of your handicapping with something else. Or if not, then a void has been created which you have apparently decided to not fill.

Please elaborate.

Thank You,
Handi :)

[ranchwest]

I'd ask you to quote the highlighted part but then you would never find it. OVER-reliance on the subjectiveness of database conclusions are well debunked in the literature, READ IT FOR YOURSELF

So if I put pace numbers in a database, you can debunk those numbers. How did those numbers suddenly go south when they went into a database?

Do you have any clue as to how absurd your assertion is?

[ranchwest]

Once you comprehend its significant limitations as a predicting tool you don't rely on it as heavily ever again.

I was taken aback with what I discovered here within.

Couple that with that the studies exposed in the book Smart Choices by Hammond, Keeney and Raiffa (psychological aspects in decision making) and you might recognize the mass of erroneous influences go into many decisions
I have no idea of what you are saying. Am I supposed to magically discern what "its" references?

[46zilzal]

So if I put pace numbers in a database, you can debunk those numbers. How did those numbers suddenly go south when they went into a database?

Do you have any clue as to how absurd your assertion is?
read a review of the book and then maybe you might understand that position I learned from it.
http://www.nytimes.com/2008/06/08/books/review/Johnson-G-t.html

"The brain, no matter how well schooled, is just plain bad at dealing with randomness and probability. Confronted with situations that require an intuitive grasp of the odds, even the best mathematicians and scientists can find themselves floundering."

[ranchwest]

read a review of the book and then maybe you might understand that position I learned from it.
http://www.nytimes.com/2008/06/08/books/review/Johnson-G-t.html

"The brain, no matter how well schooled, is just plain bad at dealing with randomness and probability. Confronted with situations that require an intuitive grasp of the odds, even the best mathematicians and scientists can find themselves floundering."

I love discussing things with you. It's just like when I was a kid on the merry-go-round. A long ride that never gets anywhere new.

You never answer a question. You just ramp up the merry-go-round.

[raybo]

Handiman,

Continue with your path. Don't let this "thread hijacker" dampen your beliefs. He only has one opinion, his. There are many of us, most of us, I'm sure, who do not share his beliefs, and no amount of "hijacking" will change our experiences or knowledge.

[ranchwest]

Handiman,

Continue with your path. Don't let this "thread hijacker" dampen your beliefs. He only has one opinion, his. There are many of us, most of us, I'm sure, who do not share his beliefs, and no amount of "hijacking" will change our experiences or knowledge.

I just wish I could identify his opinion.

[46zilzal]

I love discussing things with you. It's just like when I was a kid on the merry-go-round. A long ride that never gets anywhere new.

You never answer a question. You just ramp up the merry-go-round.
well if never READ dissension how can you comment on it?

Astounding

[raybo]

I just wish I could identify his opinion.

I'll try to help.

His "opinion" is that his opinion is the only one that counts, and, to him, his is.

Good thing he is he and we are we, huh?

The track's full of them, which is a good thing!

[46zilzal]

AMAZING, no one ever reads the dissenting opinion of a world renown professor as to the psychological aspects of putting too much faith into the conclusions of statistics and without as so much of a foray into HIS published position, it has to be incorrect?

amazingly moronic

He is in print mind you not you fellows

look at this http://fora.tv/2009/06/01/The_Drunkards_Walk_Leonard_Mlodinow

or this
http://network.nationalpost.com/np/blogs/wealthyboomer/archive/2009/09/22/the-drunkard-s-walk-how-randomness-amp-luck-affects-our-investments.aspx

[ranchwest]

AMAZING, no one ever reads the dissenting opinion of a world renown professor as to the psychological aspects of putting too much faith into the conclusions of statistics and without as so much of a foray into HIS published position, it has to be incorrect?

amazingly moronic

He is in print mind you not you fellows

look at this http://fora.tv/2009/06/01/The_Drunkards_Walk_Leonard_Mlodinow

or this
http://network.nationalpost.com/np/blogs/wealthyboomer/archive/2009/09/22/the-drunkard-s-walk-how-randomness-amp-luck-affects-our-investments.aspx

Nobody has made any assertions about the author. The problem here is that you never actually discuss his ideas. You just point to the book and suggest that it is somehow germaine.

[46zilzal]

Another source is Taleb, same work same conclusions.
http://www.fooledbyrandomness.com/

from wikipedia:The book was selected by Fortune as one of the 75 "Smartest Books of All Time." Fortune is not a slouch of a periodical either.

The book's name, Fooled by Randomness, has also become an idiom in English used to describe when someone sees a pattern where there
is just random noise.

Taleb sets forth the idea that modern humans are often unaware of the existence of randomness. They tend to explain random outcomes as non-random.

Human beings:

1. overestimate causality, e.g., they see elephants in the clouds instead of understanding that they are in fact randomly shaped clouds that appear
to our eyes as elephants (or something else);

2. tend to view the world as more explainable than it really is. So they look for explanations even when there are none.

Other misperceptions of randomness that are discussed include:

* Survivorship bias. We see the winners and try to "learn" from them, while forgetting the huge number of losers.

* Skewed distributions. Many real life phenomena are not 50:50 bets like tossing a coin, but have various unusual and counter-intuitive
distributions. An example of this is a 99:1 bet in which you almost always win, but when you lose, you lose all your savings. People can easily be fooled
by statements like "I won this bet 50 times". According to Taleb: "Option sellers, it is said, eat like chickens and go to the bathroom like elephants", which is to say, option sellers may earn a steady small income from selling the options, but when a disaster happens they lose a fortune.

you can read a bit of it here:http://books.google.ca/books?id=uMwMyDQog00C&printsec=frontcover&dq=fooled+by+randomness&source=bl&ots=t4J_NIB2PZ&sig=HdCa1TgQyMASK5_wtwgUVHxk3pQ&hl=en&ei=zPZ9S77-L4ScsgPz_tz8Cw&sa=X&oi=book_result&ct=result&resnum=5&ved=0CCIQ6AEwBA#v=onepage&q=&f=false

a good review: http://plus.maths.org/issue20/reviews/book1/index.html

the importance of understanding the structure of random events, their significance and, especially, insignificance. He clearly sees that this understanding is more important than actual calculations: "Mathematics is principally a tool to meditate, rather than to compute".

There are many ways of being the fool of randomness. One, as here, is to fail to predict the rare event. Nothing can be more certain than that the unexpected will happen sooner or later, but lulled into a sense of security by the periods of relative calm between, people forget to allow for it. Another is to see significance in some random pattern. Taleb explains with crystal clarity why the more often you look at some fluctuating quantity (the value of your share portfolio, for example), the less meaning your observations have. Yet he sees traders who watch prices move up and down in real time on screen - the changes are so small as to be completely random - and think they are learning something.

another review: http://docs.google.com/viewer?a=v&q=cache%3AmwR_xyot244J%3Awww.qaiworldwide.org%2Fpd f_files%2Fbook_0209.pdf+fooled+by+randomness&hl=en&gl=ca&sig=AHIEtbSgD6_z73zXOi9VdtC6zA5GLhz1bw&pli=1

[Greyfox]

The book's name, Fooled by Randomness, has also become an idiom in English used to describe when someone sees a pattern where there
is just random noise.

Taleb sets forth the idea that modern humans are often unaware of the existence of randomness. They tend to explain random outcomes as non-random.



Most racing outcomes are not based on random chance, albeit there may be random elements within a race.

[46zilzal]

http://bexhuff.com/2010/01/book-review-the-drunkards-walk

The author wrote this like a history book about the field of statistics, and how it evolved (slowly) over they years. But, every time the author introduces a new concept in statistics, he also shares real-world situations where people made terrible mistakes because they didn't understand these basic principles. It highlights very well that in general, people are terrible at recognizing randomness, and thus will always be controlled by it!

[douglasw32]

I have often pondered something I think is on this point, but we should start a new thread for it ?

i.e, most of my "calculations" come up with a winner in the top 4 say 40% of the time, so 60% of the time they are dead wrong but the number then reaches 1-16 others, I have often thought that totally tossing the top 4 and focusing on the larger group could produce a larger ROI.

But how? not sure.

[Greyfox]

I have often pondered something I think is on this point, but we should start a new thread for it ?

i.e, most of my "calculations" come up with a winner in the top 4 say 40% of the time, so 60% of the time they are dead wrong but the number then reaches 1-16 others, I have often thought that totally tossing the top 4 and focusing on the larger group could produce a larger ROI.

But how? not sure.

40 % looks awfully low to me. Even if your took the 4 high Beyers you'd do better.

[Tom]

Agree, I get my better factors hitting top 4 75%-80%
I don't even look at those that hit under 70% unless they have a huge ROI.

[Handiman]

46,

I applaud your enthusiasm. I certainly give way to the professor's credentials. But I believe his conclusions as you have somewhat put forth here, are a bit wrong and I'll tell you why.

Acceptance of Randomness is an arbitrary tearing of a chunk out of time and space and boxing the activity therein, and labeling it Random. The reason that doesn't hold true, is because you start with a false beginning and end with a false end.

What is happening right now in this world is the result of movement by objects at the beginning of time. Granted this is an exceptionally Macro view. But for arguments sake, let's take a chunk of time and space and examine it a bit closer.

The question is, did I marry the woman who is my wife now, back in 1975 as the result of randomness. Could it have been any woman? Or on the other hand could it have equally been probable that I would have married a man.

While going my way during my daily life, I met and dated many women. Randomness would almost dictate that I would have found at least one of them prior to my wife coming along, to marry since I had reached the age of majority, was sexually driven and adhered philosophically to the idea that marriage was a desirable thing.

But that didn't happen. Did my wife to be, randomly enter into my life. The answer is No. Due to factors happening in her life she found it necessary to return to our home town. Was it randomness that I happened to be there at the same time. No, because of factors in my life, I found it prudent to return to our home town at the same time my wife did.

We crossed paths as the result of external forces exerting their influence on us. So we were both in the same town where we could not escape seeing each other, because of things that had happened to us at an earlier age in that very same town. Was it randomness that brought us together. No, because of experiences that shaped our personalities, we found it to be very desirable to spend time together and eventually get married.

To put what I just said more succinctly, we were brought together as a result of the 'Butterfly Effect' and not because of Randomness. Randomness is nothing more than the perception that things happen for no apparent reason, due to the fact we are not privy to the external forces at work that cause actions to happen.

I may be wrong, but it makes more sense to me, than giving credit to a non ordered jumbled mass.

Respectfully,
Handi :)

[ranchwest]

Another source is Taleb, same work same conclusions.
http://www.fooledbyrandomness.com/

from wikipedia:The book was selected by Fortune as one of the 75 "Smartest Books of All Time." Fortune is not a slouch of a periodical either.

The book's name, Fooled by Randomness, has also become an idiom in English used to describe when someone sees a pattern where there
is just random noise.

Taleb sets forth the idea that modern humans are often unaware of the existence of randomness. They tend to explain random outcomes as non-random.

Human beings:

1. overestimate causality, e.g., they see elephants in the clouds instead of understanding that they are in fact randomly shaped clouds that appear
to our eyes as elephants (or something else);

2. tend to view the world as more explainable than it really is. So they look for explanations even when there are none.

Other misperceptions of randomness that are discussed include:

* Survivorship bias. We see the winners and try to "learn" from them, while forgetting the huge number of losers.

* Skewed distributions. Many real life phenomena are not 50:50 bets like tossing a coin, but have various unusual and counter-intuitive
distributions. An example of this is a 99:1 bet in which you almost always win, but when you lose, you lose all your savings. People can easily be fooled
by statements like "I won this bet 50 times". According to Taleb: "Option sellers, it is said, eat like chickens and go to the bathroom like elephants", which is to say, option sellers may earn a steady small income from selling the options, but when a disaster happens they lose a fortune.

you can read a bit of it here:http://books.google.ca/books?id=uMwMyDQog00C&printsec=frontcover&dq=fooled+by+randomness&source=bl&ots=t4J_NIB2PZ&sig=HdCa1TgQyMASK5_wtwgUVHxk3pQ&hl=en&ei=zPZ9S77-L4ScsgPz_tz8Cw&sa=X&oi=book_result&ct=result&resnum=5&ved=0CCIQ6AEwBA#v=onepage&q=&f=false

a good review: http://plus.maths.org/issue20/reviews/book1/index.html

the importance of understanding the structure of random events, their significance and, especially, insignificance. He clearly sees that this understanding is more important than actual calculations: "Mathematics is principally a tool to meditate, rather than to compute".

There are many ways of being the fool of randomness. One, as here, is to fail to predict the rare event. Nothing can be more certain than that the unexpected will happen sooner or later, but lulled into a sense of security by the periods of relative calm between, people forget to allow for it. Another is to see significance in some random pattern. Taleb explains with crystal clarity why the more often you look at some fluctuating quantity (the value of your share portfolio, for example), the less meaning your observations have. Yet he sees traders who watch prices move up and down in real time on screen - the changes are so small as to be completely random - and think they are learning something.

another review: http://docs.google.com/viewer?a=v&q=cache%3AmwR_xyot244J%3Awww.qaiworldwide.org%2Fpd f_files%2Fbook_0209.pdf+fooled+by+randomness&hl=en&gl=ca&sig=AHIEtbSgD6_z73zXOi9VdtC6zA5GLhz1bw&pli=1

You still have not addressed how your devotion to pace handicapping reconciles with your devotion to these works on randomness.

Will you ever muster the fortitude to actually discuss an issue at hand? I think not.

How would YOU suggest to predict the rare event (in horse racing)?

[46zilzal]

46,

I applaud your enthusiasm. I certainly give way to the professor's credentials. But I believe his conclusions as you have somewhat put forth here, are a bit wrong and I'll tell you why.

I may be wrong, but it makes more sense to me, than giving credit to a non ordered jumbled mass.


BOTH sources I quote review the semblance of required order you discuss here. NOT everything as much as we would like to believe, has an understandable ordered outcome. READ THESE BOOKS and you will see what I mean as it is human nature to place order over everything in our life when it is actually not there.

[46zilzal]

You still have not addressed how your devotion to pace handicapping reconciles with your devotion to these works on randomness.

Will you ever muster the fortitude to actually discuss an issue at hand? I think not.

How would YOU suggest to predict the rare event (in horse racing)?

Randomness is random. it's influence changes as does day and night. Sensitive dependence on initial conditions varies out of control and it not able to be predicted. IT IS ALWAYS there in the background.

each exotic put together includes the illogical and the random. NO race is ever viewed with a singular outcome, EVER

[ranchwest]

BOTH sources I quote review the semblance of required order you discuss here. NOT everything as much as we would like to believe, has an understandable ordered outcome. READ THESE BOOKS and you will see what I mean as it is human nature to place order over everything in our life when it is actually not there.

More merry-go-round answers. Do you have any beliefs of your own?

In horse racing, order does exist -- it just often goes unrecognized.

[ranchwest]

Randomness is random. it's influence changes as does day and night. Sensitive dependence on initial conditions varies out of control and it not able to be predicted. IT IS ALWAYS there in the background.

each exotic put together includes the illogical and the random. NO race is ever viewed with a singular outcome, EVER

You still have not addressed how your devotion to pace handicapping reconciles with your devotion to these works on randomness.

[46zilzal]

typo, each exotic brings together the logical and the random

[ranchwest]

typo, each exotic brings together the logical and the random

And, this relates to databases in this manner _______________________.

[Handiman]

I still believe that the perception of Randomness is due to a lack of data which lies at the foundation of causation.


We live in a very ordered universe. The thought that when a meteor strikes the earth, that where all the parts land upon the breakup are strictly random, which is what many people believe, just isn't true. The velocity, the chemical structure and many other variables dictate where those pieces will land.

Just because we don't have all the information doesn't not make it a random result.

Handi

My wife is making me hurry so this may not make sense.

[douglasw32]

Agree, I get my better factors hitting top 4 75%-80%
I don't even look at those that hit under 70% unless they have a huge ROI.

And GreyFox

My Bad... I was just throwing out a hypothetical.
to make a point. however bad it was =(

[46zilzal]

I still believe that the perception of Randomness is due to a lack of data which lies at the foundation of causation.

We live in a very ordered universe.

.
Until Niels Bohr came along, there was that Newtonian certainty belief that ALL can be known only if you have the correct law and the correct data. Heisenberg and others crapped all over that belief. Then Mandelbrot came along, Gleick's CHAOS, Waldrop's COMPLEXITY came forth. As knowledge continues we are finding it much different than the certainties of reductionism and determinism. The sum is much different when together than can be reconstructed by looking at each part.

"if I only had MORE data" is the mantra one hears horse players spout and the studies put out in the great book BLINK counter that position with vigor. More only complicates and lulls the observer in the belief that certainty is closer when it really it is just fuzzier.

Read Capra's the Tao of Physics and you will see that it changed in the 1920's and never had such certainty again.

Once that gate opens only thousands of things can happen and often do.THAT is outside of range of any long term predictor so you must embrace the change of the I CHING embodied in the randomness of each horse race along with that which can be predicted. One without the other, knowing what we know today, is not the complete picture.

[ranchwest]

I still believe that the perception of Randomness is due to a lack of data which lies at the foundation of causation.


We live in a very ordered universe. The thought that when a meteor strikes the earth, that where all the parts land upon the breakup are strictly random, which is what many people believe, just isn't true. The velocity, the chemical structure and many other variables dictate where those pieces will land.

Just because we don't have all the information doesn't not make it a random result.

Handi

My wife is making me hurry so this may not make sense.

Yes, ignorance can easily be explained as randomness.

I don't believe we always need to know causation. For instance, I don't need to know who the drug trainers are if I understand all of the symptoms efficiently. I just need to know the relevant factors.

[Tom]

Glad the hear that, Doug! ;)

Random is a word used to describe that which you cannot understand trough lack of knowledge. Random to one might be expected to another. 46's total lack of what a database is probably makes things seem random to him.
Many suspect speed figures appeared to me to be random until I started looking at the effects of pace on final times.

Many longshot winners appear to be random, until you look at their workout ratings or their pedigree ratings. Or thier trainer's historical patterns.

46 wrongly assumes eveything about a race is known. It is not.

[ranchwest]

Glad the hear that, Doug! ;)

Random is a word used to describe that which you cannot understand trough lack of knowledge. Random to one might be expected to another. 46's total lack of what a database is probably makes things seem random to him.
Many suspect speed figures appeared to me to be random until I started looking at the effects of pace on final times.

Many longshot winners appear to be random, until you look at their workout ratings or their pedigree ratings. Or thier trainer's historical patterns.

46 wrongly assumes eveything about a race is known. It is not.

Man, this post is so good I almost spit my 7-Up. (really, no sarcasm)

[46zilzal]

http://en.wikipedia.org/wiki/Black_swan_theory one of the best ideas on how randomness is expressed.

[ranchwest]

http://en.wikipedia.org/wiki/Black_swan_theory one of the best ideas on how randomness is expressed.

I'd get a warmer feeling about your posts if you would use multiple colors, like a traditional merry-go-round.

[headhawg]

You've again hijacked another thread with your ramblings, 46z. As we here are just horseplayer simpletons I suggest you take the findings from your Google searches to a site better suited to your skills.

http://www.imvu.com/rooms/the-super-secret-pseudo-intellectual-wannabe-clubhouse

[johnhannibalsmith]

So, in essence, the key to successful handicapping is to give greater weight to statistical outliers than to statistical norms?

If I had known this was going to end up in some reference to Black Swan Theory, I'd surely have not wasted my time attempting follow the thoughts of someone that condemned the methods of others by insisting on employing a theory that assumes that the realization of certain events can not be predicted, but only recognized after-the-fact and explained at that point.

Once again, there really was no point in any of this other than to make everyone completely aware of how many offbeat writings about things totally unrelated to the context of the subject you have digested and then "randomly" applied to life.

Not everything can be predicted by using data. Some data will lead you to conclusions that are defied by the actual result of the event.

Is that about the extent of the obvious bullshit we already knew without all of the arcane references and self-congratulatory study guides?

[ranchwest]

John, I think that was established about 18,000 posts ago.

[johnhannibalsmith]

John, I think that was established about 18,000 posts ago.

My apologies, I really was holding out hope that this one time he would come with something that could actually be applied in a novel, useful manner.

[46zilzal]

title is DATABASE DELUSIONS

published data supports that same conclusion

How is that off topic?

Review of Drunkards Walk by Barron's magazine "Even if you begin the Drunkards Walk as a skeptic, by the time you reach the final pages, you will gain an understanding- IF NOT ACCEPTANCE- of the intuitively improbably ways that probability biases the outcomes of life's uncertainties."

p.212 "we miss the effects of randomness in life because when we assess the world, we tend to see what we expect to see. That is why although there is sometimes little difference in ability between a wildly successful person and one who is not as successful, there is usually a big difference in how they are viewed. (See this all the time in those who think man made class structural boundaries on somehow like mountain ranges)

Charles Perrow p. 204; "in complex systems we should expect minor factors we can usually ignore by chance sometimes cause major incidents."

Nobel prize winner max Born p. 195 "Chance is a more fundamental concept than causality." Determinism fails to meet the requirements for predictability alluded to by Laplace for several reasons. First, society is not governed by definite and finite laws the way physics is. Second, even if we could uncover the laws of human affairs, as Quetelet attempted to do, it is impossible to precisely know or control the circumstances of life. Third, human affairs are so complex that it is doubtful we could carry out the necessary calculations even if we understood the laws and possessed the data. As a result, determinism is a poor model for the human experience."

[Charlie D]

In horse racing, order does exist -- it just often goes unrecognized.


An accurate statement, I think.

[ranchwest]

My apologies, I really was holding out hope that this one time he would come with something that could actually be applied in a novel, useful manner.

If you're looking for some random act of normalcy, you've looked in the wrong direction. Not gonna happen.

Can we get back on topic?

[46zilzal]

There are many things in this life that discoveries ALTER........ things change whether we like it or not.

Of course there is order, just as there is disorder, the Yin and the Yang, the either or, the and/if the duality of all nature if you have the intellectual curiosity to notice it. Capra talks at length about it in The Tao of Physics.

This is the end product of what Dr. Sartin introduced all who cared to read about it many moons ago stating that there is a little bit of the opposite in any horse's style and defines, in algorithms, that into several of the calculations of early/late balance.

This revelation is just an extension of that.

[ranchwest]

title is DATABASE DELUSIONS

published data supports that same conclusion

How is that off topic?

Review of Drunkards Walk by Barron's magazine "Even if you begin the Drunkards Walk as a skeptic, by the time you reach the final pages, you will gain an understanding- IF NOT ACCEPTANCE- of the intuitively improbably ways that probability biases the outcomes of life's uncertainties."

p.212 "we miss the effects of randomness in life because when we assess the world, we tend to see what we expect to see. That is why although there is sometimes little difference in ability between a wildly successful person and one who is not as successful, there is usually a big difference in how they are viewed. (See this all the time in those who think man made class structural boundaries on somehow like mountain ranges)

Charles Perrow p. 204; "in complex systems we should expect minor factors we can usually ignore by chance sometimes cause major incidents."

Nobel prize winner max Born p. 195 "Chance is a more fundamental concept than causality." Determinism fails to meet the requirements for predictability alluded to by Laplace for several reasons. First, society is not governed by definite and finite laws the way physics is. Second, even if we could uncover the laws of human affairs, as Quetelet attempted to do, it is impossible to precisely know or control the circumstances of life. Third, human affairs are so complex that it is doubtful we could carry out the necessary calculations even if we understood the laws and possessed the data. As a result, determinism is a poor model for the human experience."

This is off topic because it just simply is not the topic, as evidenced by the fact that you will not address the question of how you would reconcile the contradiction of your use of pace handicapping with your belief in randomness, which would go to the heart of database usage.

Let's face it, you're a thread hijacker extraordinaire.

[ranchwest]

There are many things in this life that discoveries ALTER things change whether we like it or not.

Huh?

[46zilzal]

Huh?
I all to often forget to whom I am speaking but Heraclitus said it thousands of years ago: "you can never step in the same river twice".... because it is always changing (whether we like it or not)

Oriental philosophers understand that and embrace it as the basis for the Tao.

Maybe their positions of long long ago will triumph one day

[ranchwest]

I all to often forget to whom I am speaking but Heraclitus said it thousands of years ago: "you can never step in the same river twice".... because it is always changing (whether we like it or not)

Oriental philosophers understand that and embrace it as the basis for the Tao.

Maybe their positions of long long ago will triumph one day

Your sentence did not make sense from a grammatical perspective.

So, what you are saying is that horse racing is always random and always different and thereby completely, undeniably unpredictable? Thus, your contention that databases are useless. Is that about it?

[46zilzal]

Your sentence did not make sense from a grammatical perspective.

So, what you are saying is that horse racing is always random and always different and thereby completely, undeniably unpredictable? Thus, your contention that databases are useless. Is that about it?
you do not read well do you? Never once hinted any such thing.

One must accept the duality of nature in both the predictive and the random as partners in wager creation.

Seems I always wind up talking to "the world is either black or white" republicans when I am suggesting a middle of the road position.

[ranchwest]

you do not read well do you? Never once hinted any such thing.

One must accept the duality of nature in both the predictive and the random as partners in wager creation.

Seems I always wind up talking to "the world is either black or white" republican when I am suggesting a middle of the road position.

I seem to read very well. You never hinted of any basis for a position related to the topic at hand, thus my effort to entice you to go somewhere with your posts.

So, now, you have acknowledged that there are predictive elements of horse racing. So, why are databases a bad idea? Just a reminder, the topic is databases, not wager creation.

And, FWIW, I'm not a Republican.

[46zilzal]

The race in front of you is specific

The conclusions of a database are averaged

The influences of any one factor on any one contest are lost in averaging...often the Black Swan is what happens.

[ranchwest]

The race in front of you is specific

The conclusions of a database are averaged

The influences of any one factor on any one contest are lost in averaging...often the Black Swan is what happens.

:lol: :lol: :lol: :lol:

You know absolutely nothing about databases, which is what I suspected all along, thus your avoidance of the topic.

Databases contain sets of data. Period. There is no averaging required or implied.

We had to go through 7 pages of posts to get to this point. Whew!

[46zilzal]

:lol: :lol: :lol: :lol:

You know absolutely nothing about databases, which is what I suspected all along, thus your avoidance of the topic.

Databases are sets of data. Period. There is no averaging required or implied.
I took database courses at universities and used dBase III, IV and Access for years in both my practice and handicapping. I came to the conclusion that it worked well for the former and not at all for the latter. Relational database were far more productive than the flat ones.

Statistical analysis told me who I treated, their age ranges, what billing were the most prominent, but the numerical basis of racing evaluations only allowed me a slight understanding of what were phenomena times on the training track as compared to the main oval and the more I learned about work outs from actually attending them, that was completely negated as well.

There are many things in this world that have to be evaluated holistically and do not allow some construction by factors from the ground up. A horse performance has as much to do with today's match up as it does any historical references to said animal.

I could throw all the physical characteristics and labs of many many patients into any database but would prefer to evaluate what I saw first hand in the office. The current mania for cholesterol levels for example to more based upon pharmaceutical standards than the reality of patients specific histories.

[ranchwest]

I took database courses at universities and used dBase III, IV and Access for years in both my practice and handicapping. I came to the conclusion that it worked well for the former and not at all for the latter.

That's because you knew the questions for the former and not for the latter. Never did you find that dBase or Access required averaging.

FYI, I've been a professional programmer for over 20 years and have been programming for nearly 40 years. I started with punch cards and now I'm working in Oracle. I have worked with the dBase file structure for many years and still do.

[ranchwest]

I took database courses at universities and used dBase III, IV and Access for years in both my practice and handicapping. I came to the conclusion that it worked well for the former and not at all for the latter. Relational database were far more productive than the flat ones.

Statistical analysis told me who I treated, their age ranges, what billing were the most prominent, but the numerical basis of racing evaluations only allowed me a slight understanding of what were phenomena times on the training track as compared to the main oval and the more I learned about work outs from actually attending them, that was completely negated as well.

There are many thing in this world that have to be evaluated holistically and do not allow some construction by factors from the ground up. A horse performance has as much to do with today's match up as it does any historical references to said animal.

So, you're going to examine today's match up by _______________.

[DeanT]

There's a cool Black Swan in Lost.

Also, I once saw a Black Swan near Niagara Falls, but after thinking about it now, I think he was just a white swan with mud on him.

[46zilzal]

So, you're going to examine today's match up by _______________.
Incremental velocity deceleration ratios as compared to a projected fulcrum pace.

[ranchwest]

Incremental velocity deceleration ratios as compared to a projected fulcrum pace.

And those are calculated from ________________.

[ranchwest]

Where's Greg?

[46zilzal]

And those are calculated from ________________.
The cpu in my laptop

[Charlie D]

And those are calculated from ________________.



Running lines in a horse's past performance record.

[ranchwest]

And those are calculated from ________________.

I have to go to sleep now, so I'll just go ahead and fill in the blank.

Incremental velocity deceleration ratios can be calculated from past performances, which can be found in the data furnished by most data providers. That data can be stored in a, tah dah, database. The incremental velocity deceleration ratios could then be stored in a, tah dah, database.

Projected fulcrum pace could be measured through data stored in a, tah dah, database and once determined could be stored in a, tah dah, database.

By also storing selections based on these criteria and the results of races in which selections were made, one could also evaluate the effectiveness of the selections being made.

So, again, why is there something inherently wrong with a database?

[ranchwest]

The cpu in my laptop

The CPU only knows 1's and 0's.

[Greyfox]

46Zil doesn't see past performance lines as "historical references."

[Charlie D]

I do believe the Pirco members like Doc Sartin, Tom Brohamer made great use of data, in fact i do believe Doc Sartin was a great believer in studying past records of not only horses, but racetracks and wagering.

[Charlie D]

Data mining is a particular data analysis technique that focuses on modeling and knowledge discovery for predictive rather than purely descriptive purposes




http://en.wikipedia.org/wiki/Data_analysis



We as handicappers are all data mining from databases, I think.

[Handiman]

I understand what 46 is saying. But again I propose that all he says is true in a static state that is all at once sent in motion. Such would be the Big Bang theory. BANG....the universe is created and where all that mass goes or ends up is just random. The same to a certain extent would be the dispersal of a rack of pool balls upon the break.

If you notice, the balls never, as far as we can discern, end up in the same place after a break. Is that the result of randomness. Of course not. It is the result of many factors interacting ....the speed and force behind the cue ball, the proximity of each ball to one another, the thickness of the felt, the makeup of the rail cushion and so forth.

On top of that, we do not live in a static state. We are in flux all the time. The time my mailman comes by tomorrow is not random, although at first glance it may seem so, if all you look at are the times he delivers the mail over any recorded time period. But his family interactions in conjunction with his personal habits along with what may take place at any given moment while he readies for his route all dictate what time he will arrive at my mailbox. And that's leaving out anything that might happen before he gets to my neighborhood.

You can not cut out a piece of time and space and then draw conclusions that what happens in the period is nothing more than randomness. Because the beginning of that period is the result of what came before.

My head hurts from thinking. I very seldom try and think more deeply than what I am going to eat and who will win the next race. So I must stop and leave the rest to great thinkers.

Handi

[PaceAdvantage]

Like I've stated many times in the past, it's my opinion that 46zilal is an experimental A.I. project stuck in beta stage.

It explains a lot.

[Jeff P]

These questions are actually from a chapter for a book I've been working on...

Do you believe that race outcomes are determined by measurable cause and effect factors? Or are race outcomes simply the result of random chance? Or could race outcomes be determined by some combination of the two? Something like X percent causal factors (whatever they might be) along with Y percent random chance?

I'm not looking for answers so much as I'm trying to get you to think. Don't be afraid to think outside the box a little. Within the context of these questions, the meaning of race outcomes isn't necessarily limited to which horses finished first second or third.


-jp

.

[Tom]

There's a cool Black Swan in Lost.

Also, I once saw a Black Swan near Niagara Falls, but after thinking about it now, I think he was just a white swan with mud on him.

So you are suggesting the smoke monster might be white smoke? ;)

[Tom]

Running lines in a horse's past performance record.

Which are a database. :D

[Tom]

I do believe the Pirco members like Doc Sartin, Tom Brohamer made great use of data, in fact i do believe Doc Sartin was a great believer in studying past records of not only horses, but racetracks and wagering.

The Brohamer model and the track profile are databases.

[raybo]

As I said before, if you rely on skills and knowledge, acquired over time, to determine your next wagering selection, then you are using a database, the human mind.

[Dave Schwartz]

Raybo,

Yes, but it is a database that is not well-organized.


Dave

[ranchwest]

Raybo,

Yes, but it is a database that is not well-organized.


Dave


Or well-managed.

[ranchwest]

These questions are actually from a chapter for a book I've been working on...

Do you believe that race outcomes are determined by measurable cause and effect factors? Or are race outcomes simply the result of random chance? Or could race outcomes be determined by some combination of the two? Something like X percent causal factors (whatever they might be) along with Y percent random chance?

I'm not looking for answers so much as I'm trying to get you to think. Don't be afraid to think outside the box a little. Within the context of these questions, the meaning of race outcomes isn't necessarily limited to which horses finished first second or third.


-jp

.

Rather than "cause and effect" and "randomness", I prefer to think of the broader context of whether a factor has some predictive qualities.

For me, randomness has no place in the prediction phase of handicapping. Randomness can have a place in wagering.

I've seen horses get hit by animals during a race. A big goose comes to mind for one. There's no way I should factor a big goose hitting a horse into my prediction process. I guess you could say that the front-runner might be slightly more likely to be hit by a goose than the other horses, but such an exercise seems rather silly to me. However, if the odds shape up right, I might want to take a chance on some possibilities because "stuff happens" -- I don't have to predict specific stuff, such as a goose getting in the way because "stuff" could be any one of millions of things. I think taking a stab on "stuff" can be especially useful for exotics.

Still, I think there are plenty of races that are predictable and sound handicapping techniques can overcome both takeout and "stuff". If I didn't find that believable, I'd quit.

[ranchwest]

Wagering against bridge jumpers is also an example of where it sometimes pays off to hope stuff happens. In a really good situation, you might also find the bridge jumper's horse to be suspect from a sound handicapping perspective. Either way, the odds are there.

[SchagFactorToWin]

These questions are actually from a chapter for a book I've been working on...

Do you believe that race outcomes are determined by measurable cause and effect factors? Or are race outcomes simply the result of random chance? Or could race outcomes be determined by some combination of the two? Something like X percent causal factors (whatever they might be) along with Y percent random chance?

I'm not looking for answers so much as I'm trying to get you to think. Don't be afraid to think outside the box a little. Within the context of these questions, the meaning of race outcomes isn't necessarily limited to which horses finished first second or third.


-jp

.

My research on 1/2 mile harness tracks has led me to the hypothesis that ~40% is available in the PP data, ~60% to what I call Race Dynamics- how the horse feels today, track/weather conditions at the start of the race, how the various entries interact during the race... I guess one could add ~1% for the Big Goose effect (see previous posts). ;)

[Warren Henry]

I all to often forget to whom I am speaking but Heraclitus said it thousands of years ago: "you can never step in the same river twice".... because it is always changing (whether we like it or not)

Oriental philosophers understand that and embrace it as the basis for the Tao.

Maybe their positions of long long ago will triumph one day

Until the oriental philosophers show me how to pick a horse, I find this of little interest.

[mountainman]

I still believe that the perception of Randomness is due to a lack of data which lies at the foundation of causation.




Deep man. DEEP. I do believe that all occurrences are physical and a reaction to other occurrences. Predicting most happenings, however, would require data and intelligence beyond human means. Let me pose a question that I think cuts to the heart of it: If God exists, can he pick every winner? By scientific means, without looking into the future, that is. I wrote a blog (never posted) on just that question, but the tone was humorous. I'm serious here. Sort of.

[Greyfox]

Physicists and philosophers can prove that everything in life is an illusion.
For practical purposes though I'll treat the chair that I'm sitting on and the horses that I'm betting as real.

[johnhannibalsmith]

These questions are actually from a chapter for a book I've been working on...

Do you believe that race outcomes are determined by measurable cause and effect factors? Or are race outcomes simply the result of random chance? Or could race outcomes be determined by some combination of the two? Something like X percent causal factors (whatever they might be) along with Y percent random chance?

I'm not looking for answers so much as I'm trying to get you to think. Don't be afraid to think outside the box a little. Within the context of these questions, the meaning of race outcomes isn't necessarily limited to which horses finished first second or third.


-jp

.

Jeff, I could probably write a response that would fill seven chapters in this book.

Suffice it to say, and it might surprise El Zilllllly to hear, I have transformed to a player that limits his play to races where the intangible leads me to a stronger conclusion than the tangible. Or, more clearly, I attempt to find value in wagering by playing races where I think I can defy the conclusions that others will draw from available data.

I began as a numbers/speed type handicapper early on, transitioned to something of a trip-esque type player, and now I'm something of a 'subtlety' player. I don't play much at all any more and have gone through long periods of abstinence, in large part because of the nature of this type of analysis - it gets a little, I don't know, demoralizing, maybe depressing - trying to read motives and evaluating the likelihood of the apparently unlikely.

[Tom]

Deep man. DEEP. I do believe that all occurrences are physical and a reaction to other occurrences. Predicting most happenings, however, would require data and intelligence beyond human means. Let me pose a question that I think cuts to the heart of it: If God exists, can he pick every winner? By scientific means, without looking into the future, that is. I wrote a blog (never posted) on just that question, but the tone was humorous. I'm serious here. Sort of.

The question is moot because while God exists, science does not. ;)

[Handiman]

Mountain,

To strike at the heart of it, is in fact close to the answer to your question. Man has created the theory of Randomness to explain occurrences for which he does not have all the required information to formulate an explanation for that occurrence.

So he believes if 'I can't explain why this happens, it must just be a random event' and so let's leave it at that. Lack of data required to solve a problem doesn't mean that the problem can be explained away by attaching a random label to it.

So if lack of knowledge is the culprit...then that is all it is...a lack of the needed skill set or data set to form an answer.

Not some random label.

Handi :)

[46zilzal]

Mountain,

To strike at the heart of it, is in fact close to the answer to your question. Man has created the theory of Randomness to explain occurrences for which he does not have all the required information to formulate an explanation for that occurrence.

So he believes if 'I can't explain why this happens, it must just be a random event' and so let's leave it at that. Lack of data required to solve a problem doesn't mean that the problem can be explained away by attaching a random label to it.

So if lack of knowledge is the culprit...then that is all it is...a lack of the needed skill set or data set to form an answer.

Not some random label.

Handi :)

BOTH authors quoted go into the psychological need to CREATE and OVERLAY order when no order is there through multiple examples and experience. In several experiments detailed in the book BLINK, accounts that have NOTHING to do with the reality of situations are created to satisfy this mania about having more and more and more and more data which only serves to blur the lines of understanding while bolstering the belief that a process has been explained to the reality of the situation.

In the book Smart Choices they also report this psychological NEED which discounts the ability to understand with a false explanation that just "seems" to fit.

[sjk]

I don't understand the negative feelings about randomness. Randomness is a good thing.

If I bet a horse at 20-1 (which is often) it is usually not because I think he is the best horse. It is because I think that at least 1 time out of 10 the random elements at play will cause him to outfinish the other better horses.

[headhawg]

In the book Smart Choices they also report this psychological NEED which discounts the ability to understand with a false explanation that just "seems" to fit.Ahh....that explains it. I hope you read that chapter more than once.

[ranchwest]

I don't understand the negative feelings about randomness. Randomness is a good thing.

If I bet a horse at 20-1 (which is often) it is usually not because I think he is the best horse. It is because I think that at least 1 time out of 10 the random elements at play will cause him to outfinish the other better horses.

Obviously what some call randomness others call value. IMHO, there is a big difference.

[Jake]

I have never understood this debate here. Yes, every race has different unique racing dynamics--this doesn't make it unpredictable or immune to modeling. Yes, every race has random or chaotic elements, again that doesn't mean you can't identify those random elements and seek to model the data. Sorry, database modeling will indeed reveal key factors for both contenders and value plays. Handicapping is about finding both an edge and value relative to predictive factors, nothing more. If you don't believe that is possible, then you shouldn't be playing. That seems clearcut to me. If you have a strong edge relative to value, then you should make a play. Again, crystal clear.

PA, re Mike, in another thread, showed his reasoning for a 5-minute event catalyst trade. I didn't agree with his reasoning, but it didn't matter. He made the trade based upon his personal insights and caught a nice upturn leg. That ends the argument right there, regardless of whether he is right or wrong in his a priori reasoning, because he made the play and has money in his pocket. You have to honor that. He may be wrong next time, but he wasn't wrong here. Nice hit, he should crow a bit. He saw it, we failed to see it.

Back to handicapping. This is no different. You find factors or angles that are predictive long term, you look to make adjustments for individual races, and then you work to find wagering advantage in the pools. If it is there you make the plays, and understand that it incorporates a large random element. That's the nature of being a good player. It's not that damn complicated or even difficult.

Winning wagers determine validity. No other debate here.

Jake

[douglasw32]

Is this not why we have this...

http://en.wikipedia.org/wiki/Monte_Carlo_method

[Tom]

Ahh....that explains it. I hope you read that chapter more than once.

He wrote it! :lol:

[Robert Goren]

Science is at best an attempt to explain God and each attempt is almost aways proved wrong when new and/or better data becomes available. The problem with most computer models is the data on which they are based. They tend treat all data the same. A lot of horse racing data is just plain wrong. Take lengths back after a half in 6f race. You do not have to watch very many replays and compare them to the charts to realize that there is lot of inaccuracy there. Yet there are people who have gotten a model from this flawed data and applied it to even more flawed data are making a bets based on what appears to be a few hundreds of a second difference. GIGO. JMO

[Robert Goren]

Sorry about the above post. I am not sure even I understand it. If someone does "get it", perhaps they could explain it to me.;)

[chickenhead]

Until the oriental philosophers show me how to pick a horse, I find this of little interest.

you orda numba 2, you like.

Eh?

No, no spicy.

you wan rice?


(oops, that was picking lunch :lol: )

[headhawg]

He wrote it! :lol: :lol:

[Tom]

you orda numba 2, you like.

Eh?

No, no spicy.

you wan rice?


(oops, that was picking lunch :lol: )

Wan exacta?

One from column A
One from column B

[SchagFactorToWin]

Sorry about the above post. I am not sure even I understand it. If someone does "get it", perhaps they could explain it to me.;)

GIGO?

[headhawg]

Garbage In, Garbage Out.

[Overlay]

Let me pose a question that I think cuts to the heart of it: If God exists, can he pick every winner? By scientific means, without looking into the future, that is.

As with life in general, I believe that if God wanted a certain outcome to occur, He could do it by continually governing conditions (seemingly "random" events) in a manner that did not interfere with the free will of any of the participants, but produced the result that He intended. If He were completely disinterested in the race outcome (from the context of His overall will for the universe), He would at least know what each entrant's chance of winning was (through having perfect knowledge of all the variables affecting the race outcome), but would exert no direct effect on making any particular entrant the winner. (Just my opinion.)

[PaceAdvantage]

PA, re Mike, in another thread, showed his reasoning for a 5-minute event catalyst trade. I didn't agree with his reasoning, but it didn't matter. He made the trade based upon his personal insights and caught a nice upturn leg. That ends the argument right there, regardless of whether he is right or wrong in his a priori reasoning, because he made the play and has money in his pocket. You have to honor that. He may be wrong next time, but he wasn't wrong here. Nice hit, he should crow a bit. He saw it, we failed to see it. Only problem with all of this is that I never made such a trade, and never stated or even implied as much (and I apologize if something I said made you think I actually traded on the Scott election).

For starters, I never trade in the overnight hours...secondly, my trades are all technically based...I never trade based on news...

[Secretariat]

Interesting thread.

Say you had a race with the following.

1. A horse who shows on a 1000 race database profit.
2. A horse who shows the best speed last race.
3. A horse who shows a top trainer statisitc applicable for the coming race.
4. A horse who shows a recent 10-20 race Brohamer decision model
5. A horse who shows the best early and looks like there is no front speed to challenge.
6. A chronic loser but a horse who exceeds your estimated oddsline.

Which of these would make the best wager in your opinion if they all appear to show similar edge in terms of profitability.

[Overlay]

Interesting thread.

Say you had a race with the following.

1. A horse who shows on a 1000 race database profit.
2. A horse who shows the best speed last race.
3. A horse who shows a top trainer statisitc applicable for the coming race.
4. A horse who shows a recent 10-20 race Brohamer decision model
5. A horse who shows the best early and looks like there is no front speed to challenge.
6. A chronic loser but a horse who exceeds your estimated oddsline.

Which of these would make the best wager in your opinion if they all appear to show similar edge in terms of profitability.

Given your stipulation of a similar edge in terms of profitability for all entrants (which I interpret as the relationship between the actual winning percentage associated with each of the above factors, and the odds of the horse possessing the factor), I'd go with whichever factor showed the highest winning percentage/probability (which I can't determine based just on the information provided).

[ranchwest]

Interesting thread.

Say you had a race with the following.

1. A horse who shows on a 1000 race database profit.
2. A horse who shows the best speed last race.
3. A horse who shows a top trainer statisitc applicable for the coming race.
4. A horse who shows a recent 10-20 race Brohamer decision model
5. A horse who shows the best early and looks like there is no front speed to challenge.
6. A chronic loser but a horse who exceeds your estimated oddsline.

Which of these would make the best wager in your opinion if they all appear to show similar edge in terms of profitability.

Overlay nailed it. You always want the safest and most consistent performer, all other things being equal.

One of the hazards of attempting to use a database is having pre-conceived notions of what is best. Overlay's answer places the emphasis on having the database speak to the user, not the other way around.

[mountainman]

These questions are actually from a chapter for a book I've been working on...

Do you believe that race outcomes are determined by measurable cause and effect factors? Or are race outcomes simply the result of random chance? Or could race outcomes be determined by some combination of the two? Something like X percent causal factors (whatever they might be) along with Y percent random chance?

I'm not looking for answers so much as I'm trying to get you to think. Don't be afraid to think outside the box a little. Within the context of these questions, the meaning of race outcomes isn't necessarily limited to which horses finished first second or third.


-jp

.

I've spent my life trying to hone my skills as a handicapper. Not with complex databases, but by feeding practical experience, books, ideas aquired from other players, the perspective of numerous positions in the industry, the product of some personal research, and lots of pure intuition into the inferior computer between my ears. And still, my mind invariably drifts to horseracing when Joni Mitchell comes on the radio singing: 'I've looked at clouds from both sides now, from up and down and still somehow, it's cloud's illusions I recall. I really don't know clouds at all.'

[Greyfox]

I've spent my life trying to hone my skills as a handicapper. Not with complex databases, but by feeding practical experience, books, ideas aquired from other players, the perspective of numerous positions in the industry, the product of some personal research, and lots of pure intuition into the inferior computer between my ears. And still, my mind invariably drifts to horseracing when Joni Mitchell comes on the radio singing: 'I've looked at clouds from both sides now, from up and down and still somehow, it's cloud's illusion's I recall. I really don't know clouds at all.'

True, but your opinions and picks are well respected in our O.T.B. :ThmbUp:

[mountainman]

True, but your opinions and picks are well respected in our O.T.B. :ThmbUp:

Thanks man. All the best.

[DeanT]

BOTH authors quoted go into the psychological need to CREATE and OVERLAY order when no order is there through multiple examples and experience. In several experiments detailed in the book BLINK, accounts that have NOTHING to do with the reality of situations are created to satisfy this mania about having more and more and more and more data which only serves to blur the lines of understanding while bolstering the belief that a process has been explained to the reality of the situation.


The thesis in Blink works very well with databases. If you see a track playing to late for example, and you subconsciously know that the highest ranked late horse with a decent speed fig win at a higher rate, you bet those at will. Just like you can see a horse with a swishing tale that you pitch through your gained knowledge thru the years with a blink moment, you can with numbers from a dbase as well.

[hcap]

Interesting thread.

Say you had a race with the following.

1. A horse who shows on a 1000 race database profit.
2. A horse who shows the best speed last race.
3. A horse who shows a top trainer statisitc applicable for the coming race.
4. A horse who shows a recent 10-20 race Brohamer decision model
5. A horse who shows the best early and looks like there is no front speed to challenge.
6. A chronic loser but a horse who exceeds your estimated oddsline.

Which of these would make the best wager in your opinion if they all appear to show similar edge in terms of profitability.
So the question is really HOW predictive are any of these 6? In all cases a database test could be done on all, assuming the necessary data has been recorded. The question then would be were they arrived at using 1 part of your database, and then tested on another going forward? Or was the entire database used to come to (premature?) conclusions? I would divide my database into 2 or 3 parts. Arrive at some testable conclusions such as 1 thru 6. Pick another period or set of races part 1 of the database has not seen yet, ---- a segment not used to arrive at these 6 conclusions.

[Secretariat]

I've been a big Database supporter, but some methodologies rely on short samples (jockey-trainer statistics), track bias, recent decision models. And trends can change at any time. Also some races seem to create conflicts. Like the above race where there is an abundance of positive indicators in horses, and some where there are none to only a few. should the handicapper go at it more like each race is a unique set of circumstances that is chaotic, or look at recent trends, or specialized circumstances, or rely on long term databases. Just curious on most viewpoints here.

[Tom]

I love short term trend. They get lost in the long term but you can cash on them today if you know about them.

[raybo]

The ability to query both long term data and short term data, allows one to recognize current trends, that might otherwise be missed. That's one of the beauties of knowing how to query databases, you can specify how much data and what period of time, about which, you're interested in finding something out.

[mountainman]

I love short term trend.
Absolutely. Long term trends become stale and unprofitable through exposure. And as more and more information gets churned through computers and disseminated, the more this will be the case. Short term trends that the public hasn't picked up on afford much better value. The trick it to know which short term cycles are random and which are real. In my opinion, experienced knowledge and good intuition can determine that better than any database.

[DeanT]

In my opinion, experienced knowledge and good intuition can determine that better than any database.

For some I think that's true, but for a lot I think it is pretty tough (guilty as charged). You would not believe Mark, how many folks (like myself many times before) speaking of track bias and things like that based on anecdote are wrong on their assumptions. When I was finally able to read a database for it, it saves a ton of time and you get dispassionate data. You can also be broad in your betting, because it allows you to look at five or ten tracks and know immediately what's winning, when there was a dead rail or a live one etc. You'd have to follow ten tracks live with a pen and paper to get such a snapshot, and that takes a ton of time, taking away from other points of handicapping.

[Tom]

In my opinion, experienced knowledge and good intuition can determine that better than any database.

No db limits the number of tracks you can follow. I can look at any track running and in a few minutes, know if it worth playing.

[mountainman]

For some I think that's true, but for a lot I think it is pretty tough (guilty as charged). You would not believe Mark, how many folks (like myself many times before) speaking of track bias and things like that based on anecdote are wrong on their assumptions. When I was finally able to read a database for it, it saves a ton of time and you get dispassionate data. You can also be broad in your betting, because it allows you to look at five or ten tracks and know immediately what's winning, when there was a dead rail or a live one etc. You'd have to follow ten tracks live with a pen and paper to get such a snapshot, and that takes a ton of time, taking away from other points of handicapping.

Dean , I was talking strictly about short term trends. I keep a crude (excel) database for mnr that profiles the surface's intrinsic(long term) nature. And I also maintain one that keeps tabs on all invading horses and trainers. I'm not a caveman..lol. Just a mountainman.

[acorn54]

46 i am still waiting for your answer to how you handicap if you do not rely on past performances.