Sports Forum   |   Scoreboard   |   Contact Us   |   WinPicks Software   |   Talkstock   |   Advertise   |   Follow Us on Twitter   |   Register  |   Login 
Register to post in the sports forumDownload the ScoresLine App for Android 

 
 
Forums   Register   Profile   Inbox   Address Book   Subscribe   My Forums   Search   FAQ   Login   Log Out

Formula Theory --- Need some input   Logged in as: Guest
Users viewing this topic: none
  Printable Version
All Forums >> [Sports Handicapping Resources] >> WinPicks Software Forum >> Formula Theory --- Need some input Page: [1]
Login
Message << Older Topic   Newer Topic >>
Formula Theory --- Need some input - 9/20/2004 9:28:25 AM   
mzldr2

 

Posts: 1
Joined: 9/20/2004
From:
Status: offline
Hi all,
New to the site and software. I wanted to briefly explain my thoughts on what I'm trying to do, and see if anyone has any input.


Basic premise is to create formulas using x number of dates and ten test the formula over at least 3 future years to determine valididty of formula.

Example:

1) Set Date to August 1, 2001. Create formula that looks at last 200 dates (example).

2) Tune the formula is tuned to get satisfactory results (Open for debate, but I try and get formuals to start with that are above 65% ATS)

3) Now, I test these formulas over the next 3 seasons to see the results.

4) The issue I'm having is that I'm finding it very difficult to find formuals that stand the test of consistency over seasons. I'm looking for formuals that have produced results over several seasons, not just 1, which is what I see most do. If you optimize a formual for 2003 and get 60% ATS and then ru it against 2003 you'll get 47% (example)

5) I use various other homemade formulas and have been very successful utilizing between the last 3 and 5 games performance numbers to generate an expectation. It seems that in the NFL package, I cannot come up with anything consistent. It's easy to optimize formulas to show good results from past data, the key is to use the data to provide soem forecast for future results. This is why I utilize steps 1-3, but have been unsuccesful.

On another note, the College program lacks any statistics other than Hoe Field, a Blow Out Limiter, etc. Has anyone had any success trying to build formuals that are based on such limited factors including strecgth of opponent based strictly on conference?

I look forward to the feedback

MZ


Post #: 1
RE: Formula Theory --- Need some input - 9/20/2004 10:51:42 AM   
endzoneantics

 

Posts: 23
Joined: 1/26/2004
From: Winston-Salem
Status: offline
I am no expert.

One thing I would add, however, is that the game changes over a course of time, and if you test over a period of 4-5 years, it will be hard to gain consistency.

(in reply to mzldr2)
Post #: 2
RE: Formula Theory --- Need some input - 9/22/2004 10:17:15 AM   
Redak

 

Posts: 10
Joined: 8/24/2004
From: Fenton
Status: offline
also in college it will be hard to stay consistent because in college teams change so dramatically year to year. here is what i am currently doing for college football:

i have formulas like this:

roll average 3 games tested over season to date
roll average 3 games tested over 236 dates (4 years)
balanced average 3 games tested over season to date
balanced average 3 games tested 236 dates (4 years)

and then i do the same set of four for rolling & balanced average of 4 games, then 5 games, then 6, and so on all the way through 30 games. That will equal out to 28 sets of 4, or 112 formulas. i get a consensus from those each week and so far i have been plus each week. i also usually break down my picks three ways i test the upcoming games with the season to date formulas, then the 4 year formulas, then the top warehouse formulas, then if all three sets choose the same game it is usually hammer time and those picks have been hitting pretty solid. also the regular plays have been doing well also.

that is what i do at least i am sure there might be a better way to do it but this is how i do it. i would be open to any and all suggestions.

_____________________________

First year with this software. It is very confusing but I am trying to figure it out. Hope you guys can put up with a newbie.

(in reply to mzldr2)
Post #: 3
RE: Formula Theory --- Need some input - 9/22/2004 11:01:02 AM   
endzoneantics

 

Posts: 23
Joined: 1/26/2004
From: Winston-Salem
Status: offline
That sounds like a pretty solid analysis, redak. I hope you plan on posting those high percentage plays!

(in reply to mzldr2)
Post #: 4
RE: Formula Theory --- Need some input - 9/22/2004 11:28:56 PM   
spreadbuster

 

Posts: 250
Joined: 8/28/2003
From:
Status: offline
Redak

When you make a play and take the consensus of 112 formulas.
What is the least number of formulas to pick that team?
58?
10?

(in reply to mzldr2)
Post #: 5
RE: Formula Theory --- Need some input - 9/29/2004 10:55:46 AM   
oldedb

 

Posts: 10
Joined: 9/3/2004
From: Atlanta
Status: offline
Redak - Nice job here..... I tried this this weekend and went an amazing 8-1 on my picks. I basically developed a strong formula with a rolling 3 game avg. From there I used that same 3 game (tuned) formula as my base formula to add the balanced 3 game, 4 game, 5 game etc...

Once you have a solid 3 game formula all you need to do is change the default database and use the "save as" feature. Do this each time you create a new formula. I tuned every formula independantly and disabled the tuning on HFA, betting window, and spread multiplier. All formulas are hitting aroun 66% and my consensus is 4 majority votes. Currently I have 8 formulas in the consensus

MZ - hit me up on PM in PPP (FatBa$tard) and we can talk about how I use ncaa stats to create predictions

(in reply to mzldr2)
Post #: 6
RE: Formula Theory --- Need some input - 9/30/2004 9:05:31 PM   
spreadbuster

 

Posts: 250
Joined: 8/28/2003
From:
Status: offline
oldedb


Can you elaborate how did you ONLY come up with 8 formulas?

If you started with formula rollback 3 then you get balance 3
all the way 4,5-30

wouldn't that be around 54 formulas???


thanks

(in reply to mzldr2)
Post #: 7
RE: Formula Theory --- Need some input - 10/4/2004 3:06:38 AM   
BOOKIE ASS KICKER

 

Posts: 32
Joined: 9/1/2004
From: Philadelphia
Status: offline
mzldr2,


You are looking for a formula with limited factors including stregth of opponent based strictly on conference?


The problem with that theory is that the turn over rate with college players are extremely high in such a way that you can't establish one set number to apply for every school in a single conference. No matter how high or low a value set, it wouldn't apply properly to several schools within that conference. As a result, you would get inaccurate numbers reading which would in turn give you an inaccurate selection, that had no business being a selection in the first place.


A better estimator, would be for you to manually check the strenghth of schedules individually for the all teams involved.


One approach would be a checks and balances approach, meaning a consensus formula selection wouldn't neccessarily be something to bet on until it has passes certain manual criteria you have preset.


Remember, a formula looks at numbers, not players. A formula has no idea what a players strenth and weaknesses are ,who is injured , what position will be a missmatch for the opposing team, who is motivated or unmotivated, etc. Are you starting to get the picture. No line maker can place any line on a team to make up for an extremely over motivated or under motivated team.


Remember, a formula pick is based on picking a team that has played well or going against a team that hasn't played well. What the formula's do not understand, is what motivation factor is involved against the team your betting against that didn't play well the last couple of outings, or what kind of let down factor or recent injury factor is involved for the team that has been playing well you are betting on.




_____________________________

My mission is to beat the living crap out of each and every Sports Book in the Universe.

(in reply to mzldr2)
Post #: 8
RE: Formula Theory --- Need some input - 10/5/2004 6:12:33 AM   
oldedb

 

Posts: 10
Joined: 9/3/2004
From: Atlanta
Status: offline


<< oldedb


Can you elaborate how did you ONLY come up with 8 formulas?
If you started with formula rollback 3 then you get balance 3

all the way 4,5-30
wouldn't that be around 54 formulas???


thanks
>>



Sorry If I mislead anyone... I actually stopped at the 6 week mark based on time that weekend. So that would be 2 formulas for each week running from roll/bal 3 wk ro rol/bal 6 week. It did so well that week I didnt add to it. I got crushed this week though so I will be headed back to the drawing board

(in reply to mzldr2)
Post #: 9
RE: Formula Theory --- Need some input - 10/7/2004 9:19:45 PM   
spreadbuster

 

Posts: 250
Joined: 8/28/2003
From:
Status: offline
oldedb
thanks for the Update

(in reply to mzldr2)
Post #: 10
Page:   [1]
All Forums >> [Sports Handicapping Resources] >> WinPicks Software Forum >> Formula Theory --- Need some input Page: [1]
Jump to:





New Messages No New Messages
Hot Topic w/ New Messages Hot Topic w/o New Messages
Locked w/ New Messages Locked w/o New Messages
 Post New Thread
 Reply to Message
 Post New Poll
 Submit Vote
 Delete My Own Post
 Delete My Own Thread
 Rate Posts


Forum Software © ASPPlayground.NET Advanced Edition 2.4.5 Unicode

0.029