Sports Forum   |   Scoreboard   |   Contact Us   |   WinPicks Software   |   Talkstock   |   Advertise   |   Follow Us on Twitter   |   Register  |   Login 
Click here to register for the sports forumClick here to download the FREE ScoresLine App for Android 
 
 
Forums   Register   Profile   Inbox   Address Book   Subscribe   My Forums   Search   FAQ   Login   Log Out

GOLD04_AVA1 NCAAF   Logged in as: Guest
Users viewing this topic: none
  Printable Version
All Forums >> [Sports Handicapping Resources] >> WinPicks Software Forum >> GOLD04_AVA1 NCAAF Page: [1]
Login
Message << Older Topic   Newer Topic >>
GOLD04_AVA1 NCAAF - 9/21/2013 5:32:52 PM   
davebyers

 

Posts: 23
Joined: 12/19/2009
Status: offline
These 2 formulas have been tossed up in the last month on the warehouse and I have looked at them and back tested them up and down. They both boast a winning pct. of 58-60% over the last 28 years! They are both based on database rolling avg 2, a very short window. What is happening is the program analyzes itself backwards. The power rankings of week 4 are used to analyze week 3. With only a 2 game window this makes a HUGE difference in past results. The program already knows who won the game, and credits itself with a win. But if I move forward thru the schedule the same results do not occur.

Conclusions: Ignore any formulas less than R7-B7, that will at least diminish this flaw. The higher the better it seems.

Microbros: What say you?
Post #: 1
RE: GOLD04_AVA1 NCAAF - 9/26/2013 9:07:41 PM   
WinPicks Support

 

Posts: 96
Joined: 11/26/2006
Status: offline
quote:

ORIGINAL: davebyers

These 2 formulas have been tossed up in the last month on the warehouse and I have looked at them and back tested them up and down. They both boast a winning pct. of 58-60% over the last 28 years! They are both based on database rolling avg 2, a very short window. What is happening is the program analyzes itself backwards. The power rankings of week 4 are used to analyze week 3. With only a 2 game window this makes a HUGE difference in past results. The program already knows who won the game, and credits itself with a win. But if I move forward thru the schedule the same results do not occur.

Conclusions: Ignore any formulas less than R7-B7, that will at least diminish this flaw. The higher the better it seems.

Microbros: What say you?


Hi Dave

Missed this post earlier. Sorry to hear you think WinPicks works like that, because it doesn't, regardless of what the rolling average setting is. All back testing uses previous data to make picks for the next date when running a formula analysis. You could roll back 22 years to every date on the schedule, verify the team stats for all teams at that point in time, and run Predict Sides, and then add it up. The results would match what the formula analysis says.

Besides, if WinPicks really works like you say it does, why wouldn't it just predict at 100%?

Thanks
WinPicks Support

(in reply to davebyers)
Post #: 2
RE: GOLD04_AVA1 NCAAF - 11/26/2013 12:43:54 PM   
davebyers

 

Posts: 23
Joined: 12/19/2009
Status: offline

Thanks for the reply, Winpicks.
First of all, I have been a loyal customer of yours for 5 years or more and plan on continuing. Your software is a fascinating and useful handicapping tool.
My goal in pointing out the discrepency in analyzing past results is not to claim that your software doesn't work as claimed. It is to educate users as to what is truly ocurring at the heart of the software. This could help people avoid misleading results.
I claim that the NCAAF formulas "AVA 1" and "Gold 04" yield misleading past results. Both are based on a database of only 2 games. Both have home field advantages in the 80 point range. These formulas are stretched to the extreme. They yield power rankings that fluctuate wildly from week to week. Both formulas back test to 60% winning rates (3554-2365-95 for AVA) dating back to 1985.
My first clue is this: If I save a copy of AVA and run a consensus test with itself back to 1985, it tests out to 51% winning rate. (2959-2844-85) Could it be that the consensus test is moving forward in the power rankings and yields more accurate results?
Clue #2: Go to MENU>> Predict>>Sides and see the AVA 1's "Recommended" picks for today. Now using the date button go to previous date. Move back through several dates. Note the results of these previous dates. Note how outstanding they are. Now using the date button go to next date and move forward in time. Note the results of these previous same dates. Note how different and less outstanding they are.
Clue #3: Go to "Formula Warehouse" for NCAAF. Check the best formulas ATS of all time. 10 of the top twenty formulas of all time are based on a database of two games. That leads to this question: Could it be that analyzing the last two games is the magic number to success for handicapping college football? Answer: Of course not.
Again, I am a loyal customer who has spent way to much time exploring your software. I have yielded very positive results this season in pro and college hoops using your software. My past record is on Covers.com at http://contests.covers.com/sportscontests/recordsbydate.aspx?ur=466418&sportID=0 . Just hoping to enlighten others as to how this program really works.
--Dave
 

(in reply to WinPicks Support)
Post #: 3
Page:   [1]
All Forums >> [Sports Handicapping Resources] >> WinPicks Software Forum >> GOLD04_AVA1 NCAAF 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

0.111