Trend Tracking Experiment

[kroyrunner89]

Trends are a widely debated topic among handicappers in today's day and age. While trends come in many forms and shapes, there are some cappers who believe that they can be an incredibly useful tool when it comes to picking winner, then there are other cappers like myself who believe that they aren't useful at all. Listed below are 32 trends, one for each team, that have stunning winning percentages (or losing percentages if it's one you're supposed to fade). While some of them have reasoning that makes sense and others make no sense at all, many cappers new to betting NFL will see a trend with a high winning percentage and immediately jump on board. My goal here is to show that trends such as these offer no advantage to a handicapper whatsoever. These 32 trends below have accumulated a record of 447-67 by playing or fading them in the situation mentioned. That's an 87% winning rate. This season we'll track these trends going forward, and show that they'll come nowhere near that 87% mark, and possibly not even make you a profit. Hope you all enjoy the thread, and feel free to help out identifying when one of these trends is in play if I happen to miss it!

"Trends to Play"

• BUFFALO is on a 13-2 UNDER the total (+10.8 Units) run on the road vs. teams giving up 17 or less PPG

• MIAMI is on a 11-0 UNDER the total (+11 Units) run at home revenging a home loss against opponent

• NY GIANTS are on a 11-0 UNDER the total (+11 Units) run as road favorites of 3.5 to 7 points

• NEW ENGLAND is on a 21-5 ATS (+15.5 Units) run vs. marginal winning teams (Win Pct. 51% to 60%)

• PHILADELPHIA is on a 9-0 ATS (+9 Units) run vs. teams scoring 29 or more PPG

• NY JETS are 1-11 ATS (-11.1 Units) vs. poor passing defenses - allowing 235 or more PYPG since '07

• Over the L2 seasons, WASHINGTON is 10-1 UNDER the total (+8.9 Units) as favorites

• Over the L2 seasons, BALTIMORE is 9-1 ATS (+7.9 Units) vs. teams with a losing record

• CLEVELAND is on a 8-0 UNDER the total (+8 Units) run against NFC South division opponents

• GREEN BAY is on a 12-1 ATS (+10.9 Units) run at home vs. good teams (teams outscoring opponents by 6+ PPG)

• PITTSBURGH is on a 30-12 ATS (+16.8 Units) run vs. top-level teams (Win Pct. > 75%)

• HOUSTON is 9-0 UNDER the total (+9 Units) vs. teams with a losing record since '07

• INDIANAPOLIS is on a 23-4 OVER the total (+18.6 Units) run on the road coming off an upset loss as a favorite

• CAROLINA is on a 9-0 ATS (+9 Units) run at home vs. excellent offensive teams - averaging >=375 YPG

• Over the L2 seasons, NEW ORLEANS is 9-0 ATS (+9 Units) in non-conference games

• TENNESSEE is on a 26-9 OVER the total (+16.1 Units) run vs. poor defensive teams - allowing >=350 YPG

• TAMPA BAY is on a 27-8 UNDER the total (+18.2 Units) run in the first two weeks of the season

• ARIZONA is 14-3 ATS (+10.7 Units) vs. good passing teams averaging 7 or more PYA since '07

• KANSAS CITY is 11-1 ATS (+9.9 Units) vs. awful passing defenses - allowing a comp. pct. of 64% or worse since '07

• SAN FRANCISCO is on a 16-3 OVER the total (+12.7 Units) run on the road vs. poor rushing teams - averaging <=3.5 YPR

• SAN DIEGO is 14-1 ATS (+12.9 Units) vs. good passing teams with a completion pct. of 61% or better. since '07

"Fade Trends"

• DALLAS is on a 3-20 ATS (-19 Units) skid on the road after covering the spread in 3 out of their last 4 games

• CHICAGO is on a 2-16 ATS (-15.6 Units) skid after scoring 25 points or more in 2 straight games

• CINCINNATI was 0-8 ATS (-8.8 Units) as favorites last season.

• DETROIT is on a 0-9 ATS (-9.9 Units) skid vs. teams scoring 14 or less PPG

• MINNESOTA is on a 5-19 ATS (-15.9 Units) skid vs. poor passing teams averaging 5.7 or less yards per pass attempt

• ATLANTA is on a 2-14 ATS (-13.4 Units) skid at home vs. poor passing defenses - allowing a comp. pct. of 61% or worse

• Over the L2 seasons, JACKSONVILLE is 0-10 ATS (-11 Units) as home favorites

• DENVER is 1-14 ATS (-14.4 Units) vs. excellent passing teams with a completion pct. of 64% or better since '07

• SEATTLE is 0-11 ATS (-12.1 Units) on the road vs. teams scoring 24 or more PPG since '07.

• OAKLAND is 1-12 ATS (-12.2 Units) after covering the spread in 2 out of their last 3 games since '07

• ST LOUIS is 1-11 ATS (-11.1 Units) in the first month of the season since '07

[kroyrunner89]

Week 1 Plays According to Trends:
Indy/Houston OVER 47 (Indy Trend)
Tampa/Cleveland UNDER 36.5 (Tampa Trend)
Washington +3.5 (Fading the Dallas Trend)
Detroit +7 (Fading Chicago Trend)
Denver +1.5 (Fading Jacksonville Trend)
Tennessee -7 (Fading Oakland Trend)
Arizona -3.5 (Fading St. Louis Trend)

The three in bold are the clear cut ones, the other four go back to last season, chime in guys should we count these four as trends for Week 1 or does the slate wipe clean at the start of the new season?

[ojs69]

you want trends?
SportsDataBase.com: Content for winning sites.

[dvb02]

i don't think you should use those that go back to last year.

[dvb02]

kroy, i was wondering why you had about the same # of plays in 2008 (starting with week #8) as you did for all of 2009? just curious if you added filters or changed something.

[officepoolguy]

Good post but you can't count the trends from last year as games at the end of the season are so different then at the beginning (especially in Oakland's case). Also I am not a big fan of "team trends". I think teams change to much to count them.

I am a big fan of league trends (such as home field advantage). I wrote an E-book on NFL office pool strategy that uses many historical trends that I think are good bets. Download it for and let me know what you think (you just have to sign up for my free newsletter). I will be tracking the trends in the book also. Here is the site.

[kroyrunner89]

kroy, i was wondering why you had about the same # of plays in 2008 (starting with week #8) as you did for all of 2009? just curious if you added filters or changed something.

Well in 2008 what prompted me to start posting plays in the first place was that I had developed an excel formula for predicting games that had done quite well. Of course being new to it at the time I thought that two winning weeks in a row was quite impressive, so I dove in and decided to post every single play the system recommended all year, plus threw in a few of my own. Last year I knew I had to take a different approach, so I got a lot more selective about what I played from the system and then also realized that value in the NFL is much tougher to find than I thought in 2008. So I scaled back a little bit and went on the conservative side, and it turned out well for me, especially because last year was a year the system struggled a little bit (although it did wind up about 60% in the end, hit a nice streak with sides weeks 10-15). This coming year, we'll see, hopefully I can find a few more plays I like and release more but ultimately it'll come down to what I think looks good from week to week

[dvb02]

Well in 2008 what prompted me to start posting plays in the first place was that I had developed an excel formula for predicting games that had done quite well. Of course being new to it at the time I thought that two winning weeks in a row was quite impressive, so I dove in and decided to post every single play the system recommended all year, plus threw in a few of my own. Last year I knew I had to take a different approach, so I got a lot more selective about what I played from the system and then also realized that value in the NFL is much tougher to find than I thought in 2008. So I scaled back a little bit and went on the conservative side, and it turned out well for me, especially because last year was a year the system struggled a little bit (although it did wind up about 60% in the end, hit a nice streak with sides weeks 10-15). This coming year, we'll see, hopefully I can find a few more plays I like and release more but ultimately it'll come down to what I think looks good from week to week

I am looking forward to it.

[kroyrunner89]

i don't think you should use those that go back to last year.

I think I agree with you, so here are Week 1 plays with that in mind:

Week 1 Plays According to Trends:
Tampa/Cleveland UNDER 36.5 (Tampa Trend)
Denver +1.5 (Fading Jacksonville Trend)
Arizona -3.5 (Fading St. Louis Trend)

[Wrecktangle]

All these "trends" are variants of data snooping.

a data set is like a prisoner of war, if you torture it enough, it will admit to anything.

[balman]

KR89.. you come up with some good stuff.. can wait to tail u in 2010... lets make so money...

[kroyrunner89]

All these "trends" are variants of data snooping. a data set is like a prisoner of war, if you torture it enough, it will admit to anything.

That's exactly what I'm trying to demonstrate with this thread, that stuff like this is totally unreliable

[kroyrunner89]

Bumping this up for the start of the season, again here would be the plays if you followed these trends. I'm not recommending these plays! Read through the whole thread again if you're wondering what this is about, I want to show that trends like these should play no role in selecting games from week to week.

Week 1 Plays According to Trends:
Tampa/Cleveland UNDER 36.5 (Tampa Trend)
Denver +1.5 (Fading Jacksonville Trend)
Arizona -3.5 (Fading St. Louis Trend)

[kroyrunner89]

Bumping this up for the start of the season, again here would be the plays if you followed these trends. I'm not recommending these plays! Read through the whole thread again if you're wondering what this is about, I want to show that trends like these should play no role in selecting games from week to week. Week 1 Plays According to Trends: Tampa/Cleveland UNDER 37 (Tampa Trend) Denver +2 (Fading Jacksonville Trend) Arizona -4 (Fading St. Louis Trend)

Fixed

[kroyrunner89]

After Week 1: 1-1-1 ATS

Will look through and post Week 2 in the next few days

[kroyrunner89]

• NE -1.5 (Fading NYJ Trend)
• Wash/Hou UNDER 44 (Washington Trend)
• Bal -1 (Baltimore Trend)
• Pitt +5 (Pittsburgh Trend)
• TB/Car UNDER 38 (TB Trend)
• KC +1 (KC Trend)
• Miami +5.5 (Fading Minn Trend)
• Jacksonville/Denver Trends cancel out - No play
• Oakland -4 (Fading Stl trend)

[Firefox14]

I am 6-9 ATS after I eat dinner and take a shower, then brush my teeth, after flossing....

[RANDAZZO]

[Firefox14]

but I'm also 9-3 SU after feeding my dogs, shutting the garage, and leaving the light on all night long.....

[kroyrunner89]

Haha my point exactly. But for some reason people still love these things, so I'm just going to have to show through the numbers that these are ridiculous

[Firefox14]

kroy where are you getting these numbers from?

[bombCanada]

Sounds like statfox to me, which is a bunch of pretty meaningless slice and dice... but interesting nonetheless.

Watch this thread go 80% for the year...

good luck in week 2, kroy.

[kroyrunner89]

• NE -1.5 (Fading NYJ Trend)
• Wash/Hou UNDER 44 (Washington Trend)
• Bal -1 (Baltimore Trend)
• Pitt +5 (Pittsburgh Trend)
• TB/Car UNDER 38 (TB Trend)
• KC +1 (KC Trend)
• Miami +5.5 (Fading Minn Trend)
• Jacksonville/Denver Trends cancel out - No play
• Oakland -4 (Fading Stl trend)

4-4 on the day, 5-5-1 ATS YTD. This is going just as I expected, it's all 50% going forward...

[VegasVixen]

How far are these trends dating back to? I would think limiting them to maybe 3-4 years for NFL & maybe 5-6 years for college may have more validity?