Originally posted on 08/28/2013:

The Short Version


I will be running a 10 point teaser experiment this year. It will
start in week two because the way I choose teams is based on the
results from the prior week's games. I will attempt to place 10
different 10 point teaser bets per week at 1 unit a piece.

This is a continuation from last year's “2012 chase experiment”. I
have made some adaptations. The biggest change is I have dropped the
chase system in favor of what I call a “matrix”.

As long as my percentage remains above the break even point my profits will be fairly large due to the large amount of bets being placed.
In my backtests I made 107% ROI for 2011 and a staggering 1,039% ROI for 2012. While I would hope for those kind of results again that is
obviously quite extreme.

I will also run an experiment using 2 team parlays. I am applying
similar principles as the teaser strategy. The risk of ruin is
higher but the backtest for the last 2 years produced ROIs so big I
will not even mention them.


Long Version


The chasing experiment I did last year was an incredible learning
experience. My thread did not garner a lot of attention because
chasing is such a controversial topic. By studying sports betting
through the mindset of a Martingale player instead of the grind it
out percentage players which are the pros it opened my mind to seeing
connections that I otherwise would have missed.

I created what I call the “Pendulum Cycle Theory”. Instead of
seeing each game as a single random result I posit that the spreads
are reactionary. The point spread is a pricing system subject to the
forces of the marketplace. Bubbles occur in marketplaces and thus
boom and bust cycles take place on a regular basis.

The theory I propose is if you can understand the psychology of the
betting public you can find value in the spreads. The biggest
knee-jerk reactions occur when upsets take place. If an underdog
beats the spread but does not win I consider this to be an upset.
The magnitude of the upset is very important. If the Oakland Raiders
beat the spread by 5 points it is no big deal, but when the STL Rams
beat the the 49ers in their second meeting last year a phenomenon
takes place. The Rams were 8 point underdogs in this game. I
suggest that the spread in the next game was weak so you should fade
the Rams inside a 10 point teaser. They were 3 point underdogs
against the Bills next week. They won by 3 points. If you had taken
the Bills in a 10 point teaser you would have won that leg of the
bet. The 49ers were beaten by an underdog. I suggest that the
spread in their next game was weak. They faced the Dolphins as -11
point favorites. They beat the Dolphins by 14 points. Had you bet
the 49ers you have two legs of a 10 point teaser covered.

This is the primary thing I look for when I am searching for team to put
in a teaser. I have several other factors I look for as well. I
will take an elite team coming off a bye week for example. If an
elite team blows out an opponent, say they beat the spread by more
then 22 points, I expect them to come back down to earth. I will
fade them in this case. Each of these indicators is weighted.

Last year I used chasing systems. I did this primarily because I did not
have faith in my ability to pick at a winning percentage. I also had
a small number of bets so even if I was nailing a high percentage my
ROI would have been small. As the season progressed I realized that
my 10 point teaser system was outperforming everything else I was
doing. About midway through I decided to focus on it exclusivlely.
My picks for individual legs of the teasers were at 93%. If my math
is correct you need to hit 81.8% to break even. (That is why they
are called sucker bets). I backtested 2011 and my picks were at 91%.
Keep in mind that even though this percentage is high, when you mix
one loser into the 3 legs it produces a losing bet. If that one loser
is mixed into several bets it produces several losers.

My success rate was so high I decided to try an experiment. I took my
picks and I started to mix and match them. Ideally speaking I needed
8 teams per week. I could not always get 8 teams but that was my
target. I would mix and match these 8 teams into 10 different 10
point teasers. The idea was to minimize my exposure to any single
team and maximize profits when all 8 hit, which was not an uncommon
occurrence. I assigned each team a number; 1 - 8. Below is the matrix I use to create the teasers.


11172 22274
23344 73485
37555 66868


In week 2 my picks were:







  1. PIT
  2. DEN
  3. GB
  4. BAL
    (fade)
  5. CLE
    (fade)
  6. ARI
    (fade)
  7. STL
    (fade)
  8. NYJ
    (fade)






You take these teams and plug them into the matrix I posted above. PIT, DEN, GB would be your first teaser. PIT, GB, STL (fade) would be
your second teaser, etc.


ARI beat the spread an created a losing leg. This caused 3 bets to lose,
while the rest of them won. I went 7-3 and made a small profit.

Had ARI been in my 1 spot instead of 6 the matrix would have gone 6-4. This will be the most common occurrence.

I had 6 weeks where all 8 teams hit. Each time this happens the bankroll increases by 41%. A 1 team loss is a small profit. My profits were huge in 2012 because I compounded six 41% gains. I had several
smaller gains as well.


2 team losses can vary from 5-5 to 2-8. and it goes down from there.

Back tests have shown this system to work but the back testing is flawed. It does not account for the emotion of putting real money on the line. There are decisions I might have made differently in a real
life situation. Because of this I opted to stop backtesting and try
this experiment in real time.

This is the maiden voyage of this new system. Best of luck to all and may 2013 be your best year yet.