TCB NFL Picks for Week 14 - MNF

[TCB in a Flash]

2 - 3 yesterday (actually 3 - 4 since GB and SF were both play on and played each other). And since the model is not sophisticated enough to distinguish between them - pretty linear here - I left them both out.
Anyway, the big bounce theory/system has another play tonight.

St Louis was a 6 pt fav last week and lost by 14, a 20 pt difference. Thus the system says - play on St L.

Play
St. Louis + 6

[TCB in a Flash]

Here are the calculatinos for the system to date...
From Weeks 5 - 13 (except for St Louis tonight)

25 - 18 Play Against Big Cover Winner the following week (1 -3 in Week 13)
28 - 14 Play On Big Cover Loser the Following Week (2 - 1 in wk 13 - on St Louis pending)
53 - 32 total.

This theory is based in human nature and economics. But it's cooler than it sounds. Really.

Here is my theory...
When teams do much better or worse than their expectations in one week, they will tend to come back to the average perfomance expectation level, as long as that level was reasonable.
So teams that do really well will tend to perform a bit worse the next time out and teams that do very poorly will tend better the next time out.
Part of this is psychological - i.e. it's human nature to feel too good about yourself when you overachieve and conversely, to focus more intensely when you underachieve.
And you get the additional benefit of the public overpricing teams that just over achieved and underprice those that have underachieved.
Thus, we buy low and sell high.

Which games make the model...
Results need to be at least 13 pts different than the spread.
For now that number is arbitrary, it just felt big enough to me. But eventually I'll figure out if 13 is too big or small and adjust it.

Why aren't weeks 1 - 4 included?
The thought is that the lines are much farther off early in the season - and that the big differences versus expectations are due to poor information/incorrect perception. Teams haven't found their level yet.
By week 5, teams seem to found their level.

How much data is this based on?
Just this year.
Thinking I'll add the past few years soon, maybe this weekend.
I'm expecting a big hangover Saturday - so if I'm not too miserable I'll do some of it then.

And that's all I have to say about that.

[molta02]

looks good. gl tonite king

[kbsooner21]

Hope the system is right Let's ca$h this bitch!