Thursday, September 27, 2007

Week 4 - The Averager, a Gambling Superhero?

Ok, so I did something pretty cool I think. I cross averaged the boxes
best winning percentage picks over the last year and a half for the
moneyline and spread. Meaning, I found out I win 90% (that's right,
NINETY) of my 11th ranked spread pick, but I only win 42% of my 11th
ranked moneyline pick - so I've started averaging the two picks to get
the highest percentage combos I can get. This way I either get a decent
hedge in a win/lose scenario, a great week on a win/win, and a tough one
on lose/lose.

So here is my list:

Top 4 Spread/Moneyline Combos (spreads first, then moneylines):

BUF+3.5 (90% pick historically!), NYJ -190
STL+12.5, DAL (which is currently OFF, so I might not be able to bet
this one)
CIN +7, NE -330 (terrible payout on NE win, but it's almost a lock)
ATL +2.5, HOU -150

First impressions: I obviously haven't figured in the lay properly yet,
but it's a start. Plus I like the picks - all but the Atlanta pick, but
frankly, that one's not so bad either. I just no very little about
either teams' playing style. Typically, AFC over NFC (which bears out
in that pick), but home dog on the spread. Either way, I'm fairly hedged.

Top parlay combos:

PHI, MIN, SF (with Plaxico possibly injured, could be a lock on Philly -
this is the 1/2/3 conviction picks)
BUF, ATL, SF (this is the 1/2/3 win percentage picks)

Frankly, I like this new system picks. I avoid the terrible games
(KC/SD weirdness, ARI/PIT weirdness, CLE covering any spread weirdness).

--------------------------------- crappy disclaimer: I just looked up
the spreads, and CAR is -3 rather than OFF. This changed all my picks
and convictions. Here are the NEW picks after entering the spread so
you can see how the Box readjusts.

Top 4 Spread/Moneyline Combos (spreads first, then moneylines):

OAK+4 (90% pick historically!), MIA -230
TB +3, TB (apparently, the Box hearts TB)
STL+12.5, DAL
ARI +6, PIT -260 (damn, I do NOT heart this game)

Top parlay combos:

TB, PHI, MIN (this is the 1/2/3 conviction picks)
OAK, STL, SD (this is the 1/2/3 win percentage picks)

So, all my previous commentary is pointless. The Box switches
conviction as spreads change, it's part of the danger of moving
averages. I actually don't hate these picks, and I liked TB to win
before the spread was announced anyway.

My thought on yours is it's very very favorite heavy on the top
spreads... historically, this does not pan out as the season goes on.
That said, this season has been a favorite heavy season, so who knows.
As to the moneyline, the whole key will be the lay I think. The payout
on NE is 330 wins 100. That's TERRIBLE. Basically, you get 30% return
for your money. So that's the real game - there's no money in easy
favorites. Which means I have to teach the system to base it's
convictions on lays as well - maximum payouts for minimum risk. For
instance, I think PHI is basically a lock this week, and it's paying out
140/100. That's more than DOUBLE the NE payout, and should therefore be
ranked higher. I'll have to program it... so much to do.

1 comment:

Juli GS said...

I really like your writing. Thanks so much, finally a decent website with good information in it.