And now something completely different. CFB Wk2 5-1 +3.96 units

[TRWIII]

Im going to try something a bit different this week. As a rule, I generally distrust "systems", since theyre generally based on biased data and dont accurately predict furture results. That being said, Im good friends with a professor in out Statistics department, and he has "honored me" [img]/images/graemlins/smirk.gif[/img] with the details fo his system. It has won consistently for him practically unaltered for 6 years (or so he says). As a disclaimer, Im a discrete mathematician, not a statistician, so Ive got no screaming clue how this thing works... all I know is its huge, complicated and ugly, and it supposedly works. I figure Ill give it a whirl for a few weeks at a quarter unit a bet.

My picks come in two sections for this week: first are the ones Im going to bet at a full unit which were the ones I arrived at after my own analysis, the rest are quarter unit bets that this system pops out. We'll see how this works.

My picks and analysis:

Viginia Tech -20 at Duke -108

Duke lost to ECU, struggled to stop the passing game, and was only mediocre against the run. And their offense was at best mediocre against ECU. Did I mention they lost to ECU? Virginia Tech, on the other hand, beat one of the best defenses in the nation on the road, and Marcus Vick started showing flashes of Vickness. Duke wont stay close for more then the first couple of posessions.

Iowa -8.5 at Iowa State -106

This line I must admit confuses me. Granted, its a "rivalry", and Iowa is on the road, but Iowa State wasnt all that spectacular against Illinois State, giving up 21 points against an offense consisting of 11 guys named Fred. I honestly cant see this being within two touchdowns.

UNC at Georgia Tech -12 +100

GT is coming off a blowing up Auburn's inexperienced QB, and now they get a QB who has never started a college game. John Tenuta is having wet dreams about what the GT defense is going to do to this poor kid, especially since UNC's offensive line isnt nearly as good as Auburn's. Meanwhile, PJ Daniels, Calvin Johnson, and Reggie Ball are a good choice for the most dangerous triple threat in the ACC on offense, and UNC is returning starters from a middling defense last year. This one is likely over by the end of the first half.

Wake Forest at Nebraska -6 -109

The Deacs lost at home to Vanderbilt. That should be sufficient, but Ill give a little detail [img]/images/graemlins/grin.gif[/img]. Wake allowed nearly 5 yards/rush attempt and lets Vandy's QB pass for 276 yards at nearly 70% accuracy. Nebraska's squash of Maine isnt too significant, although holding them to -6 rushing yards on 37 attempts is pretty neat, and running is Wake's bread & butter. Although Im still leery of Callahan's west coast offense in Licoln, I think we might be surprised with the Husker's this year.

And oh yeah, take the under 44.5 -105... should be pretty low scoring.

--------

Now for the completely different part (no analysis, b/c I dont understand it, but Ill let you know what the system thinks the line ought to be).

UL Lafayette +6.5 at Eastern Michigan -101
(System says ULALA should be a ppoint favorite)

Central Michigan at Miami(OH) -22 -105
(System says Miami(OH) -30)

Army at BC -27.5 -101
(System says BC -36)

Notre Dame +7 at Michigan even
(System says Michigan -1.5)

Kansas State at Marshall +9.5 -105
(System says Marshall -4... thats not a typo)

Texas at Ohio State -1 -106
(System says Ohio State -7)

South Carolina at Georgia -17.5 -107
(System says UGA -28)

TCU -14.5 at SMU -105
(System says TCU -33)

Hawaii +32 at Michigan State -107
(System says Michigan State -7... and thats not a typo either)

Oh and 2 things if youre curious:

This thing also spit out every game I picked above as a recommended bet (except the totals... it doesnt do those [img]/images/graemlins/grin.gif[/img]), so maybe Im doing something right.

It also spit out another half dozen or so bets that it said were "marginally" profitable [img]/images/graemlins/crazy.gif[/img]. Ill post em if anyone's interested.

Good luck for the upcoming weekend.

TRWIII

[Easy E]

[ QUOTE ]
Im going to try something a bit different this week. As a rule, I generally distrust "systems", since theyre generally based on biased data and dont accurately predict furture results.

UL Lafayette +6.5 at Eastern Michigan -101
(System says ULALA should be a ppoint favorite)

Central Michigan at Miami(OH) -22 -105
(System says Miami(OH) -30)

Army at BC -27.5 -101
(System says BC -36)

Notre Dame +7 at Michigan even
(System says Michigan -1.5)

Kansas State at Marshall +9.5 -105
(System says Marshall -4... thats not a typo)

Texas at Ohio State -1 -106
(System says Ohio State -7)


[/ QUOTE ]

So, do I read this correctly as recommendations for
UL Lafayette
Miami(OH)
BC
Notre Dame
Marshall
Ohio State

among others?

[TRWIII]

Yep you read them correctly. (Im lazy, thats the format the thing's suggestions come out in so I just copy and pasted them). The picks it recommends are the one in the top line, the line in parentheses are what it thinks the line ought to be.

Example:

Notre Dame +7 at Michigan even
(System says Michigan -1.5)

means the system says that ND +7 is a good bet, the even is the odds I laid on the bet, and the (System says Michigan -1.5) says the System thinks that Michigan should only be a point and a half favorite.)

Im more or less doing this for sh*ts and giggles and a bit of academic curiosity. Some of those picks look terrible to me, but Ill play the system for a few weeks just to see.

Moral of the story: Dont let a statistician get a Master's degree in computer science... you get wacky stuff like this instead of something beneficial to the academic community. [img]/images/graemlins/grin.gif[/img]

TRWIII

[ftball0000]

Thanks for the bets, I'll look further into them, and perhaps a few bets

[scalf]

trw: whooosh; i am certainly glad to hear you are a discrete mathamatician; buoy, those statisticians are looney birds...

lol

butcha know;

weird;

the harmon forecast is hawaii winning in east lansing, str8 up...hmmmm.

maybe the numbercrunchers are up to sumpin'

the one thing i know about number-crunching systems: sometimes they work; and sometimes they don't..

lol


nice post; thanks for the info; now you'll have to perform an elegant proof for this guy to "even up"

lol

gl

[img]/images/graemlins/wink.gif[/img] [img]/images/graemlins/club.gif[/img] [img]/images/graemlins/grin.gif[/img]

[CCx]

i smell the biggest $10 parlay coming up this saturday - go system! (nice post, good luck)

[TRWIII]

[ QUOTE ]

nice post; thanks for the info; now you'll have to perform an elegant proof for this guy to "even up"


[/ QUOTE ]

LOL, even if he'd let me give you the details of of the system, I strongly doubt an elegant proof exists [img]/images/graemlins/grin.gif[/img]. This system strikes me as an instance of making somthing sufficiently complicated that nobody but its creator understands whats going on... he tried to explain it to me, but I got lost after about three sentences.

But this guy isnt all that prone to exaggeration, and at least he understands statistics, so I hope he knows what he's talking about, for his bankroll's sake.

TRWIII

[mosuavea]

Nice post, since its getting a little closer to Saturday, I figured I would bump it up as I may be playing a couple fo these and may throw a little bet down on a parlay of the system games for giggles.

[TRWIII]

I was thinking the same thing as far as the parley goes. I practically never bet parlays, but just for fun I put one down today at sportsbook.com. Here she is:

$10 bet to win $30660
12-wager parlay:

VT -20.5
GT -12.5
Nebraska -6.5
ULALA +7
Miami(OH) -22
Marshall +9.5
Hawaii +31
TCU -14.5
BC -27.5
UGA -18
ND +7
Iowa -9.5

[img]/images/graemlins/grin.gif[/img]Here's to hoping statistics are actually useful for something [img]/images/graemlins/grin.gif[/img]

TRWIII

[TRWIII]

Eh my picks went 3-2 +0.94 units for the week, 8-3 +4.9 units for the year.

The system picks went a mediocre 4-5 for -0.31 units (they were quarter unit bets).

Eh, heres to hoping next week is a bit better for the system. As for myself, Ill take a 60% win rate every week [img]/images/graemlins/grin.gif[/img].

TRWIII