PDA

View Full Version : Me (11-6 +4.74 YTD) & the "System" (7-14 -1.88 YTD) Wk 3


TRWIII
09-22-2005, 07:21 PM
Im giving my friend hell for the poor performance of the System over its first two weeks use. He thinks I'm serious, so, as a statistician, he felt it neccessary to explain that it was variance and the sample size wasnt big enough yet /images/graemlins/grin.gif.

Same format as last week.

My picks/analysis:

Maryland +2.5 @ Wake Forest -106

Im afraid this is a sucker bet and Im giving money away with this pick. I cant for the life of me figure out why Maryland isnt at least a field goal favorite here. Both teams are 1-2, but Marylands 2 losses have come at the hands of two quality teams, and they've been able put up around twenty on both Clemson and West Virginia. Wake lost to Vanderbilt and Nebraska, two teams noticably south of quality, and nearly got beaten at home by ECU. I cant see Wake winning this game.

Troy +20 @ South Carolina -108

South Carolina has shown no tendency for having the talent or ability to run up the score on anyone, nor are they talented enough on defense to stop the other side from scoring a decent amount. USC would have to hold Troy to either a field goal or a touchdown to cover this Im thinking, and they still might not cover then.

UNC +10.5 @ NC State -101
under 45

UNC has looked much better than expected in their first two games. Granted both were losses, but they were to Georgia Tech and Wisconsin, two very good teams. They also showed much more defensive prowess then many expected. NC State looked good against VT, and smushed E. Kentucky (which isnt all that important). The NC State defense has shown the ability to frustrate opponents scoring ability. The home field and overall talent lead NC State to win this game, but UNC's defense is enough to keep it a low scoring and close slugfest.

Marshall -3.5 @ UCF -101

Marshall lost last week to a very odd twist of fate (not to mention curious playcalling). Regardless of the outcome of the game against Kansas State, Marshall showed that they have the talent and ability to play with big conference schools. UCF, on the otherhand lost to South Florida convincingly. Even though this is a road game, I cant see this even being close.

Georgia Tech +12 @ Virginia Tech -109
under 40 -107

I though about parlaying these two, since if one wins the other is very likely. The only time this season that VT got scared was their only game against a legit defense (NC State). GT is one of the best defenses in the ACC and will make scoring difficult for VT. Reggie Ball is probable for the game, and the GT offense will threaten the VT secondary with Calvin Johnson and Bilbo to toss the ball to downfield and PJ Daniels to pound the rock. GT's d-line and fast secondary will limit Vick's scrambling ability as much as anyone can. I dont see the Yellowjackets winning this game, but I dont see either team scoring north of 20.

System picks:
UL Monroe @ Fla. Atlantic -7.5 even
ECU @ WVU -21.5
Kent @ Ohio -3.5 -108
Tulane @ SMU +3.5 -105
TCU +3.5 @ BYU -108
Michigan @ Wisconsin +2.5 +101
Hawaii +1.5 @ Idaho -108

Here's to hoping the thoeretical long term gets here soon /images/graemlins/grin.gif.

TRWIII

Last Week: Week 2 (http://forumserver.twoplustwo.com/showflat.php?Cat=&Number=3417659&page=1&view=colla psed&sb=5&o=93&fpart=1)

09-22-2005, 08:46 PM
Even though I don't play CFB, it seems this is one of those situations when a YTD 7-14 historically successful system is due for a serious correction. I may play small on these picks. Thanks for the post and GL.

TRWIII
09-22-2005, 09:54 PM
Heh, thats what Im banking on, but even though Ive seen the betting history Im still not sold on this thing yet (thats why Im only betting quarter units on it).

The main reason Im not sold on it is it's based on math that I cant even begin to understand /images/graemlins/grin.gif. Thats not saying a whole lot, since Im a discrete mathematician. I know a good many discrete mathematicians, some of them unquestionably geniuses, and none of us that I know of have anything more then a basic working knowledge of statistics /images/graemlins/grin.gif.

Im still plugging along with it, trusting the law of large numbers (one of the few statistics maxims I understand /images/graemlins/smirk.gif.

Cheers

TRWIII

09-24-2005, 08:06 AM
I played the remaining 6 games straight up and in cheap 5 of 6 round robin parlays. I was actually encouraged by the loss of Fla. Atlantic last night -- significant correction due at some point between now and December.

scalf
09-24-2005, 10:19 AM
/images/graemlins/smile.gif.

gl

i like wake here to run regularly; and terps just have not showed me much....pinny w.f. -1 -113

i like va tech.ga tech over..remember special teams usually give an extra 7-10 pts at blacksburg

i like your troy bet, but cannot pull trigger yet; spurrier has changed about 5-6 starting positions due to poor performance last week against bammy

gl

/images/graemlins/wink.gif /images/graemlins/spade.gif

TRWIII
09-24-2005, 12:17 PM
You know, after further pondering, Ill add one last bet. I think I like GT on the money line at +410. So Ill put down .25 units to win 1.02. I think GT pulls this win somewhere between a quarter and a third of the time, which is enough of a margin to make this profitable. But then again, I stink at moneyline wagers /images/graemlins/grin.gif.

TRWIII

09-24-2005, 05:41 PM
"System" 7-18 YTD up to the minute! POS...

TRWIII
09-24-2005, 07:44 PM
Upon further pondering, Im an idiot. Did anyone realize VT was that good? /images/graemlins/confused.gif

*shrugs* Nobody gets em all.

TRWIII

TRWIII
09-24-2005, 11:47 PM
You arent kidding. This POS made a late push towards mediocrity, but still finished 3-4 losing .31 units (wagers were in quarter units). I cant say much, b/c I had a very poor week myself, going 2-5 losing 3.35 units.

YTD: Me: 13-11 +1.39 units
System: 10-18-1 -2.19 units

Ill give this system two more weeks putting money on it. If it flops again, Ill just keep track of its results and cut my losses. Here's to getting back on track next week.

*wanders off mumbling about knowing why he hates statistics*

TRWIII

P.S. As a side note, part of me wants this thing to flop terribly so I can get back on my soap box trashing "systems". I kinda like it up there /images/graemlins/grin.gif.

09-25-2005, 07:49 AM
It finished nicely enough and probably is due for a correction, but was disapponting nonetheless.

A system is only as good as its real applicability across seasons. If the system started in 1980 and is using the same principles to analyze games now, it's hard to see how it's still useful because the college game has changed so much since -- far more passing today, overtime rules etc.

lorinda
09-25-2005, 08:28 AM
If it flops terribly enough, it would be worth fading. It's not in that range yet.

Lori

09-25-2005, 08:35 AM
By the time it's in that range, it will be due for a correction to near 50/50. Even systems that lose their applicability choose games at a 48-52% clip. There's not a chance I fade this thing at the rate it's going.

lorinda
09-25-2005, 08:39 AM
What do you mean by a correction? Surely you don't mean that if it loses 10 in a row then it is "due" to win 10 in a row?

Lori

09-25-2005, 08:45 AM
Not 10 in a row. But there's a decent chance it will accumulate back those 10 losses with 10 wins over the long-run if it is now a system with zero applicability. If it has negative applicability (in the range of 45-48%) then it's not due for any consistent winning, but I don't consider 10-18 ATS YTD a sign it's at that point.

lorinda
09-25-2005, 08:47 AM
[ QUOTE ]
Not 10 in a row. But there's a decent chance it will accumulate back those 10 losses with 10 wins over the long-run if it is now a system with zero applicability.

[/ QUOTE ]

Excuse me for being stupid here. It seems to me that you're saying that if I flipped a coin to make the picks, and it went 14-24 that it would, at some point, be favored to get ten extra wins than losses.

I can't see how this can be possible /images/graemlins/confused.gif

Lori

09-25-2005, 09:03 AM
A system with aggregate zero applicability doesn't flip a coin each time. It contains some principles/trends that are applicable to successfully choosing games, others that are neutral (like flipping a coin), and others that have negative applicability. If among the 28 games it's chosen a large majority of these were chosen predominantly with negatively applicable principles, and there are an equal number of games chosen using both the positive and negative principles (ed: in the long run), then among the next 28 there is an increasing chance these games will be chosen using positive principles. Just to reiterate, I am assuming this system has zero applicability. I am not arguing for it otherwise.