Week 9 -- A modest system projection

[cyanyde]

Date & Time Favorite NFL Spread Predicted Spread Underdog Total Difference Difference Strength Homefield Strength Bet?
11/7 1:00 ET NY Jets -3 -5 At Buffalo 36.5 2 18 12 NY Jets
11/7 1:00 ET Philadelphia -1.5 -2.5 At Pittsburgh 43.5 1 7 18 Philadelphia
11/7 1:00 ET At Detroit -3.5 -1 Washington 38 -2.5 30 10 Washington
11/7 1:00 ET Dallas -1.5 -3.5 At Cincinnati 43.5 2 18 12 Dallas
11/7 1:00 ET At Carolina -7 -2.5 Oakland 41.5 -4.5 46 4 Oakland
11/7 1:00 ET At Miami -3 -5 Arizona 34 2 16 9 At Miami
11/7 1:00 ET Kansas City -3 -2.5 At Tampa Bay 44 -0.5 10 42 At Tampa Bay
11/7 4:05 ET At NY Giants -9 0.5 Chicago 35.5 -9.5 95 11 Chicago
11/7 4:05 ET Seattle -6.5 -4 At San Francisco 39.5 -2.5 30 27 At San Francisco
11/7 4:05 ET At San Diego -6 -2.5 New Orleans 48.5 -3.5 40 3 New Orleans
11/7 4:15 ET New England -2.5 -2 At St. Louis 48 -0.5 9 43 At St. Louis
11/7 4:15 ET At Denver -6 -4.5 Houston 41.5 -1.5 19 3 Houston
11/7 8:35 ET At Baltimore -6 -5 Cleveland 35 -1 13 2 Cleveland
11/8 9:05 ET At Indianapolis -6 -3 Minnesota 60 -3 32 1 Minnesota
Total Total
383 197

---
Last week, the system produced a 7-7 record, with a 5% return based on the difference strength. Not satisfied, i Calculated a better difference, that had it been in play, would have returned 50% on the entire system. Good luck this week.

[lorinda]

Keep at it, and keep posting the predictions in advance.

It will help you to keep honest to yourself and after a while you'll get more people here, including me, interested and helping to tweak the system.

Edit: You could do with working out what bet level = no bet.

If you have a probability under 52.5%, you should factor this in and get the prediction to say bet = 0.

HOWEVER for now, whilst tweaking, leave the no bets in there, just don't bet them.


Good luck

Lori

[TRWIII]

[ QUOTE ]
Not satisfied, i Calculated a better difference, that had it been in play, would have returned 50% on the entire system. Good luck this week.

[/ QUOTE ]

*sighs* This is why this is doomed to failure. This isnt how statistical projections work. I strongly suggest you either pick up a collegiate level stats book or Wong's Sharp Sports Betting as it addresses this issue succinctly. Using the same data as a test for eficacy that you used to derive the system answers nothing and is essentially worthless. If you keep this up, you will lose money in the long run.

If you really think this system works, test it on three or four hundred games that you never looked at to derive the system, say NFL games from 1995-2000.

The way your going is bound for disappointment, as you are essentially using statistically biased data to justify coinflip picks.

Dont mistake me, this is not negative towards you, just a friendly warning from someone who has studied a good bit of statistics and knows what perils lie in using biased data.

Best wishes (and research)
TRWIII

[Easy E]

One, use the Code function for your charts, it's really hard to follow numbers correctly. It takes some tabbing work to get things to line up (see my YTD result posts), but it's worth it if you want us to follow more accurately

Second- you're 4-8 with two games left (both traveling significant dogs) and, if I'm reading this correctly, down 95 units.

Third- as TRWIII said, post-evaluating results to "improve" the system should be a big flag that the system isn't as good as you think it is.

I think bets on every game are a bad idea. Your "system" should concentrate on a few games alone as recommended bets- everything I've seen says that's the only way to long-term financial victory.

But as Lori said- keep posting. We'll criticize results, not you, if you can keep it impersonal for yourself. Keep plugging away! [img]/images/graemlins/laugh.gif[/img]

[bugstud]

I think his system is based on $, not units, but I might be wrong. I think the advice on the thread is on target, though. The system needs to be tested on unbiased info from previous years. I think running the system for all the games this year and see how things work out would be interesting, though.

[Easy E]

call it 1 unit = $1 then.

[bugstud]

I was thinking in one post he said 10 = 1 unit...time to go run search and see.

[Easy E]

It was his suggestion, I believe.

My point was, it didn't really matter.