PDA

View Full Version : NFL Teasers by weeks

mikevegas
09-04-2010, 09:50 PM
I posted this here and over there, too.
When breaking the NFL season down into quartiles I don't think that the sample sizes are large enough to draw significant conclusions.

[url=http://www.flickr.com/photos/30485932@N08/4958740117/][img]http://farm5.static.flickr.com/4147/4958740117_29f9edd968_b.jpg
The Chaperone
09-05-2010, 02:21 AM
What does this meeeeeean?

http://www.youtube.com/watch?v=OQSNhk5ICTI
donovan
09-05-2010, 05:46 AM
Non-BS looks like it is the best of all!
Missouri Girth
09-07-2010, 07:04 AM
Are the Non BS 4.5 to 6 based on road dog data?
mikevegas
09-07-2010, 08:35 AM
Are the Non BS 4.5 to 6 based on road dog data?

Road dogs only.
Home dogs +4.5 thru +6 are not included in the above data.
Missouri Girth
09-07-2010, 08:50 AM
Road dogs only.
Home dogs +4.5 thru +6 are not included in the above data.

Thanks Mike.
2w2p2s
09-08-2010, 08:04 AM
Pinny not fearing the BS teasers anymore?
JoeOffsuit
09-12-2010, 05:38 PM
Awesome! Now when I absolultely have to get action down on a game lined at 5, I have something I can pull...
trixtrix
09-13-2010, 09:44 AM
Road dogs only.
Home dogs +4.5 thru +6 are not included in the above data.

why are you not including home dogs +4.5 thru +6?
kgbted21
09-13-2010, 05:24 PM
i recall a big fight over at the old lva with bill the cop who played only away dogs in teasers because of a sampling of historic data but i don't know if this is related.
mikevegas
09-13-2010, 05:43 PM
why are you not including home dogs +4.5 thru +6?

When BTC did his research the home dog +4.5 to +6 subset did poorly.
The Chaperone
09-13-2010, 05:58 PM
i recall a big fight over at the old lva with bill the cop who played only away dogs in teasers because of a sampling of historic data but i don't know if this is related.

Data mining at its finest imo.
ComptrBob
09-13-2010, 07:53 PM
i recall a big fight over at the old lva with bill the cop who played only away dogs in teasers because of a sampling of historic data but i don't know if this is related.

I think this is the thread between BTC, myself and others at the old LVA.

NFL Teaser thread (http://www.lasvegasadvisor.com/forum/messageview.cfm?catid=36&threadid=199806&FTVAR_MSGDBTABLE=)

Still is a good read for those who are interested in data mining.
WarDekar
09-13-2010, 08:24 PM
FWIW though there *can* be a difference between the 2 (home/road 4.5-6 point dogs) because the scoring distributions can be (slightly) different. Note I'm not saying either one of those subsets are good to tease- you'd be better off just betting the road teams straight or on the ML in the sample he datamined since I'm guessing they hit ~55% and that's why they were 75% teased or w/e.

And LOL at that thread, I like how he separated out the odd/even years after you suggested he do that to develop his hypothesis LOL.
GroovinMahoovinPartDeux
09-14-2010, 03:20 PM
I haven't looked at the data in a while but when I checked circa 2006 or so, there was more variance in the results for home dogs vs road dogs, making the home dog MLs having more value and the road dog teasers having more value. FWIW I usually try to put part of my bet on the ML when I take a home dog.
The Chaperone
09-15-2010, 02:35 PM
I haven't looked at the data in a while but when I checked circa 2006 or so, there was more variance in the results for home dogs vs road dogs, making the home dog MLs having more value and the road dog teasers having more value. FWIW I usually try to put part of my bet on the ML when I take a home dog.

This however, is kind of a chicken and egg scenario. Just by knowing that the road dog teasers have done better than home dog teasers, I could guess that the distributions are probably tighter for the road dogs.