Teaser question regarding Wong's SSB.

[Micturition Man]

For people who do teasers the most valuable reference I know of is Table 20 in Wong's SSB.

It is Wong's semi-subjective attempt to quantify how often the NFL favorite wins by various scores. (Famously they win by three 10% of the time and by 7 6% of the time.)

To the extent that these figures are reliable you can calculate the EV of any teaser possibility at any price, assuming you are confident you have a 50/50 market line point spread to start with. (And even if you don't you can always assign a reasonable margin of error to your market line and see whether the teaser is still +EV).

But I have some lingering doubts I haven't been able to get over about using table 20 to calculate cover % of big favorites (let's arbitrarily say -9.5 or higher).

Wong's methodology in compiling the table was to look at how often there was a push against the 3 when the favorite was from -1 to -5. Against the 7 when the favorite was from -5 to -9, etc.

My problem with this is that surely as you move up a point the distribution of scores moves too. For example if a -3 team wins by three 10% of the time, a -9.5 probably wins by 3 less often (which is offset by winning by say 7 or 11 more often).

This difference could be very relevant when looking at the EV of some nonstandard teasers like 10-pointers or or a 6 pter that pushes on the 3.

On a separate note where can I get a ton of NFL data?

[ImStillBen]

Your doubts are well-founded, and that's not a separate note.

[Micturition Man]

[ QUOTE ]
Your doubts are well-founded, and that's not a separate note.

[/ QUOTE ]


Thanks for the reply... do you have any further opinions on this? I would like to get a discussion started on this if possible.

Also any idea about where I can get that NFL data?

[whipsaw]

www.nfl.com

[Micturition Man]

[ QUOTE ]
www.nfl.com

[/ QUOTE ]


I mean past data in excel format. If it's on the NFL.com site I don't see it.

[whipsaw]

What kind of data do you want? There are lots of places to get final scores.

[Micturition Man]

[ QUOTE ]
What kind of data do you want? There are lots of places to get final scores.

[/ QUOTE ]

As much as possible, final scores for as many years as possible would be nice.

Final scores versus point spreads would be even better, though I'm not sure how exactly 'official' point spread would be determined.

[whipsaw]

http://www.goldsheet.com/historic/histnfl.html