A priori odds against Miami winning by exactly 1 point?

[pzhon]

Ignore the current score. What's a good ballbark estimate of the odds against Miami winning by exactly 1 point?

[rjp]

Depends on how often you figure Miami wins...

At 50% Miami wins by 1 point approximately 2.14% of the time.
At 40% Miami wins by 1 point exactly 1.71% of the time.

Just multiply how often you expect Miami to win by 4.27% (0.0427) and then subtract this from 100%

OK, so I was close the first time. [img]/images/graemlins/cool.gif[/img]

[pzhon]

[ QUOTE ]

Just multiply how often you expect Miami to win by 4.27% (0.0427)

[/ QUOTE ]
Thanks. That's much lower than I expected. Is the magnitude specific to the 1 point difference, or is a similar figure valid for any small difference?

[rjp]

Everyone will give you a different answer for that question, but I personally use the recent history of actual margins of victory to calculate a rough approximate to the odds of a team winning by a specific number of points, and I seperate home and away data sets. A home team winning by 1 point figures to occur a little less (3.66% of the time).

The most important number is 3, as home teams win by exactly 3 points roughly 15.14% of the time, and away teams win by exactly 3 points roughly 17.54% of the time.

[pzhon]

[ QUOTE ]
A home team winning by 1 point figures to occur a little less (3.66% of the time).

The most important number is 3, as home teams win by exactly 3 points roughly 15.14% of the time, and away teams win by exactly 3 points roughly 17.54% of the time.

[/ QUOTE ]
Thanks for your help. That makes sense.

In case you were wondering, a friend hedged the Mansion bet with a gap at 1 point. To help him hedge, I bet a few hundred that Miami would not win by exactly 1 point, so he wouldn't lose in that case. We weren't sure what the right odds should be.

[rjp]

Cool. [img]/images/graemlins/smile.gif[/img]