OK here is a question.

**warriorfan707** · 08-24-11, 11:50 PM

not possible to answer, it depends on if the dog or fav won the game by 3

**Justin7** · 08-25-11, 12:04 AM

The 16% isn't that releveant. As warrior fan alluded to, the push rate is what matters.

I think the breakdown is about 10%/6% (a home fav of -3 will push about 10% of the time; a 3-point dog will win by 3 about 6-7%).

**wantitall4moi** · 08-25-11, 12:37 AM

Originally posted by Justin7

The 16% isn't that releveant. As warrior fan alluded to, the push rate is what matters.

I think the breakdown is about 10%/6% (a home fav of -3 will push about 10% of the time; a 3-point dog will win by 3 about 6-7%).

I have all the break downs, and there really is no one answer because so many books post too many lines to say for certain when a game is -3 in the first place, I posted some of the numbers over at EOG, but not too specific. But home favs closing -3 (in one of my databases) pushed 39 times out of 447 games, roughly 8.7%. Road -3 favs pushed 26 of 306, 8.5% so not much of a difference. Those results are from 1994 to 2010. But that is just the closers at that one book. I can break down line moves back to 2001, and see where some of those games that closed -3 may not have started there, or sme that didnt close -3 did start at -3 and moved. And other lines that closed something else hit -3 for at least a flash.

I just thought there might be a way to determine the dogs chance to win by 3 versus the favs chance to win, and go from there. But I figured it would be a stretch since even though games are all posted -3 they obviously arent all going be weighted the same. At least without looking at results and try and call that the probability.

Basically I was wondering if there was a neutral way to predict the probability of something happening without using the past results or known scores.

Because like I have always said, and when I started saying it it was 13%, that a fav that is lined -3 winning by 3 wasnt 13% probability. Nor is it now 8.6% ,as the findings now show, (at least for that book). I just wanted to see if someone had a guestimate of what it should be and how t was matching up to the results we can see.

This is related to the question the other guy had about variance. I just slipped in some known factors to see if it could be plugged into any know probability formula.

**That Foreign Guy** · 08-26-11, 04:46 AM

bout tree fiddy.

**wiffle** · 08-26-11, 07:49 AM

50/50

either it does or it doesnt

**wantitall4moi** · 08-26-11, 07:58 AM

OK how about this, a hypothetical to see if there is a formula....

16% of games finish with a 1 pt differential
100% of games are lined-1
60% of the games will be won by the favorite.

chances/probability the fav wins by 1 pt.

**WendysRox** · 08-26-11, 08:45 PM

I'm not proficient enough in statistics to come up with a confident response to the probability question, but my search for the answer led me to this website: http://www.westgard.com/lessons-5.htm

So, thanks for that.

edit: Confidence Interval is the phrase I was trying to remember. Here is a lesson about it from that same website: http://www.westgard.com/z-11-confidence-intervals.htm. And, just to be clear, I'm sure most of the think tank regulars are familiar with these formulas. I post these articles for those users like me that are trying to improve their understanding of math as it pertains to gambling.