prop question

**Boner_18** · 03-08-11, 05:17 PM

I believe the 1.7 is average points scored by kobe per trip to the ft line (assuming all trips are on a non-scoring shooting foul, or 2pts max).

**antifoil** · 03-08-11, 06:31 PM

wouldn't the number of points scored per trip to the line assuming 2 shot simply be his ft percentage multiplied by 2? in this year he was shooting 81 percent. so he would have an 81 percent chance on each shot making it 1.62 expected points per trip to the line.

**antifoil** · 03-08-11, 06:43 PM

maybe i was looking at the wrong year. he shot 86 percent the year prior which comes to be 1.72. thanks.

anyone have an idea on my other two questions.

**smartbets** · 03-08-11, 09:27 PM

I think he computes it by 2 because the formula is figuring out points scored per possession. So if kobe shoots 71 percent, each poss would result in 2 free throws. So you divide free throw attempts per game by 2, then multiply by 1.71 to compute what he averages scoring via the free throw line per possession. I think. Lol

**Boner_18** · 03-09-11, 06:57 AM

As for the addition of the two numbers I think you are a little off base.

There is a chapter in Justin's book about aggregating poission numbers. Your approach would work for two numbers that were singular, assists + rebounds for example. However, because you are converting points scored to "successful scoring chances" so to speak you have to find a way to convert that part of the aggregated prop (notice j7 divides the prop line by 2.03 as well).

**Boner_18** · 03-09-11, 07:59 AM

Sorry didn't completely read your understanding of the problem. You are on it. I don't think using your own generated number to back out rebounds is the best approach. I think a better idea would be to back out a widely available rebound number... however, if this is out there you might just want to bet this instead.. just thinking out loud here.

**antifoil** · 03-09-11, 02:12 PM

yeah, my problem is how do i convert points to, as you call it "successful scoring chances", to be able to use with poisson so i can add it to rebound since they both now have an increment of 1. however, since the points and rebounds are added together for the book line i need to know how many the book considers to be rebounds and how many the book considers to be points so i can do the possion on the book number.

i understand how to do it by just adding rebound and assist. my question is how do you do it with points and rebound. so i guess essentially my question is when points and rebounds are added together in a prop line, how do i determine how many of those are rebounds and how many points?

of course if the linemakers makes separate lines for that players points or rebounds i just take that number, but what if this is not the case. actually these number may be wrong if the odds are presented at -150, +130, i would have to take the no vig line to figure out the percentage chance of the event to occur. shit this is a new problem.

**antifoil** · 03-10-11, 06:51 PM

its disappoint no one here has the answer to these problems..

is there another site where i can get the assistance i seek? thanks.

**Data** · 03-10-11, 11:06 PM

Originally posted by antifoil

its disappoint no one here has the answer to these problems..

is there another site where i can get the assistance i seek? thanks.

<hat on>
F#ck, how true! People should be expecting every (unreadable) question answered, it's a basic forum poster's right! Thinking about this made me realize another gross injustice in the sportsbetting and gambling in general.

While many (hardworking) people lose their asses out, a few undeserving individuals are raking in the big wins. This should NOT be happening in a fair society. This whole gambling business must be regulated by the UN. Every gambling operation must strip the winners of 90% of the winnings and redistribute those winnings among the rank and file gamblers who are the backbone of this very industry. CHANGE! Or, face the REVOLUTION! Gamblers unite!
<a commie hat off>

**smartbets** · 03-10-11, 11:41 PM

That was funny

Antifoil: its right in the video-get the points per scoring possession for the players average then get one for the prop-compares these numbers with poisson and u can see if u have a wager- just add adjusted numbers for rebounds, asists, blocks, etc. to each number then compare them using poisson

**Justin7** · 03-10-11, 11:51 PM

Originally posted by Data

<hat on>
F#ck, how true! People should be expecting every (unreadable) question answered, it's a basic forum poster's right! Thinking about this made me realize another gross injustice in the sportsbetting and gambling in general.

While many (hardworking) people lose their asses out, a few undeserving individuals are raking in the big wins. This should NOT be happening in a fair society. This whole gambling business must be regulated by the UN. Every gambling operation must strip the winners of 90% of the winnings and redistribute those winnings among the rank and file gamblers who are the backbone of this very industry. CHANGE! Or, face the REVOLUTION! Gamblers unite!
<a commie hat off>

Your post cracked me up. Sports betting is the opposite of socialism. Money flows from the dumb to the smart. Ouch, did I say that?

**antifoil** · 03-11-11, 01:05 AM

so you didn't understand what i was asking?

**antifoil** · 03-11-11, 02:26 AM

i thought of a simpler way so maybe you can understand.

when a sportsbook puts out a prop line of points + rebounds, how do i determine how many of those are points and how many are rebounds?

that's one sentence that seems to me to be easily readable; however, if you still find that unreadable i will try again. thanks.

**Justin7** · 03-11-11, 02:15 PM

Originally posted by antifoil

i thought of a simpler way so maybe you can understand.

when a sportsbook puts out a prop line of points + rebounds, how do i determine how many of those are points and how many are rebounds?

that's one sentence that seems to me to be easily readable; however, if you still find that unreadable i will try again. thanks.

Do distributions for both. One for points, one for rebounds.

When doing points, you need to convert points to "scoring possessions". Maybe each score nets about 1.5 points... so if a player is projected to have 15 points, his Poisson is based on 10 scoring possessions.

So what do you do if you have 15 points, and 10 expected rebounds? I don't a perfect mathematical solution, since you have two different basis... But you could add the curves up for different sums if you're a decent programmer.

**antifoil** · 03-11-11, 07:00 PM

thank for your response. if i am reading what you say correctly, it is a restatement of the video. i fully understand the concepts described in the video. my question is related to the books line not the line expected line created by me.

using your numbers, i would come up with an expected line of points + rebounds of 25. in single trials it would be 20. now suppose a book puts up a line of 27.5 that leaves 2.5 available points/rebounds that i need to get to single trials. of course i am making the assumption the book is automatically given 15 points and 10 rebounds to the player, which may or may not be correct. but for this example lets say it is correct. how do i determine of the 2.5 remaining points if all three are rebounds, all 2.5 are points or some combination of both?

**Justin7** · 03-11-11, 07:23 PM

Guesstimate.

use the average differential per event. If 10 rebounds and 10 scores result in a "25", the conversion is about 1.25 per event. So I would compare a Poisson of 20 vs 22 to guess your edge. There are all sorts of problems in doing it, but it should at least get you in the ball park.