how to calc props in basketball

[spongerat]

How do you cap a prop like largest lead in the game for basketball?

And individual player props such as points and rebounds?
Do you have to just estimate based on what % of the teams rebounds player X gets and extrapolate from there? Then use average number normalized against the other team with the overall average etc. Its just that when all the players are capable of making rebounds and points it is hard to figure out how many that exact player will get.

[Justin7]

Post one early in the morning, and I'll do an example.

[spongerat]

ok heres an example:

mizz v. uconn

largest lead of the game: 14.5
over
under

or if you prefer:
total rebounds plus points for Hasheem Thabeet 27.5
over
under

[spongerat]

for the first one i have no clue whatsoever on how to figure this out.

the second one i'm guessing you figure out the proportion of the teams rebounds and points the player has then go from there?

[Peep]

for the first one i have no clue whatsoever on how to figure this out.

I know how to do it, but no one but me would like the solution.

1) Copy 10,000 game logs, sort the scores for differentials.
2) Look at all the games where a team is favored by (-5).
3) Get the median for the largest differential.
4) Bet accordingly. If you have a two point advantage either way, I would feel confident. In this case, if you found the median differential was <12.5 or >16.5 for example.

[spongerat]

haha, idon't care about liking the solution, just if it creates a Ev+ bet

[Peep]

Well, why I don't like my own solution is I have to find logs, download them, sort and format them. A LOT OF THEM.

And then I don't know if the books will even keep this prop bet around. They do whine a bit when they get beat on props.

[Justin7]

Sorry, getting stats on college players is tough. NBA is much easier.

[sallysnax8]

how about lebron james pts rbds assists over under 45.5?

[Justin7]

how about lebron james pts rbds assists over under 45.5?

First, you need to normalize. Cleveland -9/190 suggests Cleveland will score about 100 points. Cleveland averages 100 points, so this game is likely to be typical of James' stats.

I didn't check the lineup, but has Cleveland's lineup changed in the last 20 games? If not, I'd just look to see how many times LJ has gone over 45.5 in the last 20 games (or even the last 10). In the last 10 games, it's gone over 6 times... so it's not horrendously priced.

Prices with higher numbers tend to be better priced. The difference between 44.5 and 45.5, while significant, won't give you nearly the advantage of a "small mistake like pricing a 6 as a 6.5.

[spongerat]

but do you need to factor in if they played easier teams in the past 10 as compared to next game? i guess it must be late enough in the season to just use averages but earlier in the season this might be a factor

[so im zach]

but do you need to factor in if they played easier teams in the past 10 as compared to next game? i guess it must be late enough in the season to just use averages but earlier in the season this might be a factor

I believe location (whether the game is home or away) and past performances over the last 30-40 days should be weighed more heavily than just taking season averages. Injuries, past performances against the opposing team this past season (check for minutes played, too), and time between games are other important factors.