BuzzerBeater Forums

Altering our model to more closely match the shot attempt distribution for the Celtics result in a much closer matchup, with LA finishing on top by an average of 94-90, winning just 61% of the time. We believe this more accurately reflects what is likely to happen, and suggests that maybe the Celtics know better than our virtual coach how to utilize their team best. Not surprisingly we predict Kobe leads the way for the Lakers, as he has in 24 out of 28 games this season. In fact he scores more than 30 points 63% of the time. We expect the rest of the Laker scoring to be spread around. If fact the probability of no other Laker scoring more than 18 points is 70%.

For the Celtics, Ray Allen is most likely to have a breakout game, scoring more than 20 points, 67% of the time. Paul Pierce will likely have an average game, scoring in the upper teens. Kevin Garnett is likely to have trouble scoring against the large Laker interior, scoring less than 15 points 75% of the time. Derek Fisher will keep Rondo from scoring a lot of points, but he’s likely to play a pivotal setup role, handing out 7 or more assists 62% of the time.

This really is a close matchup, so the Celtics don’t have to do much to push it over the top and no single factor sticks out when looking at Celtic victories. They rebound a little better, they shoot a little better, Kobe scores a little less, Ray Allen a little more and so on. So the advice from the virtual coach is simply to go out and play hard!

I don´t like Celtics nor Lakers, but GO LA hehe!!! XD

we cant see it?

(http://www.buzzerbeater.com/BBWeb/reportmatch.aspx?matchI...)

sorry about that..here is the example game

the game is going on now. hope you are right, haha go lakers!

Good prediction that La lakers won :) Haven't gone through the rest of the stats, but i'll check that later ;)

Very nice again.


May i see the Kobe`s skill ? please.

at the moment i have not made any modulations of form, effort, or tactics... i am planning on setting up fined tuned tactics which use real nba shot distributions to distribute shot types inside and outside etc...

form i dont know how i would measure statistically so it would just be by the seat of my pants. Maybe I will do something that will look into back to back games and modulate form down for those and up for longer rest... but first I'd like to see whether i can measure that this is a real effect in the nba.

effort I dont know if i could justify setting higher or lower for either team.

form in sim = 3 or 4 past games for the player and % of shoot (for example: 1st game=7/8 ; 2nd game=3/6 ; 3rd game=6/9 } = 16/23 = strong form for example

i think the form its only reduces or increments the attack power, i dont think its necesary to reduce the other skilss the player have..

its just an opinion.

i dont think thats a good option because if you are a defender the shootingpercentage has nothing to do with your stamina