BuzzerBeater Forums

But the player does take into acount the history (shots that did or did not fall) of shots he has been taken on a specific place on the court. Right?

no. independent events have no "memory"

its complicated... and i don't think i should go into TOO much detail of how the GE works, but players do have a memory in the sense of they understand what is a good shot and what is a bad shot this game and this influences the decisions they make once they decide to shoot.. but once they decide to shoot, there is a probability that that shot will go in... lets say 65%. sometimes it goes in, sometimes not. The next shot they take.. remember all the skills are likely to be different, that shot is going to be from a different spot on the floor.. a different situation in the game, he might be more tired, he might have a different defender.. who knows.. a million things could change... but say at the end of the day, this shot has a 30% probability of going in and he takes it. Should we take the fact that he missed his 65% opportunity and have it influence the 30% probability upward in an attempt to "reduce the randomness"... I think not, that would defeat the purpose of all the calculations that went into that 30% shot ending up being only a 30% opportunity.

I understand people think about it as.. my guy is a 10 his guy is a 5 that means, my guy should shoot 65% (or whatever, dont take these numbers seriously) so when i take 100 shots i should hit 65 of them..

the problem is that this ignores all the rich dynamics we consider when simulating a real game, we don't want to simplify these dynamics down.. its what makes the game engine interesting. It is inherently probabilistic, but all the evidence suggests that real life is probabilitic too... statisticians cannot find correlations between shots when they look for them. "Getting hot" actually has no evidence in fact. There is a great paper on this analyzing basketball.. maybe if i find it later i will post it here.

"Getting hot" actually has no evidence in fact. There is a great paper on this analyzing basketball.. maybe if i find it later i will post it here.

Oh where did I read a summary of this ?? I thought it was in a recent Reader's Digest, which quotes all it's contents. They talked about "hot hands" and studied NBA records that a player who missed his previous shot was MORE likely to make the next, than someone who had scored on one or two last shots.

here is a report on a presentation that was given at a sports analytic conference.

bottom line, what you said is true... it is also a very small effect and could be explained by effects on shot selection.

http://myespn.go.com/blogs/truehoop/0-39-90/hot-and-heavy--about-nba-shooting.html

They have also debunked the "clutch player" myth. With the excwption of Larry Bird.

Random coin flipper.....very coooooooollll

You flipped 16 coins of type Postumus - Bronze/Silver Antoninianus - Roman Empire - that one was seriously cool :D

If you do that enough times then there really is no need to worry about getting a formula right or not!!!!!

The randomness of the result far outweighs any logical planning for tactics and training. I feel from my match experiences that ID is much more effective blocking IS than OD is at blocking JS.

More a hunch than anything else so dont shoot me!! I'm going back to that coin flipping site.....oh and dont forget to add in another variable Ned.....

Player X is called for a shooting foul..... is this the equivalent of the coin landing on its side??!!

The randomness of the result far outweighs any logical planning for tactics and training. I feel from my match experiences that ID is much more effective blocking IS than OD is at blocking JS.

Regardless my own opinion about how defeses work in BB, I would say that this is true in real life. A good "outside" defender might be weaker against an "outside" shot than a good "inside"defender against an "inside" shot.

This might be also part of the GE.

Last edited by Zero, the Magi. at 9/21/2009 12:22:09 AM

I don't know if this is what the original poster was referring to but I have been having discussions with some pretty experienced managers about the impact of JR. Basically, the theory is that the max jump range you should have on a SG is 10-11. After that, you're better off training JS.

So why have a skill whose usefulness caps out at a certain level?

Well, the simple answer is that I don't believe the theory is right.

In that case, the majority of the user base, including the current world champion, are assembling their teams based on a flawed theory (yes, I did a little non-randomized poll - perhaps it wasn't representative of the BB population and wouldn't be publishable in a peer reviewed journal but I'm going with it).

I don't know what their evidence is and frankly I didn't even know this theory was out there until my country's NT manager told me about it. But these are managers with lots of wins under their belts. I can't believe that they would all be wrong.