Odds of Rockets' Miracle Comeback: < 1 out of 1 million!

I saw a stat that the last time an NBA team came back from double digits in the final minute to win was Dallas-Chicago 1998.

But that was an overtime game.

Does anybody here know: When was the last time an NBA team came back from double digits in the final minute to win in regulation? (I don't; I'd love to know.)

 

who takes algebra in college?

 

My dad used to teach college algebra back when he was a grad student. It's mainly for those in arts and sciences who needs to fulfill some amount math credits. It's basically the same as Algebra 2 in high school.

Yeah, I know the class is lame. Anyone who actually needs to take it most likely would never need any of the stuff taught. Just as I'll never use any of the crap in my literature classes.

And as mentioned above, the one millionth chance only applies to this particular scenario. The actual chance of a team getting back from a 10 point defecit in ~55 seconds or so should be a bit higher. Although quite impossible to calculate given the large number of factors involved.

 

I guess my comments weren't "legit", eh?

Sorry to tell you, but what you said you were trying to calculate (the "Rockets' odds to come back to win in regulation time since T-Mac missed his last layup with 52 seconds to go") has to do with what you actually did calculate (which is the likelihood of 16 specific plays, which you outline in your post, involving specific players from the Rockets and Spurs, occurring consecutively in a game between the two teams with no regard to the time left on the clock, or the score of the game).

Given that you have a stats minor, I don't understand why you're failing to grasp this, even after it's pointed out.

This is very basic probability - flipping a coin and seeing heads (or tails for that matter) has a 0.5 probability of occuring each time, right?

If you want to flip a coin 16 times and say, "I want to figure out the probability of 16 specific flips occuring consecutively", let's say:

1. Heads
2. Tails,
3. H
4. T
5. T

etc, etc.

your answer is going to be the probability of the first event occuring, multiplied by the probability of the second event occuring, multiplied by the probability of the third event occuring, etc, the result of which, in this case, is (0.5)^16, or 0.0000153, a very low probability. This is similar to what you did in your study, the results of which are meaningless to what you're intending to prove.

If you don't believe me, take any game that occurs tonight, handpick 16 consecutive plays, and use your same method to test the probability of those events occuring again in that order, and peek at the results (this may surprise you, but the likelihood of any specific 16 plays occuring in a row is going to be very low).

You could either do that, or just ask your stats professor, he'll tell you. I'm rusty myself, but I know at least the gist of what I said is right.

 

lancet,

i understand what you're trying to calculate, and i thank you for it - your post made for an interesting read and you obviously put a lot of effort into it.

again, i don't want to be difficult, but i believe the test you ran shows no significance for the reasons i outlined in the previous post. i don't consider myself to be an expert in stats, but i'm almost certain i'm right about this one.

anyways, have a good night man, i'm out.

 

Can your calculation include JVG as an intangible factor? Because I am sure he makes things even more difficult than it appears.

 

algebra in college??? gawd... you must be a product of the california school system.

 

hey lancet can you calculate the probability of the houston rockets winning the championship this yr? i think if it is 1 in a million we might have a chance

 

lancet, I'm agreeing with varunan here. I know you are just trying to get an estimate, but IMHO, this estimate has too much of an uncertainty to be considered anything useful.

I'll modify varunan's example a bit: take a coin flip 16 times, and determine the possibility that head shows up at least 8 times more than tails -- that probably resembles the rox's need to do a lot more things right than wrong to come back.

if you want to calculate the possibility of a specific sequence that would make the above requirement, e.g. HTHHHTTHHHTHHHHH, the possibility would be 1/65k

but if you actually calcuatle the possibility of 8 more heads than tails, it should be a straight forward bell curve application and probably will land somewhere like 1/100 or 1/1000. see the difference? we don't need the exact same sequence to come back and win by 1. and as it is against everyone's instincts, there are just soooooo many more ways statistically to make the come back viable than we would think.


nevertheless, I agree, another miracle are unlikely to happen again tho.

 

And the bbs finally starts to show its nerd side now where is my pocket protector...

 

Can you calculate the odds of Rockets' winning a championship? hope it doesn't require a miracle comeback in the process which will happen 1000 years later.

 

Urgh, I'm sorry, but these events are not independant. Tmac got "hot" after he hit the 4 pointer. his shooting probability should be much higher after that.