What's new
By:

(17-10) Jazz @ (18-10) Grizzlies - Gamethread

DutchJazzer said:
thats not how it works, you're model is FLAWED!
Click to expand...

It actually a pretty good approximation. It only requires the assumes that each free throw is an independent event. How would you estimate it better without going uber-analytical and taking days of research?
 
DutchJazzer said:
because there are more fatcors tyhan just 76%.

if you miss a ft, the following ft has a different change of going in. same could be said for the 3rd one after missing the first 2. because your model lacks those stats and other real world factors your model is flawed.
like for example it being a road game. or for example. players have for example different ft% for back to back or 1 day rest or 3 days rest! or on the road ft's have different changes of going in


your statiscal value fails because it is a simplified version of very complex, real world complex thing.


EDIT: i dont blame u. I blame the educational indoctrination system. thought u that .24^3 is SCIENCE.
Click to expand...

Simplifying the real, complex world is what modeling is all about. Models that try to be highly precise when the added precision does not materially add to the insight are wasteful "academic" exercises.

Good statisticians/ engineers/ economists understand this balance.

Lets say that if you miss the previous shot, your chances of making the next shot drop by 5%. Now your answer is 0.24*0.29*0.29 ~ 2 % probability.

You could of course model the **** out of this, run Monte Carlo Simulations, etc. But I bet you dollars to donuts that you'd wind up somewhere between 0.5-5%. Adding precision adds nothing to the discussion since the ballpark estimate with reasonable assumptions is close enough.

I bet if you tested the overall probability of missing 3 in a row, across the NBA and across the last 20 years, the answer would be very close to (1-league average FT%)^3, to within +-1%.

Saying "it fails" reveals a lack of sophistication. The right way to describe your view would be to challenge the assumption that the answer presumes 3 independent events. And you would then state your hypothesis that consecutive free throws are NOT independent, and test this hypothesis with data, trying to disprove it using the scientific method. You can never prove your hypothesis to be correct, but you can state with a defined probability that your hypothesis is not untrue.
 
Last edited:
silesian said:
Simplifying the real, complex world is what modeling is all about. Models that try to be highly precise when the added precision does not materially add to the insight are wasteful "academic" exercises.

Good statisticians/ engineers/ economists understand this balance.

Lets say that if you miss the previous shot, your chances of making the next shot drop by 5%. Now your answer is 0.24*0.29*0.29 ~ 2 % probability.

You could of course model the **** out of this, run Monte Carlo Simulations, etc. But I bet you dollars to donuts that you'd wind up somewhere between 0.5-5%. Adding precision adds nothing to the discussion since the ballpark estimate with reasonable assumptions is close enough.

I bet if you tested the overall probability of missing 3 in a row, across the NBA and across the last 20 years, the answer would be very close to (1-league average FT%)^3, to within +-1%.

Saying "it fails" reveals a lack of sophistication. The right way to describe your view would be to challenge the assumption that the answer presumes 3 independent events. And you would then state your hypothesis that consecutive free throws are NOT independent, and test this hypothesis with data, trying to disprove it using the scientific method. You can never prove your hypothesis to be correct, but you can state with a defined confidence that your hypothesis is not untrue.
Click to expand...


ok first the model thing was a subtle jab at the climate people maybe out of place in the game thread. those man made climate change lovers says the science is settled etc etc. but the science is not settled because it is based on models!


concerning the free throws model.
.24^3 is a simplified accepted model.
but with research we can get a far more detailed model(will that model be perfect HELL NO) modeling is a approximation that how complex the real world phenomena is the more you could get wrong!

for example the .24 miss is overall. but throughout macks whole career he has different percentages on the road vs home. he has different percentages depending on how many days rest he had.
so looking at road percentages might be better. and even if there is somehow a way to get all stats combined and make a super accurate model.

the moment he misses 2 fts on the road with the crowd booing, depending on his state of mind/psychology it might be even more likely than normal he misses the 3rd one.
or for some other people it might be less likely.


sorry i brought this in the game thread i was being vindictive at the "settled" science crowd.
not saying my model settles the 3fts missed science. who knows

i am not a fan of this kinda science, because it is severly flawed and gets touted as ultimate truth. because hey "sciewntist" like bill nye en neill de grasse tyson stated science=truth and the masses gobbled it up
 
Saint Cy of JFC said:
Why surprised? He has pretty much the best reputation as far as mentoring players off-court.
Click to expand...

Which is perfect for the Wolves right now. It is about building the necessary habits and routines that success can be built on. They are where Utah was 3 years ago.
 
You must log in or register to reply here.

Share this page
Back
Top