Quote: Originally posted by dogget on 4/18/2010This game was a 3pt spread. For one team to be leading by 13 at the half means that the ratings/performance schedule had a significant deviation imposed upon it. (Again, not because one team got "hot", because there is no such thing in HD; and not because one team got "cold", because there's no such thing in HD.)
The ratings-yields-performance model endured an even greater, 22pt, deviation in the second half. (Not because some guys got "hot".....)
Then there was the peculiar(?) offensive rebounding to points ratio thingy. (Was that a third disparate deviation?)
So, just where was the "normal" part of that game? You know, the ratings yield performance part.
Or does HD really just stand for the proposition that two deviations make no deviation, i.e., two wrongs make a right.
But I think you're forgetting some of your Stats 101. First of all, measurement matters. Why are you picking halftime like there's something magical about it? Why not 30 minutes , or 15 minutes?
Secondly, all events are probabilistic in nature. "Real life" or not, any time a person takes a shot there is a probability of hitting and missing. If you flip a fair coin 100 times, very rarely are you going to have 50 heads. The problem with colonels is that he looks at a result of only 20 heads and says "that shouldn't happen!" when probabilistically it would happen more than he thinks.
There's the old trick stat professors pull on the first day of class. The professor divides the class into two groups. The first group of students are told to flip a coin and write down the heads and tails observed. The second group is told to create their own list of random heads and tails. The professor leaves the room and makes a wager-- "if I can't tell which one is the one invented by students, you'll all get an A for the semester."
The professor always wins his wager. The actual random series--the actual coin flips -- always has a lot more patterns in it and, say, 10 straight heads, than the student-invented list.
You guys seem to forget that a lot in playing this game. The RNG is simply following a bell curve. Programming it to do otherwise situationally would be a lot harder.