Post#65 » by spolgar » Sun Oct 3, 2010 6:38 am
Koopa, I understand the allure of well published ratings. However, I have a few reasons why I don't like video game numbers for comparative purposes outside of the video game itself.
It is not where the rating is from that bothers me. Plenty of good video games come close to the reality they are suppose to purport for the sake of fantasy. It's their lack of transparency. I don't know how the hell their ratings work. I don't really enjoy the argument of ceding to authority, I enjoy the notion of comparative advantage most of the time, but that's really because I can't be bothered to learn how to fix my own plumbing or make my own car. If however, when I am interested in how something works, curiosity is not really something I can outsource very well.
I also am not in love with aggregate scores of players as well:
According to the video game, NBA2k11, the ability y of player X, say shot blocking, is first normed* to a number, and then taking a vector of abilities that are already normed*, the vector is further normed* to _one_ integer which is our rating. Even if the system was relatively fair in accounting for the aggregate ability of a player, I would think that a) it would penalize specialists, and therefore lowering the value of roleplayers despite their contributions b) it would downplay the role of chemistry c) it makes high number producers on bad teams much better than they otherwise would be.
I hope you find my explanation satisfactory as to why I disagree with your assessment about my favorite team in professional basketball right now.
As per the point where if there exists things that are mathematically impossible, I am sorry to inform that at least in the land of Mathematics, there are many, many things where the probability of an event is decidedly, zero...
Case in point. Suppose you stand in a town square blind folded. Within radius y of you, there is a point on the ground. Suppose you randomly pick a direction and point away from you. What is the probability that the line from your index finger will cross the point?
It's very much zero. Here's the reason why:
You have a point on a circle that is an arc that covers infinitely many points. Uncountably many as a matter of fact. You sit in the middle of the circle. For each line you draw from the center of the circle to the circumference itself, you hit only one point. Now take the series 1/n, n=1,2,3,4,5.... That's the probability of picking the one thing you want out of a pool of countable things if you only get one try. The more things you have to pick through, the less likely you are going to hit it randomly. Now suppose N is really really big, your probability of hitting the one thing you want tends to zero. And that's just with things that you can count out using integers as your index.
With points on an arc, the set of points becomes _uncountable_. That means for every set of points you can index using the integers, there would points between the points you index, no matter finely grained your index is. So you are trying to pick out something that is countable in a sea of uncountably many things, and basically it is impossible. In fact, if you tried infinitely many times (countable) with infinitely many arms (countable) with infinitely many fingers (countable), the probability of finding a pre-picked point on a circle is still zero.
Or for a more relevant example that just came to mind. If you play in a basketball game for however long, the probability of you showing up as a contributor in any way on a box score sheet, other than minutes played, that is accurately recorded, is zero if you managed to never touch the ball.
Oh I could've saved myself so much typing if I just wrote that. Schmuck a duck. That was a two birds with one stone reply... *sigh* There goes my short post streak...
Oh woops, I managed to not even answer the topic's question... I'm too tired to edit this post. Yes he should. Here's why:
1) We have to lock up Yao in 2011. We have no idea how much that's going to cost. All it takes is one good year from him to bump up his monetary value.
2) The cba is gonna tighten up. We might have a hard cap. Who else is expiring soon? We might need the money.
3) With the onset of one and dones, it is very difficult to draft based on need when you don't have much history to go on. The quality of statistical evaluation is correlated on the size of the sample space. Since the rookie contract has now became the talent evaluation period, might as well let other folks train and groom the youngster and yank him latter if he's a good deal.
4) With the talent level of New York right now, they have decent players at the 3,4, and a serviceable 1. I see them as the 8th to 10th team, with a high pick to be unlikely.
On a side note, if there is a player strike, how do draft picks work? Do they get rolled over to the next year? I don't know either.
*N.B. I use the word "norm" loosely here for those mathematically endowed.
**Wow... How'd I do that?**
first recorded use of kerfluffle by a non-Brit/Oklahoman!
