Statistics, or Sabremetrics as they are referred to here, are great in a vacuum. You take an action, monitor it, add a variable with no other outside influences, then monitor the new outcome. This gives you a hypothetical predictor model. As you increase the number of trials or data points, the predictor has less deviation, or in other words, becomes more accurate.
The difficulty with predictor models is that in live circumstances, where there are many other factors involved that are not contollable, it is extremely difficult to model accuracy. For example, this is why it is still difficult to predict the weather and climate change, even thoughthere are hundreds of millions of dollars and countless scientists and climatologists poured into this research.
In sports, Sabermatricians (meaning anyone with half a brain who is able to gather statistical data and run it through an excel spreadsheet or simple database) and those who follow them have become a strong influence that has started to change the face of sports. There is good reason for this, Statistics, when employed properly with other factors such as context, intuitiveness and resources are a great tool in decision making.
HOWEVER......statistics can also be dangerous. If they are used improperly (such as without equalizers or context) their ability to predict often becomes meaningless. In baseball, which is mainly an individual battle and individual skill sport......saber metrics can offer a fair predictor model. The most notable use was the Oakland A's. However a common misconception about the A's is that they used statistics as the only tool to help put them near the top. The fact is that they utilized statistics as a tool to spend their payroll disadvantage more wisely. Simply put, they used statistics to make themselves more efficient compared to other teams that were being wasteful on salaries.
Basketball is not baseball. Any statistician will tell you that it is a much more kinetic game with many more factors or outside influences involved. The pace is quicker, the outcomes often based on other players involvement more than baseball, defense is more of a direct factor on the outcome....etc. In a more kinetic system, the predictability of statistics is severely hampered...much more so than The more static and individual baseball dynamic, and almost worthless when compared to a vacuum system.
Let's take a look at Melo as an example.......our resident know it all claims that Melo must improve on his approximately 43% shooting percentage which makes him at or below an average shooter. The problem with this line of reasoning is that it is based in a vacuum.......meaning that all factors which may effect Melo's shooting percentage have been falsely equalized or that every situation for every player is considered the same when compared to other players percentages. If you are going to dig into stats and claim a hypothesis or outcome....then you need to dig deeper.
How many of Melo's shots were desperation? Meaning end of period, end of shot clock? These rare outliers and need to be removed.
What is the percentage breakdown of uncontested layups or dunks based on set ups from team mates....both for Melo and for the players he is compared to? These will certainly effect the outcome.
What level of defender is assigned to Melo when compared to other players?
Are the rims at MSG more or less forgiving then at other arenas and what effect does this have on shooting percentage?
What is the breakdown of shots and associated percentages for Melo vs other players?
How often does Melo throw up a ball against the rim and rely on his relatively quick second jumping ability to get a put back?
How often do Melos team mates draw the double away from Melo when compared to players on other teams?
How much are defenses of other teams devised around stopping Melo compared to other players?
How often is the help defender hidden or makes to hamper Melo's recognition when compared to other players.
Etc.......there are many more.
Upon the eyeball test, and intuitiveness, both which also can be good tools when making decisions, Melo does not appear to be a player such as MJ, Kobe or Lebron.....meaning a player who's skills transcend the game. So in that sense, it would be folly to compare or benchmark him to them.
Stats are a good tool, but using them without context or not equalizing them in a proper way is really doing a disservice to what they are.....just a tool fairly good for situational context in sports like basketball, not an answer.
Sorry for the dissertation. I'm a biochemist by education and have worked in he past as a scientist in the food and pharmaceutical industries....often employing statistics to come up with predicted outcomes for food products and drug delivery systems.
In the real world there are too many variables that go into Melo's performance to state with confidence that he is a starPhuck and will continue to be so. Bottom liners will claim he hasn't won a championship. Last I checked, there were other players and other teams on the court.
8/2/2012 5:55 AM
