How to Use Critical Thinking with Sports Statistics

Uzu Ometu wrote an article entitled, Is Stephon Marbury Headed for a Second Straight Year in Dipped Statistics? On the whole, the article is somewhat interesting but it contains a fatal flaw. Mr. Ometu uses two common tricks with statistics. He uses multiple endpoints and non-equal comparisons to make a player, in this case Stephon Marbury, appear better than he really is.

Everyone loves to use numbers to prove their point. In the article, Mr. Ometu says, "But the numbers don't lie. Marbury is the only player since Oscar Robertson to average 20 points and 8 assists for his career."

There's a famous saying that goes, "figures don't lie, but liars can figure." Let's start by making it clear that Mr. Ometu is not lying and to the best of my knowledge he is not a liar. His statement is true. However, the comparison makes it seem like Marbury is one of the best players around, because he shares a bond with Robertson, one of the greatest players in basketball history. But are the two players really that comparable, or is this just an illusion, or as the quote above might say, a lie?

Multiple endpoints is the use of one of many valid reference points to look like a single reference point. Here, Mr. Ometu uses 20 points per game and eight assists. But there is nothing magical about those numbers. What if we change the numbers to 19 points and nine assists? Then the only two players in NBA history to accomplish the feat are Robertson and Isiah Thomas. What if we use 27 points and six assists? Then the only two players are Allen Iverson and Jerry West. We can play this game all night long.

Fans and sportswriters use the multiple endpoints game to make a case supporting their favorite player. And that argument can sound mighty impressive on the surface if you don't think too much about it. But I've seen both Marbury and Robertson play, and let me tell you they don't belong in the same sentence.