When the stats lie in poker Part 1

The saying that you are entitled to your own opinions but not your own facts is often backed up by statistics, but I have news for you: Statistics can, and often do, LIE! Maybe lie is too harsh a word; perhaps a better way to understand statistics is to say that they can be easily manipulated to prop up a poor argument.

There are two reasons for this:

• Statistics can be cherry-picked
• Some statistics only tell half of the story

Before I get into the poker aspects of statistics lying, let me first give a few clear examples of how statistics do not tell the entire story.

In his terrific book Moneyball, Michael Lewis details how errors in baseball are a particular statistic that only tells half of the story, since in order to be charged with an error a player has to get to the ball to begin with. For instance, one of the greatest fielding shortstops in the history of the game, Ozzie Smith, was perennially among the league-leaders in errors. But to fans, errors are the measure of a fielder’s abilities (in fact team errors are the only defensive category shown in the starting lineup on TV).

Another example of statistics not telling the whole story is in Basketball. When the topic of the best scorer in NBA history comes up people often use statistics to back up their choice:

Person 1: Kareem Abdul Jabbar is the all-time points leader.

Person 2: But Michael Jordan has a higher points per game average.

Person 3: Wilt Chamberlain scored 100 points in a single game and averaged 50 points for his career.

But what if we add a fourth person, an MIT educated economist, a person who knows nothing about basketball and has no horse in the race. A person chosen by the other three to make the determination on who is right? What statistic will he choose as the winning argument?

He will likely say they are all incorrect. He could back this up with RELEVANT statistics. This is an example of how a real statistician would look at the data so there is quite a bit of guesswork on my part in the following example since the important part is not the actual stats but how they are utilized:

Jerry West averaged 27.3 points per game before the three point line was introduced, and averaged 4 shots per game past this distance, while Michael Jordan was able to play with the three-point line AND during a time when the average number of fouls called in an NBA game had gone way up. He might then look at the number of games played, shooting percentage, touches, average team points per game during the era the person played, average points per game per player during that era, and minutes per game to come up with an algorithim determining who the best scorer actually was.

The point is, none of the stats cited by the fans can give you a clear example of the best scorer in NBA history, the statistics are cherry-picked to make their case, and they don’t tell the whole story as to what actually took place. Now let me get on to how this applies to poker…

Read Part 2 of this series HERE