### What can Batting Averages tell us?

It's the bottom of the 9th, 2 outs, bases loaded in the 7th game of the World Series. On the mound is the opposing team's left-handed pitcher trying to close out the game. As the Head Coach you have a decision to make: let your left-handed 9th batter hitting .270 for the season go up and take his hacks, or pinch hit with your young, recently called-up rookie batting .350?

The first question we need to answer before making a decision is: What do the batting average numbers mean?

Batting averages are a simple decimal that approximates the number of hits per at-bat, or more simply the probability that a batter reached first base on a hit during his previous at-bats. The equation used to calculate batting average is simple: # Hits/# At-Bats.

A batting average is written in decimal form using 3 digits after the decimal point. Avid baseball readers read these as large numbers, so .400 would be read as "four-hundred" and .283 would be read as "two eighty-three." Each individual thousandth is called a "point," so .400 would be considered 117 points higher than .283.

But not all batting averages can be read equally. Two players can have the same batting average, take .300 for example, and have very different statistics. Player 1 could have 3 hits in 10 at-bats while player 2 may have 120 hits in 400 at-bats.

So which is a more accurate description of a player's ability? Let's take a look at what happens to the players after their next at-bat.

If they were to both get a hit in the next at-bat, their averages would indicate that Player 1 is much more likely to get a hit, yet if they both made an out the numbers would swing heavily in favor of Player 2.

The key to this discrepancy lies in the number of total at-bats. With more at-bats, the denominator for the fraction becomes larger and is less affected by adding 0 or 1 to the numerator. Referring to the chart, the next at-bat for Player 1 will either increase his average by 64 points or decrease it by 27. Player 2 will see either a 2 point increase or a 1 point decrease. So batting averages are less affected with larger numbers of at-bats, and can more accurately describe a hitter's tendency over a period of time.

Now, looking back to the original question, I will add more context to the problem. In an average 162-game season a player might amass about 450 at-bats, and back-ups could see 100 at-bats. Rookies and recent call-ups (players invited to the major-league team from the minor leagues) will usually be on the team for the final 50 games of the season.

Knowing this information and having seen the chart from above, does this change your original decision for what to do? Why or why not? There is no definitive correct answer to this question, but I do ask that you use numbers to support your reasoning. Please post your decisions in the comments.