Being a top goalie in the NHL takes more than quick reflexes and nerves of steel, it also requires a firm grip on the numbers. Namely, the key averages and statistics of goaltending. "Science of NHL Hockey" is a 10-part video series produced in partnership with the National Science Foundation and the National Hockey League.

Science of NHL Hockey - Statistics and Averages

LESTER HOLT, reporting:

Being a top goalie in the National Hockey League takes more than quick reflexes and nerves of steel. It also requires a firm grip on the numbers, namely, the key statistics of goal-tending.

JAROSLAV HALAK (Goaltender, St. Louis Blues): More shots you have won on you and more shots you stop, the better chance you have a bigger save percentage.

HOLT: For Jaroslav Halak of the St. Louis Blues and Pekka Rinne of the Nashville Predators, stats are one of the ways they prepare for the competition.

PEKKA RINNE (Goaltender, Nashville Predators): Before the games, I usually try to learn the other team's stats and, you know, how their players are doing and who's their goal scorers and stuff like that.

Dr. EDWARD BURGER (Williams College): A statistic really is a measure of a collection of data. So lots and lots of numbers, lots and lots of figures, lots and lots of information.

HOLT: Dr. Edward Burger is a professor in the Williams College Math Department, whic has been funded by the National Science Foundation. He notes that one of the most important statistics in hockey is a goalie's “save percentage.”

BURGER: It's just his success rate. So you take the number of saves that the goalie makes and divide it by the total number of times that someone tried to get the puck by.

HOLT: If ten shots are taken at the goalie, and nine are saved, to find the save percentage, simply take the nine saves and divide by the ten attempts. In this case, the save percentage is .90, or ninety percent.

HALAK: Over ninety percent is good and whatever is above like ninety-two, ninety-three, it's just really good in NHL.

HOLT: During the 2010-2011 season, Pekka Rinne's save percentage was an astounding .930, or ninety-three percent, meaning for every one hundred shots taken on goal, he stopped ninety-three of them.

PEKKA RINNE (GOALIE): How many saves you make and how many goals you let in, for a goalie, that's one of the most important statistics.

HOLT: To compare goalies from different teams, fans often point to another statistic, the “goals against average,” or GAA. This is the average number of goals allowed per sixty minutes played. The GAA is determined by multiplying the number of goals scored by sixty. Then dividing that number by the total minutes played.

For example, let's say a goalie allows five goals over two games. To find the GAA, take the five goals, multiply by sixty, then divide the product by the total amount of minutes played, 120. The GAA is 2.5. In other words, this goalie has an average of 2.5 goals scored against him in a sixty-minute game. Unlike save percentage, the lower a goalie's GAA is, the better.

RINNE: Save percentage and goals against are really important for goalies individually.

HOLT: Many of the stats in goaltending involve estimates of the middle of a data set. This can be calculated by using three different methods: the mean, median, and mode. The mean is the computed middle value between a range of numbers.

BURGER: You take up all the numeric data you have and you add them up and divide by the total number of data points that you have.

HOLT: For example, let's take the GAA's of the top five GAA leaders in the NHL from the 2010-2011 season. To find the mean, add up the five GAA's, then divide by five, the number of values in the set. In this case, a goalie needs a GAA of 2.14 to be considered elite.

In statistics, the median is used to describe the middle of a set of numbers that is arranged numerically. It is the dividing line between the top half and the bottom half.

BURGER: You just numerically order them from the smallest to the largest. And the middle one is actually going to give you the median.

HOLT: If there are an even set of data values, then to find the median in this case, take the mean of the middle two numbers.

BURGER: So you take the mean of those two numbers, you add those two numbers up and divide by two, and that's how you get the median.

HOLT: The median is useful because it gives us another way of understanding how the data is distributed, especially if a value, such as a bad GAA, is significantly different from the rest of the data. In this case, the median is just slightly lower than the overall mean.

The mode in a list of numbers gives us a third way to understand the data. It is the most frequent value in a set of numbers. To find the mode, simply identify the number that is repeated most often. In this case, it is significantly higher than both the mean and the median.

BURGER: It's in some sense the one that wins the popularity contest. So what you would do there is you would take your data collection and you would see how many times each particular value or each particular thing occurs in your data set and the one that occurs the most often is the mode.

HOLT: While mean, medium and mode are important to helping create our statistics, to goalies like Renne and Halak, there's really only one stat that matters: wins.