Skill Level Appropriate for all
- [Narrator] Sports analysts, sports fans, sports executives, and coaches. All of them, in our modern age, are dependent on statistics. It's how they make, criticize, and defend decisions. Statistics are the fuel for sports arguments, and, of course, sports statistics are vital for making predictions. But not all sports statistics are created equal. Some statistics are more likely to be the result of luck.
So just like flipping heads with a coin four times in a row may be a lucky anomaly, for certain, luck-based sports stats, a player or team streak of good statistics is unlikely to continue. If you are relying on statistics to make decisions or predict outcomes, we need to consider sample size. In many professional sports leagues, teams have dozens of games in a season. For example, 82 games in both the regular NBA and NHL season or 162 games in a regular major league baseball season.
But in American football, the NFL, each of the 32 teams plays only 16 games. Some opponents they'll play twice, but many of the teams they'll never play. With such a small and unbalanced sample size, many football statistics will be fairly unreliable in making certain types of predictions. The number of opportunities each team gets to score is another form of sample size. Once again, we turn to American football.
Each team, on average, gets about 13 possessions in a game. When we compare basketball to football, we can see that basketball provides a much bigger sample size for both games and possessions. It's no wonder that NBA games are generally much easier to predict than NFL games. Next, let's consider issues that increase randomness. These are things we don't think about but often have an effect on the outcome.
For example, speed of projectiles. Hitting a baseball thrown at over 90 miles per hour is extremely difficult. When the ball is hit, the hitter has very limited control over where the ball will land. Whether a ball lands for a base hit or is caught by an opposing defender might be related to luck versus hitter skill. In hockey, the puck can travel at nearly 100 miles per hour. This happens in a small playing area.
Thus, many players can barely see the puck move plus wherever the puck goes, it will likely bounce in a fairly random direction. Penalty kicks. The ball is kicked towards a keeper only 12 yards away at speeds up to 80 miles per hour. Keepers often rely on luck and shooter error to increase their chances of preventing a goal. Next, let's consider how shapes, surfaces, and environmental elements might impact the statistical outcomes.
Basketball is played indoors on a flat court with a round ball. Hockey is played at high speeds with a thick, rubber disk on a chipped, icy surface. And American football is often played outdoors during fall and winter months in rain and snow on uneven grass surfaces with an oblong ball that bounces around randomly. As you consider statistics like goals scored, football fumbles recovered, baseball hitting average, or any other number of statistics, consider the conditions that contributed to the outcome.
Is a team that has recovered a high percentage of fumbled footballs good or lucky? Is a player that has scored 60% of their goals on penalty kicks truly a great player? Since pucks travel so fast, does shot selection matter in hockey? Is a baseball player with a low batting average bad or perhaps are the balls just landing in the wrong spot right now? Calculating sports stats isn't too difficult.
Figuring out which stats provide meaningful information requires a foundation in statistics and an understanding of how those statistics were generated.