Read on Twitter
Adjusting for Opportunity
When using advanced metrics, it's critical to "adjust for opportunity" in a thoughtful way that accounts for context and avoids biases.
A thread
When using advanced metrics, it's critical to "adjust for opportunity" in a thoughtful way that accounts for context and avoids biases.
A thread
1/ In the simplest of cases, adjusting for opportunity means dividing a measure of total output by a measure of the opportunity the player or team had to achieve that total output:
Opportunity Adjusted Metric = Total Output / Opportunity
Opportunity Adjusted Metric = Total Output / Opportunity
2/ "Opportunity" is often captured by playing time (e.g. games played, minutes played), but can take other forms as well.
3/ Let's look at a simple 
example:
- Isaiah has scored 30 goals
- Tobey has only scored 15 goals
Who's the better scorer?
example:- Isaiah has scored 30 goals
- Tobey has only scored 15 goals
Who's the better scorer?
4/ It’s impossible to say for sure which of these two players is a better goal scorer without some critical context: opportunity.
5/ If Isaiah & Tobey played the same # of minutes, it might be fair to conclude that Isaiah is a better scorer.
But, if Isaiah took 3x as many minutes of play to score 2x as many goals as Tobey, the answer might be different.
But, if Isaiah took 3x as many minutes of play to score 2x as many goals as Tobey, the answer might be different.
6/ How might we adjust for opportunity to ensure a fairer comparison?
One method is to look at the opportunity-adjusted Goals per Minute metric:
Goals per Minute = Total Output / Opportunity = Total Goals Scored / Games Played
One method is to look at the opportunity-adjusted Goals per Minute metric:
Goals per Minute = Total Output / Opportunity = Total Goals Scored / Games Played
7/ And in case you hadn’t connected the dots, adjusting for opportunity is synonymous with measuring efficiency.
So, when we say that Tobey scores more Goals per Minute than Isaiah, we are effectively saying that Tobey is a more efficient goal scorer than Isaiah.
So, when we say that Tobey scores more Goals per Minute than Isaiah, we are effectively saying that Tobey is a more efficient goal scorer than Isaiah.
8/ In almost all analysis situations, you'll want to make sure a metric is adjusted for opportunity.
There are, however, some important factors to consider in doing so.
One of those factors is opportunity sample size.
There are, however, some important factors to consider in doing so.
One of those factors is opportunity sample size.
9/ Larger sample sizes of opportunity are better for analysis and smaller sample sizes can be misleading.
10/ Let's say Abby is a 
player and only plays one minute in a particular game, but hits a three-pointer during that time.
Abby's Points per 36 Minutes = (3 points / 1 minute) * 36 = 108 Points per 36 Minutes!
player and only plays one minute in a particular game, but hits a three-pointer during that time.Abby's Points per 36 Minutes = (3 points / 1 minute) * 36 = 108 Points per 36 Minutes!
11/ In this case, Abby's opportunity-adjusted Points per 36 Minutes is misleading because the sample size of opportunity (minutes) is too low.
Thus, we should always be careful about using opportunity-adjusted metrics when there isn't much sample size.
Thus, we should always be careful about using opportunity-adjusted metrics when there isn't much sample size.
12/ In all cases, it's typically good practice to cite opportunity-adjusted metrics alongside the total units of opportunity.
For example:
Instead of "Gavin averaged 6 Yards per Carry" use "Gavin averaged 6 Yards per Carry over 1,000 carries"
For example:
Instead of "Gavin averaged 6 Yards per Carry" use "Gavin averaged 6 Yards per Carry over 1,000 carries"
13/ When we’re adjusting for opportunity, we also need to be cognizant of any potential biases in the measure of opportunity we’re using.
One of the most common examples of bias in a measure of opportunity surfaces when using games played as a measure of opportunity.
One of the most common examples of bias in a measure of opportunity surfaces when using games played as a measure of opportunity.
14/ Because players play different portions of games, a single game for a starter might add up to a lot more opportunity than a single game for a player coming off the bench, because the starter plays more minutes.
15/ Thus far in this thread, we’ve only looked at examples where opportunity was expressed in terms of playing time (e.g. minutes played), but that's not the only way opportunity manifests.
Opportunity can come in a number of forms.
Opportunity can come in a number of forms.
16/ For example, metrics that are expressed as percentages are often opportunity-adjusted metrics.
Free Throw % in, for example, is opportunity-adjusted:
Free Throw % = Total Output / Opportunity = Free Throws Made / Free Throw Attempts
Free Throw % in
, for example, is opportunity-adjusted:Free Throw % = Total Output / Opportunity = Free Throws Made / Free Throw Attempts
17/ Many of the advanced metrics you see today are adjusted for opportunity in some way, as they should be.
However, even if a metric is adjusted for opportunity, be aware of how the opportunity measure might be impacted by sample size or otherwise biased.
However, even if a metric is adjusted for opportunity, be aware of how the opportunity measure might be impacted by sample size or otherwise biased.
18/ For more on adjusting for opportunity, check out my longer blog post on the topic: https://brendankent.com/2021/01/14/sports-analytics-101-adjusting-for-opportunity/
