Sports coverage is full of numbers: shooting percentages, win streaks, charts that seem to prove a team is on the rise. Many of them are accurate and useful. Others are technically true but leave out the context that would change how you read them. This checklist gives you a few questions to ask before you accept a statistic at face value. It is educational and for readers 18 and older.

1. Where does the chart's axis start?

A bar chart whose vertical axis starts at zero is straightforward: bar heights are proportional to the values. When the axis is cut off, the picture changes. Imagine two teams with win rates of 50% and 52%. If the axis starts at 48%, the first bar is 2 units tall and the second is 4, so the second looks twice as tall even though the real difference is two percentage points. Always check the numbers on the axis, not just the shapes.

2. Why this date range?

A claim like “the team is 9-1 since the trade” depends on where the clock starts. Choosing a starting point just after a bad stretch, or ending right before a slump, can make almost any trend look better or worse. Ask what the numbers look like over a full season or over several seasons, and whether the chosen window has a clear reason.

3. How big is the sample?

A player who hits 5 of 8 three-pointers is shooting 62.5%, but eight attempts say very little about skill. Small samples swing wildly. A good habit is to translate a percentage back into the number of attempts. If a headline does not mention attempts, games or minutes, be cautious. For a deeper look at why short runs mislead, see our explainer on variance and sample size.

4. What is it being compared with?

A team that scores 112 points per game sounds impressive until you learn the league average is 114. A number only means something relative to a baseline, whether that is the league average, the previous season or a similar player. Percentages need a base too: a 50% increase in a rare event may be a very small absolute change.

5. Is it a rate or a total?

Totals reward playing time. A player with the most rebounds in the league may simply play the most minutes. Rate stats, such as per 36 minutes or per possession, make comparisons fairer. Conversely, a rate based on very few minutes can be misleading. Both views are useful when used together.

6. Correlation is not causation

If teams that take more of a certain shot also win more, it does not prove the shot causes the wins. Better teams may take those shots because they are better. Look for explanations that go in the other direction, or for a third factor that drives both.

7. Beware of survivorship and selection

Success stories get attention. A list of top draft picks who became stars ignores those who did not. Ask what is missing from the data: who was left out, and would the result still hold if they were included?

8. Watch for double axes and mixed scales

Charts that plot two measures on two different vertical axes can be arranged to make them look strongly connected. The same goes for indexed series that start at different points. If a chart uses two scales, treat any visual connection as a question to test, not an answer.

A quick routine

Before you share or act on a stat, run through five questions: what is measured, over what period, from how many events, compared with what and by whom. If a claim fails two or more, wait for better evidence. Reading data this way is also useful for understanding betting prices and how they relate to probability, and it is a good reminder that no stat guarantees a result.

Frequently asked questions

Are advanced stats always better than simple ones? Not necessarily. Advanced metrics can capture more context, but they can also hide assumptions. Use them alongside basic numbers and understand what they measure.

How many games are enough to trust a number? There is no universal threshold. It depends on how noisy the stat is. Rare events and percentages based on few attempts need much larger samples.

Can charts ever be trusted? Yes. Charts with clear labels, honest axes and sources are valuable. The goal is not distrust but informed reading.

Steps

Check the axis: Look at where a chart's axis begins. A truncated axis can make a small gap look enormous.
Look at the date range: Ask why the data starts and ends where it does, and whether a different window would change the story.
Find the sample size: Note how many games, shots or attempts support the claim before trusting it.
Ask for the comparison: Make sure the number is compared with something meaningful, such as a league average.
Separate correlation from cause: Consider other explanations before accepting that one factor caused another.