On Pace Fallacy
The on-pace fallacy treats an early performance rate as a reliable forecast of a full period. It overlooks the selection of unusually strong starts, random variation, and other reasons the rate may change.
THE IDEA
On Pace Fallacy
The "on pace" fallacy is a common error in forecasting that assumes a linear continuation of early results. It is especially prevalent in sports, where a player's hot start is extrapolated over an entire season, leading to eye-popping but unrealistic projections. The fallacy arises because it ignores regression to the mean: early leaders are likely to have benefited from luck, and their performance will tend to drop toward their true ability over time.
Example: Matt Kemp in 2012
Matt Kemp hit nine home runs in the first 17 games of the 2012 season. A naive linear projection would predict:
H = G × (9 / 17)
With 162 games, this gives about 86 home runs. That is the result of continuing the early rate, not an estimate that accounts for uncertainty. Kemp finished with 23 home runs after missing time due to injury.
Evidence from Baseball History
Analysis of first-half home run leaders in the American League from 1976 to 2000 shows that, on average, they hit only 60% as many home runs in the second half as in their league-leading first half. Selecting players for unusually strong first halves creates an expectation of regression toward their underlying performance. The comparison alone does not rule out fatigue, injury, or changes in playing time.
The Home Run Derby "Curse"
Participants in the Home Run Derby are selected because they had a great first half. A later decline can reflect that selection effect without requiring a curse or an effect of the contest. Other causes would need separate evidence.
The fallacy is not limited to baseball; it applies to any situation where early results are used to predict long-term outcomes without accounting for randomness and reversion to average.
When to use it
Projecting quarterly sales based on first-month performance
Evaluating student test scores from the first few exams
Assessing a startup's growth trajectory after an early funding round
Predicting a political candidate's election chances from early polls
FROM THEORY TO EVERYDAY LIFE
In practice
Matt Kemp's 2012 Season
In April 2012, Matt Kemp’s early rate projected to 86 home runs after he hit 9 in the first 17 games. He finished with 23 after missing time due to injury. The contrast illustrates the limits of extending an early rate; it does not isolate how much of the difference was due to regression.
First-Half Home Run Leaders
Analysis of AL home run leaders from 1976 to 2000 shows they averaged only 60% of their first-half home runs in the second half, consistent with regression to the mean.
A MOMENT TO REFLECT
Put it to work.
Identify a time when you assumed a trend would continue linearly. What factors caused it to revert?
How can you distinguish between a true breakout and a lucky streak in your own performance data?
When evaluating early sales figures for a new product, what statistical adjustments should you make to avoid the on pace fallacy?
References
- Natural Inheritance — Francis Galton (1889)
CONTINUE THE CONVERSATION
Source texts
- How Not to Be Wrong The Power of Mathematical Thinking (Ellenberg, Jordan)