Run differential and rough win estimates
Run differential sketches a baseball season better than wins alone. A checked example shows the estimate, and its limits.
Run differential is a blunt instrument with a useful habit inside it. Add up the runs a baseball team scores, subtract the runs it allows, and you have a clue about how the season has gone that is wider than the win column. A rough formula can turn that clue into an estimated win total. The estimate is a sketch. It is not a prophecy, and it is not a wager.
This is for adults 18 and older. Baseball is the example because the sketch is famous there. Other sports have cousin formulas. None of them is a betting system. Laws differ by place.
Why wins are a thin summary
A one-run win and a ten-run win count the same in the standings. Over a long season the luck in close games can push a team's wins away from the quality of its run production. Run differential keeps the margin. It still ignores opponent quality, park effects and injuries. It is a better single number than wins alone, and it is still one number.
Sample size applies. A two-week differential is chatter. A full season is more serious and still not destiny. Win streaks can decorate a modest run total or hide a strong one.
A squared sketch, checked
A common baseball approximation says the share of games a team "should" win is roughly runs scored squared, divided by runs scored squared plus runs allowed squared. The exponent is a modeling choice. Researchers have used nearby exponents. Treat two as a classroom version, not as a law of nature.
Try 800 runs scored and 600 allowed. Squared, that is 640,000 and 360,000. The denominator is 1,000,000. The estimate is 0.64. In a 162-game season, 0.64 times 162 is 103.68, about 104 wins. If that team has 96 wins, the sketch says the record sits a bit behind the runs. If it has 110, the sketch says the record sits ahead. Neither gap has to close on your timetable.
What the gap is allowed to mean
Some of the gap is luck in one-run and extra-inning games. Some is a schedule that run totals do not fully describe. Some is a change in the roster that the season-long total has not digested. Regression to the mean is the reason people expect extremes to cool. Expectation is not a ticket. The next month can widen the gap.
Do not apply the baseball exponent to basketball points or hockey goals and pretend you were careful. The shape of scoring is different. A cousin formula needs its own sport. If you do not know it, say you do not know.
Leave the price alone
A moneyline is not required to match your sketch. Public run totals are public. The sketch can be a nice reading of the season and a bad reason to stake the next series. Close games really are noisy. That noise is not a product you need to buy.
Use the estimate to talk more humbly about "best record in the league" when the run gap points elsewhere, and more humbly about a disappointing record when the runs are healthy. Humility is the product. If the worksheet is becoming a staking habit, close it. Local help beats a finer exponent.
Frequently asked questions
Is the squared formula exact? No. It is a rough baseball approximation. Different exponents exist. Parks, schedule and luck sit outside the two run totals.
Does a team with better runs always finish with more wins? Not always. Close-game luck can separate the win column from the runs for a surprisingly long time. The sketch is a comparison, not a guarantee.
Can I bet the difference between wins and the sketch? This page does not recommend that. The gap has several causes, and the next games are not obliged to close it. A price may already reflect the same public runs.