Strength of schedule in baseball

Strength of schedule, usually shortened to SOS, is a measure of how good a team’s opponents are. It answers a question a won-lost record can’t: was this record earned against contenders or against the bottom of the league?

It comes in two flavours, and the distinction matters. Past SOS grades the opponents a team has already faced, which tells you how much to trust its record. Remaining SOS grades the opponents still to come, which tells you what to expect. In a pennant race the second is usually the more useful number.

Baseball’s unbalanced schedule is what makes this necessary. Teams play their own division far more often than the rest of the league, so a club in a strong division genuinely faces a harder road than one in a weak division — with the same record meaning different things.

The basic method is an average of opponents’ winning percentages, weighted by how many times each is played.

If a team plays 19 games against a .600 club and 6 against a .400 club, the .600 opponent counts for far more in the average. Sum that across the whole schedule and you get a single number, usually expressed as a winning percentage around .500 — say .512 for a hard schedule and .488 for an easy one.

More careful versions add two refinements. Opponents’ opponents are folded in, because a team’s record is itself a product of who it played. And true strength estimates — from run differential or projections — replace raw records, since a rival’s record is a noisy measure of how good it really is. That second adjustment matters a lot early in a season, when records are mostly noise.

Strength of schedule is nearly irrelevant in April and can decide a division in September.

Early on, every team has most of its schedule ahead, and those schedules are more alike than different. By late August the remaining slates diverge sharply: one contender closes with three series against last-place clubs while its rival faces two division leaders and a cross-country trip. Over the final three weeks, that difference is worth real games.

This is exactly why playoff-odds models simulate the actual remaining schedule rather than assuming every game is a coin flip. A team two games back with an easy close can genuinely be favoured over the team ahead of it — and the odds will say so, even though the standings don’t.

A few honest limits.

• It’s an average, so it hides shape. Two identical SOS figures can mean “uniformly mediocre opponents” or “three brutal series and a lot of easy ones.” The distribution matters and the average erases it.
• Opponent quality moves. A team that was .400 in June may be .550 after a deadline overhaul. SOS built on season-long records lags reality.
• It ignores context — travel, rest days, doubleheaders, and who happens to be pitching. A weak opponent starting an ace is a harder game than the average implies.

Read it as one input among several, not a verdict.

TwentySeven includes schedule strength in its season view, so the standings come with the context that explains them — alongside power rankings, run differential, Pythagorean over- and under-performers, and Monte Carlo playoff odds that already fold the remaining schedule into a single number.

It also feeds the week’s must-watch games: the matchups that matter most are usually the ones where two strong teams meet with something at stake.