Introduction
Our Strength of Schedule pages show, for every team, how many games are left, the home/away/ neutral split, how many remaining opponents currently rate above or below average, and one summary number: "Expected Wins at 1500." This article explains what that number actually measures and why we built it the way we did.
The problem Strength of Schedule is trying to solve is simple to state and easy to get wrong. You want to know whether a team's remaining games are hard or easy. But if you just look at, say, a team's projected win total for the rest of the season, you're measuring two things at once tangled together: how good that team is, and how hard their remaining opponents are. A great team can put up a good number against a brutal schedule; a mediocre team can put up the same number against a soft one. That comparison tells you nothing about the schedule itself.
The Core Idea: Give Every Team the Same Rating
To isolate schedule difficulty from team quality, we remove team quality from the calculation entirely. For every remaining game, we replace the team's own real rating with a flat, perfectly average rating of exactly 1500, the same number for every team in the league. Then we ask: how would a team with no strengths and no weaknesses of its own, exactly average in every respect, fare against this specific remaining schedule?
"Every team gets the same hypothetical opponent: themselves, if they were exactly average."
Because every team is measured against the identical constant, the only things left that can make one team's number higher or lower than another's are the two ingredients that actually define "hard schedule": who's left to play, and where those games are played. That's what makes the result comparable across teams. A last-place team and a first-place team with the identical remaining schedule get the identical Strength of Schedule number, which is exactly the point.
Two Real Ingredients We Keep: Opponent Strength and Location
The team's own rating is fictional for this calculation, but nothing else is. Each remaining opponent's rating is their real, current SharpModels rating, the same one used everywhere else on the site, updated as results come in. And each game's location is real too: home games add the standard home-field boost to the hypothetical average team, away games subtract it, and true neutral-site games, like international series games, get no boost either way, exactly as described in how the NFL season simulator works.
That matters because schedule difficulty isn't just about who you play, it's about where. Nine home games against good teams is a different proposition than nine road games against the same good teams. Folding location into the actual probability calculation, rather than just listing it as a separate fact next to the difficulty number, is what lets the two effects combine properly instead of being left for the reader to somehow weigh up themselves.
From Win Probability to Expected Wins
For each remaining game, we take the hypothetical average team's rating (1500, adjusted for home field where it genuinely applies) and the real opponent's rating, and run them through the same win-probability formula used across the whole site: the bigger the rating gap, the more lopsided the probability, but it never reaches certainty. That gives a single number for that one game, such as "an average team would win 62% of the time in this matchup, at this venue."
Add up that per-game probability across every remaining game on the schedule, and you get Expected Wins at 1500: literally, how many of these specific remaining games an exactly average team would be expected to win. A team whose remaining schedule shows 9.2 expected wins at 1500 has a meaningfully easier run-in than one showing 7.8, independent of how good either real team actually is.
Reading the Other Columns
The rest of the table is the same idea broken into its raw ingredients, rather than compressed into one number. Games Left, Home, Away and Neutral show exactly what's left and where it's played. Opponents >1500 and Opponents <1500 count how many remaining opponents currently rate above or below the league-average mark. Average Opponent Rating is the plain mean of every remaining opponent's current rating, the classic, simplest version of a "strength of schedule" number. Expected Wins at 1500 is the one that folds all of it, including venue, into a single comparable figure.
Why Not Just Use Opponents' Combined Record?
A common shorthand elsewhere in sports media is to describe a schedule's difficulty by adding up opponents' win-loss records, often from the previous season. It's an intuitive number, and it doesn't require trusting any particular rating model, but it has real weaknesses we designed around.
Last season's record is stale by the time it's being used to judge a schedule played months or a year later. Rosters turn over, coaches change, young players develop, and teams that were bad can be genuinely good by the time you actually play them, and vice versa. Our version uses each opponent's current rating, continuously updated as this season's results arrive, rather than a snapshot frozen months in the past.
A combined record also doesn't fold location into the difficulty figure itself. It might get reported alongside a note like "four of these are road games," but that's a separate fact left for the reader to weigh up, not something the difficulty number itself accounts for. And a win-loss total is a count, not a probability, so it can't be combined cleanly with anything else the way an expected-wins figure can.
"Opponents' record from last year tells you who they were. Expected Wins at 1500 tells you who they are, and where you have to beat them."
We think that trade-off is worth it, with one honest caveat. Our version depends on the current ratings being reasonably well-calibrated, and early in a season those ratings carry more uncertainty than a full prior season's record does, simply because there's less current-season evidence behind them yet. A raw win-loss method is easier to sanity-check by hand from box scores alone. We think the gain from being current and venue-aware outweighs that cost, especially once a few weeks of games have accumulated, but it's a real trade-off, not a free win.
Kept Up to Date, Every Day
Strength of Schedule is built from the same daily-refreshed ratings and schedule data as every other page on the site, so it updates automatically as games are completed and ratings move, without any separate calculation being needed to keep it current.
You can see the full league output on the NFL Strength of Schedule page and the MLB Strength of Schedule page, both split out by conference and division.