Close games and lineup calls
A bad lineup only costs you a game when the game is close enough. This page looks at every matchup of a real season to see how often that happened.
The numbers come from one real 12-team Sleeper league, half PPR, across the 14 weeks of the 2025 regular season: 84 head-to-head games. It is the same league as the sample report, with teams shown as Team 1 to Team 12. All scores are real and were worked out by the same code that builds a live report.
Most games are not close
The median winning margin was 23.9 points and the average was 28.5. Only 21 games (25%) were decided by less than 10 points, and just 9 (11%) by less than 5. In most weeks the result was settled by who had the better week overall, not by one lineup call.
That matters for how we talk about lineup mistakes. The average team left points on its bench almost every week (see start/sit mistakes by the numbers), but a mistake only changes a result when it is bigger than the margin. In a 30-point loss, no single swap was going to save you.
How many losses a better lineup would have flipped
For every loss we asked a simple question: if the losing team had started its best possible lineup, from the same roster and the same points, would it have won? In 21 of 84 games, 25% of all losses, the answer was yes. Those are the losses that were really decided on the bench.
These flipped games were lost by 10.7 points on average, while the losing team left 19.3 points on its bench, against a league average of 14.4 a week. So these were somewhat worse lineup weeks than usual, but not wildly so. What really set them apart was a margin small enough for the bench to count.
Close games are where this happens most. Of the 21 games decided by under 10 points, the loser's best lineup would have won 14 of them, 67%. Lose a close game and there is a good chance the difference was sitting on your bench.
The narrowest and the widest
The closest flipped game came in week 2: Team 4 lost to Team 10 by 0.7 points with 2.6 points on the bench. Almost any small change would have done it.
The widest was week 8, when Team 6 lost to Team 9 by 43.5 points. That looks like a blowout on the scoreboard, but Team 6 left 46.6 points on the bench, more than the whole margin. It is a good reminder that a lopsided score does not always mean a lopsided matchup. Sometimes it means the losing team's best players were watching.
What if both teams had been perfect?
There is a fair objection to flipped losses. The winner left points on their bench too. If we give the loser a perfect lineup, should we not give the winner one as well? We checked. Of the 21 flipped losses, the loser's best lineup also beat the winner's best lineup in 8. In those games the losing team had the stronger roster that week, and the result really was decided by lineup calls. In the rest, both teams made mistakes and the loser's were simply more costly.
Mistakes on the winning side are common. In 13 of the 21 close games, the winner left more points on the bench than the loser did. Winning a close game does not mean you set a good lineup; it often means your opponent set a worse one.
Team by team
Team 3 had the most flipped losses, 3. Their real record was 7-7; with those games turned around it would have been 10-4. Read that last column with care: it changes one team at a time and leaves every other result alone, so it is a measure of what each team's lineup calls cost them, not an alternative set of standings.
| Team | Record | Losses under 10 | Flipped losses | Record with flips |
|---|---|---|---|---|
| Team 3 | 7-7 | 2 | 3 | 10-4 |
| Team 11 | 6-8 | 2 | 3 | 9-5 |
| Team 7 | 5-9 | 2 | 3 | 8-6 |
| Team 6 | 3-11 | 3 | 3 | 6-8 |
| Team 4 | 12-2 | 2 | 2 | 14-0 |
| Team 1 | 6-8 | 2 | 2 | 8-6 |
| Team 12 | 11-3 | 1 | 1 | 12-2 |
| Team 2 | 9-5 | 1 | 1 | 10-4 |
| Team 10 | 8-6 | 2 | 1 | 9-5 |
| Team 8 | 6-8 | 3 | 1 | 7-7 |
| Team 5 | 2-12 | 1 | 1 | 3-11 |
| Team 9 | 9-5 | 0 | 0 | 9-5 |
Flipped losses spread across almost the whole league. Good teams and bad teams both lose games on the bench. What differs is how often a team plays a close game, and that is partly luck: see how much the schedule decides a season.
What to take from it
- Most losses are not your lineup's fault. A large share of games are decided by more than any lineup change could cover.
- Close losses are different. Most of them could have been won from the bench, so a close loss is the week worth reviewing.
- A flipped loss is hindsight. It shows what the roster could have done, not what a manager could have known before kickoff.
- Count the flipped losses over a season, not one game. A team with several is leaving wins on the table, not just points.
Every Shoulda Started report flags flipped losses, so you can see them for your own league. For how the bench tax is split by position, read which positions cause the most bench tax, and for the method behind the best lineup, lineup efficiency and bench tax. The sample report shows a flipped loss on a real receipt.