Open a row for every game that position has played against them, with the stat line and what it scored.
| # | Defense | Pts allowed/g | Yds/opp | This week | Player | Proj pts | Total |
|---|
DvP is at its most useful on the players you are genuinely undecided about. It will not talk you off a first-round back, and it should not: volume beats matchup almost every week at the top of a roster. Where it earns its keep is the flex spot, the streaming tight end and the second quarterback in a superflex, where two options are close enough that the defense across from them is the tiebreaker.
Read it with the game total next to it. A defense can be last in the league against receivers and still be a poor place to start one, because the game is a 37-point rock fight nobody is throwing in. That is why the total sits in the same row rather than on another page.
DvP is short for defense vs position, and it answers a narrower question than a normal defensive ranking. A team can be stingy overall and still be the place receivers go to eat, usually because it is built to stop the run or because its slot corner is the weak link. Total yards allowed hides that. DvP splits it out, so you can see that a defense is fine against backs and a problem against tight ends, which is the version of the information you can actually act on when you set a lineup.
Switch the table to TE and the top of the list is your answer under your own scoring.
Tight end is where this matters most, and it is not close. The position is shallow enough that most teams are starting somebody they do not really trust, so a matchup swing of five points a game is the difference between a startable week and a punt. Quarterback is the opposite case: the spread between the softest and toughest defense is real, but every starting quarterback plays every week, so you are rarely choosing between two of them.
There is no single answer, which is the point of the page. Sort each position and you will find the same handful of names near the top of two or three lists and nowhere on the others. A defense that is last against receivers and ninth against backs is not a bad defense, it is a specific one, and knowing which kind it is tells you who to start against it.
They work in opposite directions, which trips people up. Kicker points are charged to the defense: a unit that holds drives to field goals is conceding kicker points by doing its job, so a good red-zone defense can be a great kicker matchup.
Team defense is the reverse. Fantasy points for a DST come from sacks, turnovers and keeping the other side off the scoreboard, so the number measures how generous a team's offense is to the defense lined up across from it. When you are streaming a defense, you are really picking an offense to attack. The team named in that row is still the team, but the side of the ball has flipped.
Usually not. Volume beats matchup almost every week at the top of a roster, and a back getting twenty touches against the league's best run defense will normally outscore a committee back in a soft spot. Nobody has ever regretted starting their first-round pick.
Where DvP earns its keep is the flex, the streaming tight end and the second quarterback in a superflex, where two options are close enough that the defense across from them is the tiebreaker. If you are within a point or two, take the matchup. If you are not, start the better player.
Because three games rank the offenses a defense happened to draw more than they rank the defense. A team that has played two backup quarterbacks looks elite in September and average by Halloween, and nothing about the defense changed.
The blend weights last season in on a fitted curve that puts % on this season right now and rises every week until the current year carries it alone. This season on its own is still the default, because it is what has actually happened, and some teams genuinely are different from last year. Use the toggle: if the two views disagree sharply about a defense in September, that disagreement is itself worth knowing.
Yes. The matchups, totals and player projections rebuild with the betting lines every morning, and the points-allowed numbers update as soon as the week's games are in the books. The date under the headline is the last rebuild.
Per game the defense played, not per game in which the position happened to record a stat. Those differ exactly when a defense shuts a position out completely: no stat line gets written, and dividing by the games with a row would quietly reward the shutout by dropping the game entirely.
Scored under your settings, not ours. Every figure is stored as components, so changing the reception point or the passing touchdown re-ranks the table rather than relabelling it. A defense that is eighth toughest on receivers in full PPR is not necessarily eighth in standard.
The blend. This season on its own is the default because it is what has actually happened. But in September three games rank the offenses a defense happened to draw more than they rank the defense, so the blend weights last season in on a fitted curve. It currently puts % on this season and rises every week until the current year carries it alone.
Kickers are charged to the defense, which concedes the field goals. Team defense is the other way round: fantasy points for a DST come from sacks, turnovers and keeping the other side off the board, so that row measures how generous a team's offense is to the defense it faces. The team named is still the team, but the side of the ball is reversed.