GPP Leverage: Week 2

Welcome to Week 2 GPP Leverage! Last week, we spent a lot of time asking “what if” as we considered which popular assumptions about teams going into the season might turn out to be wrong. We now have a week of actual football to work with, but that doesn’t mean we suddenly have all the answers.

We questioned whether Buffalo’s passing game could succeed against Houston’s defense, and Josh Allen went nuts. We also considered whether Cleveland’s offense could be more functional than expected, and the Browns were somehow even worse. Meanwhile, Chicago’s offense gave us reason to believe the leap we questioned might actually happen. Some of the alternatives we considered showed up, and others didn’t. That was always going to be the case. The challenge now is deciding how much weight to put into what we just watched.

That’s what Week 2 in the NFL is always about. Are the teams that overperformed really better than we thought, or did a favorable matchup and a few big plays make them look that way? Are the teams that disappointed actually in trouble, or are we putting too much stock into one bad afternoon? We should update our expectations when we learn something meaningful, but one game still leaves plenty of room for competing explanations.

From a leverage perspective, we also need to consider how our opponents are answering those questions. A disappointing offense can be an appealing bounce-back candidate without offering much leverage if everyone else expects the same rebound. Similarly, a team that played well can still be overlooked if the field dismisses that performance too quickly. This week, I want to consider those potential overreactions and underreactions through the lens of preseason playing time.

One possible explanation for an uneven start is that an offense simply hasn’t had many recent game reps together. That got me interested in whether preseason playing time had any relationship with how players and teams performed relative to our Week 1 projections. We’re dealing with an extremely small sample here — one week of one season — but I wanted to see whether there was anything worth considering as we work through the Week 2 slate.

First, I looked at players we projected for at least five DraftKings points in Week 1 and compared their actual production with those projections. As shown below, there wasn’t much separation overall between players who participated in the preseason and those who didn’t. The direction also varied by position: Quarterbacks and tight ends who played in the preseason performed better relative to projection, while running backs and wide receivers who sat out performed better. But overall, there wasn’t a consistent case for treating preseason participation as a reason to expect an individual player to finish above or below his projection.

Next, I considered team-level tendencies. Could sitting the starters throughout the preseason help explain why some offenses fell flat in Week 1? Here, the results were more interesting. Six of the nine teams that rested their starters fell short of projection. Those nine teams averaged 6.5 combined DraftKings points below projection, while the 23 teams with some starter participation averaged 2.1 points above projection.