If you fall into the school of thought that considers “wins” as a quarterback statistic, Lamar Jackson had an excellent rookie season. The Baltimore Ravens had a 10-6 record and won the AFC North, which is impressive for a rookie, but he didn’t exactly set the world on fire as a passer.

Although he played in all 16 regular-season games and started seven of them, Jackson only managed 1,201 total passing yards and 6 touchdowns. Many people questioned whether or not he would ever develop enough as a passer to win games with his arm, rather than his legs.

Jackson struggled according to PFN’s metrics in his rookie year

These doubts were backed up by PFN’s own metrics. His 2018 PFN Offensive Share Metric (OSM) grade was 18.85, 34th out of the 39 qualifying quarterbacks last season. The OSM measures a player’s performance based on how much they contributed to their own production. Jackson’s grade indicates that he had among the lowest impact on his passing statistics of any quarterback.

What hurt Jackson’s grade the most was his completion percentage. He only completed 58.2% of his passes during his rookie season, the fourth-worst rate of any quarterback in 2018. Put bluntly, that is not very good. But it also doesn’t quite paint the full picture. The NFL tracks, based on several factors, the expected completion percentage of every pass a quarterback attempts.

Jackson’s expected completion percentage was 63.5%, meaning that he completed 5.3% fewer passes than he should have, the fourth-worst differential in the NFL. He missed on a number of throws that he should have made, a reflection of the accuracy issues that he has struggled with throughout his career. Considering how strong a trend this was and has been, it wouldn’t be unreasonable to assume that it would carry over into the current season.

Week 1 appeared to show dramatic improvement for the sophomore quarterback

Early on in 2019, Jackson appears to have shown dramatic improvement, exceeding the expectations of even his more ardent supporters. Through two games, he has the second-highest OSM grade of any quarterback at 40.32. In Week 1, he torched the hapless Miami Dolphins, throwing 5 touchdowns on his way to a perfect passer rating. He gained those statistics while only rushing the ball three times for six yards. As mentioned in last week’s article, his OSM grade of 58.62 was the highest received by any quarterback in Week 1 by far. And it was significantly higher than his grade from 2018 as well (39.77 points higher, for those too lazy to do the math).

Easily the biggest reason for this improvement was the change in his expected completion percentage differential. Against Miami, Jackson’s completion percentage was 24.8% higher than expected, the best mark in the NFL. Based purely on that number, Jackson appears to have transformed from a player who missed throws on a regular basis to one of the most accurate passers in the NFL over the course of a single offseason.

That level of improvement, if Jackson can maintain it, would catapult him into the status of an elite quarterback. He would join players like Patrick Mahomes and Carson Wentz as one of the best young quarterbacks in the NFL. However, we have already seen evidence which suggests that Jackson’s Week 1 performance will not be easy for him to replicate.

Jackson regressed back to the mean in Week 2

This past week against the Arizona Cardinals, Jackson showed that he remains perfectly capable of taking over games on the ground by rushing for 120 yards on 16 attempts. But as a passer, he regressed significantly, looking more like the player we saw last year than the one we saw in Week 1. His OSM grade dropped to a 22.02, 20th in the NFL, and his expected completion percentage differential returned to approximately where it was last season at -4.2%, a change of 20.6%.

He still played better than he did as a rookie; he completed 6.4% more of his passes, and his intended and completed air yards were both slightly higher as well. Even so, his statistics still represented a significant drop-off compared to what he did in the opening week.

All of these metrics imply that we probably shouldn’t expect Jackson to repeat his Week 1 performance with any degree of regularity. Maybe it was a fluke, or perhaps his statistics were inflated by the fact that he was playing a historically bad Dolphins team. Either way, it is unlikely to happen again anytime soon. To be fair, we only have two weeks of data so far. You expect to see some level of variation from players from week to week.

Theoretically, it is possible that Week 2 was actually the outlier, and that Jackson will return to his Week 1 form next week against the Kansas City Chiefs. But given his history, it seems much more likely that his performance against the Dolphins was the exception, rather than the rule. Of course, it might also be the case that his actual level is somewhere in between the two. Once again, even in Week 2, he played better than he did last season. Perhaps he will continue to improve. Or maybe he is still the same, inconsistent passer we all saw last year. We’ll just have to wait and see.

Lucas Ellinas is a writer for the Pro Football Network covering the PFN Data Lab. You can follow him @Lucas_Ellinas on Twitter.