Kai Havertz is much maligned among Chelsea fans, this this season proving the final straw for many of them. The German forward has played as a striker for much of the season, but had frustrated until the last few games, where some penalties and some good finishes have improved things.
In fact, if you look purely at the stats, there’s not much to complain about. FBRef have him at 14.1 expected non-penalty goals and assists for the season in a competition, miles clear of Raheem Sterling’s second plaved 9.5.
Joao Felix edges him with a slightly higher expected goal and assist rate per 90 minutes, but the Portuguese is working with a smaller sample size, and probably wouldn’t be carrying the same numbers if he had played as many games.
So while Havertz is never going to be a goal machine, there’s no denying that his recent upturn in form is forcing some of the criticism to quieten down a little.
SuperFrank, you’ve committed analysis malpractice once again [slaps forehead in bewilderment].
First, off congratulating Havertz for leading the team in npxG+xAG is like handing out a trophy for participating. He OUGHT to have the highest numbers in the squad simply as a product of having been played in front of goal the most! There’s absolutely no reason for Sterling (or any other Blue) to be near him (let alone ahead of him) when they’re playing in areas of the pitch that lead to fewer dangerous looks at goal. That’s just statistics 101.
Second, you completely miss the point in evaluating whether Havertz’s team-leading xG is evidence of him serving us well as a #9 (Hint: He’s NOT.) You totally fail to compare his xG to his ACTUAL goals, lol! Once you deduct his penalties, his actual non-penalty goals (7) are WELL BELOW his EXPECTED goals (11)—meaning he’s doing a subpar job of finishing his opportunities. By comparison, a real #9 like Harry Kane (I won’t even touch Haaland cuz the guy isn’t human) has 12.5 npxG and has ACTUALLY scored 19 non-penalty goals. Now that’s a striker! (Havertz is a 10 pressed into duty a position for which he’s not particularly well suited.)
It’s great to bring stats into the conversation, but please spend a little more time doing real analysis next time before leaping to unsupportable conclusions