Chelsea: xStats kind to Frank Lampard this year while actual stats are not

Expected stats are a good barometer of how a team should be doing. Chelsea has good numbers but the actual stats have been less than kind.

Yes, this is an expected stats article. No, those stats ultimately do not mean much when the only stat in the game that matters is the final score. But they can offer insight for those willing to humor them.

They also make for prime fodder in arguments for or against something. For instance, when Frank Lampard was hired there was much ado about his Derby County team massively outperforming their expected stats. Of course, a swing of plus or minus one or two is not really statistically significant. Three through six implies that there is over performing going on. Anything more and there is something the xStats are not picking up which was probably the case for Lampard’s Derby County.

So given those stats, it would be expected that Lampard’s Chelsea would follow suit and would be outperforming their expectations. Except the exact opposite is happening. Chelsea’s xStats are very good this season but the reality is much bleaker. That can imply certain trends going forward.

Understat is not a perfect source but it is consistent when compared to itself and perhaps more importantly it is free. It only accounts for the leagues and the sample size is small but already telling about Lampard’s Chelsea.

Goals wise, Chelsea is at about the number they should be on. There is only a 1.41 difference between goals scored and expected goals scored so it is not particularly significant. The real issue comes on the other end of the pitch where Chelsea has conceded a whopping 4.06 more goals than they should have this season.

Overall, that does not really change the amount of points Chelsea is expected to have but it does point at some issues. More or less, it means the defense is making simple errors that are turning low probability shots into goals. And those goals, of course, are costing Chelsea points.

Theoretically, this means Chelsea will eventually revert back to the mean as the players stop making so many mistakes. But that is by no means a guarantee and that is why Lampard has already opted to change the formation to a 3-4-3 to shore things up. That worked, though not perfectly against Wolves as the Blues still conceded twice.

This is, as mentioned, new territory for Lampard as his Derby County side were seemingly worse behind the scenes than in reality. And while xStats are no guarantee of anything, it does imply that Lampard’s team has simply been more unlucky than bad and things will swing the other way in time. That does not mean stop trying to fix things, but it does mean things are not nearly as bad as they appear at times.

Of course, stats are only as good as the person telling you what they mean is. Chelsea can claim unluckiness all they like but that has to turn into real points. But the implications are that Lampard is a good if not great manager who is getting sucked into a rut by mistakes his players are making. While Lampard could wait it out, he could also opt to remove the players most guilty from the firing line.

As the season goes on, these stats will get more refined as the sample size grows. But on a game by game basis, things are looking okay. It just needs to be put together finally and then the team will start cooking.