As you may have heard, the Caps re-signed Logan Thompson on Monday, their now-goalie of the future (and recent past and literal present, as he’s slated to start tonight). Thompson has been staggeringly good of late, which may color some perception of the new deal, but with the Caps’ next window for contention seemingly opening earlier than expected, a rising cap and the state of the goalie market, the team opted for stability at the most important position on the ice.
Reasonable minds may differ on the value that the deal represents – a debate that won’t be settled for quite some time, but will undoubtedly be argued ad nauseum between now and then – and whether the team bet on the right horse in their two-man rotating goalie stable, but those are debates for other times and places (or the comments to this post, if the spirit moves you). But rather than looking forward, let’s take a moment to look back on an interesting observation that Micah McCurdy made in the wake of the Thompson extension:
It is striking how different the shots allowed by the Caps as a team are in Thompson's minutes versus the minutes of their other goalies, at least at 5v5.
— Micah McCurdy (@hockeyviz.com) January 27, 2025 at 2:56 PM
[image or embed]
Huh. Clear as day, in over 1,000 minutes with each netminder, the Caps’ expected goals against (xGA) rate at five-on-five is way higher (worse than League average) with Thompson in net than without L.T. between the pipes (where their defense has been quite a bit better than League average). Conveniently, “other goalies” here is only Charlie Lindgren, so let’s take a look at some possible explanations for this curious split.
1. The Timing of Lindgren’s Injury
Lindgren only missed a five games (and change) with his recent injury, but it coincided with some of the worst hockey the Caps have played at fives all season. The Caps’ xGA rate over those five games was a whopping 2.87/60, up from the team’s 2.59 up to that point, per NST. Visually, via HockeyViz, the injury came at the tail-end of a bad run of form where xGA outpaced xGF and had the Caps rockin’ the wrong kind of red:

Oh, and Thompson went 5-0-0/.969 over that stretch, with two shutouts, four total goals allowed (three at five-on-five) in 302 minutes. Not bad.
2. Score Effects
As we know, teams teams that trail in a game tend to create more offensive opportunity and teams that lead allow more offense. The Caps are not immune to that axiom (and, in fact, may be worse than they should be here, a discussion for another day):

That’s a big spike in xGA with a lead (including in a not-tiny sample at plus-three or more). And that’s particularly noteworthy for a team that has spent more time per game protecting leads than any other club on the circuit. By now you’ve probably guessed where this is going – Thompson has played a disproportionate amount of those minutes:

Lindgren, on the other hand, has played nearly three-quarters of the minutes the Caps have played at fives while trailing. (Lindgren has actually played more 5v5 minutes trailing than with a lead this season (320 to 298), which is remarkable for a team that has spent nearly a thousand more five-on-five minutes with a lead than trailing.) Point being, Thompson has spent far more time under comeback siege than Lindgren, and that has an impact on their respective xGA rates. (Sidenote: those “Down 1” numbers are a bit concerning.)
3. Splits
There are other contextual elements to consider here. For example, home/road splits – Thompson has played 62% of his total minutes at home, Lindgren has played 33% of his total minutes at Cap One, and the team’s 5v5 xGA/60 split? It’s 2.76 at home and 2.42 on the road. Thompson has also played more “second of back-to-back” games, though only one more. A deeper dive might also reveal something on the quality of the teams they’ve faced, though nothing on that front stands out on a quick glance.
4. The Team Hates (and/or Loves) Thompson
Obviously the team plays better defense in front of Lindgren because they like him more OR the team is more confident with Thompson in net, so they sacrifice defensive attention for more offense – take your pick on which narrative better fits your priors.
There’s interplay, of course, among many of the above factors. And maybe there’s an aspect of the goalies’ actual play contributing (for example, per MoneyPuck, Thompson faces more rebounds per shot against, which gooses the xGA and may be because his rebound control is inferior to Lindgren’s… or the defense is more lax with him in net because they’re more often leading, etc.). But these samples are still not huge and probably will regress towards each other over time. Best guess on what’s happened so far is that Thompson has simply been better than Lindgren and, as a result, has faced more score effect-boosted offenses. And I don’t think any Caps fans would trade “massive amounts of time playing with the lead” for a slightly lower expected goals against, even if the numbers with Thompson in net (and with a lead generally) are a bit of a yellow flag.