Aaron Dell or Martin Jones? Sharks need to defend 'a little tighter' for both


After the Sharks' second straight loss, a 3-2 defeat at the hands of the New Jersey Devils on Sunday, head coach Peter DeBoer summed up his team's problems succinctly. He told reporters that San Jose simply isn't scoring enough.

But he didn't stop there. 

“And we’ve got to find a way to get another save a night,” DeBoer continued. “Part of that’s defending a little tighter, and part of that’s [just] getting another save.”

Goaltender Martin Jones didn’t make that additional save in Sunday’s loss -- or in two other losses this season -- but still had his best performance of the year. He stopped 33 of the 36 shots he faced, including 16 of 18 in the third period as the Devils took control of the game.

Backup Aaron Dell has been the better shot-stopper so far, posting a .917 save percentage in all situations, and a sterling .957 during five-on-five play, compared to Jones’ respective marks of .880 and .881.

That’s a legitimate disparity, but the Sharks are asking a lot of their goaltenders -- no matter who’s in net. Through six games, San Jose is controlling play as well as anyone, sitting at or near the top of the NHL in the major, five-on-five possession metrics. Yet, they’ve still given up a lot of quality chances. 

Below, you’ll see a graph that Sean Tierney, a writer for Hockey-Graphs and The Athletic, created. It breaks Corsica Hockey founder Emmanuel Perry’s “K rating” -- a metric used to evaluate team performance -- down to offensive and defensive numbers, and sorts the league based on each. The Sharks ended their six-game road trip as the league’s best team offensively by this measure, but in the bottom-third of the league defensively.


San Jose has limited shots well overall, but not quality shots, and the latter element is largely driving their place on the above graph defensively. That’s also reflected in Jones and Dell’s expected save percentages.

Expected save percentage (xSV%) works just like save percentage, except it uses Corsica Hockey’s expected goals against in place of goals against, and determines what a league-average . Jones (.891) and Dell (.881) are 43rd and 46th by the measure across all situations, respectively, out of the 47 goalies that have played at least 60 total minutes this season.

Jones’ actual save percentage, then, is a tick below what would be expected of a goalie based on the quality of chances they’ve faced, while Dell’s is significantly higher. Dell’s stopped a lower percentage of shots than Jones on the penalty kill, but has outperformed him at even strength.

The table below shows both goalies’ performances at five-on-five, and how Jones and Dell stack up among the 46 goalies that played at least 50 five-on-five minutes entering Monday. It also shows their delta save percentage (dSV%), or the difference between each goalie’s expected and actual save percentages.

5v5 performance (min. 50 minutes)
  SV% (rank) xSV% dSV% 
Jones .881 (43) .915 (31) -3.46 (42)
Dell .957 (9) .899 (42) 5.74 (1)

Jones has underperformed his expected, five-on-five save percentage and Dell has overperformed his, but both goalies have faced a lot of quality shots. Among that same group of 46 goalies, the Winnipeg Jets were the only team other than the Sharks that had two goalies in the bottom third of the league by the measure on Monday.

Before San Jose can get better goaltending from Jones, the team will need to limit the quality chances he and Dell are facing in all situations. Even though both goaltenders are likely to regress to their career means, putting them in a position to do so will help. 

Then, the Sharks will be in a spot to get that extra save.