Blackwood’s Strong Play Makes the Devils’ Offseason Goalie Decision Tougher

At the time now-Predators coach John Hynes was fired from NJ, the Devils were getting the worst goaltending in the league — allowing almost a full goal per hour more (3.22) than expected (2.26) at 5v5. Schneider continued to be one of the worst goalies in the NHL, Blackwood was raising concerns that his AHL struggles were the signal and the NHL competence was the fluke, and TB cast-off Louis Domingue was showing why they were willing to give him up for almost nothing.

Shero’s decision to run with an injury-prone and declining Schneider and a recent AHL-demotee in Blackwood seemed, in retrospect, to be a massive oversight. This oversight, by the way, was more than likely a deciding factor in his own dismissal from the GM position. Building a team to compete now and going in with a goaltending situation like this is a good way to lose your job when it doesn’t work out.

But, since then, a wrench has been thrown into the works. The goaltending (Blackwood, in particular) has improved! Since Hynes’s dismissal, at 5v5, the Devils are actually allowing fewer goals per hour (2.74) than expected (2.80). This, coming from the team that had given up the most goals in the league relative to their expectation immediately before. Why is this? Nas improvements? Luck? Goalie voodoo? It’s tough to say (though I lean towards voodoo), but regardless of the reason, the Devils management enters next offseason in the same situation Shero entered last time. The question is, how should they deal with it?

Is Blackwood an NHL Starter?

I should make a blanket statement at the beginning of this section that analytics really struggles to assess goaltenders. There’s a lot of randomness in hockey already, and we don’t have where in the net-face a shot went, how fast it got there, what happened before the shot, if there were screens, etc. Blackwood has 62 games played which is one full season for an top-shelf starter — that’s not nearly enough to be confident in anything. As an example, the 5 worst goalies with 1000+ EV FA this season are Bobrovsky, Rinne, Holtby, Quick, and Fleury. Over the previous 3 seasons, none of those 5 goalies were worth less than 30 GAR (10 goals saved / season) and combined they were with 277 GAR in that time.

Allow me to rephrase. The 5 least efficient goaltenders so far this year are composed exclusively of players who were in the top half of starters over the previous 3 seasons. So, yeah, goalies are voodoo.

With that said, let’s lean in hard to the numbers to figure out exactly and unambiguously how good Blackwood is.

Interestingly, these charts from Evolving-Hockey show a little more consistency than his raw numbers would indicate. His xSv% has radically decreased since last season which means we expected more goals per shot. But his performance relative to that expectation has remained pretty constant. He’s been about a half a standard deviation better than the average goalie at 5v5. His shorthanded performance has cratered since last season, but for as little as we know about goalies in general, goalies on the PK are even noisier.

In other words, if we look to the most stable metrics available for goaltenders, Blackwood is about a half a standard deviation above the average goalie (about 70th percentile). That makes him pretty comfortably an NHL starter.

To have another datapoint, Cole Anderson is one of the go-to voices on goaltending analytics publicly. He created a tool that simulates seasons based of their year-to-date performance and variability to get a sense of how confident we can be in their performance.

Over the past two seasons, you can see that his average result is in the 15-30 range (very similar to the 70th percentile I quoted above) and, importantly, only 2 out of the 100 simulations put him outside the top 45 goalies. This means that we expect he is a back-end starter and it would be surprising if he weren’t at least a top-end backup.

One final flavor to add to this recipe is the goaltender aging curve. My recent research on NHL aging indicates that the trajectory for NHL goaltenders may be a little later than previously anticipated — my model has them peaking around 26 years old. This means that Blackwood, who is already an NHL starter, should be this good or better through his early 30s. Again, we need to observe the caveat here that goalies are difficult predict, and the error bars in their aging curve include a really wide array of possibilities. But if he has a textbook aging progression, then what I’ve said stands.

Do the Devils Still Need More Goaltending?

Yes.

Let me unambiguously say that at the beginning. The Devils should get goaltending help. Why do I think this if Blackwood is expected to be a competent if not flourishing NHL goalie? Put simply, goaltending is a franchise-deep problem. The New Jersey Devils goaltending problems are clear, but the Binghamtom Devils are bottom 5 in the AHL in goaltending, and even the Adirondack Thunder are saddled a starter and primary backup goalie that are each sub-900 Sv% performers. This is a top-to-bottom franchise issue that needs to be addressed from the outside.

I like Cory Schneider, and I’ve written enough bad words about him for one lifetime, so I’ll just assume everyone agrees he’s not relevant to this discussion and move on. Louis Domingue has probably been the most encouraging backup, but he has been below replacement-level for 3 consecutive seasons, so he’s not a viable NHL backup. And remember, Blackwood is likely a back-end NHL starter (15th-31st best) so if we want league-average goaltending, we need a top-half backup. Where can we get that?

The most desirable name you’ll see thrown around (at least among the analytics community) is Robin Lehner. In a field of study fraught with constant variation, Lehner has been remarkably consistent. His GA%- has been 95 or lower (he’s been an above average goalie) with all 4 teams he’s played for — and he’s registered .920 Sv%-seasons with every one of them as well — culminating, finally, with his Jennings-winning and Vezina finalist 2018-19 season last year with the Islanders. And he’s only 28.

It’s tough to know exactly what a goalie with his recent success would demand on the market. Raanta was an analytics darling for 4 years with 3 different teams before signing with Arizona for 3 years at 5.7% of the cap as a 29-year-old. Devan Dubnyk played with 4 different teams and was good with 3 before signing for 6.07% of the cap for 6 years as a 28-year-old. Ben Bishop got 6.74% of the cap as a 30-year-old with very good success scattered across 3 teams and 5 years that included two Vezina-finalist seasons. I think Lehner is the next rung in the ladder but before the likes of Jonathan Quick (9.67%) and Corey Crawford (9.33% , cup heroes) and well short of Carey Price (14%) and Sergei Bobrovsky (12.27% , generational). I think 7-8% of the cap (~$6M) for 3-4 years is likely what Lehner will go for. With Schneider still on the payroll, and an ELC goalie looking like a serviceable starter, Lehner may not be on the Devils wish list. Let’s look older and more temporary, then.

The best old goalie on the market is Jaroslav Halak — he’s succeeded everywhere he’s been, sometimes being among the best in the league and NEVER ... in 13 YEARS ... been below replacement-level. And yet, he’s somehow never made more than $4.5M. He could probably be had for around $4.5M AAV on a short deal that would help bridge the gap until Blackwood’s ready for the full job. That seems more reasonable, but may still be unaffordable.

If it’s not him, the goalie list become 34-year-old Thomas Greiss (success with 3/4 different teams, great last two years), 33-year-old Anton Khudobin (5/7 last seasons better than league-average), or 32-year-old Cam Talbot (5/7 good years, awful in 2018-19). I think these guys could all be had for less than $4M. After that, the list becomes people that I’m not significantly more confident in than the likes of Louis Domingue.

Concluding Thoughts

The Devils do not have a John Gibson on their hands. They don’t have someone that you can throw on the ice night-in, night-out and log 60+ GP. Of the top 5 teams in the NHL in team Save% (BOS, DAL, CBJ, ARI, NYI), ALL of them have backups that have played 20+ games already and we’re less than 60 games into the season. If the Devils want to compete anytime in the near future, they need to get Blackwood some help, and they may need to spend a little bit to help get him there. Whether it’s long-term (Lehner), short-term expensive (Halak), or short-term middling money (Talbot), the answer to the Devils goaltending woes is likely going to need to be found on either the trading block this deadline, or in the UFA pool this offseason.

What do you guys think? Is Blackwood a long-term starter? Do we need a goalie that’s a 1A a 1B, or a clear backup to Blackwood? Which name are you looking for in the UFA pool? Are there any trade targets you have?

Thanks as always for reading, and leave your thoughts in the comments section below.