It's hard to tell the story of the 2014-15 Arizona Coyotes and discount the poor play of Mike Smith. Yes, the offense struggles to score goals, and yes, the defense is young and inexperienced. But Smitty is now 11-32-5 with a 3.25 GAA and a .897 save percentage. The only goalies (greater than 20 appearances) with worst save percentages than Smith are Ray Emery and Viktor Fasth. That's not good company to keep.
When we look at Smith's entire NHL career, we see only two seasons with worse save percentages than this one. Brace yourself, they're pretty bad.
His worst year was the 2010-11 season, when he posted a 5v5 save percentage of .899 (interestingly, he would have a .941 save percentage in the playoffs that same season). The inconsistency is concerning, but in neither 2007-08 or 2010-11 did Mike Smith take the majority of his team's starts (he played 34 games in 07-08 and 22 games in 2010-11).
Now, let's isolate Smith's percentages during his time in Arizona (both graphs courtesy of War On Ice):
Smith's even-strength save percentage is an incredibly low .902, which is dead last in the league among players who have played 1,200 minutes or more this year. In his next worst year as a Coyote, the lockout shortened 2012-13 season, Smith's 5v5 save percentage was .924.
What does this mean? It means Smith has been awful this year, without a doubt. But he's been abnormally awful compared to his numbers as a career starter and in his time in Arizona. None of this means that we can guarantee a big rebound in the 2015-16 season; stats do not provide clairvoyance, merely probability.
But there is reason to believe that Smith can and will likely play better next year. In his past five games Smith has posted a 5v5 save percentage of .935 or better in four of them. He is clearly still capable of playing above league average, but both he and his team will need to improve their defensive play and their penalty killing if Smith's numbers are to truly rebound.