These days, there is more than enough chatter about Ryan Miller to go around.  Some people want him traded before this year’s deadline because they see him walking at the end of next year when he becomes a free agent, while others think that is the worst decision this team could possible make.  They feel like the team in front of him has been a flaming pile of dog-do and without him, the Sabres would be lucky to win a game.  I honestly don’t have a definitive position on the subject, although I would lean towards the argument of keeping him. 

While I do admit he lets in soft goals on occasion (then again every goalie does), I can see the argument that he has single handedly won dozens of games over the last few years where the team in front of him got outplayed significantly by the opposing team.

Having said that, I wanted to come up with a way to measure Miller’s consistency and compare it to other Sabres netminders, and other goalies around the league.  I thought about using standard deviation of save % per game.  I played with the idea of a complex stat that would take into consideration the win % of the opponent and stats of the opposing goalie.  In the end, I decided to go with a much simpler, yet I feel fairly telling stat.  I liken it to the “Quality Starts” stat in professional baseball.  Though it was developed in 1985, it has really gained steam and become more of an official stat in recent years.  For those that don’t know, a quality start in baseball is when a starting pitcher pitches at least 6 innings and gives up no more than three earned runs.  The concept is that said pitcher has given his team a chance to win the game. 

That’s what I wanted to come up with for my Goalie Quality Starts stat.  I wanted to be able to measure games that the goalie has given his team a chance to win.  Keep in mind, in baseball if a pitcher pitches 7 innings and gives up 2 earned runs it is considered a quality start, but the opposing pitcher may pitch a shutout and he loses the game.  Both pitchers could be awarded Quality Starts in the same game.  The opponent’s performance is not taken into consideration for the baseball stat, and therefore it is not in mine either.

What I decided to use as the benchmark of when a goalie gives his team a chance to win is whether or not his save % for the game is above the league average.  If he gives his team above average goaltending, his team has a chance to win.  What I have done is gone back and looked up the average save % in the NHL each season and used that as the benchmark for each game within that season.  For the current season, because the average is constantly changing, I will use the average from the prior completed season.  That will lead us to the ultimate consistency stat of percentage of games started that resulted in a Quality Start.  The higher the number, the more consistent the goalie.  I will refer to that as the Quality Start %.

This will be a fairly lengthy discussion so I will break it down into three different articles:

  1. Today we will look at Ryan Miller’s career numbers in Quality Start % and discuss the ups and downs and trends that we see.
  2. The second part will compare Miller’s Quality Start % numbers to other Sabres goalies over the years.
  3. The third and final piece will compare Miller’s Quality Start % to other goalies around the league for the 2013 season and the 2011-12 season.

So let’s start by looking at Ryan Miller’s career numbers.


League Average Save%:  This is the average save% across the league for that season.

Games Started: Total number of games that the goalie was in net to start the game. (we’ll look at some relief appearances in part two)

Quality Starts:  a Quality Start is awarded if the goaltender’s save% for that game exceeds (not equals) the league average for that year.  In other words, he has provided above average goaltending for his team in that particular game.

Quality Start %: Simply the total number of Quality Starts divided by the Games Started.

The first thing that you may notice is that the average save% in the league was highest in 2003-2004 before the lockout and then dropped significantly with the league’s adoption of stricter rules on defensive play that resulted in less clutching and grabbing.  However, since then, the average save% has crept up almost every year and in the last 2 years has actually surpassed that of 2003-2004. 

Either those rules are more lax today or teams have figured out a way to be a bit stingier on defense within the confines of the same rules.  Another thing to note about these stats is that in each of the last two seasons, a goaltender has set and all time NHL record for save% in a season.  In the 2010-11 season, Tim Thomas broke Dominik Hasek’s record from 1998-99 with a .9382 save%. 

The very next year, Brian Elliott eclipsed that with a .9403 save%.  This season, Craig Anderson is sitting on an incredible .9521 save%, which would obviously smash Elliott’s mark.  Interestingly enough though, the average save% across the league this year as of the writing of this article is .909, down from last year’s .914. 

Looking at Miller’s numbers, the obvious jumps out.  He was most consistent during the 2009-10 season, the year that he was awarded the Vezina Trophy as the league’s best goaltender.  We could have all probably guessed that one.  One that may be a surprise is that he was apparently slightly, and I mean slightly, above average in 2006-07, the year that the Sabres won the Presidents Cup for the best record in the league.  Clearly he had a team in front of him that could cover up his mistakes and put big numbers on the board themselves.  This, to me, suggests the possibility that good goaltending isn’t as crucial to winning as one might think.  When Miller was at his best, the Sabres finished with less points than when the team was at its best.

We can also see that over the past two seasons, Miller has been slightly above 50%, we’ll see in part three of this discussion where that lands him in the NHL ranks.  For now, what we notice is that he just hasn’t been as consistent as he was before that.  His Quality Start % has declined each of the last four years and is nearly 20% off its peak.  Surely the defense in front of him has something to do with that, but there’s obviously a decline in Miller’s consistency alone as well. 

So where do we go from here?  Do we trade Miller and stock up on future prospects and draft picks, knowing that goaltending would be a question every night?  Would we even be that much worse off with Enroth in net vs. Miller?  Granted, Enroth got off to a really rough start this year, but in his last two starts he has been fantastic.  We’ll look at Enroth’s Quality Start % in my next article.  Spoiler Alert: His numbers aren’t much different than Miller’s. 

Before doing this study, I would have been vehemently against the idea of trading Miller.  After looking at some of the numbers, I’m still against it, but I wouldn’t have to cry myself to sleep if it happened.  Yes, we remember Miller saving the team’s bacon on a number of occasions, but we might be remembering the old Miller, the Vezina winning Miller, not the one that has been less consistent over the last two seasons. 

I love Ryan Miller, I always will, but all good things come to an end at some point.  With the trade deadline just a few weeks away, and the team clearly not challenging for a playoff spot, we may be seeing the end of the Ryan Miller era in Buffalo.  Darcy has a tough decision to make, a decision that could change the course of this franchise completely.  Let’s hope he makes the right one.

For more Sabres stats nuggets in 140 characters or less, follow me on Twitter (@SabresStats).