Certain bad vs. uncertain good: the Steve Mason story

Jonathan Willis
April 08 2013 05:29PM

Photo: PicktownGolfer/Wikimedia

Steve Mason signed a $1.5 million contract with the Philadelphia Flyers for the 2013-14 season on Monday. It wasn’t a surprising development after the Flyers were willing to spend a draft pick acquiring him in trade, but it remains a poor decision for a hockey club with a long history of making poor decisions in net.

It’s bad enough that there’s a pretty good argument to be made that the Flyers would have been better off promoting a good goalie out of the AHL or the top European leagues instead of signing Mason.

The argument is essentially this: Steve Mason is a known quantity at the NHL level, and based on five years of data he’s a very poor goaltender by the standards of the world’s best hockey league. A good goaltender playing in the AHL or Europe is a known quantity in those leagues, but an unknown at the NHL level; he might be bad but even if he is the team using him is no worse off than if they had used Mason, and there’s a chance that he isn’t bad.

For that argument to be true, Mason needs to be a well-below average NHL backup. Is he?

Mason and the three tiers of NHL goaltenders

Steve Mason played his first NHL season in 2008-09. From 2008-09 on, I’ve gone through the performances of all NHL goaltenders and divided them into three groups in each season: starters, backups, and reserve goaltenders. “Starters” in each season were defined as the 30 goalies with the most minutes played, “backups” as the next 30 by minutes played, and “reserve” as everything left over.

Here’s how each group fared by total save percentage:

  • Starters: 0.914 save percentage
  • Backups: 0.907 save percentage
  • Reserve: 0.899 save percentage

In other words, the average starting goaltender over the last four and a half seasons (data was current as of today for 2012-13) has managed a 0.914 save percentage, the average backup a 0.907 save percentage, and the average third-stringer a 0.899 save percentage.

The interesting thing is that save percentages on the whole appear to be rising (this is consistent with previous findings). Leaving out the abbreviated 2013 campaign, which suffers from sample size problems, and allowing for the fact that the “reserve” goaltenders also have sample size issues, this is what a chart of these players’ performance looks like:

In every case, the 2011-12 numbers are better than the 2008-09 numbers.

What happens when we plot Steve Mason’s numbers on that chart?

Mason has rebounded to the 0.901 save percentage level in 2012-13, and the best way to judge him is probably his career number – he currently holds a 0.903 save percentage over 233 games at the NHL level. His career numbers are well below “average backup” level, so it isn't a surprise that he ranks 53rd in total save percentage among NHL goalies since 2008-09. His career numbers are better than the numbers of reserve goaltenders over the same span, but if the trend of the last few years - which saw reserve goalies post a 0.905 save percentage between 2009-12 (and a 0.903 save percentage including this year's numbers) - holds up then Mason belongs firmly in the middle of the "reserve" class.

The implications of that should be obvious. The "reserve" goalies are likely to be cheaper than Mason's $1.5 million contract, they're likely to put in a comparable performance, and there's always a chance that the team signing one lucks into a useful NHL player who has been buried. That's how Minnesota found Dwayne Roloson and Niklas Backstrom, how Boston found Tim Thomas, how Nashville found Chris Mason. It's better to go with the cheap and uncertain good than the (relatively) expensive and proven bad. 

Recently around the Nation Network

74b7cedc5d8bfbe88cf071309e98d2c3
Jonathan Willis is Managing Editor of the Nation Network. He also currently writes for the Edmonton Journal's Cult of Hockey, Grantland, and Hockey Prospectus. His work has appeared at theScore, ESPN and Puck Daddy. He was previously founder and managing editor of Copper & Blue. Contact him at jonathan (dot) willis (at) live (dot) ca.
Avatar
#1 Robert Vollman
April 09 2013, 02:11PM
Trash it!
1
trashes
+1
0
props

I see where you're going but I have two points.

1. There's some bias in your division of starter, back-up and reserve since one's save percentage has considerable influence over future minutes played.

2. Based on a .9143 even-strength save percentage and 2763 shots against (from 2008-09 to 2011-12), we can only say with 95% confidence that Mason's "true" save percentage is between .9065 and .9221. It is possible, therefore, that Mason actually is a good goalie.

The trick is to figure out whether it's more likely Mason is a good goalie, or this AHL/KHL/SEL/whatever goalie is a good goalie.

Avatar
#2 Baalzamon
April 09 2013, 10:40AM
Trash it!
0
trashes
+1
0
props

"That's how Minnesota found Dwayne Roloson and Niklas Backstrom, how Boston found Tim Thomas, how Nashville found Chris Mason"

How Calgary found Miikka Kiprusoff.

Avatar
#3 Ty
April 09 2013, 04:01PM
Trash it!
0
trashes
+1
0
props

Are these percentages based on 5-on-5 play?

Avatar
#4 NateBaldwin
April 09 2013, 04:18PM
Trash it!
0
trashes
+1
0
props

@Robert Vollman

This fits my intuition that part of what's plagued mason's counting numbers is the fact that he played on arguably one of the worst teams in the NHL (or at least the western conference) for the majority of his career.

Avatar
#7 ehren
April 10 2013, 11:09AM
Trash it!
0
trashes
+1
0
props

Perhaps there has been no proof that shot quality differs all that much, but it does intuitively follow, does it not? A team prone to defensive breakdowns leading to odd man rushes and/or a persistent inability to clean rebounds or limit shooters' time and space deep in the offensive zone is going to allow a higher number of higher quality chances than a defensively responsible team. I would suggest, in fact, that allowing quality opportunities is a more common symptom of defensive shortcomings than simply allowing a larger volume of shots. I have this season's Flyers in mind. By all measures, they have struggled defensively...by all measures other than shots allowed. As of today, they have allowed the 7th FEWEST shots against in the league...but the shots they do give up tend to be in prime scoring areas and in situations highly favorable to the opposing offense.

I do not have proof of this, per se. I would be shocked to find that the quality of chances a given goalie faces is in no way related to the quality of the defense in front of him, however.

Avatar
#9 EV Sv%?
April 11 2013, 10:08PM
Trash it!
0
trashes
+1
0
props

So you're #'s are based on overall SV%? What about EV SV% - that' probably a better indicator. Though I'm pretty sure both will tell us exactly what you've already told us - Mason is not good.

Comments are closed for this article.