clock menu more-arrow no yes

Filed under:

The incredible, and ineveitable, resurgance of Drew Stafford

New, comments

Sometimes regression to the mean takes longer than expected, but huge variations in shooting percentage usually have a way of working themselves out.

Chris Trotman

When the Buffalo Sabres re-signed Drew Stafford after the 2011 season, and after a year that saw him score 31 goals in 62 games, they expected him to continue producing at a high level and fill the role of #2 forward, behind Thomas Vanek.

But we all know how that plan turned out.

Since signing that contract, Stafford's production has sharply declined. Actually, sharply might not be a strong enough word; Stafford's goal production has been cut by more than 50% - those 31 goals in 62 games Stafford scored in 2010-11 equate to a goals per game average more than twice what happened over the following two and a half years combined, where the Sabres right winger scored just 38 goals in 183 games.

However, over Buffalo's last 15 games, Stafford has 14 points, eight of them goals. That goal scoring average puts him right back on his torrid pace from the 2011 season. But we've always known Stafford to be a "streaky" player, so the eternal question is "will it last"?

We've always known Stafford to be a "streaky" player, so the eternal question is "will it last"?

To answer that, let's take a look at Stafford's shooting percentage over the course of his career. During his first five seasons in the NHL, culminating with his breakout 2011 season, Stafford took 713 shots on net, and scored 94 goals for a shooting percentage of 13.2%. For comparison's sake, the league's average shooting percentage for forwards from 2005-2013 was 10.8%, while the league's top scorers can average around 16-23%  in any given season. While Stafford's was a bit higher than the average, he had five years of production to back it up, and in his role as a scoring forward, a 13.2 shooting percentage was not unusual in the least.

Over the next two seasons, Stafford would see his shooting percentage plummet from 13.2% to a measly 7.5%, despite seeing his shot totals increase on average. You can look at this a number of ways - Stafford was unlucky, he was in a slump, he simply lost his mojo, or, he was taking part in an extended statistical anomaly.

There's a phrase that statisticians use called "regression toward the mean", which essentially states that when presented with repeated examples of predictable data, extreme points in variance will eventually start to move towards their average. Think of it this way - even if you flip 100 coins and get tails 80 times, you'll eventually start to see the average work towards 50/50 the more coins you flip.

Over the past two seasons, Drew Stafford has "flipped tails" far more often than is the norm, as his extreme drop in shooting percentage indicates. He was due, at some point, to start moving back toward his career average in shooting percentage - it just may have taken longer than Sabres fans (and certainly Stafford) would have liked.

During these past 15 games, he's begun to move in that direction, and I for one like to think it's a karmic reward for seeing Stafford play some of his best overall hockey this year. But will it last? I think it will, because despite his torrid pace recently, his increased production has only had a 0.1% change on his shooting percentage since 2011. If you believe the numbers, there should be much more to come from Stafford, and not just this season. If you'd rather ignore the stats and focus on what you see on the ice, then you've seen a Stafford who, in the second half of the season, has turned into Buffalo's best and most reliable scoring forward while playing top line minutes in all situations.

Ted Nolan was impressed recently with Stafford taking over a leadership role in the locker room for the departed Steve Ott and Ryan Miller. At only 28 years old, Stafford's growth as a hockey player certainly isn't over. Let's hope the growth in his game we've seen over the last two months isn't over either.