What VUKOTA projects for the 2012-13 San Jose Sharks

The impending NHL lockout, slated to begin less than a week from today should the two sides not reach a miraculous last-minute agreement, has mostly dampened enthusiasm for every hockey fan's favorite September pastime of predicting the standings and projecting player performance to brace themselves for the coming year or just get a leg up on their Fantasy Hockey opponents. Player expectations generally present more interesting questions: is this the year that young player has a breakout season? Will that veteran scorer finally start to decline? Can that player who disappointed a season ago bounce back?

In the absence of a crystal ball, the default answers to these questions are usually positive, sometimes unrealistically so, as training camp breeds optimism in nearly every NHL fanbase. But what if we had a crystal ball? Hockey Prospectus writer Tom Awad's VUKOTA system is far from that but it can still provide some valuable insight into how players might perform in the coming year. VUKOTA uses traditional and non-traditional statistics to compare active players to those from NHL history in order to predict their output in the following season. After the jump, we'll take a look at what it projects for Sharks players in 2012-13 and try to gauge just how accurate VUKOTA's predictions are by comparing its 2011-12 Sharks projections with what actually transpired last year.

First, a look at the forwards. All tables in this post are sortable by column.

• Thornton and Marleau are predicted to record substantially lower point totals than our previous article that looked at historical comparables for them projected. I'd be surprised to see either player drop off quite as much as VUKOTA is predicting.
  That said, projecting Couture to lead the team in scoring seems much more realistic even though I'd bet the over on his numbers above. Couture potted 34 assists last season despite the fact his linemates had a dismal time finishing their scoring chances at even-strength, something that's likely to rebound. Logan also hasn't had to rely on an anomalous shooting percentage to top 30 goals each of his first two full seasons in the NHL. /

What struck me as noticeable, especially when looking at the points per 82 games column, is the fairly massive gulf between the team's top six and the remainder of their forwards. While the exact numbers will probably differ, it's a good bet to make that the Sharks' bottom six will struggle mightily to produce offense as currently constructed.
I'd be shocked if Frazer McLaren scores more points per game than T.J. Galiardi. Then again, if McLaren appears in anywhere close to 27 games for the Sharks, they're in trouble.
In terms of raw numbers, VUKOTA predicts the Sharks to have zero 30-goal scorers on the team next season, a projection I'd find a lot more reasonable if it was accounting for the inevitable work stoppage, but it isn't. San Jose was only without a 30-goal scorer once between lockouts: the 07-08 season. I'd be surprised if at least two of Couture, Pavelski and Marleau didn't top 30 goals over a full year.
And now the defensemen:

Free from the shackles of being paired with Douglas Murray, Burns should probably score at a greater rate than he did last season although without a spot on the power play's first unit I can't see him overtaking Boyle. VUKOTA's projections for those two seem reasonable.
Jason Demers and Justin Braun are both well ahead of Brad Stuart in the pecking order for power play time, so I'd be surprised if they didn't both outscore the former Wings blueliner, even though the expectation they'll draw into the lineup for fewer games than Stuart is a correct one.
Defensemen are definitely harder to project than forwards since even-strength point scoring for blueliners tends to be highly variable from one year to the next but, overall, these estimates seem to make sense for the Sharks' defense.
Now that we've caught a glimpse of what the system projects for Sharks players in the coming year, the most important question to answer is one of accuracy. How much stock should we really put into these predictions? While this isn't exactly definitive, we can compare the system's predicted point totals for Sharks players in 2011-12 with what those players actually did last season. I rounded VUKOTA's raw numbers to make for easier comparisons to reality. Predictions are on the left, actual scoring totals on the right.

Across the board, the biggest misses are in the games played column. There appears to be some sort of serious systemic issue with the way VUKOTA predicts player health (or, more accurately, lack thereof). Despite that, VUKOTA's projections were fairly on the money as far as raw scoring numbers go. The only forwards' point totals VUKOTA was more than 6 points off on were Havlat (who was injured), Clowe (who played through injury), Jamie McGinn (who exploded after being traded to the Avs) and a few of the AHLers who ended up having short or nonexistent stints in the big leagues. On defense, the big misses were Demers, Murray and Vandermeer (as well as predicting Mike Moore would play 27 games).

VUKOTA definitely has issues to sort out in projecting games played totals and improving its hit rate with depth players but it's clearly a valuable tool. Exactly how much predictive value it provides is very much an open question; it's likely better than simply expecting a player to repeat his performance from a season ago but it's still certainly not the elusive crystal ball. The thing to keep in mind is that it's painting with a fairly broad brush (which makes it a better predictor of team performances as a whole than individual players) and, as a result, doesn't take into account specific circumstances that may have changed for some players. In the coming weeks, we'll try to use finer strokes to figure out what we should expect out of certain Sharks next season, whenever that happens to begin.