October 22 2012 07:14AM
Point projections you find in hockey magazines are based on the intuitive guesses of analysts who follow the teams closely, and can therefore occasionally be skewed by personal biases or wishful thinking. That's why it's best to supplement (not replace) them with purely statistically-based projections.
For the third season in a row over at the Flamesnation sister site we're using the two most popular statistical projection systems, Tom Awad's VUKOTA and my own Snepsts67, to anchor our expectations. The former is more established, regularly winning David Staples' comparative analysis, and has the advantage of also estimating games played while the latter defines lower and upper ranges, and finds good historical examples within.
Given Edmonton's well-established reputation as being among the most stats-savvy fans in the NHL, it makes sense to start covering this here as well – and give Staples another system to throw into the mix.
For a spreadsheet with complete results and a bit more of an explanation head on over to Hockey Abstract. Speaking of which, figuring out how a player is going to be used is the trickiest part of a projection, and that context can generally be found in the Player Usage Charts also available on that site, so have that handy too.
Today we'll start with the top-six forwards here today, analyze the defensemen next time and then the remaining secondary forwards in a third and final piece. Let's dig in!
Aided by the league's fifth best shooting percentage that leapt from 11.4% to 18.9%, a team shooting percentage jumping from 7.9% to an amazing 12.8% when he was on the ice, and an offensive zone start percentage vaulting from 49.3% to 60.7%, Jordan Eberle's scoring surged forward from 18 goals and 43 points to 34 goals and 76 points.
Defensively Eberle was used even more infrequently on the penalty kill than the year before, but his personal goals-against average remained exactly 3.08, against average competition.
Obviously Eberle's scoring totals due for a fair deal of statistical regression but fortunately his own development will cancel most of that out, leaving him once again just above those 30-goal, 70-point plateaus.
GP G A PTS Last Year 78 34 42 76 VUKOTA 79.4 31.7 42.3 73.9 Best 82 37.6 56.7 94.4 Worst 82 18.6 20.6 39.2 Average 82 31.2 40.5 71.7
Four of his ten closest historical matches were on pace for 78-81 points, including his closest match Jeremy Roenick. Though already far grittier than Eberle at that age, Roenick was shooting and scoring at a comparable rate (when adjusting for era).
Eberle GP G A PTS Age 20 69 18 25 43 Age 21 78 34 42 76 Roenick GP G A PTS Age 20 78 19 29 48 Age 21 79 31 42 73 Age 22 80 40 39 79
Roenick enjoyed his highest scoring seasons at this early stage of his career. In absolute era-unadjusted terms, Roenick averaged 50 goals over the next three seasons, topping 100 points each time and, most importantly, was absolutely unstoppable in NHL 94. This is all good news for Eberle, both on the ice, and in NHL 2014.
Context is everything, and Gabriel Landeskog edged out Ryan Nugent-Hopkins for the Calder trophy partly because of Nugent-Hopkins' injury, but also because he played the toughest minutes on the team while Edmonton used their sensational rookie almost exclusively in the offensive zone, and against more second-line competition.
While his even-strength assignment will no doubt get more difficult as the years progress, he's already proven himself amazing with the man advantage, scoring at an amazing rate of 7.4 points per 60 minutes. Unfortunately the historical similarity system requires at least two years of data, but VUKOTA was able to confidently project a modest improvement over-all.
GP G A PTS Last Year 62 18 34 52 VUKOTA 67.3 21.8 36.8 58.6
As a side note, Nugent-Hopkins was perhaps the league's worst at faceoffs, winning just 37.5% of his 605 draws.
Perhaps my favourite of all Edmonton's young stars, Taylor Hall has had fantastic double-digit possession numbers in each of his two seasons, meaning the Oilers have at least a 10-shot advantage (per 60 minutes) when he's on the ice. In fairness this is partly from an offensive zone start boost, but he does face opposing top lines.
He's also enjoyed a consistent even-strength scoring rate of around two points per 60 minutes, and his power play scoring jumped from 3.7 to 6.7 points per 60 minutes thanks to a little luck and his great chemistry with Ryan Nugent-Hopkins. His style of play has also quickly made him one of the league's best at drawing penalties.
GP G A PTS Last Year 61 27 26 53 VUKOTA 67.4 28.2 29.3 57.5 Best 82 31.8 41.3 73.2 Worst 82 17.0 7.5 24.5 Average 82 25.9 31.4 57.3
Though the average scoring expectations for Taylor Hall seem quite modest, his closest historical matches include Hall-of-Famer Pat LaFontaine and future Hall-of-Famer Jarome Iginla, both of whom developed into highly prolific scorers. Three of his ten closest historical matches topped 30 goals, 40 assists and 70 points in this upcoming season and personally I think he can too.
For all these amazing young talents to get such advantageous ice-time in the offensive zone and against second or third line competition, someone had to start holding the short end of the stick, and that someone has been Ales Hemsky. Hemsky has lost his offensive zone boost, and has finished top-3 among Oilers in quality of competition these past two seasons, and his numbers reflect it.
Despite the change in playing conditions, Hemsky remains a great talent with some of the strongest possession numbers on the team. His page on hockey-reference is humourously adorned with Tyler Dellow's message “Wondering if your team should get him? The Oilers have the worst record since the lockout - they didn't get there by winning trades. Your team will rip them off and he'll play well. Enjoy - great player to watch, a genius of sorts.”
Unfortunately when it comes to the box score numbers, Hemsky didn't look very good. His tougher playing conditions (and a little bad luck) led to a team shooting percentage dropping from 12.1% to 7.0% when he was on the ice, and his own shooting percentage dropping from 14.0% to 7.3%, which led to an even-strength scoring rate dropping to just 1.5 points per 60 minutes after two years in the 2.7 range. Furthermore, the changing role on the power play has him scoring just 3.6 points per 60 minutes these past two years, after regularly scoring over 6.0. Assuming these conditions persist, so will his deflated scoring levels.
GP G A PTS Last Year 69 10 26 36 VUKOTA 61.9 12.8 23.5 36.3 Best 82 31.7 43.5 75.2 Worst 82 5.3 14.0 19.3 Average 82 17.7 32.9 50.7
His closest historical match is similarly-styled (and injury-prone) playmaking artisan Michal Pivonka, who played his entire 13-season career with the Washington Capitals. Thanks to the opportunity to feed Peter Bondra, Pivonka scored the modern-day equivalent of 30 points in 46 games in the lockout-shortened 1994-95 season, followed by 70 points in 73 games in 1995-96.
Four of his ten closest historical matches scored at least 20 goals, and seven topped 30 assists (over 82 games). Pair Hemsky up appropriately and he could no doubt do the same.
With five straight seasons scoring between 13-18 goals, and 41-49 points, you might consider Sam Gagner's scoring the most predictable – but no. While the average projections are certainly within that range, half of Gagner's closest historical matches scored 20 goals, and four of ten topped 58 points. Given the potential he showed with his amazing eight-point night, Gagner could very easily break out of his established scoring range.
GP G A PTS Last Year 75 18 29 47 VUKOTA 67.3 16.1 25.0 41.1 Best 82 27.2 43.1 70.2 Worst 82 9.6 15.2 24.8 Average 82 18.9 28.8 47.7
An Oiler who is terrible at faceoffs, which I suppose is a redundant statement, Gagner has consistently strong possession numbers, as you might expect from someone who generally gets a solid offensive zone boost and just faced above average competition for the first time. His even-strength scoring rate has been in that lower top-six range of 1.6-2.0 points per 60 minutes his entire career and though he had one strong season on the power play in 2009-10 he's generally a secondary option only.
Last year Gagner also went 6 for 9 on the shootout after going 3 for 20 over the previous three seasons – do you still think the shootout is a predictable, reproducible talent closely related to hockey, rather than an arbitrary way to simulate parity through randomness?
While you shouldn't generally sign 30-year-old secondary centers to long-term $5.5M deals, you have to admit that the Oilers are fortunate to have a strong, veteran two-way pivot like Shawn Horcoff to play the truly brutal minutes until their younger players develop.
The least horrible of Edmonton's faceoff men, Horcoff has been the Oilers' top penalty killer up front for quite some time, and though he finished with the highest quality of competition among their forwards for the third year in a row, and among the bottom three in offensive zone starts for the third time in the past four seasons, his possession numbers consistently remain at least even.
Offensively Horcoff remains a useful secondary power play option but at even-strength his scoring level last grazed the lower top-six level way back in 2009-10. Unless he's given softer minutes alongside one of Edmonton's talented young superstars, Horcoff's scoring will remain in the 30s.
GP G A PTS Last Year 81 13 21 34 VUKOTA 65.8 10.2 16.7 26.9 Best 82 20.5 47.1 67.6 Worst 82 7.7 11.2 18.9 Average 82 11.3 25.8 37.1
While four of Horcoff's closest historical matches scored 27 points or less, half of them were in a tight range between 37-45 points – his likely maximum upside this season.
Receiving a 78.3% stamp of approval from a recent ESPN survey of top statistical analysts from around the league (including Jonathan Willis) of all this off-season's UFA deals, Ryan Smyth was signed to a bargain two-year, $2.25M/season deal. A dominant two-way top-six forward most of his career, the accomplished 36-year-old veteran will now be wind up his career in a tough minutes support role not dissimilar from last year's.
Though Smyth's lower top-six even-strength scoring rate of 1.6 to 1.9 points per 60 minutes has held strong for quite some time, his over-all scoring levels have consistently dropped every season, in pace with decreases in ice-time at both even-strength and the power play. Meanwhile, after two seasons in Los Angeles where he ranked 9th among their forwards in Quality of Competition, he finished 3rd in Edmonton last year, while simultaneously finishing with the 5th lowest offensive zone start percentage – quite possibly his lowest ranking ever. He also became one of the team's top line penalty killers after years of depth usage only.
GP G A PTS Last Year 82 19 27 46 VUKOTA 69.8 15.1 20.1 35.2 Best 82 20.5 26.8 47.3 Worst 82 9.4 18.4 27.8 Average 82 15.5 21.4 36.9
Given his age and his usage, you'll have to agree with the realism of a 16 goal, 37 point projection. Though there were only two 20-goal scorers among his ten closest historical matches, there's actually still a lot of reason to predict something closer to the upper end of this projection. Half of his closest matches, for instance, were in a tight band between 42-47 points, including his closest historical match Claude Lemieux.
It's been a pleasure to put this analysis together for one of the more receptive fanbases, and I hope you enjoyed it. If you did, stay tuned for the next part, which will feature the defensemen, before we wrap it up with a final article covering the remaining, secondary forwards.