Stats endorse, warn on D-man prospects

P.K. Subban's numbers painted him as a high-potential prospect. Tom Szczerbowski/US Presswire

As noted in Tuesday's post regarding statistical insights on draft forwards, there has been an increased emphasis on analytics at the prospect level in the hockey world.

To review, many studies have shown there is notable predictive value in stat-based analysis for prospects. The main issues facing analytics for junior, collegiate and other levels is adjusting for context and trying to separate the relationship between production and the typical uncertainty/miss rate of projecting prospects. After all, stats can't tell you everything about prospects. They need to be adjusted for a variety of factors, they can't replace scouts, and they are subject to statistical noise. However, if used correctly for the NHL draft, stats can help teams find some overlooked sleepers and may prevent them from making a glaring mistake.

Just as it applies to how teams are assessing forwards, it also applies to how they are assessing defensemen. Here are some notable recent examples of defensemen whose pre-draft stats told a much different tale than their draft position:

Andrew Ference, Pittsburgh Penguins

Goalies may be the toughest position to project overall, but in regard to using statistics, defensemen may be a bit tougher because the majority of their value doesn't come from points or other offensive production. There is a relationship between scoring and future NHL value on both ends, but as one can imagine, it isn't as strong for blueliners as it is for forwards.

However, scoring can be a proxy for ice time in some regards and thus how much responsibility a coach gives a player. Ference's 44 points in 72 games as a 17-year-old in the WHL placed him second in scoring for under-18 defenders and 20th overall in WHL defenseman scoring. That alone is good but not overly impressive. Ference's standout numbers are as a 16-year-old, when he notched 40 points in 77 games, good for fourth overall among all under-17 players in the WHL and over 20 more than the next best defenseman.

A player's production as a 16-year-old tends to separate those whom coaches have identified as top talents, but they get bunched in CHL production at age 17. Given that Ference was known for his great physical play and potentially having great value on the defensive end, his high scoring was a potential indicator of two-way value and extensive ice time. He has ended up developing into a rough NHL defenseman in his career, with his value tilted toward defense, but he was certainly better than a 208th overall selection. One wonders where Ference would have been selected in the NHL draft if he were taller than 5-foot-11.

P.K. Subban, Montreal Canadiens

Subban was a mid-second-round pick by Montreal in 2007. Touted as a highly skilled offensive defenseman, he fell due to his significant issues on the defensive end. He scored 57 points as a 17-year-old defenseman in the OHL, which is a significant total. Since Subban's 2007 draft, the only 17-year-old defensemen to hit that total in the OHL have been Zach Bogosian, Drew Doughty, Michael Del Zotto, Ryan Ellis, Dougie Hamilton and Ryan Murphy.

That group includes several lottery picks, and all are first-round picks. Players like Ellis and Murphy are smaller than Subban, and that pair and Del Zotto all carried the same defensive issues in their draft seasons as Subban. Ellis and Murphy scored notably more, but there's a sliding scale based on size. Given that scouts praised Subban's skating and skill, with his average size and high scoring numbers for a defenseman, he likely did not merit his second-round draft slot even if he carried some other issues. NHL teams have shown a willingness to take on that risk for upside in the draft. They should have done so for Subban as well.

Boris Valabik, Atlanta Thrashers

Whenever a team drafts a huge defenseman high in the draft, one often hears the lazy comparison, "He could be the next Zdeno Chara." The alternative argument usually is, "He could be the next Boris Valabik."

As a 17-year-old in the OHL in 2003-04, Valabik scored 15 points in 68 games. He was 13th on his team in scoring and third among defensemen. Taking a defenseman who exhibits that level of scoring ability is like drafting a forward who scored 0.5 points per game. The only difference is that the defenseman will generally have more value in his defensive zone capabilities. The issue here is that in order for Valabik to merit the 10th overall selection, teams would have to be extremely confident in his defensive skills, because he brings almost no offense to the table. Unfortunately, defensive skills tend to have a lot of uncertainty when projecting to the NHL level. Chara didn't score a ton either in the WHL, but in his one season (combined with the playoffs), he scored 30 points in 64 games. In 188 OHL games across three seasons, Valabik put up only 31 points.

Valabik was a 6-7 defenseman who was extremely physical and had a fine defensive projection, but given his poor skating and clear low offensive ability, it was quite debatable at the time if Valabik was a top-15 prospect or even a first-rounder.

Fedor Tyutin, New York Rangers

Tyutin played in Russia during the final years of the Russian Super League, an era during which Russia's top league was 85 to 90 percent as strong as the quality of the NHL. Tyutin was a 17-year-old during the 2000-01 draft season, and Hockey Prospectus' Robert Vollman has indicated that for a player in that position -- based on his age and the RSL during that time period -- this was just about as challenging as if he had been playing in the NHL. Tyutin was getting a regular shift for SKA St. Petersburg, and in the two seasons before and after 2001, only one other under-18 defenseman was a regular all season in the league: Kirill Safronov, the 19th overall pick in 1999 (Phoenix). When teams give significant responsibility to an under-18 player in a top pro league, especially a defenseman, it usually is a sign of advanced two-way play.

While Tyutin hasn't put up gaudy numbers, he's been a tough-minutes, all-situations player and logged 22:13 of ice time per game during his NHL tenure. His all-around value was identified early by SKA, and he should have been in the conversation as a first-round selection instead of where he landed (40th overall).

Sasha Pokulok, Washington Capitals

If you're drawing a blank on just who, exactly, Pokulok is, it's hard to blame you. Here's the rundown: Pokulok was the 14th overall selection in the 2005 NHL draft. He was drafted out of Cornell after notching 10 points in 26 games, numbers that made him 32nd in under-19 NCAA scoring, seventh among defensemen. He was not in the top 100 in NCAA defenseman scoring or PPG. One notable statistical adjustment to his numbers is that the ECAC is the weakest of the major conferences. Players with identical numbers in the CCHA and Hockey East end up producing 20 percent more in the NHL, and for those from the WCHA, it's 35 percent more.

Pokulok was drafted because he was a 6-5, mobile defender with good defensive sense. The issue was that his offense was questionable. When a player can't outscore Ryan O'Byrne on his own team, it's a bit of a red flag. Defensive skills can be tougher to translate to the pro game, which adds uncertainty to the projection.

There's no issue in drafting a guy for his defensive skills, but at 14th overall, he needs to be super high-end defensively or have a good two-way game. Pokulok ended up being average to below-average defensively as a North American pro with no offense outside his final ECHL season. He is playing this upcoming season in Kazakhstan.