New-Look HP! Unfiltered Articles Stats Glossary Contact Us
Hockey Prospectus home

New Look Hockey Prospectus is available with new Premium Content!
Limited time offer: $9.99 for an entire year!

<< Previous Article
Zamboni Tracks (02/26)
<< Previous Column
Premium Article From Daigle To Datsyuk (02/11)
Next Column >>
Premium Article From Daigle To Datsyuk (02/28)
Next Article >>
Premium Article From Daigle To Datsyuk (02/28)

February 27, 2013
From Daigle To Datsyuk
Seth Jones

by Corey Pronman


The two main questions that had to be asked at the top of the 2012 NHL Entry Draft were how much of a concern is on-ice work ethic, and how much risk is there in taking a player who has been injured for the entire draft season. The two prospects in mind for those questions were Mikhail Grigorenko (BUF) and Alex Galchenyuk (MTL). This time around the question is, "do you take a defenseman first overall?" and it applies to star defense prospect Seth Jones. I ranked him third in my midseason rankings due to my cautious approach on defense prospects. However, this is a debate worth more research, which may or may not change my stance.

During this debate, past first overall selections like Erik Johnson, Roman Hamrlik, and Bryan Berard will be referenced commonly. The decision for whatever team ends up with the top pick is complex. This article will attempt to tackle the different elements of the debate.

Historical overview

There is some pretty compelling evidence that defensemen, as a whole, tend to be riskier selections in the NHL Entry Draft than forwards, for instance as explained in The Edmonton Journal by Jonathan Willis, who also writes for Hockey Prospectus. Similar analyses can be found elsewhere as well.

However, you simply cannot look at the population; you also have to focus in on the elite prospects. Sure, top defensemen tend to come from outside the first round more often than forwards, but for the purpose of this debate, the focus is on how elite defense prospects pan out versus elite forward prospects—so the top of the draft will be keyed in on. The sample used for this study was the 1989-2003 drafts. That 15-year sample was chosen to keep the era relevant while allowing for enough production from the drafted players. Goals Versus Threshold (GVT) is used as the metric of evaluation. GVT encapsulates a player's total value. One of GVT's best utilities is to analyze a large group of players from different positions over a large period of time. The data used is as of the end of the 2011-12 season.

Total GVT is the average GVT of the player's career GVT in a given sample and Peak GVT is the average of player's combined top five season totals in a given sample.

Pick # No. Total GVT Peak GVT
1st 9 147 79
2nd-5th 39 86 53
6th-10th 46 44 29
Top 5 48 97 58
Top 10 94 71 44
Pick # No. Total GVT Peak GVT
1st 4 106 51
2nd-5th 19 67 40
6th-10th 23 34 24
Top 5 23 74 42
Top 10 46 54 33

As you see, forwards have higher numbers than defenders in every category although sample size is a potential issue.

The next sets of tables are for forwards and defensemen with over 500 games played that were picked in the top 10. The purpose of that criterion was to find the picks that turned into full-time NHLers. Peak Age is the average age of the player's five best seasons as determined by GVT. The purpose of it will be discussed more in the contracting rights section of this piece.

Pick # No. Total GVT Peak GVT Peak Age
1st 8 163 87 24.2
2nd-5th 31 106 65 26
6th-10th 26 71 44 25.2
Top 5 39 118 69 25.7
Top 10 65 99 59 25.5
Pick # No. Total GVT Peak GVT Peak Age
1st 4 106 51 25.40
2nd-5th 16 74 43 25.10
6th-10th 12 60 40 26.20
Top 5 20 81 44 25.10
Top 10 32 72 42 25.60

The trend repeats here, where forwards outproduce defenders at just about every point. The major takeaway is the seemingly small difference in Peak Age between forwards and defensemen.

Finally, the following table shows skaters with a career GVT of over 100. While not always true, for the most part these tend to be the true "big hits" at the top of the draft.

Forwards vs. defensemen
Position N Total GVT Peak GVT Peak Age
Forwards 28 160 87 26.1
Defensemen 7 159 73 26.8

This table is the most interesting, because of how close the Total GVT categories are. Still, no matter how you slice the data, the top forwards never seem to trail.

The issue with this method is that GVT's ability to evaluate individual defensemen can be suspect due to the Defensive GVT component, although that argument does not seem as convincing on a mass level. Hockey Prospectus' Tom Awad, the creator of GVT, acknowledged that GVT can undervalue the defensive aspect of defenders in some cases. In an email exchange, Awad said that he feels the league's elite defensemen should actually be worth around 20 GVT, whereas he notes Zdeno Chara and Shea Weber's average GVT the last four seasons have been 15 and 14 respectively. Gabe Desjardins went even further to estimate with a WOWY analysis (With Or Without You) that a top pairing defenseman is worth almost four wins (roughly 21 GVT).

The actual answer to how much elite defenders in this current NHL player pool should be worth relative to the league's top forwards would be a discussion for a whole other article, but for now, acting on the assumption the difference does not seem that notable seems reasonable, when you put generational players like Evgeni Malkin aside.

Based on this information, we can reach the following conclusions:

1. In general, forwards tend to be wiser picks than defensemen at the top of the draft.

2. In general, when you get a top-tier skater out of the top of the draft, forwards and defensemen seem to be relatively close in terms of the value that player provides over the course of their careers.

3. The value of having an elite, yet not generational, NHL defenseman in today's NHL seems roughly equal to having an elite, yet not generational, NHL forward.

4. There is some uncertainty in the results based on the method chosen (GVT) but much less so due to the study involving a sizeable amount of players.

5. The small sample size in several of the table rows leaves room for notable error, which is the result of taking a subset of a subset of data.

Contracting rights

With the new CBA agreed to, contracting rights in the NHL will remain the same. That means a three-year entry level contract and unrestricted free agency occurs after seven accrued seasons or if a player is 27 years old as of June 30 at the end of the league year. While you could extend a player beyond that timeframe, per this bargaining agreement, those are your mostly-guaranteed controlled years for a player you've drafted (putting aside the rare offer sheet). It's hard to know exactly what the odds are of a team extending a top player they drafted high. Based on an eye test, there seems like a decent chance you can expect more than seven years of production for your team, but there is also a notable enough chance the player won't stay with the team that drafted them. Thus, it is important to know how a player will produce in that controlled timeframe. We will do so using the same set up as in the previous section, the 1989-2003 drafts and top 10 picks only.

The seasons that will be looked at for production will be the first seven seasons of a player's NHL career, or up until their age-27 season. In any instance where a player under 20 years of age played under 10 games, that season would not be counted under this CBA. These GVT numbers are adjusted for the 1994-95 lockout.

For the purpose of trying to capture how successful picks progress, the sample was limited to players with a career GVT over 100 for defensemen and over 150 for forwards (for ease of viewing).

Here are the table versions for this analysis, in this case for forwards and defensemen with a career GVT over 100.

Players with career GVT over 100
Age N Forwards GVT N Defensemen GVT
18 8 3.49 1 3.0
19 19 7.14 7 5.3
20 25 8.48 7 6.4
21 28 8.79 7 8.2
22 28 11.87 7 8.9
23 28 13.66 7 9.9
24 28 14.37 7 13.8
25 22 15.16 6 16.5
26 11 9.61
27 4 14.00

The next table is for players with Career GVTs over 150. Wade Redden was added as well, as he just barely missed the 150 mark. Redden had a stellar first seven seasons.

Players with career GVT over 150
Age N Forwards GVT N Defensemen GVT
18 6 3.57 1 3.00
19 9 10.04 4 6.58
20 11 10.65 4 8.58
21 12 12.98 4 9.75
22 13 15.93 4 9.95
23 13 15.85 4 12.55
24 12 16.88 4 14.75
25 9 18.61 4 14.50
26 4 11.08
27 2 14.05

The things to note from this section are:

1. In general, forwards picked at the top of the draft who succeed in their careers tend to have much better early production than similar defensemen.

2. There is only a small sample of defensemen who were picked high and who had successful seasons, which may mean that more data is needed to make any conclusions of how early defensemen development works or that so few of the 46 top-10 defensemen picks from 1989-2003 ended up being successful that it's an argument against drafting the position. When players like Drew Doughty and Alex Pietrangelo are done their first seven seasons, this kind of analysis may look different.

3. Even when you take best case scenarios like Chris Pronger, Scott Niedermayer, and Wade Redden, they are usually below the best cases for forwards.

4. There are possible issues with the method chosen (GVT) and the sample size, but given the wide discrepancies, it is likely not much of an issue towards the conclusions.

A related analysis was done by Benjamin Wendorf of SB Nation. He took the % change from year-to-year in ice time for forwards and defensemen. Per that research, Ben shows that defenders' peaks last longer than forwards' peaks. You will recall that I noted that top forwards and defensemen drafted high tend to very close in Career GVT and Peak Age, but that forwards had a better five-year peak. Ben did not use just top draft picks, but when you tie his research together with what we have learned, a fair assumption would be that defenders tend to have longer peaks, but less outstanding (25+ GVT) years. If you look at the careers of top picks like Chris Pronger, Scott Niedermayer, Wade Redden, or even a non top pick like Nicklas Lidstrom, this seems like a trend.

It's certainly legitimate to make the argument that Chris Pronger or Scott Niedermayer were worth more in the first 15 years of their career than an elite forward. In the context of the controllable first seven seasons, though, forwards seem to be favored.

The individual case

To this point, this column has discussed drafting forwards vs. defensemen at the top of the draft on a general level. The purpose is to provide important context as to whether a team with the first overall selection should select a forward like Nathan MacKinnon or Jonathan Drouin, or the top defense prospect Seth Jones. However, an important part of the discussion is that teams are not interested in how the top forward prospect in the next 30 draft classes will do relative to the best defense prospect, it is only worried for now about this one specific case.

And in this case, Seth Jones is a very special prospect, one who you do not see come out of the NHL Entry Draft often. The specifics of his talent analysis are something I will go into detail on in my Draft Preview, but as a summary, he has high-end to elite abilities across the board. If someone were to make the case that he has a slightly higher overall talent level than Nathan MacKinnon, that is an argument I could buy, and I may even agree with.

In general, defensemen taken at the top of the draft tend to pan out less as prospects relative to forwards and/or take longer to provide significant output. This usually is due to the fact that the transition for the position to the NHL is much harder than it is for forwards. If someone wanted to make an argument that due to Seth Jones' immense hockey sense and physical tools that such a transition risk is much less than it is for most top defense prospects, that would seemed to be a reasonable argument as well.


If I were a general manager, these are the questions I would be asking my staff regarding potentially picking Seth Jones first overall:

1. Does Jones project as a more valuable player than Nathan MacKinnon or Jonathan Drouin for the duration of his career?

2. If the answer to question #1 is yes, is the difference between Jones and the next best alternative slight, or moderate to significant?

3. Is the development/projection risk on Jones substantially lower than it usually is for top defense prospects?

4. If the answer to question #3 is yes, how much so?

5. How much do you see Jones producing relative to the next best alternative forward on your team for the next seven seasons, assuming both start in the NHL in 2013-14?

6. Finally, given all of the historical evidence in this debate—and noting the Erik Johnson draft and the Ovechkin/Malkin/Barker draft were excluded from the sample—is it worth the risk to select Jones?

There are a mix of possibilities to these answers I would deem appropriate to getting my green light to draft Jones. For instance, if there was only a slight gap between him and the next best player, and he seemed like a very low risk who would produce a lot, soon or immediately. Or if there was a moderate to significant gap and the risk was fairly low.

So how do I answer these questions, and what is my final call on whether to draft Seth Jones first overall? You'll have to wait for my 2013 Draft preview!

Corey Pronman is an author of Hockey Prospectus. You can contact Corey by clicking here or click here to see Corey's other articles.

0 comments have been left for this article.

<< Previous Article
Zamboni Tracks (02/26)
<< Previous Column
Premium Article From Daigle To Datsyuk (02/11)
Next Column >>
Premium Article From Daigle To Datsyuk (02/28)
Next Article >>
Premium Article From Daigle To Datsyuk (02/28)

Annoucements: Where Are The New Articles?
Zamboni Tracks: Who's That Guy? Special Edmo...
Hall Of Fame: My 2014 HHOF Inductees
On The Horizon: Four Nations And Junior "A" ...
A Closer Look: MacArthur-Turris-Ryan Keeping...

2013-03-16 - Premium Article From Daigle To Datsyuk: NCAA Prospects Ready...
2013-03-01 - From Daigle To Datsyuk: To Match Or Not To M...
2013-02-28 - Premium Article From Daigle To Datsyuk: Top AHLers Ready To ...
2013-02-27 - From Daigle To Datsyuk: Seth Jones
2013-02-11 - Premium Article From Daigle To Datsyuk: Tampa Bay's Well-Oil...
2013-02-05 - From Daigle To Datsyuk: Names To Know, Janua...
2013-01-12 - From Daigle To Datsyuk: Top 50 NHL Prospects...

2013-03-16 - Premium Article From Daigle To Datsyuk: NCAA Prospects Ready...
2013-03-01 - From Daigle To Datsyuk: To Match Or Not To M...
2013-02-28 - Premium Article From Daigle To Datsyuk: Top AHLers Ready To ...
2013-02-27 - From Daigle To Datsyuk: Seth Jones
2013-02-11 - Premium Article From Daigle To Datsyuk: Tampa Bay's Well-Oil...
2013-02-05 - From Daigle To Datsyuk: Names To Know, Janua...
2013-01-12 - From Daigle To Datsyuk: Top 50 NHL Prospects...