jwhouk
Former Cheesehead, Always a Preds Fan
I realized that I never posted this anywhere. It's been a little thing I've been working on for years. Stats are through the end of last season. Sorry in advance for the bad formatting. - JH
Every year during the NHL draft, fans and pundits alike have opinions on the available players and the draft strategy of certain NHL teams. A lot of these opinions and discussions break down to arguments like, "Who's done better in the draft – Colorado (who had the first pick in 2013) or Tampa Bay (who had the third pick)?" Or, "Is there really a significant difference between picking someone in the first half of the first round, as opposed to the last half?" Or even, "What kind of talent is there in the last few rounds of the draft?"
Believe it or not, these questions can be answered by using statistical analysis. And we don't even have to get into a deep philosophical discussion about "drafting for needs" or "filling a hole on your roster" or even "drafting for the future". That is because there is one overwhelming commonality between every pick that has been made in the NHL Entry Draft, since the very first time it was held in the 1960's.
NHL General Managers are trying to select players who they think are going to play in the NHL. They're not drafting "a backup goaltender", or "a second-pairing defenseman", or even "a two-way winger". They are trying to pick players who will play games in the NHL – hopefully with their team, and hopefully in as short as possible a time frame from the date they were selected.
Thus, to measure "value" of a draft pick – whether it's the #1 overall pick or the last selection in the last round – we should use the yardstick of NHL games played to help estimate (or, in the language that we're going to use in deriving it, "approximate") that value.
Granted, not all games played are equal. 40 games from two defensemen that were picked in same round – but one year apart – are drastically different in value. And 40 games played from a first round pick, compared to someone picked late in the fifth or six round, is also a drastic difference in value.
How do you measure this difference, without strictly comparing games played to games played? Let's take a look at the top five picks of the 2010 NHL Entry Draft:
[FONT="]
[/FONT]
Believe it or not, this is actually easy to figure out. For someone drafted in 2012, that total is 48. For someone drafted in 2011, it's 130. You add the number of games played per team to each round prior (except, of course, for those drafted in 2004), until you reach 1,760. When the maximum number of possible games played is the same as Gordie Howe's totals as a player, you've pretty much hit the top end for player value. As of 2013, all seasons prior to the 1990 NHL Entry Draft are measured versus the 1,760 game maximum.
So, taking the player's total number of NHL games played, and dividing it by this "Max Games Played" total, you get a number between 0 and 1 – one that can be easily expressed as a percentage, by simply multiplying the result by 100. We call this result a player's Approximate Draft Value.
It's not unusual for there to be one or two players with an ADV of 1.000 one year after a player was drafted. As of right now, Nail Yakupov and Alex Galchenyuk have 1.000 ADV's, because they played all 48 games for their teams this past season. A high ADV from an earlier draft indicates a player who has played in a LOT of NHL games since he was drafted.
Other than the two players listed above, only 30 players since 1988 have posted ADV totals of .90 or more:
Year Rd OA FRN Player Name Pos GP MAXG ADV
2008 1 1 TAM Steven Stamkos C 373 376 .9920
2009 1 1 NYI John Tavares C 291 294 .9898
2003 1 2 CAR Eric Staal C 690 704 .9801
1997 1 2 SAN Patrick Marleau C 1165 1196 .9741
2007 1 1 CHI Patrick Kane RW 446 458 .9738
2002 1 3 FLA Jay Bouwmeester D 764 786 .9720
2008 1 2 LOS Drew Doughty D 364 376 .9681
2004 1 1 WAS Alex Ovechkin LW 601 622 .9662
2010 1 2 BOS Tyler Seguin C 203 212 .9575
2008 1 5 TOR Luke Schenn D 357 376 .9495
2011 1 8 PHI Sean Couturier C 123 130 .9462
1997 1 1 BOS Joe Thornton C 1125 1196 .9406
2001 1 1 WPG Ilya Kovalchuk LW 816 868 .9401
1995 1 7 PHX Shane Doan RW 1246 1326 .9397
1988 4 67 PIT Mark Recchi RW 1652 1760 .9386
2006 1 5 BOS Phil Kessel RW 504 540 .9333
1998 1 1 TAM Vincent Lecavalier C 1037 1114 .9309
1995 1 11 DAL Jarome Iginla RW 1232 1326 .9291
2000 1 6 NAS Scott Hartnell LW 875 950 .9211
2010 1 12 ANA Cam Fowler D 195 212 .9198
2010 1 8 WPG Alexander Burmistrov C 194 212 .9151
2002 1 1 CLB Rick Nash LW 718 786 .9135
1999 1 3 VAN Henrik Sedin C 940 1032 .9109
2003 1 13 LOS Dustin Brown RW 641 704 .9105
1997 1 3 LOS Olli Jokinen C 1087 1196 .9089
2011 1 2 COL Gabriel Landeskog LW 118 130 .9077
1995 1 10 FLA Radek Dvorak RW 1200 1326 .9050
2009 1 3 COL Matt Duchene C 266 294 .9048
2007 1 6 EDM Sam Gagner C 414 458 .9039
2009 2 33 COL Ryan O'Reilly C 265 294 .9014
FRN: Franchise ID; YR: Draft Year; RD: Round Drafted; OA: Overall Pick; GP: NHL Games Played; MAXG: Maximum Possible Games; ADV: Approximate Draft Value
Notice two names on that list – Mark Recchi in Pittsburgh, and Ryan O'Reilly in Colorado. Recchi's draft value went up because of who he played for and with; O'Reilly has been consistent for the Avs because they haven't had much other talent around him.
This brings up an answer to one of our questions at the beginning: What's the difference between a first round pick and a later round pick, in terms of draft value? As you can pretty much guess, there's more ADV among players selected in the first round than there is in later rounds. Below is a table of the combined ADV's of each round of the draft, from 1988 to 2012:
RD ADV ADV/D ADV/P
1 243.6282 9.7451 .3562
2 97.9810 3.9192 .1339
3 61.5084 2.4603 .0874
4 48.0341 1.9214 .0676
5 34.7222 1.3889 .0481
6 36.7977 1.4719 .0532
7+ 84.7155 3.3886 .0418
ADV/D: Approximate Draft Value per Draft (ADV / 25); ADV/P: Approximate Draft Value per Player (ADV/# players drafted in round from 1988-2012).
You'll notice that the first round produces a score of about 244. When we're talking about groups of draftees like this, think of combined ADV as a measure of "complete" player games. Thus, the first round was where the NHL drafted the equivalent of 244 players over the last 25 years – or just under 10 players per year. That draft value drops each round until you reach the sixth round, where it bumps up slightly – but not significantly. It's at that point that GM's likely start looking at the quality of the players left on their draft board, and try to find one player who perhaps impressed them with their work ethic. Or, of course, they could just take a flyer on someone who they think has NHL skill, but no one apparently believed they could play in the NHL.
So where exactly does draft value start to drop dramatically? Believe it or not, right after the first pick. I know, I know – I can hear you saying, "Next thing you're going tell us is that the sky is blue." The truth is, though, that the 25 players selected first overall since 1988 have a combined ADV of 19.18, or an average "per draft" (that ADV/D figure) of .7676. The players picked second overall have a 16.88 ADV, or .6753 per draft – which is 14% less than the previous pick.
The ADV's of each of the top 10 draft picks from 1988 to 2012 are listed below:
OA ADV ADV/D DPS D1S
1 19.1888 0.7676 .00 .00
2 16.8823 0.6753 -.14 -.14
3 16.4981 0.6599 -.02 -.16
4 12.4453 0.4978 -.33 -.54
5 13.4077 0.5363 .07 -.43
6 10.5575 0.4223 -.27 -.82
7 11.5978 0.4639 .09 -.65
8 10.3220 0.4129 -.12 -.86
9 9.6069 0.3843 -.07 -1.00
10 8.8433 0.3537 -.09 -1.17
DPS: Percentage difference of ADV/D between pick number and previous pick; D1S: Percentage difference of ADV/D between pick number and #1 overall pick.
In case you can't figure it out from that list, the 9th player selected overall has an ADV of about half that of a number one overall pick. It doesn't get any better when you start looking at the draft in terms of groupings of 15 players:
From To ADV #PK ADV/P
1 14 164.5803 350 .4702
15 29 88.6675 375 .2364
30 44 49.8499 375 .1329
45 59 42.9250 375 .1145
60 74 42.3412 375 .1129
75 89 28.6731 375 .0765
90 104 25.9692 375 .0693
105 119 18.9467 374 .0507
120 134 19.8437 375 .0529
135 149 14.8779 375 .0397
150 164 19.7008 375 .0525
165 179 18.3312 375 .0489
180 194 14.2743 375 .0381
195 209 11.4120 375 .0304
210 224 15.5153 285 .0544
225 239 11.9733 256 .0468
240 254 8.6951 213 .0408
255 269 5.5697 144 .0387
270 284 2.4601 108 .0228
285 293 2.7809 40 .0695
From and To are the boundaries of the picks (1-14 is picks 1 through 14, etc.); #PK is number of picks. ADV/P is the ADV per pick.
There's some variance by a couple of percentage points after pick 119, but draft value tends to decrease, on a whole, after you reach the equivalent of the end of the fourth round (picks 120 and higher). The anomaly of the .0695 ADV/P for picks 285-293 can be explained simply: two players (Kim Johnsson and Hans Jonsson) who were picked 286th overall in the early 1990's weren't expected to play in North America – and yet they did.
What about our comparable players we mentioned a bit earlier? Which of them actually provided more ADV to their teams? Let's sort them by ADV:
·
Is there a positional difference in the NHL draft? Yes:
The difference between forwards and defense is significant, but not that great – both are about the same as an early third round pick. The ADV for Goaltenders is at a point where you'd have to say it's not worth it to take one until the late rounds.
In case you're wondering, the 2003 NHL draft is looking more and more each year like a once-in-a-generation quality draft. The total ADV from players selected in Nashville that year is currently at 43.1932, or an AADV of 1.4398. What that essentially means is that every team in the NHL drafted at least one, possibly two players who have played a significant amount of time in the NHL since then. The 1998 draft is right behind – and that may be more because of who was making the selections (it was Nashville's first draft) and the quality of some of the picks (players like Scott Gomez and Karlis Skrastins, who weren't top selections).
It takes about five or six years for a significant number of players in a particular draft to see NHL playing time. This can easily be proven by noting that the last six drafts have yet to reach an AADV over .8000 (the 2006 draft has the highest with .7629).
What's eye opening is when you see a draft from more than 10 years ago behind a more recent draft; the 2005 draft had a 27.0257 ADV (.9009 AADV); the 1999 NHL draft was a relatively poor 23.5068 (.8395 AADV). Strangely enough, the 1999 draft was significantly worse than the draft prior – and that's not all the fault of the Atlanta Thrashers.
Speaking of which: what team has done better drafting since 1988 – Colorado, Florida or Tampa Bay? The answer might surprise you: Colorado leads the entire NHL with a 30.2683 ADV (1.2107 AADV) in talent over the last 25 years. Florida comes in at 12th overall (18.277, .9139 AADV); Tampa – with only two #1 overall draft picks – has done poorly the rest of the time. They have a 13.6189 ADV, which translates to .6485 AADV over 21 years.
FRN ADV #DR AADV
COL 30.2683 25 1.2107
BUF 27.9896 25 1.1196
SAN 23.7801 22 1.0809
OTT 22.3092 21 1.0623
MON 26.1461 25 1.0458
NJD 25.7940 25 1.0318
NYI 25.6349 25 1.0254
EDM 23.5120 25 .9405
CLB 12.1081 13 .9314
NAS 13.9265 15 .9284
LOS 22.9955 25 .9198
FLA 18.2772 20 .9139
NYR 22.8018 25 .9121
BOS 22.5469 25 .9019
PIT 22.3519 25 .8941
ANA 17.2312 20 .8616
WAS 21.0709 25 .8428
TOR 20.2984 25 .8119
STL 20.0625 25 .8025
PHI 19.9270 25 .7971
CHI 19.7435 25 .7897
DET 19.3684 25 .7747
MIN 10.0544 13 .7734
WPG 10.8192 14 .7728
CAR 19.2576 25 .7703
DAL 19.1192 25 .7648
PHX 19.1091 25 .7644
VAN 18.7570 25 .7503
CGY 18.5076 25 .7403
TAM 13.6189 21 .6485
NHL 20.2462 684 .8861
#DR: Number of Drafts
Of the expansion teams since 1991, the Ottawa Senators have performed the best in the draft, with 1.0623 AADV. Of the younger teams, Columbus has the better AADV (.9314), but Nashville has produced more draft value (13.9265).
Meanwhile, Calgary has not had much luck in drafting players since they won the Cup in '88. They have the lowest ADV (18.5076) and AADV (.7403) of any NHL team that participated in all 25 drafts in the last 25 years.
If you look at the list of the players with the highest ADV for each NHL franchise since 1988, many of the names won't surprise you. (Nail Yakupov isn't on this list, because he has played less than 100 games in the NHL.)
FRN Player Pos YR RD OA GP MAXG ADV
TAM Steven Stamkos C 2008 1 1 373 376 .9920
NYI John Tavares C 2009 1 1 291 294 .9898
CAR Eric Staal C 2003 1 2 690 704 .9801
SAN Patrick Marleau C 1997 1 2 1165 1196 .9741
CHI Patrick Kane RW 2007 1 1 446 458 .9738
FLA Jay Bouwmeester D 2002 1 3 764 786 .9720
LOS Drew Doughty D 2008 1 2 364 376 .9681
WAS Alex Ovechkin LW 2004 1 1 601 622 .9662
BOS Tyler Seguin C 2010 1 2 203 212 .9575
TOR Luke Schenn D 2008 1 5 357 376 .9495
PHI Sean Couturier C 2011 1 8 123 130 .9462
WPG Ilya Kovalchuk LW 2001 1 1 816 868 .9401
PHX Shane Doan RW 1995 1 7 1246 1326 .9397
PIT Mark Recchi RW 1988 4 67 1652 1760 .9386
DAL Jarome Iginla RW 1995 1 11 1232 1326 .9291
NAS Scott Hartnell LW 2000 1 6 875 950 .9211
ANA Cam Fowler D 2010 1 12 195 212 .9198
CLB Rick Nash LW 2002 1 1 718 786 .9135
VAN Henrik Sedin C 1999 1 3 940 1032 .9109
COL Gabriel Landeskog LW 2011 1 2 118 130 .9077
EDM Sam Gagner C 2007 1 6 414 458 .9039
DET Nicklas Lidstrom D 1989 3 53 1564 1760 .8886
OTT Andrej Meszaros D 2004 1 23 533 622 .8569
MIN Nick Schultz D 2000 2 33 811 950 .8537
CGY Dion Phaneuf D 2003 1 9 600 704 .8523
NJD Scott Gomez C 1998 1 27 941 1114 .8447
STL Rod Brind'Amour C 1988 1 9 1484 1760 .8432
BUF Thomas Vanek LW 2003 1 5 585 704 .8310
NYR Todd Marchant C 1993 7 164 1195 1494 .7999
MON Mark Streit D 2004 9 262 491 622 .7894
It's probably no shock to anyone that Rick DiPietro of the Islanders (.3347 ADV) is the biggest "bust" among #1 overall draft picks in the last 25 years. And among players picked in the top five (that actually made it to the NHL), Alexandre Volchkov of the Washington Capitals (1996, 4th overall, .0023 ADV) is probably the biggest bust among first rounders. There's also Thomas Hickey (Los Angeles, 2007, 4th overall) and Ryan Sittler (Philadelphia, 1992, 7th overall), both of which never played a single NHL game.
One last thing about draft value: there have been 6,270 picks made in the NHL entry draft between 1988 and 2012. Three were invalid picks (two in 1990 – Edmonton's 227th overall and Calgary's 220th overall; the third was Edmonton's 123rd pick in 2002), and one forfeited pick (New Jersey's 69th pick in 2011).
Of the rest, 3,863 of the selections have never played a single game in the NHL – 2,998 of those were chosen before the 2008 Entry Draft. Of those who have played at least one NHL game, their average ADV is .2528.
Every year during the NHL draft, fans and pundits alike have opinions on the available players and the draft strategy of certain NHL teams. A lot of these opinions and discussions break down to arguments like, "Who's done better in the draft – Colorado (who had the first pick in 2013) or Tampa Bay (who had the third pick)?" Or, "Is there really a significant difference between picking someone in the first half of the first round, as opposed to the last half?" Or even, "What kind of talent is there in the last few rounds of the draft?"
Believe it or not, these questions can be answered by using statistical analysis. And we don't even have to get into a deep philosophical discussion about "drafting for needs" or "filling a hole on your roster" or even "drafting for the future". That is because there is one overwhelming commonality between every pick that has been made in the NHL Entry Draft, since the very first time it was held in the 1960's.
NHL General Managers are trying to select players who they think are going to play in the NHL. They're not drafting "a backup goaltender", or "a second-pairing defenseman", or even "a two-way winger". They are trying to pick players who will play games in the NHL – hopefully with their team, and hopefully in as short as possible a time frame from the date they were selected.
Thus, to measure "value" of a draft pick – whether it's the #1 overall pick or the last selection in the last round – we should use the yardstick of NHL games played to help estimate (or, in the language that we're going to use in deriving it, "approximate") that value.
Granted, not all games played are equal. 40 games from two defensemen that were picked in same round – but one year apart – are drastically different in value. And 40 games played from a first round pick, compared to someone picked late in the fifth or six round, is also a drastic difference in value.
How do you measure this difference, without strictly comparing games played to games played? Let's take a look at the top five picks of the 2010 NHL Entry Draft:
[FONT="]
[/FONT]
- Taylor Hall, LW, EDM – 171 NHL games
- Tyler Seguin, C, BOS – 203 games
- Erik Gudbranson, D, FLA – 104 games
- Ryan Johansen, C, CLB – 107 games
- Nino Niederreiter, RW NYI – 64 games
- Joey Crabb, RW, NY Rangers, 2002 draft – 170 games
- Cam Stewart, LW, Boston, 1990 draft – 202 games
- T.J. Brodie, D, Calgary, 2008 draft – 104 games
- Troy Bodie, RW, Edmonton, 2003 draft – 107 games
- Ryan Ellis, D, Nashville, 2009 draft – 64 games
Believe it or not, this is actually easy to figure out. For someone drafted in 2012, that total is 48. For someone drafted in 2011, it's 130. You add the number of games played per team to each round prior (except, of course, for those drafted in 2004), until you reach 1,760. When the maximum number of possible games played is the same as Gordie Howe's totals as a player, you've pretty much hit the top end for player value. As of 2013, all seasons prior to the 1990 NHL Entry Draft are measured versus the 1,760 game maximum.
So, taking the player's total number of NHL games played, and dividing it by this "Max Games Played" total, you get a number between 0 and 1 – one that can be easily expressed as a percentage, by simply multiplying the result by 100. We call this result a player's Approximate Draft Value.
It's not unusual for there to be one or two players with an ADV of 1.000 one year after a player was drafted. As of right now, Nail Yakupov and Alex Galchenyuk have 1.000 ADV's, because they played all 48 games for their teams this past season. A high ADV from an earlier draft indicates a player who has played in a LOT of NHL games since he was drafted.
Other than the two players listed above, only 30 players since 1988 have posted ADV totals of .90 or more:
Year Rd OA FRN Player Name Pos GP MAXG ADV
2008 1 1 TAM Steven Stamkos C 373 376 .9920
2009 1 1 NYI John Tavares C 291 294 .9898
2003 1 2 CAR Eric Staal C 690 704 .9801
1997 1 2 SAN Patrick Marleau C 1165 1196 .9741
2007 1 1 CHI Patrick Kane RW 446 458 .9738
2002 1 3 FLA Jay Bouwmeester D 764 786 .9720
2008 1 2 LOS Drew Doughty D 364 376 .9681
2004 1 1 WAS Alex Ovechkin LW 601 622 .9662
2010 1 2 BOS Tyler Seguin C 203 212 .9575
2008 1 5 TOR Luke Schenn D 357 376 .9495
2011 1 8 PHI Sean Couturier C 123 130 .9462
1997 1 1 BOS Joe Thornton C 1125 1196 .9406
2001 1 1 WPG Ilya Kovalchuk LW 816 868 .9401
1995 1 7 PHX Shane Doan RW 1246 1326 .9397
1988 4 67 PIT Mark Recchi RW 1652 1760 .9386
2006 1 5 BOS Phil Kessel RW 504 540 .9333
1998 1 1 TAM Vincent Lecavalier C 1037 1114 .9309
1995 1 11 DAL Jarome Iginla RW 1232 1326 .9291
2000 1 6 NAS Scott Hartnell LW 875 950 .9211
2010 1 12 ANA Cam Fowler D 195 212 .9198
2010 1 8 WPG Alexander Burmistrov C 194 212 .9151
2002 1 1 CLB Rick Nash LW 718 786 .9135
1999 1 3 VAN Henrik Sedin C 940 1032 .9109
2003 1 13 LOS Dustin Brown RW 641 704 .9105
1997 1 3 LOS Olli Jokinen C 1087 1196 .9089
2011 1 2 COL Gabriel Landeskog LW 118 130 .9077
1995 1 10 FLA Radek Dvorak RW 1200 1326 .9050
2009 1 3 COL Matt Duchene C 266 294 .9048
2007 1 6 EDM Sam Gagner C 414 458 .9039
2009 2 33 COL Ryan O'Reilly C 265 294 .9014
FRN: Franchise ID; YR: Draft Year; RD: Round Drafted; OA: Overall Pick; GP: NHL Games Played; MAXG: Maximum Possible Games; ADV: Approximate Draft Value
Notice two names on that list – Mark Recchi in Pittsburgh, and Ryan O'Reilly in Colorado. Recchi's draft value went up because of who he played for and with; O'Reilly has been consistent for the Avs because they haven't had much other talent around him.
This brings up an answer to one of our questions at the beginning: What's the difference between a first round pick and a later round pick, in terms of draft value? As you can pretty much guess, there's more ADV among players selected in the first round than there is in later rounds. Below is a table of the combined ADV's of each round of the draft, from 1988 to 2012:
RD ADV ADV/D ADV/P
1 243.6282 9.7451 .3562
2 97.9810 3.9192 .1339
3 61.5084 2.4603 .0874
4 48.0341 1.9214 .0676
5 34.7222 1.3889 .0481
6 36.7977 1.4719 .0532
7+ 84.7155 3.3886 .0418
ADV/D: Approximate Draft Value per Draft (ADV / 25); ADV/P: Approximate Draft Value per Player (ADV/# players drafted in round from 1988-2012).
You'll notice that the first round produces a score of about 244. When we're talking about groups of draftees like this, think of combined ADV as a measure of "complete" player games. Thus, the first round was where the NHL drafted the equivalent of 244 players over the last 25 years – or just under 10 players per year. That draft value drops each round until you reach the sixth round, where it bumps up slightly – but not significantly. It's at that point that GM's likely start looking at the quality of the players left on their draft board, and try to find one player who perhaps impressed them with their work ethic. Or, of course, they could just take a flyer on someone who they think has NHL skill, but no one apparently believed they could play in the NHL.
So where exactly does draft value start to drop dramatically? Believe it or not, right after the first pick. I know, I know – I can hear you saying, "Next thing you're going tell us is that the sky is blue." The truth is, though, that the 25 players selected first overall since 1988 have a combined ADV of 19.18, or an average "per draft" (that ADV/D figure) of .7676. The players picked second overall have a 16.88 ADV, or .6753 per draft – which is 14% less than the previous pick.
The ADV's of each of the top 10 draft picks from 1988 to 2012 are listed below:
OA ADV ADV/D DPS D1S
1 19.1888 0.7676 .00 .00
2 16.8823 0.6753 -.14 -.14
3 16.4981 0.6599 -.02 -.16
4 12.4453 0.4978 -.33 -.54
5 13.4077 0.5363 .07 -.43
6 10.5575 0.4223 -.27 -.82
7 11.5978 0.4639 .09 -.65
8 10.3220 0.4129 -.12 -.86
9 9.6069 0.3843 -.07 -1.00
10 8.8433 0.3537 -.09 -1.17
DPS: Percentage difference of ADV/D between pick number and previous pick; D1S: Percentage difference of ADV/D between pick number and #1 overall pick.
In case you can't figure it out from that list, the 9th player selected overall has an ADV of about half that of a number one overall pick. It doesn't get any better when you start looking at the draft in terms of groupings of 15 players:
From To ADV #PK ADV/P
1 14 164.5803 350 .4702
15 29 88.6675 375 .2364
30 44 49.8499 375 .1329
45 59 42.9250 375 .1145
60 74 42.3412 375 .1129
75 89 28.6731 375 .0765
90 104 25.9692 375 .0693
105 119 18.9467 374 .0507
120 134 19.8437 375 .0529
135 149 14.8779 375 .0397
150 164 19.7008 375 .0525
165 179 18.3312 375 .0489
180 194 14.2743 375 .0381
195 209 11.4120 375 .0304
210 224 15.5153 285 .0544
225 239 11.9733 256 .0468
240 254 8.6951 213 .0408
255 269 5.5697 144 .0387
270 284 2.4601 108 .0228
285 293 2.7809 40 .0695
From and To are the boundaries of the picks (1-14 is picks 1 through 14, etc.); #PK is number of picks. ADV/P is the ADV per pick.
There's some variance by a couple of percentage points after pick 119, but draft value tends to decrease, on a whole, after you reach the equivalent of the end of the fourth round (picks 120 and higher). The anomaly of the .0695 ADV/P for picks 285-293 can be explained simply: two players (Kim Johnsson and Hans Jonsson) who were picked 286th overall in the early 1990's weren't expected to play in North America – and yet they did.
What about our comparable players we mentioned a bit earlier? Which of them actually provided more ADV to their teams? Let's sort them by ADV:
·
- Tyler Seguin, C, Boston, 2010 – 203 GP; .9575 ADV
- Taylor Hall, LW, Edmonton, 2010 – 171 GP; .8066 ADV
- Ryan Johansen, C, Columbus, 2010 – 107 GP; .5047 ADV
- Erik Gudbranson, D, Florida, 2010 – 104 GP; .4906 ADV
- Nino Niederreiter, RW NY Islanders, 2010 – 64 GP; .3019 ADV
- T.J. Brodie, D, Calgary, 2008 – 104 GP; .2766 ADV
- Ryan Ellis, D, Nashville, 2009 – 64 games; .2177 ADV
- Joey Crabb, RW, NY Rangers, 2002 – 170 GP; .2163 ADV
- Troy Bodie, RW, Edmonton, 2003 – 107 games; .1520 ADV
- Cam Stewart, LW, Boston, 1990 – 202 GP; .1165 ADV
Is there a positional difference in the NHL draft? Yes:
- Forwards: 390.1762 ADV, 3575 picked (about 5.2 per team per draft), .1091 ADV/pick
- Defense: 187.7982 ADV, 2044 picked (about 3.0 per team per draft), .0919 ADV/pick.
- Goaltenders: 29.4128 ADV, 644 picked (about 0.9 per team per draft), .0457 ADV/pick.
The difference between forwards and defense is significant, but not that great – both are about the same as an early third round pick. The ADV for Goaltenders is at a point where you'd have to say it's not worth it to take one until the late rounds.
In case you're wondering, the 2003 NHL draft is looking more and more each year like a once-in-a-generation quality draft. The total ADV from players selected in Nashville that year is currently at 43.1932, or an AADV of 1.4398. What that essentially means is that every team in the NHL drafted at least one, possibly two players who have played a significant amount of time in the NHL since then. The 1998 draft is right behind – and that may be more because of who was making the selections (it was Nashville's first draft) and the quality of some of the picks (players like Scott Gomez and Karlis Skrastins, who weren't top selections).
It takes about five or six years for a significant number of players in a particular draft to see NHL playing time. This can easily be proven by noting that the last six drafts have yet to reach an AADV over .8000 (the 2006 draft has the highest with .7629).
What's eye opening is when you see a draft from more than 10 years ago behind a more recent draft; the 2005 draft had a 27.0257 ADV (.9009 AADV); the 1999 NHL draft was a relatively poor 23.5068 (.8395 AADV). Strangely enough, the 1999 draft was significantly worse than the draft prior – and that's not all the fault of the Atlanta Thrashers.
Speaking of which: what team has done better drafting since 1988 – Colorado, Florida or Tampa Bay? The answer might surprise you: Colorado leads the entire NHL with a 30.2683 ADV (1.2107 AADV) in talent over the last 25 years. Florida comes in at 12th overall (18.277, .9139 AADV); Tampa – with only two #1 overall draft picks – has done poorly the rest of the time. They have a 13.6189 ADV, which translates to .6485 AADV over 21 years.
FRN ADV #DR AADV
COL 30.2683 25 1.2107
BUF 27.9896 25 1.1196
SAN 23.7801 22 1.0809
OTT 22.3092 21 1.0623
MON 26.1461 25 1.0458
NJD 25.7940 25 1.0318
NYI 25.6349 25 1.0254
EDM 23.5120 25 .9405
CLB 12.1081 13 .9314
NAS 13.9265 15 .9284
LOS 22.9955 25 .9198
FLA 18.2772 20 .9139
NYR 22.8018 25 .9121
BOS 22.5469 25 .9019
PIT 22.3519 25 .8941
ANA 17.2312 20 .8616
WAS 21.0709 25 .8428
TOR 20.2984 25 .8119
STL 20.0625 25 .8025
PHI 19.9270 25 .7971
CHI 19.7435 25 .7897
DET 19.3684 25 .7747
MIN 10.0544 13 .7734
WPG 10.8192 14 .7728
CAR 19.2576 25 .7703
DAL 19.1192 25 .7648
PHX 19.1091 25 .7644
VAN 18.7570 25 .7503
CGY 18.5076 25 .7403
TAM 13.6189 21 .6485
NHL 20.2462 684 .8861
#DR: Number of Drafts
Of the expansion teams since 1991, the Ottawa Senators have performed the best in the draft, with 1.0623 AADV. Of the younger teams, Columbus has the better AADV (.9314), but Nashville has produced more draft value (13.9265).
Meanwhile, Calgary has not had much luck in drafting players since they won the Cup in '88. They have the lowest ADV (18.5076) and AADV (.7403) of any NHL team that participated in all 25 drafts in the last 25 years.
If you look at the list of the players with the highest ADV for each NHL franchise since 1988, many of the names won't surprise you. (Nail Yakupov isn't on this list, because he has played less than 100 games in the NHL.)
FRN Player Pos YR RD OA GP MAXG ADV
TAM Steven Stamkos C 2008 1 1 373 376 .9920
NYI John Tavares C 2009 1 1 291 294 .9898
CAR Eric Staal C 2003 1 2 690 704 .9801
SAN Patrick Marleau C 1997 1 2 1165 1196 .9741
CHI Patrick Kane RW 2007 1 1 446 458 .9738
FLA Jay Bouwmeester D 2002 1 3 764 786 .9720
LOS Drew Doughty D 2008 1 2 364 376 .9681
WAS Alex Ovechkin LW 2004 1 1 601 622 .9662
BOS Tyler Seguin C 2010 1 2 203 212 .9575
TOR Luke Schenn D 2008 1 5 357 376 .9495
PHI Sean Couturier C 2011 1 8 123 130 .9462
WPG Ilya Kovalchuk LW 2001 1 1 816 868 .9401
PHX Shane Doan RW 1995 1 7 1246 1326 .9397
PIT Mark Recchi RW 1988 4 67 1652 1760 .9386
DAL Jarome Iginla RW 1995 1 11 1232 1326 .9291
NAS Scott Hartnell LW 2000 1 6 875 950 .9211
ANA Cam Fowler D 2010 1 12 195 212 .9198
CLB Rick Nash LW 2002 1 1 718 786 .9135
VAN Henrik Sedin C 1999 1 3 940 1032 .9109
COL Gabriel Landeskog LW 2011 1 2 118 130 .9077
EDM Sam Gagner C 2007 1 6 414 458 .9039
DET Nicklas Lidstrom D 1989 3 53 1564 1760 .8886
OTT Andrej Meszaros D 2004 1 23 533 622 .8569
MIN Nick Schultz D 2000 2 33 811 950 .8537
CGY Dion Phaneuf D 2003 1 9 600 704 .8523
NJD Scott Gomez C 1998 1 27 941 1114 .8447
STL Rod Brind'Amour C 1988 1 9 1484 1760 .8432
BUF Thomas Vanek LW 2003 1 5 585 704 .8310
NYR Todd Marchant C 1993 7 164 1195 1494 .7999
MON Mark Streit D 2004 9 262 491 622 .7894
It's probably no shock to anyone that Rick DiPietro of the Islanders (.3347 ADV) is the biggest "bust" among #1 overall draft picks in the last 25 years. And among players picked in the top five (that actually made it to the NHL), Alexandre Volchkov of the Washington Capitals (1996, 4th overall, .0023 ADV) is probably the biggest bust among first rounders. There's also Thomas Hickey (Los Angeles, 2007, 4th overall) and Ryan Sittler (Philadelphia, 1992, 7th overall), both of which never played a single NHL game.
One last thing about draft value: there have been 6,270 picks made in the NHL entry draft between 1988 and 2012. Three were invalid picks (two in 1990 – Edmonton's 227th overall and Calgary's 220th overall; the third was Edmonton's 123rd pick in 2002), and one forfeited pick (New Jersey's 69th pick in 2011).
Of the rest, 3,863 of the selections have never played a single game in the NHL – 2,998 of those were chosen before the 2008 Entry Draft. Of those who have played at least one NHL game, their average ADV is .2528.
