HSP Data: Primary and secondary assists

[matnor]

I have been compiling some of the Hockey Summary Project data (see http://hsp.flyershistory.com/, it's a truly great resource!) in a database and thought I might share some of the things I have found. Specifically HSP has boxscore data that can be used to compile how many primary and secondary assists players have which, as far as I know, is not available for different time periods. I would generally consider primary assists to me more important than secondary assists and a breakdown into these two categories may illuminate differences between players.

DISCLAIMER:

The data covers the season 52/53 until 09/10 with the exception of 87/88, 88/89, 89/90, 91/92 and 92/93. The data is of high quality but still contains some errors. Therefore individual point totals may not always add up to the correct number, but the difference is generally small. There's also the very real possibility that I have made some coding mistake.

Anyway, here are the results.

The game of hockey has changed over time and also the way stats are recorded. Therefore, assists and the division of primary and secondary assists may not be completely consistent over time. The key stat I will use is the percentage of all assists that are primary (i.e. primary/(primary+secondary)). Here is that stat broken down by different time periods:

Time periods | Share Primary
52/53-59/60|59.0%
60/61-69/70|57.8%
70/71-79/80|57.4%
80/81-86/87|56.9%
90/91-99/00|56.1%
00/01-09/10|54.8%

As can be seen there has been a steady decline in the share of the primary assists meaning that the secondary assists is becoming more frequent. This may say something about how the game has evolved but it may also say something about how the counting of stats have changed. Anyway, it's good to know these trends when looking at individual players.

Obviously the share of primary assists depends on what position the player plays. For instance, we may think that defenders have a lower share. Here is the breakdown by position:

Position | Share Primary
C | 61.5%
RW | 59.7%
LW | 59.6%
D | 46.9%
G | 16.5%

This is pretty much as expected Centers have in general the highest share of primary assists followed by wingers. Defensemen have a much lower share and, surprise, surprise, goalies have almost exclusively secondary assists.

Now for the more interesting stuff, namely which players have an unusually high (or low) share of primary assists. The following list is for forwards with at least 500 assists (in the HSP data with restrictions mentioned in the disclaimer) sorted by their share of primary assists.

Player|Primary assists|Secondary assists | Share primary
Wayne Gretzky | 1037 | 446 | 69.9%
Denis Savard | 401 | 181 | 68.9%
Dale Hawerchuk | 372 | 170 | 68.6%
Henri Richard | 430 | 198 | 68.5%
Darryl Sittler | 426 | 205 | 67.5%
Dave Keon | 396 | 191 | 67.5%
Rick Middleton | 347 | 168 | 67.4%
Mike Bossy | 365 | 182 | 66.7%
Teemu Selanne | 385 | 194 | 66.5%
Doug Gilmour | 404 | 206 | 66.2%
Gordie Howe | 537 | 278 | 65.9%
Joe Thornton | 426 | 221 | 65.8%
Marcel Dionne | 645 | 338 | 65.6%
Butch Goring | 335 | 176 | 65.6%
Dennis Maruk | 333 | 177 | 65.3%
Mats Sundin | 429 | 229 | 65.2%
Bobby Clarke | 552 | 298 | 64.9%
Gilbert Perreault | 529 | 286 | 64.9%
Andy Bathgate | 371 | 202 | 64.7%
Stan Mikita | 596 | 326 | 64.6%
Jean Beliveau | 430 | 237 | 64.5%
Bernie Federko | 383 | 213 | 64.3%
Peter Stastny | 359 | 205 | 63.7%
Phil Esposito | 551 | 315 | 63.6%
Frank Mahovlich | 351 | 201 | 63.6%
Peter Forsberg | 391 | 224 | 63.6%
Ray Whitney | 333 | 194 | 63.2%
John Bucyk | 488 | 285 | 63.1%
Ron Francis | 593 | 347 | 63.1%
Mark Messier | 515 | 302 | 63.0%
Paul Kariya | 359 | 211 | 63.0%
Norm Ullman | 438 | 259 | 62.8%
Mario Lemieux | 350 | 207 | 62.8%
Pierre Turgeon | 327 | 194 | 62.8%
Brendan Shanahan | 326 | 194 | 62.7%
Guy Lafleur | 459 | 280 | 62.1%
Bobby Hull | 316 | 204 | 60.8%
Adam Oates | 424 | 275 | 60.7%
Bryan Trottier | 473 | 307 | 60.6%
Alex Delvecchio | 457 | 303 | 60.1%
Jaromir Jagr | 505 | 335 | 60.1%
Jean Ratelle | 454 | 309 | 59.5%
Rod Brind'Amour | 349 | 240 | 59.3%
Steve Yzerman | 414 | 285 | 59.2%
Sergei Fedorov | 338 | 236 | 58.9%
Tom Lysiak | 322 | 225 | 58.9%
Mike Modano | 377 | 264 | 58.8%
Daniel Alfredsson | 358 | 255 | 58.4%
Vincent Damphousse | 299 | 213 | 58.4%
Rod Gilbert | 354 | 256 | 58.0%
Doug Weight | 388 | 281 | 58.0%
Wayne Cashman | 298 | 218 | 57.8%
Jari Kurri | 294 | 217 | 57.5%
Alex Kovalev | 313 | 233 | 57.3%
Jeremy Roenick | 305 | 232 | 56.8%
Joe Sakic | 439 | 338 | 56.5%
Mark Recchi | 416 | 329 | 55.8%

Note that due to the decreasing share of primary assists over time this tends to give more credit to players from earlier periods. Nevertheless, one thing stands out: The guy with the most assists also has the highest share of primary assists. There is no doubt whatsoever that Wayne Gretzky is the greatest playmaker of all time.

Turning to defensemen, the following list is for those with at least 400 assists:

Player|Primary assists|Secondary assists | Share primary
Brian Leetch | 310 | 235 | 56.9%
Paul Coffey | 425 | 352 | 54.7%
Guy Lapointe | 240 | 207 | 53.7%
Reed Larson | 216 | 192 | 52.9%
Bobby Orr | 331 | 308 | 51.8%
Borje Salming | 293 | 281 | 51.0%
Brad Park | 345 | 335 | 50.7%
Dave Babych | 210 | 204 | 50.7%
Sergei Gonchar | 240 | 235 | 50.5%
Doug Wilson | 218 | 217 | 50.1%
Phil Housley | 298 | 299 | 49.9%
Raymond Bourque | 428 | 430 | 49.9%
Larry Murphy | 319 | 321 | 49.8%
Nicklas Lidstrom | 353 | 359 | 49.6%
Tomas Kaberle | 196 | 205 | 48.9%
Denis Potvin | 339 | 357 | 48.7%
Larry Robinson | 314 | 334 | 48.5%
Roman Hamrlik | 198 | 212 | 48.3%
Rob Blake | 228 | 245 | 48.2%
Scott Niedermayer | 251 | 270 | 48.2%
Doug Mohns | 211 | 232 | 47.6%
Mathieu Schneider | 206 | 231 | 47.1%
Teppo Numminen | 190 | 214 | 47.0%
Al MacInnis | 308 | 349 | 46.9%
Chris Chelios | 242 | 276 | 46.7%
Sergei Zubov | 266 | 314 | 45.9%
Carol Vadnais | 186 | 223 | 45.5%
Scott Stevens | 208 | 251 | 45.3%
Chris Pronger | 228 | 279 | 45.0%

This list is dominated by some of the all-time greats on offence, like Leetch, Coffey and Orr.

Finally, I list the players with the highest share of primary assists by season. The following list is the top-50 of the players with at least 50 assists:

Player|Season|Primary assists|Secondary assists | Share primary
Guy Chouinard | 1978/79 | 45 | 10 | 81.8%
Butch Goring | 1976/77 | 44 | 11 | 80.0%
Dale Hawerchuk | 1983/84 | 51 | 13 | 79.7%
Dale McCourt | 1979/80 | 40 | 11 | 78.4%
Bernie Federko | 1985/86 | 54 | 15 | 78.3%
Dennis Hextall | 1973/74 | 48 | 14 | 77.4%
Joe Thornton | 2002/03 | 51 | 15 | 77.3%
Wayne Gretzky | 1983/84 | 91 | 28 | 76.5%
Paul Kariya | 1995/96 | 39 | 12 | 76.5%
Bernie Federko | 1986/87 | 39 | 12 | 76.5%
Wayne Gretzky | 1986/87 | 92 | 29 | 76.0%
Jean Beliveau | 1963/64 | 38 | 12 | 76.0%
Wayne Gretzky | 1982/83 | 95 | 30 | 76.0%
Ed Olczyk | 1985/86 | 38 | 12 | 76.0%
Denis Savard | 1986/87 | 38 | 12 | 76.0%
Doug Gilmour | 1996/97 | 44 | 14 | 75.9%
Wayne Gretzky | 1981/82 | 91 | 29 | 75.8%
Peter Forsberg | 2000/01 | 47 | 15 | 75.8%
Barry Pederson | 1982/83 | 46 | 15 | 75.4%
Bobby Clarke | 1982/83 | 46 | 15 | 75.4%
Guy Chouinard | 1980/81 | 39 | 13 | 75.0%
Pierre Turgeon | 2000/01 | 39 | 13 | 75.0%
Wayne Gretzky | 1984/85 | 101 | 34 | 74.8%
Andy Bathgate | 1963/64 | 44 | 15 | 74.6%
Mats Sundin | 1996/97 | 38 | 13 | 74.5%
Marcel Dionne | 1975/76 | 40 | 14 | 74.1%
Henri Richard | 1962/63 | 37 | 13 | 74.0%
Denis Savard | 1985/86 | 51 | 18 | 73.9%
Michel Goulet | 1983/84 | 48 | 17 | 73.8%
Craig Janney | 1993/94 | 48 | 17 | 73.8%
Wilf Paiement | 1977/78 | 42 | 15 | 73.7%
Paul Coffey | 1990/91 | 50 | 18 | 73.5%
Kent Nilsson | 1980/81 | 61 | 22 | 73.5%
Denis Savard | 1983/84 | 41 | 15 | 73.2%
Denis Savard | 1982/83 | 62 | 23 | 72.9%
Joe Thornton | 2005/06 | 70 | 26 | 72.9%
Sidney Crosby | 2008/09 | 51 | 19 | 72.9%
Gilbert Perreault | 1976/77 | 40 | 15 | 72.7%
Barry Pederson | 1983/84 | 56 | 21 | 72.7%
Marcel Dionne | 1979/80 | 61 | 23 | 72.6%
Jacques Lemaire | 1972/73 | 37 | 14 | 72.5%
Mike Bossy | 1984/85 | 42 | 16 | 72.4%
Dennis Maruk | 1981/82 | 55 | 21 | 72.4%
Mike Bossy | 1985/86 | 44 | 17 | 72.1%
Ron Francis | 2001/02 | 36 | 14 | 72.0%
Robbie Ftorek | 1980/81 | 36 | 14 | 72.0%
Don Maloney | 1979/80 | 36 | 14 | 72.0%
Darryl Sittler | 1979/80 | 41 | 16 | 71.9%
Frank Mahovlich | 1971/72 | 38 | 15 | 71.7%
Daniel Alfredsson | 2005/06 | 43 | 17 | 71.7%

I'll leave the analysis to you but I think this may help shed some light on which players are more responsible for helping their team score. If there is one particular player that you want to see the stats for feel free to ask, now that I have the data sorted it's pretty easy to get.

Edit: Overpass pointed out that the 1957/58 data was corrupted. I have therefore updated and excluded that year.

Click to read more...

 

[overpass]

Good stuff. I'll look at it and respond more later.

Data quality note: I've looked at this before and the last time I checked the HSP data for the 1957-58 season listed assists in alphabetical order, not as primary-secondary. Is that still the case?

 

[Bear of Bad News]

Very interesting analysis - thank you for putting this together.

 

[jkrx]

Don't really understand the obsession with primary assists. To me the secondary (in context) is equaly important as the primary.

 

[LeBlondeDemon10]

Don't really understand the obsession with primary assists. To me the secondary (in context) is equaly important as the primary.

Agreed. And sometimes the player who creates the scoring chance does not get an assist due to numerous rebounds, etc...Very interesting data though. Good post.

 

[reckoning]

Excellent research matnor

The most interesting part is that the spread between the percentages of the top and bottom forwards isn't that much. It looks like there isn't any player who had their assist totals artificially inflated due to an extraordinary amount of secondary assists, despite the claims of many on the main board here.

 

[matnor]

Good stuff. I'll look at it and respond more later.

Data quality note: I've looked at this before and the last time I checked the HSP data for the 1957-58 season listed assists in alphabetical order, not as primary-secondary. Is that still the case?

Thanks a lot, you're absolutely right, I should have noted that. That year is now excluded.

 

[ContrarianGoaltender]

These numbers certainly seem to support the argument that Mats Sundin had weak linemates.

The Forsberg vs. Sakic split is also kind of interesting. Do you by any chance happen to have the numbers broken down by game situation? It would be interesting to see the numbers for those two guys on the power play in Colorado, where they often would have been playing together. Without any evidence to back it up, my guess is that maybe Forsberg was more often the primary playmaker on the power play, which meant that Sakic was perhaps more likely to either score or get the second assist.

I think these numbers are perhaps most useful in comparing players who played together, to try to determine who was really driving the play. For example, was Ron Francis' second assist rate higher in Pittsburgh than elsewhere? If so, that supports the idea that his scoring numbers were helped out by Jagr.

 

[unknown33]

Don't really understand the obsession with primary assists. To me the secondary (in context) is equaly important as the primary.

Okay then try to look up some random goals (e.g 2010 playoffs) and count how often the 2nd assist is more/equally important than the primary assist to create the goal.

 

[Scott1980]

THANK YOU!

Now to show some jerk I know about how many of Gretzky's assists were, "cheap, second assists".

 

[SidGenoMario]

Gretzky has 1000 primary assists... just when you think you know just how good he is, new stats keep popping up and making him more impressive. Wow.

 

[Big Phil]

Nice work, well done. I will say though I am not a fan of primary or secondary assists by any means. If you registered an assist you helped cause the play that resulted in a goal. If a guy had 3 assists in a game but they were all "secondary assists" didn't he still have a good game more than likely and was a big cause of his team's goals?

What about the tic tac toe passing plays? Or the pinpoint breakout pass a guy like Pronger or Coffey or Robinson or whoever did back in the day? More than likely that ends up being a secondary assist yet probably the most important play of the goal.

I mean here's an example. John Tonelli got the secondary assist on the famous goal from Bossy in the 1984 Canada Cup. Coffey of course originally broke up that play and skated end to end but Tonelli did the work in the corners, outmuscled a Russian and fed the puck to Coffey at the point who shot and had the puck deflected in by Bossy. Hey, give full credit to Bossy and all for the deflection but in my mind he had the least to do with that goal out of the three.

Of course we see it go the other way too, which is why I always cancel out tracking primary assists

 

[Trottier]

Don't really understand the obsession with primary assists. To me the secondary (in context) is equaly important as the primary.

Agree 100%.

Interesting statistical compilation and high marks for the work.

This poster, however, draws no conclusions from it, certainly not that secondary assists are "inferior" or less important. (Not suggesting that the thread opener is implying as such.) To be sure, in certain goal scoring situations they are less integral to the end result. But the reverse is also true. Likewise, the third pass prior to a goal, never to be registered on the stat sheet, can be the key component leading to a goal.

Play and watch the game and one tends to refrain from such silly phrases like "cheap assists" and "he leeches".

 

[Hardyvan123]

Agree 100%.

Interesting statistical compilation and high marks for the work.

This poster, however, draws no conclusions from it, certainly not that secondary assists are "inferior" or less important. (Not suggesting that the thread opener is implying as such.) To be sure, in certain goal scoring situations they are less integral to the end result. But the reverse is also true. Likewise, the third pass prior to a goal, never to be registered on the stat sheet, can be the key component leading to a goal.

Play and watch the game and one tends to refrain from such silly phrases like "cheap assists" and "he leeches".

Fully agree here and Brian "who never Leetch's" headed up the Dman list.

Interesting stuff but nothing can compare to actually watching great players in action, which is impossible to do for every player I know.

 

[RabbinsDuck]

I am not very skeptical of secondary assists from elite players, though your average player might rack them up just from playing with great linemates.

I am never skeptical of secondary assists coming from defensemen.

 

[BraveCanadian]

Very interesting to see the assists broken down this way.

Thanks for this matnor.

I don't put an awful lot of stock into separating primary and secondary assists myself because if you don't see the goal you really don't know which one was more important.. but this is still very interesting.

And man.. look at Gretzky. Wow.

I wonder if we'll ever see a playmaker even close to his calibre again.

The always underrated Hawerchuk does really well in this too. That guy was so slick its too bad his hip slowed him down and made him retire early.

I'm a little surprised to see Mike Bossy up so high considering he is always considered much more of a trigger man. That is an awful lot of assists to be explained away as rebounds.

Also strikes me as odd how far down Guy Lafleur is in comparison when he was supposed to be the much better playmaker than Bossy.

 

[Rhiessan71]

The always underrated Hawerchuk does really well in this too. That guy was so slick its too bad his hip slowed him down and made him retire early.

^This bigtime.
It floors me sometimes how badly I get ripped by some people when I state my opinion that Dale was the best center in the league in the 80's not named Gretzky or Lemieux.


Great work btw Matnor

 

[BraveCanadian]

^This bigtime.
It floors me sometimes how badly I get ripped by some people when I state my opinion that Dale was the best center in the league in the 80's not named Gretzky or Lemieux.


Great work btw Matnor

Hawerchuk is definitely right up there. He was amazing.

You could easily make the case he is up there with Stastny and Yzerman and that level of guy behind Gretzky and Lemieux.

Dale just had the misfortune of playing on a team that was weaker than the Oilers and Flames were during his best seasons.

 

[TheDevilMadeMe]

I'm a little surprised to see Mike Bossy up so high considering he is always considered much more of a trigger man. That is an awful lot of assists to be explained away as rebounds.

Also strikes me as odd how far down Guy Lafleur is in comparison when he was supposed to be the much better playmaker than Bossy.

I think t makes perfect sense in fitting in with the stereotypes of these two guys. If Bossy shoots and someone else puts in the rebound, it's a primary assist.

Likewise, a lot of the offense that is captured by secondary assists is that of guys who control the puck for a long time before ditching it off to teammates - exactly how Lafleur used to play.

 

[Derick*]

Thanks for compiling this. I'd be interested in is a measure of "goal shares," where one point is awarded for an unassisted goal, 0.5 each for a goal with one assist, and 0.33 each if it had two assists. Part of the reason I value primary assists is that some of them were the only assist, while all secondary assists are one of three points being awarded for one goal.

Pretty cool that Joe Thornton's nutty assist total in 2006 was also one of the highest primary assist shares of all time. He had more primaries that year than anyone else had assists.

 

[matnor]

Nice work, well done. I will say though I am not a fan of primary or secondary assists by any means. If you registered an assist you helped cause the play that resulted in a goal. If a guy had 3 assists in a game but they were all "secondary assists" didn't he still have a good game more than likely and was a big cause of his team's goals?

What about the tic tac toe passing plays? Or the pinpoint breakout pass a guy like Pronger or Coffey or Robinson or whoever did back in the day? More than likely that ends up being a secondary assist yet probably the most important play of the goal.

I mean here's an example. John Tonelli got the secondary assist on the famous goal from Bossy in the 1984 Canada Cup. Coffey of course originally broke up that play and skated end to end but Tonelli did the work in the corners, outmuscled a Russian and fed the puck to Coffey at the point who shot and had the puck deflected in by Bossy. Hey, give full credit to Bossy and all for the deflection but in my mind he had the least to do with that goal out of the three.

Of course we see it go the other way too, which is why I always cancel out tracking primary assists

I agree completely that you can't say that primary assists are always better than secondary assists. And one should be very careful comparing different players because a lot has to do with their style. For instance, some players are experts at putting the last pass on the tape and get a lot of primary assists. Other players may be very good at carrying the puck in the zone and establish puck possession. That may be just as effective but lead to a lower share of primary assists. These stats are not supposed to tell a complete story but rather be used in context with the environment each player played in.

That being said, I do believe that on average the primary assist is more important and I'll try to illustrate how these stats can help understanding this. Take a look at Gretzky's stats:

Season |Primary assists|Secondary assists|Share primary
1979/80 | 60 | 26 | 69.8%
1980/81 | 69 | 40 | 63.3%
1981/82 | 91 | 29 | 75.8%
1982/83 | 95 | 30 | 76.0%
1983/84 | 91 | 28 | 76.5%
1984/85 | 101 | 34 | 74.8%
1985/86 | 109 | 54 | 66.9%
1986/87 | 92 | 29 | 76.0%
1990/91 | 80 | 43 | 65.0%
1993/94 | 55 | 37 | 59.8%
1994/95 | 24 | 13 | 64.9%
1995/96 | 49 | 23 | 68.1%
1996/97 | 46 | 19 | 70.8%
1997/98 | 42 | 21 | 66.7%
1998/99 | 33 | 20 | 62.3%

During his prime years he also had his highest share of primary assists (as a sidenote, it's crazy that he has a season with more primary assists than any other player not named Mario has had total assists). As he declined with age and injury his primary assist share also went down. However, there is one exception, his increase to the record-setting 163 assists from (the previous record of) 135 assists was largely driven by an increase in secondary assists that he never came close to again. It seems to me that he had a little bit of luck achieving the 2-assists-per-game mark (definitely no discredit to him).

It can also go the other way. Take a look at Lidström's numbers:

Season |Primary assists|Secondary assists|Share primary
1993/94 | 23 | 23 | 50.0%
1994/95 | 4 | 10 | 28.6%
1995/96 | 19 | 24 | 44.2%
1996/97 | 16 | 21 | 43.2%
1997/98 | 22 | 19 | 53.7%
1998/99 | 17 | 26 | 39.5%
1999/00 | 23 | 30 | 43.4%
2000/01 | 32 | 24 | 57.1%
2001/02 | 29 | 22 | 56.9%
2002/03 | 23 | 21 | 52.3%
2003/04 | 18 | 10 | 64.3%
2005/06 | 34 | 30 | 53.1%
2006/07 | 25 | 24 | 51.0%
2007/08 | 25 | 35 | 41.7%
2008/09 | 24 | 19 | 55.8%
2009/10 | 19 | 21 | 47.5%

Again we see the pattern that during his prime years the primary assist share goes up and is stable above the .5 mark. But here we have an outlier going in the other direction. Lidström's abysmal (offensively, by his standards) 03/04 campaign was driven primarily by a decrease in secondary assists. It looks as he was, partly, a little unlucky that season.

Finally, some may find it a bit curious that Oates, one of history's finest playmakers is on the bottom half of the list. Breaking down his data by season give us an explanation of why that is.

Season |Primary assists|Secondary assists|Share primary
1985/86 | 7 | 4 | 63.6%
1986/87 | 22 | 9 | 71.0%
1990/91 | 52 | 37 | 58.4%
1993/94 | 52 | 27 | 65.8%
1994/95 | 26 | 12 | 68.4%
1995/96 | 42 | 21 | 66.7%
1996/97 | 35 | 22 | 61.4%
1997/98 | 32 | 19 | 62.7%
1998/99 | 25 | 16 | 61.0%
1999/00 | 32 | 23 | 58.2%
2000/01 | 37 | 32 | 53.6%
2001/02 | 36 | 28 | 56.3%
2002/03 | 17 | 18 | 48.6%
2003/04 | 9 | 7 | 56.3%

The reason is that the data is missing from his prime years. Around those years his share is hovering around the 65% mark. The seasons that lower his average are the seasons on the tail end of his career. And, again, his outlandish 90 in 60 assists was driven mostly by an increase in secondary assists.

 

[BraveCanadian]

I think t makes perfect sense in fitting in with the stereotypes of these two guys. If Bossy shoots and someone else puts in the rebound, it's a primary assist.

Likewise, a lot of the offense that is captured by secondary assists is that of guys who control the puck for a long time before ditching it off to teammates - exactly how Lafleur used to play.

I don't know about that. That sounds more like trying to jam this data into the stereotype.

Obviously context is very important when evaluating the primary and secondary assists but there are some strange anomalies like this one and also Sundin never struck me as an elite playmaker. I always thought of him as a better scorer than playmaker.

 

[matnor]

I don't know about that. That sounds more like trying to jam this data into the stereotype.

Obviously context is very important when evaluating the primary and secondary assists but there are some strange anomalies like this one and also Sundin never struck me as an elite playmaker. I always thought of him as a better scorer than playmaker.

I think TDMMs explanation makes a lot of sense. For instance, a sniper like Ovechkin has a high share of primary assists and I would suspect it is due to a lot of rebounds.

 

[jkrx]

Okay then try to look up some random goals (e.g 2010 playoffs) and count how often the 2nd assist is more/equally important than the primary assist to create the goal.

Comes to mind...

We also have defensemen's breakout passes which usually results in secondary assists (if they even become assists at all). Every good pass is vital and as much important as the last pass.

 

[matnor]

These numbers certainly seem to support the argument that Mats Sundin had weak linemates.

The Forsberg vs. Sakic split is also kind of interesting. Do you by any chance happen to have the numbers broken down by game situation? It would be interesting to see the numbers for those two guys on the power play in Colorado, where they often would have been playing together. Without any evidence to back it up, my guess is that maybe Forsberg was more often the primary playmaker on the power play, which meant that Sakic was perhaps more likely to either score or get the second assist.

I think these numbers are perhaps most useful in comparing players who played together, to try to determine who was really driving the play. For example, was Ron Francis' second assist rate higher in Pittsburgh than elsewhere? If so, that supports the idea that his scoring numbers were helped out by Jagr.

Splitting the data by EV and PP is a good point that I thought about. Comparing only the time they played together (94/95-03/04) Forsberg had a primary assist share of 61.1% during powerplay and 64.7% otherwise. Sakic, on the other hand, had only a 47.3% share during powerplay but 63.9% otherwise so your hypothesis was absolutely right.

As for Francis, here's his numbers:

Season|Primary assists| Secondary assists|Share primary
1981/82 | 25 | 18 | 58.1%
1982/83 | 42 | 18 | 70.0%
1983/84 | 39 | 19 | 67.2%
1984/85 | 36 | 21 | 63.2%
1985/86 | 35 | 18 | 66.0%
1986/87 | 41 | 22 | 65.1%
1990/91 | 40 | 24 | 62.5%
1993/94 | 36 | 30 | 54.5%
1994/95 | 31 | 12 | 72.1% 1995/96 | 41 | 43 | 48.8% 1996/97 | 30 | 26 | 53.6% 1997/98 | 28 | 24 | 53.8%
1998/99 | 19 | 12 | 61.3%
1999/00 | 36 | 13 | 73.5%
2000/01 | 35 | 14 | 71.4%
2001/02 | 36 | 14 | 72.0%
2002/03 | 23 | 12 | 65.7%
2003/04 | 20 | 7 | 74.1%

You seem to be right here too. His primary assist share went down during Jagr's (and Lemieux') prime years with the exception of 94/95.