Tuesday, August 22, 2006

A Helping Hand: Assists in MLS

This past Saturday in the Home Depot Center, the game between the Los Angeles Galaxy and F.C. Dallas changed in a matter of seconds. In the 26th minute, Troy Roberts sent a long ball from his own half to Landon Donovan. Dallas defender Greg Vanney and goalkeeper Dario Sala had a miscommunication, which allowed Donovan to sneak in and chip the ball over Sala's head and into the net.

What's the most amazing thing about that goal? It's not the poor defense or the opportunistic finishing. It's that Pete Vagenas was awarded an assist. This despite Roberts taking a decent run up to the ball before striking it, which again was from his own half of the field, and the ball bouncing next to Vanney and Donovan. I have no idea what Vagenas' contribution actually was on the play, because the highlight clip on MLSnet doesn't even show him. He's also not mentioned on the official recap or in the LA Times story.

Ladies and gentleman, you have now met the secondary assist, one of the last remnants of the "Americanization of soccer." Most of the complaints from that era from hardcore soccer fans were about the things that happened on the field (mainly the shootout), but the statistics were Americanized too.

I'm sure most of you remember the "catch/punch" stat for goalkeepers? MLS only stopped including it in the league stats last season. Ditto for "points" and blocked shots. They still have a "consecutive games with a point" stat, but that's not a big deal. They could word it differently (games with a goal or assist), but the important thing is that the made up stat of "points" is gone. Goals and assists are important, but they stand on their own. You don't need to create an official stat to combine them. This isn't hockey.

Which brings me to assists. My historical knowledge is shaky, but I believe the NASL was the first league to record assists in soccer. It comes from hockey, where the last two players to touch the puck before a goal get an assist. I can't find anything online about it, but I want to say that the NASL gave out two assists for goals as well. This page shows the top 50 all time assist leaders for the NASL, and most have one for every two games. That's basically the same as Carlos Valderrama, Preki, and Marco Etcheverry in MLS.

Anyway, along with the points stat, the NASL brought the assist from hockey to soccer. The difference between the two stats is that the assist is actually a useful addition to the game. FIFA now keeps track of them, and actually has started using them as a tiebreaker for the Golden Boot (which is why Eddie Johnson won it in the 2003 U20 tournament).

So the assist is good. The secondary assist is not though, because it's arbitrary and unnecessary. Rarely are there two very good passes which lead to a goal. Even then, the secondary assist doesn't directly set up the goal, which is what an assist should measure. FIFA and the rest of the world don't use the secondary assist. That's not to say that's why we shouldn't either. Quite simply, the inconsistency of its use in MLS along with the inconsistency in the quality of those passes means that it should be dropped.

I want to explain further what I mean about the inconsistency of its use in MLS. From 1997-2002, secondary assists were given out like candy on Halloween. Regular assists too. This wouldn't be a Climbing the Ladder post without some stats to back up my assertions, and here you go:

Assists Per Goal

Goals Assists A/G
1996 539 505 0.94
1997 522 620 1.19
1998 685 866 1.26
1999 549 738 1.34
2000 612 769 1.26
2001 519 697 1.34
2002 421 565 1.34
2003 433 418 0.97
2004 392 398 1.02
2005 551 557 1.01
2006 347 328 0.95

MLS started ok in 1996, but then things got out of whack. After 2002, assists went back down to a more normal rate. All of the 2006 numbers are accurate through August 20.

Primary/Secondary Breakdown

Assists Primary Secondary Prm-Pct
1996 505 397 108 78.6%
1997 620 427 193 68.9%
1998 866 551 315 63.6%
1999 738 454 284 61.5%
2000 769 498 271 64.8%
2001 697 431 266 61.8%
2002 565 343 222 60.7%
2003 418 300 118 71.8%
2004 398 271 127 68.1%
2005 557 388 169 69.7%
2006 328 238 90 72.6%

Not only were more assists being given in 1997-2002, but more secondary ones as well. Also note that 1996 had very few secondary assists.

There are some discrepancies on MLSnet. I added up the total number of assists from the team stats pages. On the overall league stats pages, there are differences in 1998 (+1), 2001 (-2), and 2004 (+1). It wouldn't really make a difference though. Oh, I also should note that since last year, it appears that MLS doesn't keep track of goalkeeper assists any more (surely less than 10 a year).

Primary/Secondary Assists Per Goal

Goals Prm/G Sec/G
1996 539 0.74 0.20
1997 522 0.82 0.37
1998 685 0.80 0.46
1999 549 0.83 0.52
2000 612 0.81 0.44
2001 519 0.83 0.51
2002 421 0.81 0.53
2003 433 0.69 0.27
2004 392 0.69 0.32
2005 551 0.70 0.31
2006 347 0.69 0.26

Now it's clear how the secondary assist got out of hand. Things have been remarkably consistent since 2003 though, which I'm happy to hear. It dropped in half in one year! Notice the rate of primary assists decreases as well.

In England (and probably FIFA/Europe), assists given out less frequently. Anyone who plays Winning Eleven knows what I mean. One difference there is that they give out an assist to the man who earns a penalty, provided it's not the man who takes it. These are probably good ideas. With all my talk about the secondary assist, it's important not to give out bad primary ones as well.

While I'm glad that MLS has tightened up the awarding of assists (I remember reading about someone who got one for hitting a penalty off the post, which a teammate finished), they need to go further. When guys like Pete Vagenas benefit (and Troy Roberts, for that matter), it hurts the players who have truly earned assists. When you give them out too easily, it makes them less meaningful and a less accurate stat. Let's eliminate the secondary assist, and only consider primary assists when looking at the historical records in MLS. Much like my pet peeve of shootouts being counted as wins, it gives an unfair advantage to those early years. Oh, and they benefited from the whole "assists that aren't really assists" thing too.

Tomorrow, in part two of this post, I'll take a look at the top individual seasons in MLS when it comes to assists, and how they're affected when considering primary assists only.

Comments on "A Helping Hand: Assists in MLS"


Anonymous MetroFanatic said ... (7:37 AM, August 22, 2006) : 

If you want to see current GK assists, see the stat pages on ESPN, such as http://soccernet.espn.go.com/team/squad?id=190&cc=5901

Look under "Outfield Statistics", not "Goalkeeping Statistics".


Blogger Eric PZ said ... (8:01 AM, August 22, 2006) : 

Brillient stuff. I've always hated the whole assist stat as it is pretty random. Yeah, setting up a goal should be recognized, but all too often assists are given like you mentioned above. Why doesn't a guy who makes a run drawing a defender away allowing a goal to be scored also get assists? Personally, I think they should dump the stat...but what do I know. Keep up the great work.


Anonymous Elninho said ... (1:59 PM, August 22, 2006) : 

The drop in primary assists probably was the result of certain things no longer being considered assists. Most notably, until 2002, a shot that was deflected to the scorer (by a defender, a goalkeeper, or the woodwork) was considered to be a pass, and constituted an assist. This is no longer the case. It also seems that a player spending an extended time on the ball (either before scoring or assisting a goal) causes the previous pass to be ignored in awarding assists, though I'm not sure how that's defined.


Anonymous acleansheet said ... (7:40 PM, August 22, 2006) : 

So, Vagenas touched the ball before Roberts I believe, if not then MLS is doing a terrible job.


Blogger scaryice said ... (8:16 PM, August 22, 2006) : 

I know Vagenas touched it. The point is that his contribution was meaningless.


post a comment