Save % Analysis - Part III

Mignolade

Coaches Are Wimps

Part I and part II offered some interesting discoveries when comparing pro-level save percentages to college-level numbers.  The numbers bear out what college coaches will tell you: kids aren't talented enough coming out of club ball to play a possession game based on huge numbers of short passes a la Man City, Arsenal, Barca, Shaktar Donetsk, Ajax and other top clubs.  And that's fair...but it's not the whole story.  It does recall the situation when the Scots traveled to England in 1878 and crushed the Three Lions with short passes, or a few years later when the Hungarians came to Wembley and laid a 6-3 stomping on the hosts, again elevating the bar in terms of technical play.  It is lovely to remind folks that while the English may have invented and spread the game, they have never been the innovators of the game.  English sensibility and pragmatism just doesn't quite cut it.



(OK, Chelsea won the Champions League in 2012 playing as negatively, reactively and ugly as it could be done, so it cuts it, but no one enjoyed watching that)

More recently,  Rinus Michels at Ajax in 1965 (with some player named Cruyff...) revolutionized the game with "Total Football," a system he and Cruyff later took to Barcelona.  And we all know how that turned out. But others have followed suit.  Mircea Lucescu at Shaktar Donetsk, Roberto Martinez and Brendan Rodgers at Swansea, and even our own Caleb Porter with Akron (NCAA) and Portland (MLS) have brought fine versions of short passing games to great heights.  The point being, if we all did what the English did, these names would be meaningless.

Where's the connection between passing and the initial inspiration for these posts, save percentage?

From the Four Four Two Stats Zone app,
powered by OPTA data

One likely explanation is sketched out in the image on the right (happily from one of the more forward-thinking EPL managers- who's French!):

Arsenal completed 84% of their passes in the final third of the field against Fulham on Jan. 18.  Just a shade under 300 attempts. They took 22 shots in the 90 minutes; that's a shot every 14 passes attempted in the final third (they attempted 684 total passes on the day).  See Appendix B at the bottom; smart teams that control the ball may well earn more shots from vital areas.  There isn't a college team in America who could pull that stat off.

More to the point, there isn't a coach in college ball who would ask their players to try (mostly).   Nor are there 18 year old kids who get the connection between patience and quality of chances or have the technical ability to put it into practice if they did.  The primary reason because youth coaches spend time trying to win, not teach...as this space has been screaming to the heavens from day one - if a team wins, mom and dad will never ask what the kid isn't learning, yet again equivalent to caring more about a high SAT score than any actual education taking place...

Let's get back to NCAA DIII.  SUNY IT would rank 7th in the EPL in shots-per-game average...But I bet they wouldn't come close to the Gunners number of completed passes (through 22 games this season - 10,386).  Swansea and Man City round out the top three, all of which are over 10,000 complete passes in 22 matches (click here for the same stat updated as of now).  Handy number, because SUNY IT played 20 games this fall, so it's a fairly apples-to-apples notion.  Indeed, the top 9 EPL clubs in passing attempts are all over 10,000 attempts through 22 matches.  Highly unlikely SUNY IT (or any DIII club) came close - 454 passes per game, on average!  That comes out to roughly 14 passes a minute when in possession...I can't recall the last time I saw a team consistently putting together 14 passes in a row (it happens...but recall the graphic of Arsenal's almost 300 passes in the attacking third V. Fulham, with an 84% completion rate: location, location, location - it matters) It seems clear that college teams shoot on sight, regardless of the quality of the chance.  If 141 GKs are saving at rates superior to some of the world's best then it's certain those chances are not high quality.

How do we make the shots-per-game number stand with the possession numbers....After fifteen years of watching men and women play the game at the college level, it's pretty clear, though my data are limited (I'll dig up some relevant numbers soon and post them - part III) that the college game is very much making watered-down lemonade from lemons.  Lousy defending offers chances, but poor attacking means those chances are not taken.  Many good looks go begging, for sure, but far too many low-probability shots are not spurned for higher probability opportunities.  Which is to say, patience is lacking.  As is soccer intelligence.  Players settle for poor shots because they can't imagine a better chance being fashioned.  So offenses take the chances they're given (rather than the shots they want/create) and hope for the best.

This graphic gives a sense of what the top level requires in this day and age.  Four of the top teams in Europe (and remember that while Arsenal are impressive here, they are merely among 8 or 9 other EPL teams who have passed the ball roughly as much) are clearly playing many, many passes.  Their Goals Against Average is very good (the other guy can't score without the ball...) but even more stark is the number of goals scored from inside 18 yards. How about Bayern at 50 goals from inside and only ONE from outside 18?  Whoa.

For argument's sake, chew on this.  Most college coaches claim to want to keep the ball, but will simply point out that their players aren't good enough (what's that telling us about the youth level?  And isn't it crazy that the college coaches who coach club are sucked up in the vacuum of pressure, perceived or real, to win...).  And that's true in some senses; Xavi and Modric and Pirlo aren't exactly the sort of players we find in NCAA games.  But it's a cop out in other senses.

One good, recent and US-based example is the Portland Timbers.  Not only did Caleb Porter have the University of Akron playing a wonderful brand of football in his time there (right, had the best players in the nation...or did he?) he moved to the MLS and suddenly a dormant, modestly competitive Portland Timbers team becomes a high energy passing team.  To the parenthetical argument a moment ago...no one would claim the Timbers had the best roster in the league.  Will Johnson was the only All-Star for the team.  We might expect their passing game to improve the longer Porter is there and bringing in better players for his preferred style, too.

Over the pond, take a Swansea, a Southampton...teams that are spending fractions of the big clubs, but are staying afloat with both players who you've never heard of and passing numbers that are impressive.  What's telling is that the big clubs don't have noticeably more athletic players (Barcelona are the smallest team in Europe, year after year, remember...) but players who are just that thinnest of margins more technical.  A fraction more accurate, cooler under pressure, better vision, speed of thought, even.  But the point is, you can build teams that mimic Sunderland and Stoke: smash and grab points via gritty defending and countering, or trying to play, to let the players express themselves and make the game a contest of technical skill.

The 10-time DIII national champs, Messiah, as well as other teams like them (Loras, most notably, though many others mimic the Falcons' approach) simply overwhelm their opponents numerically.  Messiah, for example, used 27 players in 10 or more games this past fall.  11 players appeared in 18 or more games without a start, while an additional 3 appeared in 8 or more games without a start.  Mind you, Messiah would be a title contender if they only used 14 players; they do have talent.  But they want to control the game by maximizing the chaos and making the game a test of athleticism.  And that's fine...very effective.  Perhaps they feel forced to do so because the players they get (talented though they are) still aren't good enough to keep the ball and exercise patience.

Are the players that bad, or are the coaches simply selling players short?  Put it another way...would you say your abilities (in any field) have been more over- or under-estimated over the course of your life?  There aren't many managers out there willing to bet their million-dollar jobs on the upside of their players.

The people who benefit are the GKs, who face panicky, easily saved shots and routinely put up SV % numbers that the pros would drool over!

Appendix A:

Appendix B:
Here's a cleaner look at the number of shots needed in the EPL to score a goal from a given area; shots within the width of the 6 and inside 18 yards (green area) generate goals 3X and 5.5X as often as shots from the yellow and red areas (I charted similar effects at a recruiting event in March, here):
 Recall this graphic from the top of the post:
There has to be a reason Arsenal was so patient...and look at the passes into the penalty area (yellow = assist; light blue = shot assist- a pass that leads to a shot but not a goal)...

Appendix C:
Here's the whole EPL from 2012-13...fairly low save percentages (From Here)