Thorns FC: Why You Can’t Get No Heat Maps Here

I wanted to post this just because I thought you readers might be interested in seeing behind the scenes of the glamorous soccer-blogging lifestyle.  It’s not just hanging out with the players in trendy nightclubs or hob-nobbing with the wealthy owners.

Sometimes it’s all about the boring, dry stats, baby.  Stuff like this:

That’s from the OPTA stats on the Timber’s match against San Jose back in early June.

OPTA has tons of this sort of great stuff; heat maps, player actions, average team position…while I won’t pretend that “stats tell the story” it helps to get a big picture of a team or a player’s actions, successful or otherwise, to get some sort of sense beyond just your feeling “Gee, seems like Kat Reynolds is getting beat a lot” or “Is Chara pushing up more in the second half, or an I just seeing him more there than when he drops deep?”

As you know, I’ve been using Richard Hamje’s “plus-minus” ratings this season to try and give you a sense of what the Thorns are doing on the pitch.

Actually, I keep more notes that I give you here; the basic numerical PMRs are just the Clif’s Notes version, the digested form, if you will.

(I have no idea whether you care about how I do this, but I’ve come this far so bear with me for a moment and I’ll tell you what goes into these little bunches of numbers.

While I’m reviewing the tape I try and keep track not just of significant touches, both positive and negative, but the type of significant “plus” or “minus” action.

Did Henry hit a great pass, or did it go to an opponent?  Was Long out of position on the Rodriguez goal, or did she tackle her for gain?  How effective were Franch’s goal kicks; did they go to a Portland boot and, if not, how dangerous was the loss?

My first half stat sheet for Henry, say, will have a “+” box with a series of letters, like this:


So in the first half Amandine hit three significantly good passes – that is, passes that were a part of an effective Thorns attack, or defense – including an outstanding pass (P!) in the 71st minute.  She had one good cross (X) in, a good “defensive action” (D), a tackle for gain (Tg), and a good run; +8 for Henry in the first half.

In her “-” box she might have a P (poor pass), L (loss of possession), D (poor defensive position/play), and an S (weak/off-target shot); -4 for the half.

That helps me think about the games, and, I hope, at least publishing the plus-minus numbers helps you, too.

But…I don’t have the sort of skill, I can’t provide the sort of great analytical products that the OPTA people do.  Their match information is a order of magnitude more sophisticated and informative than my little marked-up screenshots and PMRs.

I won’t kid you; I wanted that for this blog.  I wanted to give you more hard information to back up my opinionating.

I wanted heat maps.

So earlier in this year I contacted OPTA to discuss Thorns stats.

One of their staff people got back to me immediately, asking what sort of thing I was looking for.  I said;

“In my analyses I try and evaluate how the individual players (or groups of players) are affecting each match, including the tactical directions (or reactions) by the coaching staff.  The match widgets include some terrific information; pass matrix, heat map, and average position widgets would help my readers see where on the pitch the action took place and who was involved, and both the team and player match stat widgets would be terrific in evaluating both the teams and players for that match as well as the effect of changing matchups between matches.

Perfect example; the last Portland-Seattle match.  The Thorns made adjustments that led to significantly more attacking actions in the second half, but the only immediately visible change on the pitch was the substitution of Klingenberg for Cox.  A qualitative look at the difference between the two suggests that the big difference was Kling’s ability to get forward; Cox looked more like she was pinned into the left corner in the first half.  A heat map could compare the positions of the two, as could the average position widget.  The pass matrix and match stats could tell the reader whether Kling really did pass forward more, and to whom, or not.  The “attacking third” widget would show whether the Thorns’ attack shifted more to their left when Kling entered (since Raso remained largely left-sided throughout the match).

I appreciate all the help.  Thank you!”

The nice person from OPTA took this and thought about it for a while, and, then, got back to me with this:

Hi John,

We are now able to offer NWSL stats through widgets. This is the pricing structure we’ve pulled together based on what you expressed interest in (pricing is on an annual basis):

NWSL Widgets
Widget League Price Price for 2 Price for all 3
Pass Matrix NWSL $8,000 15% discount 25% discount
Heat Map $8,000
Average Position $5,000


Twenty.  Three.  Thousand.  Dollars.

Now I realize that I’d be saving almost $6,000 if I subscribed to all three widgets, but…holy hell.  I can’t afford $17,000 a season.  I have to scrape and save for the couple of hundred bucks for my Thorns GA pass.  Thousands of dollars?  Unpossible.

Would I love to be able to afford these widgets?  Youbetcha.  Will I ever be able to?


So.  I’m afraid you’re stuck with me and my pad of paper with the “+(R) Raso L channel run” and “-C poor clear into touch” notes scribbled on it.

Because the good stuff?

Doesn’t come as cheap as I do.