Kolek planning to go pro
Johnson is our best and most creative passer in the final third, a guy who was essentially a Bundesliga All-Star on the wing this past season for a Champions League-bound team. There are few players in the world playing that spot at a higher level.Playing him at left back eliminated him as an attacking threat. And because Bobby Wood -- a center forward who runs, but doesn't do a lot of clever positional defending -- was on the left wing, Johnson needed a ton of help from the midfield. That dragged Jermaine Jones out of the center and reduced the US numbers in central midfield, leaving Bradley stranded on an island.Jones wasn't best pleased with the gameplan, saying afterward "We have three midfielders where I feel like if we have a good gameplan, we can put a lot of pressure on other teams. Today was more focused on Colombia and on the left side to help Fabi, so we lost one guy in midfield. So it was tougher to play, yeah."Tactics! Strategy! Formation! They all matter.
So...thoughts based on yesterday...It was certainly an interesting tactical plan. Seemed like we were happy to absorb pressure and counter attack early on, letting Costa Rica dictate the pace of play early on. I'll have to watch the replay today, but the Wood penalty looked like a complete non-foul from where I was sitting right behind that goal. My buddy and I were already yelling at Wood to get up before the ref pointed to the spot. Amazed he gave it. Costa Rica continued to dictate play after that, but needing to at least get a draw, had to go further forward.I know they are renowned for a 5-4-1, but man that defense played like a 3-5-2 yesterday once they started really pushing. The Ticos left huge gaps at the back because their wing backs were bombing forward. Dempsey exploited those holes on a few occasions, though there were numerous other opportunities where he either missed the mark or missed a guy making a run.Forwards: In general, poor, though they accounted for 2 goals. Dempsey was okay and it felt like he was basically playing the role of a distributing 10. Worked well for him, definitely better than his hold-up play. Wood was poor most of the match, but I imagine after drawing a penalty and scoring another quite a few would have him in for man of the match. While those two plays worked out great, on the whole I felt like he offered little. Though not as little as Zardes. Man, what does he do well? Not fast enough to be a winger, terrible first touch, not a great passer. His best plays on the night seemed to be when he was helping out on the back line. He's pretty clearly stuck out of position.Midfielders: Jones had a very good game, though why he's playing as our left flank midfielder still baffles me. He's just not suited to be out on the wing. I think that Jones more than Wood is blocking Fabian moving up the pitch. And it's definitely by design, as he kept returning to that side. But yesterday, he was everywhere. Started a bit slowly, but once he got that first shot that went over, he just kicked it up. Made good runs, good defending, and of course the goal. After being terribly disappointed with his performance Friday, I think he'd be my man of the match. Bradley was also solid. Not spectacular, but held his position well, made a ton of good tackles, really looked comfortable playing a more traditional #6 role. Shocking...he was comfortable in his natural position. Bedoya, on the other hand, was fairly useless as well. He has no chemistry with Zardes and just doesn't seem to grasp his role. Maybe because he's used to playing on the left, maybe something else, but it was no surprise that all our positive attacks seemed to either come off the left or from Yedlin. Zusi looked good in relief and made quite a few positive plays before the goal.Defenders: I was very, very impressed with our back line. Brooks did a great job staying at home and patrolling the center of the pitch. He's starting to take charge of the back line. Got caught out a couple times, but each time Cameron was back there to cover him. I've never been impressed by their partnership in the past, but they were solid yesterday. As were the backs. Both Johnson and Yedlin defended far better than I expected. Fab looked fantastic and had a good sense for when to push up and when to stay home. Also had a great reminder of how blazing fast Yedlin is in the second half when he came from out of nowhere to shut down a Tico on the flank that looked to have a wide open crossing opportunity.Goalkeeper: Guzan was good. Made a few very good saves and made sure his back line kept their form. Nothing overly spectacular.
http://www.mlssoccer.com/post/2016/06/07/armchair-analyst-tactical-preview-usa-vs-costa-rica-copa-americaCouldn't be more correct. Get Fabian out on the wing where he belongs. And get Wood off the wing.
Colombia result good for the US. Draw most likely gets them in.
So I had a thought. Based on your guys description of the USMNT personel and what I saw today, why wouldn't a 3-5-2 work?Dempsey and Wood up top, Bedoya in a CAM role, Bradley and Jones as holding mids and Fabian and Yedlin as wing backs. Solves the LB problem and not having to put Fabian in a deeper role. Plus Dempsey and Wood are better in a 2 striker formation than on the wings or by themselves up top.
Nice analysis, brew. I'm going to also pimp my own stuff here. http://www.lowedownstats.com/index.php/2016/06/08/second-game-changes-things-usmnt/
Henry, love the stats as always...couple of questions for you. You mentioned the offensive finishing and the US getting lucky, which I can't argue with....but how do you differentiate between a bad shot on target that goes in versus a good shot that is off target from a qualitative standpoint? As an example, there was a shot Zardes "Wondo'ed" from 10 yrs out on the right side of the goal. Everything was done perfectly to get in that position and then he squandered that, how is that not unlucky?Not necessarily rated to anything, just curious...what's the driving force for you to create this new site. I think it's interesting and something I'll need to visit more. I'm guessing it's because soccer is one of the last bastion of professional sports to essentially reject the value of advanced metrics and analytics (sabermetrics if you will). Either way, any thoughts as to how that starts to change? Will it require technology changes/acceptance to start to get telemetry on these players, etc?Thanks, I'll hang up and listen to your response.
If you're looking for some supplementary stuff, check out Ted Knutson. He's on Twitter, I think his handle is @mixedknuts. He has some fantastic in-depth analysis of individual players.
Appreciate the questions. I actually wrote about why I started the website on the site itself. If anyone wants to really know a little more about me, there's more there. Even talks a bit about Cracked Sidewalks...http://www.lowedownstats.com/index.php/my-story/The tl;dr version is that I wanted to figure out why the Fire kept getting draws in 2014 and couldn't find any data on it.Most of the data out there is poorly organized and the language of how people talk about soccer is not analytical. I'm trying to approach this to figure out some basic questions regarding teams. Like, is my team actually any good? Do they have a good offense or defense? Why? I'm purposely focusing on high-level type stats to broaden the audience. I'm targeting the tailgate crowd initially and will do random detailed analysis to keep my "nerd cred". Partially, I'm doing high-level because that's the type of data that I have available. It's also simpler to understand. Figuring out which team had better shot quality is a ratio that my fifth grader can solve (4/12 > 2/10). In addition, because soccer stats are in their infancy, we need to get people up to speed on the basics. Or even figure out the basics. Last, I focus on this because those are the types of questions I find interesting. To your question of a bad shot on target vs a good shot off target, that gets to a level deeper than I'm currently presenting. However, there are lots of people that have created expected Goal (xG) models. Most xG models work with the idea that shots from certain areas will go in the net x% of the time. A shot right in the box scores 60% of the time or whatever. One of those shots is worth 0.6 xG. Over time, goals tend to equal expected goals. xG is the "next layer of the onion". Eventually, when I'm ready to pay for the API and hire a developer, I'll probably develop my own xG model or use someone else's.Because people understand the concept here, my goal is to eventually become the Pomeroy for soccer. And again, I'm really doing this because I want the information and it's not out there. If there was already a Pomeroy for soccer, I wouldn't need to put in the effort. But I'm really making a serious go at this.Since I'm sharing, I do have a general request, which people can take offline via PM if preferred. If there's any feedback about the site or the stats or if there are suggestions, please provide them.
If I'm not mistaken, I believe expected goals vs Colombia and CR were almost the same. Crazy.
Were they really? Do you have the numbers?
Just cracking open your site, definitely in the basic mode right now but I can see the potential. Not sure what else you've looked at. I think because of the size and scope of the game, that's why there aren't as many hard analytics.Take Bobby Wood the other night. Drew a penalty, scored a goal, put at least one other "shot" on target. Based on that, most ratings I saw gave him something in the 6.5 to 7.5 range. But watching live...he looked ghastly. Quite a few squandered opportunities and lost balls. His other shot on target was a softball lob straight to the keeper that a 4-year-old could have stopped. Penalty was, in my eyes, dubious at best. The goal was excellent. A true striker's goal where he did all the right things, including keeping his head down and trusting his accuracy, not looking up at the goal until after the strike was en route (are you paying attention, Gyasi Zardes?) to the corner of the net.But all the bad things go unacknowledged so people only focus on the "he scored" aspect and think he played well. I'd love to see more detail of stuff like runs made, offside tendencies, crosses, tackles, tackles without fouling (especially in the box), pass completion percentage...put it all together to bring player ratings in along with the team ratings.One thing that has always kind of bugged me about soccer, and I think it started when I was first playing Winning Eleven 6 (before it became Pro Evo in the States) was how players would be given their game rating. No idea how it calculated why my CMF that I thought did well, won balls, and passed accurately ended up with a 5.5 but the striker that was off target 90% of the time but got one goal ended up with a 7.5 and man of the match.I think a true calculation on the 0-10 scale would be fantastic. Something where you could actually point to the data and not just say "Player X was man of the match" but could statistically demonstrate WHY he was man of the match. Of course, I'm not sure that would be possible without a ton of in-depth information, getting all the passing/tackle/scoring/shot/cross/possession/save statistics and finding a way to balance them between the positions, and figuring out how to balance it for defensive changes that leave guys marking players not typically their responsibility (such as Cameron losing his mark on the corner against Colombia, a defender, though he would typically be marking a forward in open play).Just a few random thoughts. Looks good. Can't wait to see it develop.
It was a tweet. This is all I have.https://twitter.com/PCarrESPN/status/740363746871148544Found this in the responses. Seems like a guy bringing advanced stats to soccer like you.https://twitter.com/MC_of_A/status/740365526304739329
I'm using your Wood comments as a spring board for a deeper thought about analytics in soccer. The Bobby Wood goal was indeed a classic strikers goal, and evidence I would argue(Taylor Twellmen disagrees I guess) that good should be playing centrally, not as a winger. From an analytics standpoint, how do you divorce formation/position from ability/performance? I think statistically(and visually for me at least) I thought Wood and Dempsey had sub par games but all of the traditional stats would say they were good and/or great....but I also think some of the sub par performance is related to the formation at play. I think there needs to be some recognition that with soccer, probably more than any sport, that where/how you are playing on the field can account for a large portion of your ability to perform.Just my rambling for the day.
Anybody going to the game tomorrow?