Possession in the CONFED Cup

[Martin Fischer]

Lots has been made of the USMNT's anti-possession style. Since I remember Nutmeg calling for the US to abandon the possession style a year ago, the question raised in my mind whether this is an accurate assessment of the USMNT's style.

Official possession numbers in the CONFED Cup are as follows:

1. Brazil 57%
2. Colombia 55.3%
3. Japan 52.3%
4. France 49.7%
5. Turkey 49.3%
6. Cameroon 48%
7. USA 47.3%
8. New Zealand 41%

In general, these numbers support the fact that the US was not good at possession in this tournament.

However, it should be noted that:

1. Every team that played Brazil had possession below 50% because Brazil is so damm good at possession -- you simply cannot pressure them as they can control the ball against almost any team's top pressure.

2. The US's posession numbers are very close to those of the two teams that advanced in its group.

3. New Zealand was far worse than the US in possession.

4. The best possession team in the tournament did not advance.

[rollo]

I think the USMNT gave up possesion as a matter of tactics. I actually found the Brazil game very exciting right up till the gaff (ironically Berhalter was the one talking about needing to keep mental concentration just before the Turkey game). I think Arena had visions of WC02 vs Mexico before the game started and it was working fairly well. I would bet that BA thinks about the Italians in Euro00 and smiles with delight and then thinks about us in WC02 and grins with satisfaction. Well enough projection.... My point being that as you correctly pointed out, possesion statistics can be a complicated thing to understand and easily taken out of context. What went wrong against Brazil, I thought keeping Ronaldinho out of plays as much as possible was a major goal and working somewhat, Adriano was dangerous and had to be limited in his chances in the box, and the gaff may be partly due to Greg's lack of confidence in working with Gibbs on the left and did a mental hesitation that was severly punished.

[Nutmeg]

Originally posted by Martin Fischer
Lots has been made of the USMNT's anti-possession style. Since I remember Nutmeg calling for the US to abandon the possession style a year ago...

Actually, Martin, I called for the US to make a shift towards more direct soccer, but never once said we should abandon possession. That seemed to work OK in the World Cup.

I now believe that we have gone too far in the other direction, but I believe that happened particularly against Turkey and Brazil because we didn't have a guy in the middle of the field, like Reyna, who would help us maintain possession. Martino started against Cameroon, and we found a good balance as a team.

I've mentioned this on other threads, too, but I strongly believe that successful teams have a good balance between possession and direct attack.

Too much in either direction and you're asking for trouble. A lot of it comes down to the players and their ability, particularly in the central midfield.

[Parmigiano]

Originally posted by Nutmeg
Actually, Martin, I called for the US to make a shift towards more direct soccer, but never once said we should abandon possession. That seemed to work OK in the World Cup.

I now believe that we have gone too far in the other direction, but I believe that happened particularly against Turkey and Brazil because we didn't have a guy in the middle of the field, like Reyna, who would help us maintain possession. Martino started against Cameroon, and we found a good balance as a team.

I've mentioned this on other threads, too, but I strongly believe that successful teams have a good balance between possession and direct attack.

Too much in either direction and you're asking for trouble. A lot of it comes down to the players and their ability, particularly in the central midfield.

I agree with Nutmeg. It's a balance. Look at Japan against Colombia. They were very entertaining, with a lot of skill and possession. But I couldn't help but think how easily they could have won that game if they had just let Colombia have the ball a little more and quit pretending to be able to play like Brazil. With their skill and speed, they could have easily surprised the Colombians on the break, and they would have advanced somewhat like the US against Mexico at the WC.

[Martin Fischer]

Nutmeg -- I never said you said the USMNt should abandon possession. I said that you said the USMNT should abandon the possession style -- in orther words, move away from a style that was overly focused on possession to one that is more focused on direct play.

[JohnR]

Originally posted by Nutmeg
Martino started against Cameroon, and we found a good balance as a team.

I didn't see the game (no Galavision) but as I understand it Cameroon didn't really give a damn about possession. Two wins, advancement to the next stage guaranteed, they were happy enough to let us back pass all day long.

I'm a defender of soccer statistics but possession in particular is a tricky statistic to interpret.

To me, the real issue is not the actual possession statistic, which depends heavily upon the score, the situation, and the other team's attitude, but a more basic question -- can we maintain it if the other team doesn't want us to maintain possession?

For Brazil, the answer almost always is Yes. Which goes a long way toward explaining 3 World Cup finals during the past 3 World Cups. For the U.S, of course, the answer is usually No against top opposition.

Take the '02 Italy friendly. Italy came out a bit sluggish and let us take the game to them in the first half. Got booed at helftime and chewed out by the manager. Came out in the second half with a different attitude. That was it for our possession -- they took the ball back, baby.

I'd like for Italy to struggle if they want the ball from us. Right now, we're not there. In all honesty, I doubt we could have held the ball very much against Cameroon if Cameroon had approached Game #3 like it did Games #1 and #2.

[appoo]

Originally posted by JohnR
I'd like for Italy to struggle if they want the ball from us. Right now, we're not there. In all honesty, I doubt we could have held the ball very much against Cameroon if Cameroon had approached Game #3 like it did Games #1 and #2.

When we had a midfield of Reyna at A-Mid, Armas at D-Mid, a spry Stewart at R-Mid, and JOB at L-Mid the Italians were basically defending for an entire half against us. They rarely touched the ball

[Karl K]

Originally posted by Martin Fischer
Nutmeg -- I never said you said the USMNt should abandon possession. I said that you said the USMNT should abandon the possession style -- in orther words, move away from a style that was overly focused on possession to one that is more focused on direct play.

Martin, first of all, thanks for those numbers. Interesting.

Second, like JohnR, I am in favor of more revealing statistics in soccer. It's arguable whether possession % is one of them.

I have been thinking a lot about this, and I am coming around to the belief that some measure of "scoring chances/opportunities" might prove particularly useful in the context of possession %. Not shot/shots-on-frame necessarily, but instances of really dangerous advancement of the ball that allow, say, and difficult-to-handle cross or a combination play inside that area that forces a desperate clearance by the defense.

This way, if a team dominates possession, but has only a few more, or basically the same number, of "scoring chances" than you had, then the fact you conceded possession is less of an issue. But if you have conceded possession, and they are constantly knocking on the door, then you have a serioius tactical problem, or simply are just not as good as the other guy.

Finally, on the issue of playing "style." After WC '02, I think we shouldn't necessarily commit ourselves to a single "style" of play. I think the manner in which we play must be predicated upon the opponents we face, and the individual player matchups those opponents present. Thus a counterattacking style may be more appropriate against a Mexico or a Brazil, a midfield possession game more appropriate against a El Salvador or a Jamaica.

Moreover, player availability and venues may also dictate style of play. A wide Saprissa, along with a hostile crowd, may forces us into a more defensive posture, particulary if we don't have O'Brien or Reyna available.

Actually, I beleive that "style flexibility" should be our hallmark. Play a certain way consistently, and you become predictable, and opponents can counter your style with their own adjustments.

Better to keep the other side guessing.

[Scotty]

Originally posted by JohnR
In all honesty, I doubt we could have held the ball very much against Cameroon if Cameroon had approached Game #3 like it did Games #1 and #2.

Exactly.

And we are seeing way too many people around here putting too much stock in our performance from this game.

We had a good possession game against Cameroon mainly because they allowed it to happen. They were already through so they fielded their reserves. And they were content to play a counter attacking game.

Look at the way Bocanegra and Califf were allowed so much time and space on the ball. They would receive it, look around, and then they could basically pass it to whoever they wanted. Cameroon was only pressuring in their own half.

Hence our possession.

[JohnR]

Originally posted by appooOnU
When we had a midfield of Reyna at A-Mid, Armas at D-Mid, a spry Stewart at R-Mid, and JOB at L-Mid the Italians were basically defending for an entire half against us. They rarely touched the ball

Yeah, and those same guys also played in the second half, except that the Italians had the ball, once they got serious.

[lurking]

Our problem wasnt a lack of possession but a lack of effective offensive ideas, specificly off the ball movement. We actually had the ball too much at times as we played pointless square ball after square ball. Note that just above us was group winner Cameroon, while at the top was Brazil, who are already going home. In fact, only Columbia saw more of the ball then its opponents and advanced.

[Martin Fischer]

I agree that possession stats are not meaningful in terms of telling you who is the better team. That is borne out by comparing possession statistics and results in this tournament.

But possession stats are very meaningful in telling you how good your team is at holding the ball. While John R is right that the score and the opponent affect this statistic as much as pure ability and desire to hold the ball, over time the score and the opponents even out which leaves the statistic as a pretty good measure of your ability and desire to hold the ball.

In fact, even in the three game first rounds, the statistics pretty well, IMHO reflected my opinion of these teams' ability and desire to hold the ball. Brazil is the best possession team in the world, and it is hardly surprising that it was by far the best team in the Confed Cup in terms of possession percentage (Colombia's stats were buttressed by the fact that it held the ball 67% against New Zealand, which was a factor of the fact that New Zealand is a bad possession team and that NZ was actually ahead for 30 plus minutes in that game).

[appoo]

Originally posted by JohnR
Yeah, and those same guys also played in the second half, except that the Italians had the ball, once they got serious.

Actually they pur in four fresh players and then basically owned the last 20 minutes of the match while our players tired out. We're not nearly as deep as the Italians are

[Nutmeg]

Originally posted by Scotty
And we are seeing way too many people around here putting too much stock in our performance from this game.

We had a good possession game against Cameroon mainly because they allowed it to happen.

Cameroon played the US largely how they played Turkey and Brazil. In all three games, they were a counter team. The difference, as you mentioned was their intensity and the players on the field.

Still, you can only rate players based on the games they played in. Would Martino have made a huge difference against Brazil? Who knows, it would lead to a very circular conversation...

[dchang44]

Originally posted by Karl Keller
I have been thinking a lot about this, and I am coming around to the belief that some measure of "scoring chances/opportunities" might prove particularly useful in the context of possession %.
karl, i like this idea; it gets away from the idea of measuring possession in terms of time, but it would be a more meaningful metric about offensive effectiveness. the possession statistic is not meaningful if the other team lets your defenders play square balls to each other.

another idea would be to measure possession only when your player is being challenged (ie, somehow measure which team can withstand challenges and hold onto the ball, or pass cleanly to a teammate while under pressure).

nah, soccer doesn't need statistics...

[Nutmeg]

Martin,

Your thread here made me wonder how these possession statistics measured against those of the last World Cup. I went on a hunt, and found the 2002 World Cup statistics page here (http://fifaworldcup.yahoo.com/en/t/st/index.html). Unfortunately, possession was not one of the statistics that was tracked, but there were a number of interesting numbers posted there.

Take a look around if you get some time. I found the “Team Attack” and “Team Defense” sections particularly interesting. Immediately I noticed that there were some categories where the US stacked up fairly well, and some where we were behind the best teams in the World Cup.

One thing I noticed right away is that we fell far behind the leader, Germany, in Total Shots (TS), but that statistically we were much better when you compare Total Shots and Total Shots On Goal (TSOG). Germany had a total of 100 shots, but only 40 of those were on goal, a rate of 40%. The US had only 54 Total Shots, but 30 of those were on goal, a rate of 55%. Brazil had a 58% rate, the best of the top 8 finishers.

What I would find interesting is to compare those statistics against possession. That would give a solid statistical view of how possession equates into quality opportunities – or at least shots on goal.

Another area that I found interesting was in the defensive department, the top 3 finishers were also the top 3 in intercepting an opponent’s pass (TTCP) during the tournament. Also, the top 4 finishers were also the top 4 in tackling an opponent who receives the ball (TTCG). The US wasn’t as strong in these areas, which leads me to believe that one of the things that separated the US from the 2002 top dogs was that we were not able to anticipate our opponent’s passing patterns and react as quickly as did Brazil, Germany, Turkey, and Korea.

I found these stats to be really interesting. Take a look around when you get a chance.

[lurking]

Here are the stats for the tournament:


Team.........Sh...SG...PK...G
Cameroon.....28...15....1...2
Turkey.......34...19....1...4
Brazil.......34...13....0...3
USA..........22....8....0...1
Japan........34...16....0...4
Columbia.....44...10....0...4
France.......37...10....2...8
New Zeland...22....8....0...1


Id say SOG are generally the best indicator normally, but in small periods you can have strange conversion rates. Typically youd expect around 20-30% depending on the quality of the goalkeeping. We gave up 17 on the tournament for reference. Cameroon 10, Brazil 12 and Turkey 16. That would make it Cameroon +5, Turkey +3, Brazil +1 and the US -9. The other group had some weird stat keeping going on, as Columbia and France were only reported as having 1 SOG against NZ which makes no sense.

[Martin Fischer]

This is already interesting. Blind reliance on statistics is bad, but if you take time to consider what the statistic means and how it was collected, it can help your analysis.

In this case, though possession was not tracked at the World Cup, I wouldn't be surprised if our possession percentage was the same or even lower in the World Cup. Could be wrong, but I don't think our possession was any worse at the Confed Cup.

In the World Cup, the US averaged 10.8 shots per game. In the Confed Cup, that dropped to 7.3 shots per game. In the World Cup, the US averaged 6 shots on goal per game, in the Confed Cup that dropped to 2.67. The percentage of shots on goal dropped from 55% to 36%.

What does this tell us? Well, what we already knew -- namely that the US didn't play well in this tourney. The statistics would be more helpful if, for example, shots on goals had decreased but total shots had remained constant. In that circumstance, there would be an indication that the problem was in the last 1/3 of the field, either poor finishing or worse chances being created. Alternatively, if total shots had dropped, but the shots on goal percentage remained the same, that would indicate that the attackers were effective, but the service was not frequent enough.

The way it was, we sucked all the way along the chain required to play good soccer -- according to what I saw and what the stats tell us.

[appoo]

Originally posted by Martin Fischer
This is already interesting. Blind reliance on statistics is bad, but if you take time to consider what the statistic means and how it was collected, it can help your analysis.

In this case, though possession was not tracked at the World Cup, I wouldn't be surprised if our possession percentage was the same or even lower in the World Cup. Could be wrong, but I don't think our possession was any worse at the Confed Cup.

In the World Cup, the US averaged 10.8 shots per game. In the Confed Cup, that dropped to 7.3 shots per game. In the World Cup, the US averaged 6 shots on goal per game, in the Confed Cup that dropped to 2.67. The percentage of shots on goal dropped from 55% to 36%.

What does this tell us? Well, what we already knew -- namely that the US didn't play well in this tourney. The statistics would be more helpful if, for example, shots on goals had decreased but total shots had remained constant. In that circumstance, there would be an indication that the problem was in the last 1/3 of the field, either poor finishing or worse chances being created. Alternatively, if total shots had dropped, but the shots on goal percentage remained the same, that would indicate that the attackers were effective, but the service was not frequent enough.

The way it was, we sucked all the way along the chain required to play good soccer -- according to what I saw and what the stats tell us.

Team statistics are indeed very telling. I think they give you a good idea about what kind trends you see in a certain team. Just glad this conversation is statying away from personal stats, something I believe has no place in soccer. Is there a statistic for SOG without sustained possesion, and SOG with sustained posession? I'd think that would be really telling of just what kind of Soccer Arena had our boys playing during world cup. It would also really tell you the value of posession.

[nobody]

I think the answers for our lack of possession are pretty easy to figure out in the Confed Cup. Take away O'Brien and Reyna, our two best possession players, then take the field against the teams that finished 1 & 3 in the World Cup and there's no way we are gonna maintain much control over the ball. Yeah, yeah, they had weaker squads, but so does everyone else and we're not as deep as more developed soccer nations yet.

I'm not a supporter of possession for possession's sake, but you really put yourself at a disadvantage if you just chase the ball all day. A good team needs to know when to slow it down and when to attack and have the players capable of doing so. We were only so-so at this aspect of the game in the World Cup and without Reyna and O'Brien, frankly we stink at it.

Wish I had gotten to see Martino play, sounds like we may have another guy with some potential in that area. Hate to say it, bit as good as Convey looks at times, I haven't seen him show much of an inclination to dictate pace and direction for a possession game.