## World Cup pseudo-science

Lee Sechrest pointed me to this news article by Vitomir Miles Raguz, “Brazil Won’t Win the World Cup. A European team will win again thanks to training and statistical analysis.”

Hmmm . . . “statistical analysis.” This Raguz character better coordinate stories with Nate; it seems that the statistical experts are disagreeing . . .

Much of the article is reasonable enough, I suppose (I’m not enough of a soccer expert to say more), but this bit raised some flags for me (and not in a good way):

Also pay close attention to how European teams make substitutions. If they do it at, say, the seemingly arbitrary 58-, 73- and 79-minute marks, you can safely bet that a quant is involved in game mapping. Empirical studies have shown that fresh legs around these specific points have a statistically significant impact on team performance.

The 58-minute mark, indeed! I could be wrong, but this looks a bit like numerology to me.

1. David says:

The paper in question is “A Proposed Decision Rule for the Timing of Soccer Substitutions” from Journal of Quantitative Analysis in Sports. I don’t think there’s an ungated version anywhere, but a gated one is here: http://www.degruyter.com/view/j/jqas.2012.8.issue-1/1559-0410.1349/1559-0410.1349.xml?format=INT

• Andrew [not Gelman] says:

I do not know of an ungated version, but it is also discussed in “The Numbers Game” (page 234-240) by Chris Anderson and David Sally.

2. Chris says:

As a player, coach and data geek, there are a few things things odd about soccer substitutes. First, is the player cannot enter the game until there is a stoppage in play (ball out of bounds or foul). So there is a delay between when the coach wants the substitute and when the player enters. If your opponent is Spain, then this delay could be 5 minutes or more. If you are loosing, then opponent will be playing more possession, also delaying your substitutions.

So there is no true difference between a 58m sub and a 60m sub. These subs are likely decided at half time, giving the sub 15m of warmup and 30m of playing time in 2nd half.

Later subs trade-off fresher legs vs less time to acclimate to the game and time to affect outcome.

• thom says:

Agreed, except they usually warm up more than one sub so they aren’t predetermined (though imagine they have a game plan).

Declaration of interest: I drew Brazil in the department sweepstake …

3. numeric says:

I could be wrong, but this looks a bit like numerology to me.

What’s wrong with numerology? Next thing you’ll be attacking astrology. Geez.

4. germo says:

I guess those 58,73,78 minute-marks are from WSJ’s earlier article: http://online.wsj.com/news/articles/SB10001424052748704364004576132203619576930 Hard to say what was exactly done, but at least this part seems rather “intresting”: (sample size 1,037 games):

“He concluded that if their team is behind, managers should make the first substitution prior to the 58th minute, the second substitution prior to the 73rd minute and the third prior to the 79th minute. Teams that follow these guidelines improve—score at least one goal—roughly 36% of the time. Teams that don’t follow the rule improve about 18.5% of the time.”

5. Giles Warrack says:

For the record no European country has won a WC held on South or Central American soil. Only Brazil (Sweden, 1958) have recorded a South American win in Europe, and they also won in the USA and Japan.

I don’t see history being made in this WC. This article makes numerology seem like solid bench science.

• Steve McKay says:

… plus Argentina won in Mexico, and Spain won in South Africa. Still hard to see beyond a South American victory in 2014.

6. Martin says:

Usually, I am completely indifferent to soccer. But now, I really really want Brazil to win, for the one simple reason to see Vitormir Miles Raguz explain where he went wrong.

• Andrew says:

Martin:

I know what you mean. That article about all those innovative Europeans, it gave off a bit of a Nicholas Wade vibe.

• DK says:

I was reminded of that book yesterday, watching Japan – Ivory Coast. For the first 30 min Japan was clearly very dominant, with great teamwork and excellent technical skills. And then they ran out of steam. In the second half, it was almost pitiful to see how badly Japanese were outrun by Ivorians. The difference in physicality between the two teams was pretty striking.

• Martin says:

Well, it gets exciting – Spain and England are out.

No surprise for Spain: this can be explained by the African admixture to those Southern people, which surely diminishes their “statistical analysis” skills. After all, as Dominique Dufour de Pradt knew already (as quoted by Coates):

«C’est une erreur de la géographie d’avoir attribué l’Espagne à l’Europe; elle appartient à l’Afrique: sang, mœurs, langage, manière de vivre et de combattre; en Espagne tout est africain. Les deux nations ont été mêlées trop longtemps, les Carthaginois venus d’Afrique en Espagne, les Vandales passés d’Espagne en Afrique, les Maures séjournant eu Espagne pendant 700 ans, pour qu’une aussi longue cohabitation, pour que ces transfusions de peuples et de coutumes n’aient pas confondu ensemble les races et les mœurs des deux contrées. Si l’Espagnol était Mahométan, il serait un Africain complet; c’est la religion qui l’a conservé à l’Europe.»

But England??? Their truly imperial statistical skills should r00l, isn’t it? Is it all this post-colonial immigration, perhaps, that perhaps made England more “urban”?

Wow, I feel I am almost at a point where I can write a book about genes and soccer, I clearly have the skills in fancy-theory making combined with a complete disregard for data or “science”.

• Andrew says:

Martin:

Are you kidding? It’s obvious. The English have been weakened by nearly 70 years of socialism. The National Health Service can’t be good for evolution, can it?

• Martin says:

There you see why I am not yet quite ready compared to Wade: While I know _some_ jargon in general wishy-washiness, I still lack a PhD in evolutionary gobbledygook.

• Martin says:

Sooo… seems that all turns out exactly as expected. The European football teams are just fine. Meanwhile, another African nation, Italy, was sent home. The true nature of the Italian is known since this online lecture by Dennis Hopper:

By the way and in all seriousness, in “Group D” the two winners are Costa Rica and Uruguay, while England and Italy are out. He. Turns out that it is already clear that this claim by Raguz won’t hold up:

“Thus, it’s likely that at least 10 European teams will move into the second round, along with about five from South America and possibly one from the rest of the world.”

If I got it right, there are currently three European teams qualified for the next round, and only four can (or not) still qualify – not all of them from a good position, though. So, already at this early stage – when half of the matches of the first round haven’t even been played – this claim is spectacularly wrong. At least ten! I’d love to see Raguz’ explanation.

I read it was the heat and humidity in Brazil, which European teams are not used to. This seem to be extra-statistical factors, though, because his claim was based on statistical analysis, amiright?

• Andrew says:

Martin:

According to Wikipedia, Uruguay is 90% white and Costa Rica is 65% white, so they count as European countries. And, sure, Spain is out, but they’re basically South American, they don’t really count as European. And we’ve already discussed England, a country where socialism has sidetracked evolution. So I think the Wade/Raguz analysis still holds up pretty well.

• Martin says:

That’s true. But to go back to a purely statistical analysis, I also realized that the European teams mostly failed to make their substitutions at the 58-, 73- and 79-minutes marks. I think that contributed heavily. For example, Italy made them at minutes the 46, 71, and 75. That just won’t cut it, so: no surprise.

Also, what’s the chances for a striker called Ciro Immobile to make any difference?

• Martin says:

So… six (Raguz: at least ten) European teams are qualified for the secound round, five South American teams (Raguz: about five), and five from the rest of the world (Raguz: possibly one).

I guess the number for South America demonstrates that his approach is completely correct!

• Daniel Gotthardt says:

We all know that reality is so messy. We can’t let what actually happens distract us from our true theories, can we?

• Martin says:

Well, damn…

7. Dan Wright says:

So it is half time in the Germany v Brazil match, the “beautiful game”. Hmmmm. I guess Raguz > Silver in this battle.

8. Martin says:

OK, so, against my expressed will, Brasil will not win the Word Cup. Now, of course, it is inevitable that everybody knows after the fact what is wrong with the Brasilian team. A particularly nasty example comes from the FT, as I read in today’s “Estado de São Paulo”:

“They [the brazilians] certainly must recognize now that the country needs, more rapidly than it believes, to adopt a new, “more European, more German” style, says the FT. The Brasilian soccer needs too to start over, ideally lead by German trainers, the journal adds.”

“Eles [os brasileiros] devem agora certamente perceber que o País precisa, mais rápido do que pensa, adotar um novo um estilo “mais europeu, mais alemão”, diz o FT. O futebol brasileiro precisa começar de novo também, idealmente, liderada por treinadores alemães, emenda o jornal.”

This, I think, is bad. And I am not even talking about the adulation for anything “European” (especially given that this still is a catastrophical World Cup for European teams in general) or “German” implied here. Rather, the FT suggestion is completely hollow. If one feels one has to come up after the fact with an analysis of what’s wrong with the Brasilian team _in general_ (contrary to an analysis of the match itself), I’d expect plenty of concrete reasoning. Instead, we get this content-free pseudo-analysis about an “outdated-style” and a need for “reinvention” and blah blah.

But anyway, this lead me to a question. I think nobody has any doubts that the German team was better, yesterday. But I also think it’s fair to say that they are not technically better to such an extent as to explain the 7:1 (if so, all the other results wold make little sense). I guess Scolari’s explanation now widely cited that, to speak losely, sh*t happens and the team got into a panic the German team profited from ideally, is at least part of the reason.

To put it differently: I think both the German and the Brazilian performances up to yesterday suggest that the 7:1 is fairly extreme (and as Nate Silver suggests, that is what basically everybody would have said before the match, rather than after it, if suggested such a result).

So, my question: What is really to learn from such an event? Is it really a reason to initiate a complete makeover, as the FT suggests? That is, should the coming actions of the Brasilian soccer federation be led mostly or even solely be yesterday’s extreme event? At the expense of all the other matches and the fact the Brazilian team had e very good run with its trainer until yesterday (not a single match lost!)? If so, why so?

Another question: Why don’t I get paid for writing content-free gobbledygook for the FT?

• Andrew says:

Martin:

I’m no soccer expert but a 7-1 victory does seem like some evidence. But one could flip it around and say that there’s no reason that Brazil (or any other team) has the moral right to be in the World Cup final, especially with an injury to one of their best players. Sure, Brazil has some great players, but so do other countries, and they’re all trying hard too. To put it another way: suppose all the teams got German trainers. They couldn’t then all make it to the final game during the tournament.

P.S. I wanted to play around with a simple model of all this in Stan but I’m having install problems.

9. Martin says:

Andrew,

I don’t suggest that the 7-1 victory doesn’t provide any evidence. But it is obviously extreme: Nate Silver suggests that even if you give more weight to psychological issues following Neymar’s injury etc. etc., there is no way one could have deemed such a result even remotely probable – he sees it as a “black swan” event.

So, my question is: How much weight to give to such an extreme event in the context of a contact sport? Discard everything else (e.g. that the team up to yesterday was so succesful as to be the favorite, so apprently they must have done _something_ right, isn’t it?) and solely conentrate on this extreme event? Obviously one cannot just ignore it, but how does one sensibly factor it in? Or is the easy answer that I see now everytwhere, namely a complete overthrow of how to think of and approach soccer, the sensible answer? Had Brazil lost 1-0, certainly changes would be suggested, too – but I don’t think abnybody would suggest a complete overthrow, especially given the long series of successes with this team and trainer. What does such a “black swan” teach us? Or is it “just” a lost game turned extreme by nothing else than bad luck? I just don’t know how to think about this…

• Martin says:

By the way, I also don’t mean to suggest that there is nothing to learn from the Germans. I just don’t think that the FT established any of this with a clear evidence and analyses, they only claim it after the fact, and that’s just cheap.

• Daniel Gotthardt says:

Martin,

I think, Brazil did not play such a good World Cup and was too dependent on Neymar and Thiago Silva but talking about some kind of germanization is really silly, especially since the recent successes and good developments in Germany were due to “ungermanaization”. North American non-soccer fitness and mental training standards were applied by Klinsmann first and Löw later, modern football tactics as played by more offensive European teams like the Netherlands and Spain were used, modern analytics (!) were used, much more players with a foreign background were allowed to play and team players instead of annoying alpha wolves were prioritized. Quite some Germans did not like that all and are still complaining. “No real man” rhetorics and all that. At least racist comments have lessened.

Brazil would probably benefit from some of it but you can just call that professionalization.

• Daniel Gotthardt says:

I really should not write as tired as I am without an edit button. Sorry.

• Martin says:

Thanks, and yes, it is all very well to suggest that Brazil’s soccer was not the best or not even good during the world cup. What bothers me is that in the FT (and lots of other outlets) this dies not spring out of any actual analyses, but allegedly from yesterday’s event, and I just don’t buy it – yesterday was not representative of Brazil’s team (or is it, in a way? I am just confused about this!), I think, even if one thinks that they were not good anyway. That’s too why I am so confused wbout how to think about yesterday’s outcome.

I thought about writing about that curious Klisman episode, too. Back then, when the German team had its problems and they chose Klinsman, he was ridiculed for his allegedly ‘esoteric’ methods by German soccer Gods like Beckenbauer. Now he is widely respected as the one who initiated the makeover that was the basis of the current team (Löw was widely seen as sharing Klinsman’s approach and was the natural choice after Klinsman left, if I remember correctly).

So, I entirely agree that there is something to learn from the Germans. And I also agree that there are specific points to be learned from the Gernamns, none of them “being” German.

• Daniel Gotthardt says:

Martin, I agree that yesterday was not representative for Brazil under normal circumstances but I think that such an astonishing event could happen is something which needs explaining and might shed light on general problems. It’s a “story event” which one could see as model-checking, I think (see Andrew’s and Basboll’s discussion about that). We could of course just take it as a random fluke but it’s quite unlikely to just be noise. I think far too much pressure is actually the main source of what has happened as I already mentioned in another comment. I’m not sure about the details of Nate Silver’s model but I think, we might want to check if the model needs to take different degrees of pressure and mental training into account more.

And yes, Löw was Klinsmann’s assistant and there is even some discussion which part of the makeover came from whom, some claiming that Klinsmann had nothing to do with tactics at all (which I don’t believe to be true).

• Martin says:

How would one sensibly account for something like pressure?

As a speculative aside, that’s one thing Brazil could learn from the Germans. When Germany lost the semi-final match against Italy at home in 2006, the public was behind the team and proud of it, as far as I remember. Now, 7-1 at this stage is nothing the loser can reasonably be “proud” of (in a ‘we were good, but they were better’ way), but the Brazilian fans and commentary are simply being awful (and were already during the match yesterday). The most disgusting example I found yet ist this here:

http://blog.jovempan.uol.com.br/quartarollo/que-grande-merda-hem-felipao/

It’s understandable they talk about changes, but I think what the team needs right now is a collective hug, not hatred and shame. Apparently, the media are now at a stage where they ask what all this “means” for the country as a whole. If so, it would be interesting to know why the the national response to such a defeat is pure hatred and cynicism instead of introspection and holding together – which can make defeats quite edifying experiences, after all, but perhaps I am personalizing this too much now…

• KKnight says:

For what it’s worth, I reckon that Thiago Silva’s suspension played a much bigger factor in the outcome than Neymar’s absence. Defensively, the Brazilians looked like amateurs – heck, even England could have beaten them yesterday! But these things happen in sports.

• I suggest a different alternative: some players on the Brazilian team had pressure to lose. Perhaps from some political source seeking to capitalize on the Anti-World-Cup-Economics sentiment, perhaps from some financial source such as threat to future contracts, perhaps from some other source. With Neymar and Silva out of the game, they had an opportunity to lose big and potentially expose this pressure.

The Brazilian team played worse than most pick-up soccer games I’ve participated in. There are a variety of possible explanations for that, but I think off-field pressure has to be in the consideration of the possibilities.

• Martin says:

@ KKnight

That might be true, but it also does not explain why they lost 7-1 rather than, say, 1:0 or 3:2, or something.

@ Daniel Lakeland

That sounds highly speculative, and though I gather that there are up-coming elections, I wonder which political force would be so powerful as to trump the influence of the incumbent president and party so decisvely. It also does not explain why 7-1 instead of something “normal”. How, exactly, would such outside off-field pressure look like?

• I don’t have a favorite off field pressure theory. My theory is speculative, but it’s part of a prior that is informed by things like:

and the fact that FIFA ran anti-fixing ads about 7 times each game on my ESPN streaming feed at least.

External pressures could be anything from what Daniel Gotthardt suggests below (general societal pressure on the team due to economic and political issues) to for example threatening the lives, families, or future careers of people on the team. See Andres Escobar Saldarriaga from Colombia, murdered in 1994 for scoring an own-goal.

The political forces could be business forces that affect both sides of the upcoming election for example.

Like I say, I don’t have a particular theory, I just think it is highly plausible that specific off-field pressures on individual team members, the coach, or the team as a whole contributed to Brazil just not playing any defense. It could have been pressure to shave a few points even that went horribly wrong when Germany started scoring and scoring and scoring, and the Brazilian team couldn’t figure out what to do (they’re supposed to not play but now it’s getting ridiculous, they have to coordinate if they want to look like they aren’t throwing the match… but they can’t coordinate until halftime, when they come back out with a clear message “don’t play like you’re trying to throw the match!” and they started playing a little more seriously). In my opinion, it was obvious from minute 1 even before the first goal that they were playing sloppy and extremely open defense, long open forward passes, things that they don’t normally do.

• KKnight says:

Let’s not forget that Germany is a very good team – at its best (and with Thiago Silva and Neymar), Brazil would have been hard-pressed to beat Germany.

• You’re absolutely right. But Germany beat Algeria 2-1, US 1-0, and tied Ghana 2-2. Those teams put up defenses, and Brazil just DIDN’T. There just doesn’t seem to be an explanation why Brazil let Germany waltz into their box without much of any cover, hardly any challenges… why did Brazil not play in the midfield? instead passing the ball on long shots to their offside players (Hulk was offside by maybe 6 body lengths on one play!)

Germany outclassed Brazil, but not 7-1. It’s this outlier score that requires explanation and the explanation that makes the data most likely is some kind of intention on the part of Brazil IMHO. Combine with a prior informed by previous match fixing scandals and you get a high enough posterior probability that you can’t ignore this possibility.

That’s all I’m saying really. It could have been just breakdown by a team with no leadership, but it could have been off-field pressure too. I give those two possibilities similar orders of magnitude of probability.

• Abdullah says:

This is why we should never let stats guys analyse matches.

ONE outlier event like this and you will come up with the craziest theories {“it’s just not possible Scotty! No way, no how!”}

Here’s a simpler explanation. Watch the match. Brazil are a mediocre team, losing two key players {Silva & Neymar} for this match. Germany are a well-co-ordinated outfit. They are not particularly physical, so they are thwarted by teams that have a strong physical presence {e.g. Ghana & the US, which is why they were not able to whitewash those teams} in defence.

The mood of the Brazilians even prior to the match was one of defeat. Even the fans were not that confident in this team. Did you watch the matches? Hulk was {to put it bluntly} throughout the tourny, summed up in one word ▬ “crap”. Punters were commenting on how it was surprising that Scolari was picking him time after time.

So, the Brazilians lacked skill, lost their best players, were lacking confidence, had doubtful managerial representation {team management}, and under an awful lot of pressure.

The Germans had fluid {but not staggering} technical skills, were full of confidence, had {at least} competent management, and good team spirit.

The Brazilians had sufficient spirit to possibly come back after going a goal down. But once it went to 2-0 so quickly, they were like a rabbit caught in headlights. It may be a statistical black swan; but in the end, footballing events like this are well within the realm of possibility {of course that’s no consolation if you’re the one on the receiving end}. All that happened was that their defence absolutely crumbled and a reasonably clinical {but not staggeringly so} Germany took full advantage. Silva was not there to co-ordinate a defence properly, and the attacking flanks lost their confidence without the symbolic talisman of Neymar and simply gave up. Let’s face it, they didn’t really have sufficient skill to get back into the match to begin with.

And that’s all she wrote, folks. No need to pull out any beard hairs over this.

• KKnight says:

Spot on, Abdullah. Only two of the Brazilian players (Neymar and Thiago Silva) would have improved the German team. I doubt that Fred would have cracked the US lineup.

• Martin says:

Abdullah,

I am not sure if I agree. Though I also have my difficulties with the idea of off-field pressure somehow doing all that, what does an extreme outlier have to do with “stats guys analys[ing] matches”?

You offer reasons that are palpable (to me), but again, none of them explains the extreme score (or, conversely, the extreme rarity of such scores in World Cup history, especially at this stage). Has the Brazilian team been the first with a loss of key players and mediocre skills in general? And why didn’t all that show much more obviously in the other matches? Again: This team with this trainer didn’t lose a single match until yesterday. And while the resons you offer maybe point to a defeat (though, I’d mention, Silver didn’t see it, and neither did anybody else, much less so with any certainty, so this smells a bit of ex post-rationalization) the extreme score is… extreme.

• Daniel Gotthardt says:

Pressure and, I assume, lack of professional mental training were an issue and it broke out when the two key players were not present who could calm down the team. I don’t think you have to look at specific players who have been pressured by financial issues. The team in general was pressured due to the economical and political problems. Soccer is akin to a religion, especially in Brazil and the team had the pressure to give hope to the people. They even said so themselves.

10. Martin says:

Another thing: Who knows something about the upcoming Brazilian elections?

Today I read in CénarioMT (a newspaper from Mato Grosso) that after the defeat of the Brazilian team against the Germans, the Palácio do Planalto (i.e. the Brazilian White House) is thinking about a change in the election strategy (for Dilma Rousseff, obviously):

What it says is that Rousseff initially planned to take advantage of a) holding the World Cup successfully, and b) the succes of the Brazilian team. Now, as plan B for the latter (e.g. Brazil would lose against Germany) they would fully embrace the defeated team, anyway. Turns out that plans might change given the extreme result: Rousseff would probably not start to be officially ashamed of the Brazilian team, but not exactly stand by and celebrate it either. Also, it states that they are reinforcing secrurity for the last week in order to keep at least the World Cup a success, whatever it takes.

Looking at the polls up to now, I am wondering:

http://pt.wikipedia.org/wiki/Elei%C3%A7%C3%A3o_presidencial_no_Brasil_em_2014

Rousseff seems to have a comfortable advantage and didn’t even come close to losing it during the most heated period of demonstrations in the last couple of months. But then, I think she needs to win a second round, as she won’t have 50 percent in the first.

Any idea what the strategy of quietly disowning the own team might bring about in this context? Is this teally a thing in Brazil?

• Martin says:

Today’s “O Estado de S. Paulo” runs with a similar story:

“One day after the humiliating defeat of Brazil against Germany, president Dilma Rousseff adjusted the narrative in order to neutralize the “[World] Cup effect” on the reelection campaign. Afraid that the bad mood w/r/t the Seleção bleeds into the campaign, the president and her team try to separate “the wheat from the chaff”, concentrating their energies on the “organization” of the World Cup.”

They also cite various sources close to Rousseff and her reelection campaign who do already shift talking points in this spirit, making her all Stateswoman rather than a close ally of the Brazilian team (aside: the idea alone makes me feel very sad for the members of the team!).

I can’t for the life of me imagine that a similar event would have much of an influence when it comes to the reelection of Merkel (or rather her party), or Hollande, or Faymann (his party), for that matter. But lest I fall into a “European countries are more rational” trap myself, I’d like to know if there is any data on this (i.e. the influence of a country’s performance on elections).

Is it possible that this is a phenomenon in relatively poor or unequal countries like Brazil, meaning that it works through other channels? For example, the above article suggests that World Cup euphoria suppressed the social unrests we have seen some months ago and the cause of which (high prices and inflation) has not been eliminated, and Rousseff hoped to carry that euphoria at least to the elction. Now, the shocking end to this euphoria might touch off these unrests again. This is a channel that might not be available in rich countries (or not to this extent). Any idea?

Andrew, sorry I am perpetuating this thread as much as I do with stuff quite far away from the original topic – it’s just that I find all this fascinating. Just tell me if you feel I should stop.

11. jrc says:

Before I give a stats-y answer – let me just say I basically agree with KKnight, and that the loss of Thiago Silva is the major reason this happened. Without him, no one was keeping an eye on Brazilian Side-Show Bob, who thought he would just dribble the length of the pitch to get Brazil back in the game, no one was yelling at the outside backs (I’m looking at you Maicon) to get back in position, or at the defensive midfielders to fill the gaps the fullbacks left, and essentially the team was left without an organizing, calming presence.

Stats-y-wise: let’s pretend (modelling!) that teams get a number of chances C() that is a function of game conditions g – a vector that includes current score, minute of the game, and both the individual talents of players and the collective energy/emotional-state of the players. So a team gets C(g) chances in a game.

The probability of converting that chance into a goal is P(), and the arguments for the functional form of P are a vector s of determinants of scoring – the attacking skill of the shooter, the quality of goalkeeper and defense, the effort of both teams in their respective roles (attack/defense).

So – Thiago Silva goes out, and Germany scores early. The talent of Brazil in defense is less (C up, P up for Germany), the emotional state of Brazil is below where it would be (C up for Germany), the focus of Brazil goes from defense to attack (C up, P up for Germany), etc….

Now, the stage is set for Germany to have more chances, and be more likely to convert those chances. Why don’t we usually see so many goals? Well, first off, P and C are usually fairly low. You maybe get 4-8 chances in a game, and that usually leads to 1-2 goals – so call P around .25.

Let’s say that, given the situation, Germany was going to get around 12 chances, and the P was up to .35. In that case, I get a probability of about .08 that Germany scores at LEAST 7 goals*.

Now – I completely made up those parameters, and the measurement of “chances” is probably tough and imperfect by nature, but I think that the basic conceptual structure makes it pretty easy to show that in these conditions, we should expect this kind of game from time to time. We don’t see it in most games even when big players are missing because a) in the world cup knock-outs, a loss is a loss and the scoreline doesn’t matter, so you have to shift your attention to attacking and off defense (exaggerating the C response) and b) because most of the time winning teams would reduce effort (lowering P for them). But Germany couldn’t risk that, and its not their style.

My point is just – this is an uncommon outcome, but not really that crazy, and not a “oh my G-d football is completely turned on its head!” type of situation. People said tiki-taka would rule forever too – you saw what happened to Spain. Some tactics work against other tactics, and people will adjust to Total Football or Power Football or whatever you want to call this German style (New Pep?), just like they did to tiki-taka (Old Pep). See – Pep Guardiola made the transition just to beat his old self!

*My crazy complicated Stata code for the simulation:

set ob 10000

gen goals = rbinomial(12,.35)

gen seven_plus = goals>=7

sum seven_plus

• Martin says:

So, does this suggest the Brazilian team is just fine, basically – some pretty feasible adjustments, like making the defence less dependent on one player – will do it? No overthrow, no ‘rethinking’ the way Brazil approaches soccer? No Brazil has to become more like a European team etc.?

• jrc says:

Brazil needs to find a new way to play the football they play. They will always be a team that relies on flair, individual moments of brilliance, clever movement off the ball, and unexpected tricks for goals. That is the way many Brazilian players learned the game – watch 8 year old Ronaldinho playing futsal.

Now – the team has succeeded in the past because it has been able to incorporate the brilliant and unpredictable attacking players with a strong spine – think Gilberto Silva holding the midfield: discipline at the back and in the middle, and flair on the wings and in the front. They are a counter-attacking team that counters so well they seem like an attacking team.

Tactically, this would still work in some form – how many top European clubs use two creative attacking wingers and a couple holding defensive midfielders in a 4-2-3-1? Arsenal, Chelsea, Real Madrid, the list goes on.

Brazil doesn’t need to change their identity, they just need to update it with the new players. Those in the middle have to allow Neymar and Oscar to be the creative players, and have faith in the ability of those two to create goals, while the central midfielders and defenders try to stop doing too much and just keep their shape, maintain discipline, and know that the front 4 will find a way through on the break.

There’s the highlights. On almost every one of those goals there is essentially NO opposition from Brazil. I think the commentator describes it perfectly. when are people going to close down? make it difficult for Germany? when are they going to stop leaving the field wide open?

I see plays where Brazilian players were just standing there or walking away from the ball while Germany was in their penalty box.

It’s that behavior that makes me wonder if something else is going on other than a football match.

A Brazilian loss to the talented Germany team I can understand. Brazilian players not even playing defense? That seems to require some other explanation.

• http://www.soccerbase.com/teams/team.sd?team_id=395

That gives the players and their current club. Almost the entire Brazil team plays in Europe for European clubs. So another option is just self-interest. Once they’re down 2-0 perhaps defensive players simply said “No way am I going to get injured challenging these German players”. That could easily be amplified by their Brazilian compensation being dependent on a win, or something like that.

Does anyone know if there were significant performance dependent compensation issues for the Brazilian players? Perhaps informal or indirect ones (like advertising contracts in Brazil, large bonuses for wins, etc etc).

If you “know” that you’re not going to win without Silva and Neymar, and you know that getting injured means you might lose significant future compensation coming from European sources, and you know that without a win you won’t be getting much value out of your participation in the Brazilian team… you might be inclined to stay the hell away from trouble and hope to go back to your club and make good money instead of challenge your fellow european players and one or both of you get an injury….

• jrc says:

My understanding is that all players get compensated in some way for playing for their national team (see the Ghana fiasco), and yes, there are often bonuses for wins. The US team structure (such as it is made public) is provided here:

http://www.si.com/soccer/planet-futbol/2014/05/20/world-cup-usa-usmnt-player-bonuses-money-ussf

And yeah – it is true that getting injured may cost you, but probably not as much as you’d think. First off – you are probably on a long-term contract already. David Luiz, (aka Brazilian Side-Show Bob) just signed a 5 year contract with PSG, and that will take him through most of his career. And beyond that, a good showing at a world cup is likely to lead to lucrative new contract (either with a transfer, or to protect against a transfer with a large buy-out clause).

But here’s the thing: Brazilian national team players are fabulously rich. And while there is a lot of money to be won fixing sporting matches, trying to bribe someone who already makes \$100,000 per WEEK gets really pricey. Plus, these guys have a huge amount of ego on the line. If you are gonna fix a game, you bribe Cameroonian players who play in Africa or second-tier leagues in Europe and make \$100,000 per YEAR, with little chance for future riches or income post-career.

I just don’t see the incentives lining up for Brazil to throw it. I think they just psychologically fell apart. That body language doesn’t strike me as “cheating” body language, it strikes me as “defeated” body language.

Come on Daniel, haven’t you ever had a seminar go bad? That feeling after 15 minutes that you’ve lost everyone, no one cares, and you know there is no coming back from it. It happened to me once (so far), and I think I literally ended the seminar 10 minutes early with “Well, there’s some more stuff, but its been a long day, so let’s just call it here.”

• Of course you’re right, it’s possible they just fell apart. But I feel like it’s worth exploring the other options. The thing is, they played crappy defense from before even the first goal. If they had fallen apart at goal 2 or 3… makes sense. but why are people standing or walking away from the ball on the very first and second German goals?

I don’t think bribes are necessarily the only pressure that could come down. Threats to your life, to your family in Brazil, or other kinds of pressure could be a lot “cheaper”.

Another option is some kind of threats or bribes or pressure directly on the coach. Did the coach instruct them “don’t spend efforts on defense, just get forward, even if Germany is in your back-field, stay forward?” and they were too inexperienced and disorganized to re-think that on the fly without Silva?

Sure, from goal 3 to goal 7 you could say “well they just gave up” but goals 1,2,3 seemed like they were giving up already, didn’t it?

12. Martin says:

And because I still want Raguz to be wrong, I now suggest that Argentina better win against Germany!

13. Martin says:

Neymar cried so much, and Luiz cried so much… and they lost again… and again without any defence… as if the 7-1 against Germany had not happened…

@#!~ it! I never cared about this stuff, I only watched because of this nonsence by Raguz – – – and voilà I am emotionally invested.

Never again!