Soccernomics (17 page)

Black players became accepted because the market in players is transparent. It is pretty obvious who can play and who can’t, who’s “got bottle” and who doesn’t. The market in players’ salaries, as we have seen, is so efficient that it explains 92 percent of the variation in clubs’

league positions.

However, the market in managers doesn’t work nearly as well. Only a few managers, like Brian Clough or Bill Shankly, consistently perform better with their teams than the players’ wage bill suggests that they should. It’s hard to identify what it is that these men do better than others, because if it were easy, everyone would copy them. Most other managers simply do not matter very much, and do not last very long in the job. They appear to add so little value that it is tempting to think that they could be replaced by their secretaries, or their chairmen, or by stuffed teddy bears, without the club’s league position changing. Even Manchester United’s manager, Alex Ferguson, who has won more prizes than anyone else in the history of soccer, has probably performed only about as well as the manager of the world’s richest club should.

Perhaps his unique accomplishment is not winning but keeping all the interest groups in the club united behind him for so long. If you manage to stay manager of the world’s richest club for nearly twenty-five years in an era when the rich are getting ever richer, you are guaranteed to stack up prizes.

112

It would interesting to see what would happen if a club stopped hiring managers, and allowed an online survey of registered fans to pick the team. We suspect the club would perform decently, perhaps even better than its rivals, because it would be harnessing the wisdom of crowds. And it could use the money it saved on managers to up those crucial players’ wages.

None of this is good news for black managers. Because it is so hard to measure a manager’s performance, it will never become painfully obvious that clubs are undervaluing black managers. That means clubs can continue to choose their managers on the basis of appearances. Any club appointing someone who is not a white male ex-player with a conservative haircut must worry about looking foolish if its choice fails.

Hiring a black manager feels risky, because as Barnes says, “Black guys haven’t proved themselves as managers.” White guys have—or at least some of them appear to have.

At last, soccer players get more or less the jobs that they deserve. If only other professions were as fair.

THE ECONOMIST’S FEAR OF

THE PENALTY KICK

Are Penalties Cosmically Unfair,

or Only If You Are Nicolas Anelka?

A famous soccer manager stands up from the table. He’s going to pretend he is Chelsea’s captain, John Terry, about to take the crucial penalty in the Champions League final in Moscow.

The manager performs the part with Schadenfreude; he is no friend of Chelsea. He adjusts his face into a mask of tension. He tells us what Terry is thinking: “If I score, we win the Champions League.” And then, terrifyingly, “But first I have to score.”

The manager begins pulling at the arm of his suit jacket: he is mim-icking Terry pulling at his captain’s armband. Terry is telling himself (the manager explains), “I am captain, I am strong, I will score.”

Still pulling rhythmically at his suit, the manager looks up. He is eyeing an imaginary, grotesquely large Edwin van der Sar who is guard-ing a goal a very long twelve yards away. Terry intends to hit the ball to Van der Sar’s left. We now know that a Basque economist told Chelsea 113

114

that the Dutch keeper tended to dive right against right-footed kickers.

Terry runs up—and here the manager, cackling, falls on his backside.

Van der Sar did indeed dive right, as the Basque economist had fore-seen, but Terry slipped on the wet grass, and his shot into the left-hand corner missed by inches.

“This really is soccer,” the manager concludes. A player hits the post, the ball goes out, and Chelsea’s coach, Avram Grant, is sacked even though he is exactly the same manager as if the ball had gone in. By one estimate, Terry’s penalty cost Chelsea $170 million.

The penalty is probably the single thing in soccer that economists have most to say about. The penalty feels cosmically unfair; economists say otherwise. Penalties are often dismissed as a lottery; economists tell both kicker and goalkeeper exactly what to do. (Indeed, if only Nicolas Anelka had followed the economist’s advice, Chelsea would have won the final.) And best of all, penalties may be the best way in the known world of understanding game theory.

DIABOLICAL: ARE PENALTIES REALLY UNFAIR?

At first sight, the penalty looks like the most unfair device in all of sports. First of all, it may be impossible for a referee to judge most penalty appeals correctly, given the pace of modern soccer, the tangles of legs and ball, and the levels of deception by players. When the Canadian writer Adam Gopnik watched the World Cup of 1998 on TV for the
New Yorker
, he as an outsider to soccer immediately focused on this problem. The “more customary method of getting a penalty,” he wrote,

“ . . . is to walk into the ‘area’ with the ball, get breathed on hard, and then immediately collapse . . . arms and legs splayed out, while you twist in agony and beg for morphine, and your teammates smite their fore-heads at the tragic waste of a young life. The referee buys this more often than you might think. Afterward the postgame did-he-fall-orwas-he-pushed argument can go on for hours.” Or decades. The fan at home is often unsure whether it really should have been a penalty even after watching several replays.

T H E E C O N O M I S T ’ S F E A R O F T H E P E N A LT Y K I C K

115

And the referee’s misjudgments matter, because the penalty probably has more impact than any other refereeing decision in sports.

Umpires in baseball and tennis often fluff calls, but there are fifty-four outs in a baseball game, and countless points in a tennis match, and so no individual decision tends to make all that much difference. Referees in rugby and football blunder, too, but because these games are higher scoring than soccer, individual calls rarely change outcomes here, either. In any case, officials in all these sports can now consult instant replays.

But soccer referees cannot. And since important soccer matches usually hinge on one goal, the penalty usually decides the match. As Gopnik says, the penalty “creates an enormous disproportion between the foul and the reward.”

No wonder the penalty drives managers crazy. As Arsène Wenger lamented at the end of the 2007–2008 season, “Every big game I’ve seen this year has been decided, offside or not offside, penalty or not penalty.” Indeed, it’s now a standard tactic for managers in England, after their team has lost, to devote the postmatch press conference to a penalty given or not given. It’s a ritual song of lament, which goes like this:
The penalty completely changed the outcome of the game. We were clearly
winning/tying but lost because of the (diabolical, unjust) penalty
.

The manager knows that most newspapers prefer covering personality clashes to tactics, and so the “match” reports will be devoted to the press conference rather than his team’s losing performance. Meanwhile, the winning manager, when asked about the penalty, recites:
It made no
difference whatever to the outcome of the game. We were clearly winning and
would inevitably have done so without the (entirely just) penalty.

These two ritual managerial chants amount to two different hypotheses about how penalties affect soccer matches. The first manager is claiming that randomly awarded penalties distort results. The second manager is saying penalties make no difference. On occasion, either manager might be right. But over the long term, one of them must be more right than the other. So which is it—do penalties change results, or don’t they? We have the data to answer this question.

116

Our guru is Dr. Tunde Buraimo. One of the growing band of sports econometricians—the British equivalent of baseball’s sabermetricians—

Tunde works, appropriately, in the ancient heartland of professional soccer at the University of Central Lancashire in Preston. As the saying goes, the plural of “anecdote” is “data,” and Tunde prefers to work with tens of thousands of pieces of evidence rather than a few random recollections. To help us with our book, he examined 1,520 Premier League games played over four years, from the 2002–2003 season until 2005–

2006. For each game he knew the pattern of scoring and, crucially, which team was expected to win given the prematch betting odds.

Our test of the two rival hypotheses about penalties is simple. We asked Tunde to divide the games into two groups:

1. Games in which penalties were awarded

2. Games in which they were not

We then asked him to compare how often the home team won when there was a penalty, and how often when there wasn’t. Figure 6.1 shows what he found.

F I G U R E 6 . 1
Comparison of home team and penalties
Penalty awarded in match?

Result

No

Yes

Total

Home win

577

142

719

46.76%

49.65%

47.30%

Away win

336

80

416

27.23%

27.97%

27.37%

Tie

321

64

385

26.01%

22.38%

25.33%

Total

1,234

286

1,520

100%

100%

100%

Look at the last column first. Taking all games in the database, 47.30

percent ended in home wins, 27.37 percent in away wins, and 25.33 percent in ties. These frequencies reflect the intrinsic advantage of home teams. Now imagine that the first manager is right: penalties change the T H E E C O N O M I S T ’ S F E A R O F T H E P E N A LT Y K I C K

117

outcome of the game. How will they do that? It might be that they always favor home teams (because referees are cowards), in which case we would expect the percentage of home wins to be greater when penalties are given. Alternatively, it might be that penalties favor away teams (perhaps enabling a team that has its back against the wall to make an escape). If so, the proportion of away wins (or ties) would rise with penalties.

But in fact, as the first two columns show, the percentages for all results barely change whether a penalty is given or not. The percentage of home wins is about three points higher when there is a penalty (up from 46.76 percent to 49.65 percent), and the percentage of ties is commen-surately lower (down to 22.38 percent from 26.01 percent). The percentage of away wins remains almost identical (27.97 percent against 27.23 percent) with or without penalties. So in games with penalties, there are slightly more home wins and slightly fewer ties.

It’s tempting to read significance into this: to think that the rise in home wins when there is a penalty is big enough to show that penalties favor the home team. However, statisticians warn against this kind of intuitive analysis. The absolute number of home wins when there were penalties in the game was seventy. Had the frequency of home wins been the same as in games when there was no penalty, the number of home wins would have been sixty-six. So the difference (four extra home wins) is too small be to be considered statistically significant. The rise is likely due to chance.

It would have been a different matter had the number of home wins when there was a penalty exceeded seventy-eight (or 55 percent of the games concerned). Then, the increase would have met the standard generally used by statisticians for confidence that there was a statistically reliable difference in outcomes depending on the award of a penalty. As it is, though, the data suggest that the award of a penalty does not affect either home wins, away wins, or ties.

But perhaps penalties have a different effect on match results. Perhaps they help favorites (if refs favor the big team). Or maybe they help underdogs (if penalties truly are given randomly, they should help the worse team more than the better one). Tunde tested these hypotheses, too.

118

F I G U R E 6 . 2
Penalty effects on match results

Penalty awarded in match?

Result

No

Yes

Total

Favorite win

633

147

780

51.30%

51.40%

51.32%

Underdog win

254

67

321

20.58%

23.43%

21.12%

Tie

347

72

419

28.12%

25.17%

27.57%

Total

1,234

286

1,520

100%

100%

100%a

aThe totals should add up to 100%, but there are rounding errors.

It’s obvious even to the naked eye that penalties have no impact at all on whether the favorite wins: favorites win 51.3 percent of games without a penalty, and 51.4 percent with a penalty. It’s true that underdogs win nearly 3 percent more often when there is a penalty than when there is not, but once again the tests demonstrate that this fact has no statistical significance. We can put the increase down to chance. Match results appear to be the same with or without penalties. Penalties do not matter.

Now, this is a statistical statement that requires a very precise inter-pretation. Penalties do matter in that they often change the outcome of an individual game. Clearly, a team that scores from a penalty is more likely to win, and so, whatever a manager says, a converted penalty will affect the evolution of almost any game.

However,
on average
, taken over a large sample of games, a penalty does not make it any more likely that home teams or away teams or favorites or underdogs win. If penalties were abolished tomorrow, the pattern of soccer results would be exactly the same.

This sounds counterintuitive. After all, we argued that penalties look like the most unfair device in sport. They are often wrongly awarded, they cause a lot of goals, and many of these goals decide matches. So surely penalties should make results less fair.