 |
| Tom Brady: Make-or-break quarterback of the New England Patriots |
Ten minutes into the 2008 NFL season, on September 7 last year, the football was snapped into play and two lines of giants collided in exquisitely choreographed violence. Kansas City were playing the Patriots, perennial favourites for the Super Bowl. The ball was in Brady’s hands. Kansas City’s Bernard Pollard, nicknamed “The Bonecrusher” at university, was trapped on the ground under one of Brady’s huge guardians. Just as the quarterback cocked his arm to throw, Pollard squirmed free and leapt headlong. Helmet hit knee, and that was the end of the Patriots’ season.
Could Brady really be so vital? Yes. He had led the Patriots to four Super Bowls and won three of them. In 2007, he was named the league’s Most Valuable Player. Bill Barnwell, managing editor at Football Outsiders, which produces an eye-watering almanac of American football statistics, says that the observed impact on a team of an injury to a superstar is about 6.75 times that of the observed impact of an injury to a reserve player. And Brady isn’t just a superstar, he’s a superstar quarterback. He leads the offence, where injuries have a statistically higher impact than they do on defensive units. Barnwell explains that the injury rate of a team’s offensive players year on year has a 0.42 correlation with team success (where 1 is a perfect correlation and 0 is no correlation). With defensive players, that drops to a 0.29 correlation. And the quarterback is crucial – his individual correlation is 0.3, twice that of any other position. An NFL offence is designed intricately around the quarterback’s strengths and weaknesses. Change the central cog, and the whole machine jams.
Medically, Brady is not so interesting. When Pollard’s helmet crashed into his left knee and he fell twisting to the ground, Brady tore his medial collateral and anterior cruciate ligaments. These are cords of tissue on the inside of the knee and running through the centre of the knee, respectively. But 30 years ago, before modern surgical techniques were available, such injuries ended professional sportsmen and women’s careers.
With teams matched as evenly as they are in the NFL, the loss of Brady had pundits guessing that the Patriots, who had won every regular season game in 2007, wouldn’t make the playoffs. They didn’t.
Brady’s story shows vividly how injury can affect a team, and in particular injury to a star player, whose performance is doubly critical. And since sport is increasingly obsessed by Freakonomics-style analysis, where economists and statisticians (and mere stats nuts) hunt for inefficiencies in the system, his case raises questions, too. Most Valuable Player isn’t some airy-fairy construct in the big-money world of US sport. The NFL uses a salary cap to ensure more balanced competition between the teams: each can spend $116m on player salaries this season, which they divide between 53 active players. New England spends $14m of this – more than 12 per cent – on Brady.
Sport is not a conventional business (Barnwell says that getting to the playoffs can cost a team more than they earn from it) but an NFL team’s raison d’être is to win, and one of the biggest constraints they face is the salary cap. In this context, the $14m spent on Brady represents a massive opportunity cost – a cost made painfully clear by his season on the sidelines as he convalesced. With so much at stake in the big money world of professional sport, the questions become more urgent. Do teams and their owners understand the economics of injury? How are decisions surrounding sports injury made? And is that changing?
. . .
Simon Kemp is head of sports medicine at the Rugby Football Union in England, an organisation striving to incorporate stats-based economic insights into its approach to injury. He says that the way professionals are treated when they are injured differs from the way ordinary people are treated. “We’re dealing with people whose priorities are very different from the man on the street. While I might advise an amateur to try conservative treatment, which will still let him lead a pretty active life, professionals can’t get by with that. Besides, they’re used to injury.” Kemp says that between 2002 and 2004, professional rugby players in England spent an average of 69 days a year injured.
How long players – and especially star players – spend recovering, therefore, becomes a matter of vital concern. Team doctors, for better or worse, are focused on helping players get back on the field as quickly as possible: in professional sport, the difference between 10 days and 12 days can be the ball game. When Jason Robinson, the talismanic England winger and full back, pulled his hamstring at the 2007 World Cup, the medical team threw everything at the problem. “We’re on the cutting edge and we have a very small window,” says Kemp. “We’ll try new things, so long as the science suggests they’ll help, and we think they’re not dangerous.” Kemp had found some success with homeopathic injections of Traumeel – a combination of various botanical bits and pieces. He told Robinson that the potential benefits exceeded the risks, and Robinson agreed to be injected. “I’m not saying it was the Traumeel,” Kemp says. “Jason was getting great rehab and placebos can be powerful. But we had him back on the pitch for the quarter-final.”
In 2004 Richard Hill, one of England’s World Cup-winning rugby heroes of the year before, faced a similar injury to Brady’s. He says now: “I just wasn’t ready to stop.” Hill was told about risks of infection and rates of success of surgery, which for this kind of procedure are 85 per cent and upwards. He had the operation. When he needed another one a year later, he was told it would probably mean he’d need a knee replacement in his forties. He had that operation, too. As it happened, the replacement took place only a year after Hill stopped playing. “It was worth it,” he says. “The last two games I played [the European Cup quarter- and semi-finals last year] are two of my best rugby memories.”
It’s clear that most serious sportsmen will risk long-term health difficulties in order to play.
. . .
 |
| Carson Palmer: Cincinnati Bengals star who returned to form after surgery |
The numbers are already being crunched – and proving useful in planning which treatments might best serve not just the player, but the team. A recent study of hamstring injuries found that every new hamstring injury costs the team an average of 14 playing days; an average recurrence costs 25 days. Furthermore, almost all the recurrences took place during matches in the first month after return, and after an hour of play. It quickly became clear that players who had sustained hamstring injuries should be replaced after an hour during their first few games back. Moreover, the two clearest risk factors for hamstring injury were age and a previous injury, and players who performed specific strengthening exercises reduced the incidence, severity and recurrence of hamstring injuries.
It’s work like this that has helped the RFU halve injury recurrence rates between 2003 and 2007. Ian Beasley, senior doctor at the Football Association, looks at the data sets with envy. “I can’t really understand why we don’t run a proper audit,” he says. One difficulty is getting clubs to agree: each team, after all, needs to devote resources to collecting the information and entering the stats, at the very least – let alone finding people to analyse the data. “Hopefully,” says Beasley, “we can get someone attached to do that.”
It is eyebrow-raising that rugby – and cricket – can afford audits but the FA cannot. According to Stefan Szymanski, an economist at Cass Business School who recently co-authored, with FT columnist Simon Kuper, Why England Lose & Other Curious Football Phenomena Explained, “You have to look at where the money is and why it’s being spent. With the England and Wales Cricket Board and English Institute of Sport, the money is being spent by people whose target is national success.” The ECB pays each county £20,000 a year to improve their medical practices. “The Rugby Football Union is more balanced between the clubs and England, but England money is still very important, which helps England set the agenda. Also, rugby is a dangerous game, so there’s a responsibility to the players. Therefore, these organisations can run sport-wide programmes.”
Some information is published – more by the RFU than the others, echoing its wider concerns and more balanced financial structure – but these audits primarily exist to help English sport maintain a competitive edge over its rivals.
The financial situation in football (and the NFL in the US) is radically different. What might be feasible for Chelsea’s sophisticated medical team is not feasible for a lower league club with one physio and a part-time doctor. And even within the Premier League, Chelsea have vastly more money than Hull. The FA can’t demand compliance because, points out Szymanski, “[the] England [team] is an economic sideshow. The clubs have the muscle. It might be great for Leyton Orient to pool medical information – the losses in competitive edge would be more than outweighed by efficiency savings – but the same isn’t true for Chelsea.”
AC Milan’s famously secretive players’ lab, which routinely coaxes great players into their mid- and late-30s, gives the club a big advantage. Why would it share? It is safe to assume the same holds true in the NFL. For all their money, the 32 clubs don’t have access to the sheer quantity of data that less wealthy sports are starting to gather.
. . .
The defining text for the new breed of sports statisticians is Moneyball, Michael Lewis’s tale of how a new understanding of scoring efficiency helped an underfunded US baseball team punch above its financial weight. In a forthcoming piece for the British Journal of Sports Medicine, John Orchard speculates that football medics might be facing their Moneyball moment.
Orchard says that differences in skill between the top clubs’ players are so slight that he’s surprised teams don’t seem to appreciate “what seems obvious from the outside: injury outcomes will be a key factor in determining who wins the Premiership each year”. It has become almost hackneyed to say that squad depth is crucial and that Liverpool don’t score if Gerrard and Torres aren’t playing, and yet still there is no published study exhaustively trying to link injury with success, and seeing if medical treatment makes a difference.
According to the RFU’s injury report for 2002 to 2004, the teams finishing in the top four places of the league lost an average of about 600 player days per season. The next four lost something like 800. The bottom four lost closer to 1,000. The situation is hideously complex – better managed, better funded teams, which you would expect to do well, tend to have better medical programmes as well – but it is hard to imagine that causation doesn’t run both ways.
Back in the US, Bill Barnwell and Football Outsiders crunch the numbers to show an even more graphic link: of the 10 healthiest teams in 2008’s NFL, seven made the playoffs. A huge 26 per cent of a team’s win rate from year to year can be chalked up to a change in injury rate. In spite of this, says Barnwell, “teams rarely credit the absence of injury for their success”.
Someone has to do this kind of analysis on football. Orchard argues that a team spending £100m on wages with an injury rate of 22 per cent will probably beat a team paying £25m with a rate of 12 per cent, but statisticians can control for that. If better medical teams really do keep players on the pitch, then management (and fans) might finally feel that £1m invested in better medical care has more marginal utility than £4m spent on another spare midfielder, “just in case”.
If injury incidences and outcomes were blind luck, then there would be nothing to do except focus on treatment. But they clearly aren’t. The RFU’s success with hamstrings is just one example. Simon Kemp knows rugby will always be dangerous, but he hopes that better conditioning might even help players come through some of the collisions which tear anterior cruciate ligaments.
“How?” I ask.
“We don’t know yet,” he replies.
Knees will never be impervious to helmets, but audits can protect future Tom Bradys in other ways. The RFU data have demonstrated so clearly that illegal tackles disproportionately cause injuries that they have helped persuade referees to apply the law more strictly. The NFL is tightening up on knee-level tackles post-Brady – but perhaps an ongoing data collection programme would have caught the problem earlier.
If so, then going by Brady’s salary – the crudest measure of his worth – it would have been at a fraction of the cost of his absence, let alone those of the other quarterbacks whose recent seasons might have been saved. As Tom Brady runs blinking into the floodlights this season, he won’t be thinking about value-for-money risk and resource management; it’s not his job. But somebody should be.
Robert Hudson is author of ‘The Kilburn Social Club’, published by Jonathan Cape
....................................................................
Rugby Union’s most common injuries
The data illustrated here, collected by the RFU (and featuring Gloucester and England’s Olly Morgan), show the highest-risk training injuries, measured in days missed, as suffered by rugby backs. Injuries are also broken down by the type of problem most associated with a particular match event, eg the commonest injury for a forward to suffer as a result of a scrum is calf strain, while a back is most susceptible to a thigh haemotoma as a result of a collision
....................................................................
He stats, he scores
Statistics have been popular among sports fans and coaches long before Freakonomics-style analysis took off, writes Alex Cardno. Walks, hits, runs … every player comes with a compendium of facts and figures. About eight years ago, however, Nate Silver, a Detroit Tigers baseball fan, began using these figures differently. Rather than judge players on batting average alone, for example, Silver developed a series of complex algorithms to predict the outcome of a player’s career and the success of various teams. His predictions, including that the usually dismal Tampa Bay Rays would win more than 80 games in the 2008 season, proved correct time and again. This consistency soon gained him a large following of baseball fans.
In 2003, Silver sold his main algorithm to Baseball Prospectus, publisher of a website devoted to detailed analysis and advanced statistics on the game. Silver received a stake in the site and stayed on as a writer and consultant until late 2007.
Then in March last year, he launched a statistics-driven politics blog, FiveThirtyEight.com – named after the number of votes in the US electoral college. Silver built a database of political polls and weighted their results against historical fact. He noticed patterns emerging and applied the techniques he had used with baseball statistics to predict elections.
The website’s profile rose dramatically as he accurately predicted Democratic primary elections in state after state. When it came to the presidential poll, he predicted the vote of 49 out of 50 states, and was the first to declare Obama’s victory, beating professional pollsters.
Silver, now 31, continues to publish FiveThirtyEight, applying statistical analysis to major policy issues. “There are some things [where] you can’t be quite as robust as you can be in forecasting an election, but, for example, there was a climate change bill that was passed, barely, by the House [of Representatives], where we can look at the characteristics of which House members voted for it and which didn’t, and try to predict what is going to happen in the Senate.”
Silver is a regular on US news channels and made Time magazine’s 100 Most Influential People List in April. Last November he signed a book deal with Penguin (owned by Pearson, the FT’s parent company). His first book will focus on his predicting methods, picking up where Freakonomics left off. “It should be a book that’s both serious and fairly rigorous but also fun,” he says. It’s due for publication late next year, by which time Silver’s name may be known round the world: as this magazine went to press he wouldn’t give details, but he did reveal that his next website venture would focus on the 2010 FIFA World Cup.
To read Alex Cardno’s interview with Nate Silver, go to www.ft.com/natesilver
