Inverting the Pyramid: The History of Football Tactics

by Jonathan Wilson

  ‘The results of this analysis were used in the playing method of all the international teams I managed between 1964 and 1974 - seventy-seven matches in all. The essence of this work was published in 1973 in a book entitled Tactics and Teamwork and a series of eleven films under the same title.’

  Well, maybe the results were, but the work is light on statistics and preaches no unified philosophy. It is hard to be sure either way, but it is easy to see why Reep was suspicious. As the Norwegian academic Øyvind Larson points out, ‘The style of play recommended immediately changed from being general in nature to being penetration-based in particular.’ Hughes had been an ordinary pragmatic coach; he became, to use the term applied by Howard Wilkinson, his successor as FA technical director, ‘a zealot’. There were differences in terminology and weighting (Hughes specified five passes or fewer rather than three), but Reep claimed in his letter to Olsen that that was because Hughes was unaware of some of the calculations he was performing.

  ‘The work of analysing football matches has continued ever since,’ Hughes went on. ‘In the early part of 1982 [Reep says 1981; Taylor can’t remember] I had the pleasure of meeting a wonderful man named Charles Reep, who had been analysing football matches for thirty years, successfully advising a number of Football League clubs.’

  Hughes claims that he then got Primus to begin analysing games, using her shorthand skills. ‘The method of analysis itself, which I devised around twenty-five years ago, is actually different from Charles Reep’s… Although Charles Reep and I had come by our strategic philosophy by different routes, there was no disagreement on the major conclusion.’

  There was significant disagreement between the three personalities involved though: Taylor denied Hughes, who denied Reep, who blamed Hughes. To both Reep and Taylor, Hughes is somebody who exploited their ideas for his own ends, gaining in reputation and selling his books and videos. Perhaps that is a familiar tale of committee-room back-biting, and of the impossibility of copywriting ideas, but it meant that by the time Taylor was appointed England manager in 1990, his relationship with Hughes was unworkable. It hardly helped either that Hughes was furious at having lost out to Graham Kelly in the battle to succeed Ted Croker as chief executive of the FA in 1989.

  Reep had written to Taylor on 4 August 1980 - by which time he was seventy-five - explaining his theory that ‘all goals come at random within a framework of probability’. They met in Exeter later that month for two hours, after which, Reep wrote in a piece in the Scottish football magazine The Punter, ‘any time Graham Taylor wanted to ring me to get further details regarding aspects of the style of play, I would be willing to talk for as long as he wished. We did in fact have several very long indoctrination talks, during which a lot of ground was covered without the need to meet each other again.’

  Reep did not see Watford play that season, but, he says, he met Taylor again on 11 March 1981 at Reep’s house in Plymouth. Receiving reports from Richard Pollard, a Watford fan who had co-authored an early article with him, Reep concluded that, for Watford, ‘only one goal in five comes from passing moves containing more than three received passes.’ In other words, Watford’s goals fitted the pattern Reep had been demonstrating for thirty years - 80 percent of goals come from moves of three passes of fewer (which, of course, is still less than the 91.5 percent of moves that, he showed, consist of three passes or fewer).

  One of Reep’s other constants also remained true: that it took roughly nine shots to produce a goal. Convinced of this, he maintained records of how many goals Watford and their (combined) opponents were ‘in credit’ or ‘overdrawn’ (so if a team had had ninety shots he would expect them to have scored ten goals; if they only managed eight, they were two in credit; if they had had scored twelve, they were two overdrawn). In the first leg of a second round League Cup tie at Southampton, Watford lost 4-0. Reep, noticing ahead of the return leg that Watford were 2.5 goals ‘in credit’ while combined opponents were 4 goals ‘overdrawn’, wrote to Taylor, advising him that this could be evened out in one game, and that Watford should therefore pursue an attacking approach.

  This, of course, is absurd. Taylor, presumably, would have worked out for himself that to overhaul a four-goal deficit, his side needed to attack. They did, led 5-1 after ninety minutes, and added two more in extra-time. ‘I was enchanted to think,’ Reep wrote, ‘that the first time one of my teams had attempted to exploit the calculated situation regarding “goals in credit” and “overdrawn”, it had succeeded brilliantly well - always with the admission, of course, that random chance had done us a huge favour. And as might be expected, random chance penalised Watford by awarding us a 1-2 defeat in the next game.’

  And so the lack of mathematical basis to Reep’s research is betrayed. ‘Random Chance’ is not a deity handing out or denying goals to level up some cosmic balance. It is just random. Toss a coin a hundred times and if the first ninety-nine land heads, the odds of the hundredth landing tails are still one in two. If there really is - and there isn’t - a one in nine chance of scoring with any shot regardless of circumstance, it is still one in nine whether a forward has scored with his last ten or missed with his last hundred. That assumes the coin is unbiased, of course. If a coin keeps on landing heads, it is probably because it is weighted; if a striker keeps on missing chances, it’s probably because he’s not very good.

  Nonetheless, ahead of the 1981-82 season, Taylor decided to employ one of Reep’s trainees - an archaeology graduate from the University of Lancaster called Simon Hartley, who had become intrigued by Reep’s ideas after seeing him taking notes one day at Plymouth. Reep did not speak to Taylor by telephone that season, but he did write him three letters, one of which dealt with the lack of goals Watford were scoring from the right wing (John Barnes scored thirteen from the left) and hinted at, but did not fully reveal, his plan for how wingers should play. As Watford were promoted that season, 93.4 percent of their goals came from moves of three passes or fewer. Reep notes this was ‘superb’, although if the number of moves consisting of three or fewer passes remained constant at 91.5 percent, of course, it is only just higher than would be expected if the number of passes in a move made no difference. Given Watford were a self-consciously direct side, it is probable that for them a greater proportion of moves consisted of three passes or fewer: even there, in other words, there is still little evidence for the greater efficacy of direct football.

  Reep claims he and Hartley were both paid a £6,000 bonus, but that he then fell out with Taylor over his fee for the following season. Taylor’s memory is that they fell out over an obscure statistical point. Reep was obsessed by ‘reachers’ - that is, balls that landed in the final third. Watford averaged 156 a game, although against Chelsea on 6 February 1982, a game they won 1-0, they achieved a record of 202 (this was later taken by John Beck’s Cambridge United, with 219). Stan Cullis’s Wolves side had managed around 180 a game, and Reep urged Taylor to try to raise Watford’s level to match that. Taylor pointed out that his side’s main strength was that they regularly won the ball back in the final third, something that did not count as a reacher under Reep’s system, and suggested the figures should reflect that. Reep refused to change and, although Hartley stayed on for another season, his personal association with Watford came to an end. ‘With Reep,’ said Taylor, ‘it was all or nothing. There was no room for compromise.’

  Reep may not have liked him, but it was Hughes, as director of education and coaching at the FA between 1983 and 1994, who ensured his principles - or at least Hughes’s version of them he set out in The Winning Formula - became enshrined at the highest level. That book, in a link-up beyond the dreams of satirists, was sponsored by British Aerospace. ‘The strategy of direct play,’ Hughes asserts in his introduction, ‘is far preferable to that of possession football. The facts are irrefutable and the evidence overwhelming.’ Some might suggest that the record of an England team largely shaped by that philosophy is in itself a refutation, but then foot
ballers are fallible: statistics are not.

  Noting that the average number of goals per game in World Cup matches fell from 5.4 per game in 1954 to 2.5 per game in 1986, Hughes passes almost immediately to the conclusion that ‘football is not as good as it was’. That a man whose authority came from the supposed application of reason and logic should be allowed to get away with such a leap is staggering. Discerning the quality of football is necessarily subjective and, anyway, there are bad 4-3 thrillers (excitement and quality are not synonyms) just as there are superb goalless draws. If goals alone were a mark of excellence, there would be thousands queuing to watch primary-school football.

  The reason for the decline in goals scored, Hughes goes on, ‘lies not in new efficient defensive strategies so much as a misguided attacking strategy, that of possession football.’ Now, certainly, as Chapman argued and as a glance at the statistics for any given weekend of Premiership football will show, there is no necessary correlation between domination of possession and winning games, but neither does that mean that possession is a bad thing. Yet Hughes, deploying that word ‘overwhelming’ again, argues that ‘the fact is that the longer a team takes to build an attack when it has possession of the ball … the more time the defending team has to recover, regroup and reorganise.’

  In The Winning Formula, Hughes uses the evidence of 109 matches between 1966 and 1986 in which 202 goals were scored. That, it might be noted, is not a huge sample, particularly not for somebody basing on it the claim that ‘world soccer has been moving in the wrong strategic direction for the better part of thirty years’. It is also tempting to ponder the significance of the fact that while Hughes rails against a World Cup that produced 2.5 goals per game, the matches in his sample produced only 1.85. But still, the results are intriguing, and they have, presumably to head off those - like Taylor - who argue that direct football is ineffective at the very highest level, been filtered to include only successful sides: Liverpool, the England Under-16 and Under-21 teams, and World Cup or European Championship matches involving Argentina, Brazil, England, Holland, Italy and West Germany.

  Of those 202 goals, fifty-three came from moves of no received passes, twenty-nine from one pass, thirty-five from two passes and twenty-six from three passes. In total, 87 percent of the goals came from moves of five passes or fewer, while fewer than 3 percent came from moves of ten or more passes (if Reep’s statistic that 91.5 percent of all moves consist of three passes or fewer is correct, of course, that is still no endorsement of direct football). Then there is the issue of how many of those goals scored from three passes or fewer are the result of breakdowns brought about by longer moves. Hughes anticipates the question and presents his figures with a misplaced sense of triumph. Of nineteen no-pass goals (that is, penalties, free-kicks struck directly into the net, or shots fired in from a rebound off the goalkeeper or after a tackle or misplaced pass by the defending team) analysed by Hughes from sixteen England international matches, only twelve resulted from moves of three passes or fewer - 63 percent: far fewer than Reep’s benchmark of 91.5. The question then becomes not whether Hughes is right or wrong, but how he got away with it for so long.

  Perhaps not surprisingly, Brazil were the side most likely to score after a long string of passes, 32 percent of their goals coming from moves of six passes or more, with West Germany next on 25 percent (given, at the time, they had won between them six of the thirteen World Cups to have been played, that might be taken as an argument in favour of possession football). Almost unbelievably, none of the ten Dutch goals under consideration came from moves of six passes or more. And that’s when the alarm bells really start to ring: why are there only ten Dutch goals under consideration? They scored fifteen in the 1974 World Cup finals alone. This is not only a small sample, but it is a selective one (‘we have extracted 109 games from all those analysed’), and nowhere in The Winning Formula is that selection process explained.

  Even assuming there is nothing sinister in that, Hughes is, at the very least, guilty of identifying a symptom and not a cause. ‘The first objective,’ he states in his conclusion, ‘is to get into the attacking third of the field more often than the opponents do, and the final objective is to achieve a minimum of ten shots on target every game… If the strategies we have proposed are adopted and the tactical objectives are achieved, the chances of winning are extremely good - over 85 percent. The chances of not losing are even better. We have never recorded a match in which a team achieved ten shots on target and lost.’ Yes, but are the shots really the reason for that? Or are they simply a natural result of one team dominating? Do teams win games because they’re having shots, or do they have shots because they’re winning games?

  Where Hughes does have a point is in his advocation of pressing. Chapman and Helenio Herrera had great success by getting their teams to sit deep, but in modern football pressing is almost universal. ‘If a team can increase the number of times they regain possession in the attacking third,’ Hughes says, ‘they will score more goals.’ His statistics show that 52 percent of goals were scored when possession was gained in the attacking third, as opposed to 18 percent in the defensive third, and that moves begun when possession is regained in the attacking third are seven times more likely to produce goals than when possession is regained in the defensive third. Now those figures are skewed, clearly, by occasions when moves break down in the attacking third and the loose ball breaks for the attacking side, but they are a striking vindication of pressing nonetheless. Pressing, of course, has the added advantage of smothering an opposing attack almost before it has begun. This is the basis of Taylor’s disagreement with Reep. Pressing is an action that can help a team win a game; getting the ball into dangerous areas is something that happens as a result of those actions.

  Hughes goes on to argue that a team should shoot whenever it gets the opportunity, noting that ‘even at the highest level, more than half of all shots miss the target, so players should never shy away from shooting for fear of missing the target’. Should a player really shoot even if a team-mate is in a markedly better position? Should he always shoot from 20 yards? From 30? From 40? Hughes argues that even mishit shots can create scoring opportunities, which is true, but why turn the situation into a lottery if a well-placed pass would significantly increase the chances of the initial shot, when taken, being a telling one? It is as though there is a distrust of technique, a fear that by adding an extra element into the move, the chances of it going wrong are increased to a level that makes relying on a lucky bounce or deflection preferable. Allen Wade, the technical director of the FA from 1963 to 1983, was no aesthete, but he was horrified by his successor’s dogma. ‘This slam-bang-wallop stuff will be the death of football,’ he said. ‘Football in which players are controlled by off-pitch Svengalis, backed up by batteries of statisticians and analysts, will never hold the magical appeal of what Pelé called the beautiful game.’

  There is, perhaps, a nod here back to the days of the head-down charging of the Victorians: where Italian paranoia led to catenaccio and a faith in strategy over ability, English insecurity led to a style that similarly distrusted ability, favouring instead a thoughtless physicality - keep battling, keep running, keep trying. As the German journalist Raphael Honigstein put it sardonically in the title of his work on English football, ‘Harder, Better, Faster, Stronger’. But not more skilful.

  Although Hughes advocates increasing shooting practice to improve accuracy - and, supposedly, demonstrates his success in that area with England Under-16s, although his sample includes only four matches - this surely is just the kind of argument that had so appalled Jimmy Hogan almost a century earlier at Fulham: buy enough tickets and you’ll win the raffle eventually.

  Quite aside from that there is a baffling lack of subtlety to Hughes’s work. He claims that by applying his ‘formula’ - which arguably just comes down to being better than the opposition anyway - a side has a greater than 85 percent chance of winning. The question then is whet
her there is a pattern to the other 15 percent. What Hughes’s statistics do not show is the possibility Taylor accepts, the possibility that feels intuitively true, that direct football can take a team only so far, that there comes a certain level of opposition capable of keeping the ball, capable of controlling possession, against which it is rendered ineffective. Brian Clough was typically unequivocal. ‘I want to establish without any shadow of a doubt that Charles Hughes is totally wrong in his approach to football,’ he said. ‘He believes that footballs should come down with icicles on them.’

  That is an exaggeration, but the flaws were obvious enough. Chapman’s direct approach worked because he lured teams out and encouraged them to leave space behind them that his team then exploited. Taylor’s Watford, with their high-tempo pressing game, were vulnerable to just that kind of attack. Hughes’s formula makes no distinction as to the style played against. Given Taylor’s direct approach foundered against a Sparta Prague side adept at retaining possessing and launching astute counter-attacks, wouldn’t Hughes’s have too? Organisation and energy, Taylor found, will carry a team so far, but only so far.

  The irony is that while Taylor was well aware of the defects of the method, it was he who was left to reap the pitiful harvest after Hughes had implemented Reep’s principles as FA policy. Yes, there were players missing through injury, but still, have England ever sent out a weaker team in a major championship than that which lost 2-1 to Sweden in the final group game of Euro ’92: Woods; Batty, Keown, Walker, Pearce; Daley, Webb, Palmer, Sinton; Platt; Lineker? To rub salt into the wound, when England then failed to qualify for the 1994 World Cup, they were knocked out by Norway, a side practising the Reepian model taken to its extreme.