15 August 1987, Upton Park. West Ham v QPR on the opening Saturday of the season. West Ham had finished 15th in the First Division the previous season and QPR 16th; no-one expected much more than the usual rough and tumble of a London derby. And yet a significant piece of English football history was about to be made. 

QPR lined up in a 3-5-2 system, with wing-backs, two man-to-man markers in central defence and a sweeper. It was the first time a major club side in England had opted for the formation as a first-choice strategy and, perhaps more significantly, it worked. QPR won 3-0, and went on to win six and draw one of their opening seven games. In a world that had been dominated by 4-4-2 since the 1960s, this was a radical departure and it took QPR to the top of the league.

"I first got the idea from watching European football on the TV, particularly the Germans," said QPR's manager, Jim Smith, who was already 18 years into his eventful management career. "I thought it was a great way to play."

When I asked whether he went over to Germany to watch matches or consult with other managers, he laughed. "At Oxford, they couldn't afford to send you to Carlisle," he said. It was when he'd been manager of Oxford United in the early eighties, though, that he first tried playing with three at the back.

"In particular games," he said, "when we were in trouble and needed a goal, we'd go three at the back, and push another man up into the attack to go 3-4-3. I can remember some games where it helped us to get a draw from a defeat, or a win from a draw." Before he left for QPR in 1985, Oxford gained successive promotions from the Third Division to the First. Smith doesn't believe the system was particularly significant in their success, as they only used it on half a dozen occasions, but he'd become convinced of its usefulness.

After finishing 13th and 16th in his first two seasons at QPR, Smith decided to take the plunge. "At the time, in England, there was such a lot of hostility about a sweeper system," he said. "I told my coach, Peter Shreeves, and the players that I wanted to go to a three, and they didn't like the idea at all. I had to promise that we'd go back to a four if it didn't work."

Before the opening league game, Smith was very aware that the new formation was unlikely to survive a defeat. He got lucky. "In pre-season, I'd bought Paul Parker from Fulham as a wing-back, but I got a bit worried because West Ham had Cottee up front, who was very fast," he said. "Parker had a lot of pace and I decided to use him as my marker instead. It turned out that he was ideally suited to the position. That game was the making of him, really, and he went on to become an England player." 

Behind Parker and the solid Alan McDonald, Smith used Terry Fenwick as his sweeper. "He was a leader and organiser, and loved that position. It's also important in the system that you have defenders who don't mind going wide, to help the wing-back if necessary. Most centre-backs don't like it, but Fenwick and Parker were comfortable."

Success bred confidence, and although they were knocked off the top by a 4-0 defeat at Anfield, Rangers maintained their form and finished the season fifth. Bewildered teams struggled to contain their wing-backs, although as the season progressed, other managers gradually developed a counter-strategy. "They'd use wingers to double up on the wing-back," Smith explained. "If you're on top of your game, one of the three can go across to help, and the other full-back just tucks in. The problem was, we weren't a major club and we didn't have a large squad, and a difficulty of the system is that you need players who are familiar with it for it to work. That's why, later on at Derby, I got the reserve team and the youth team to play 3-5-2 as well." 

Imitators quickly followed, although, somewhat to Smith's exasperation, mainly among clubs who were struggling. "Our goals against was very good, so many teams saw it as a way of staying in the First Division," he said. "I always played it as an attacking system, but they'd often end up with a five at the back, which I've never liked, because when you get the ball, there's no-one to pass to." Nevertheless, Smith's experiment was a breakthrough, if only because he had demonstrated that British players did not have to be confined to 4-4-2 or its close variants. Two years later, Bobby Robson's successful use of 3-5-2 in the 1990 World Cup was the final endorsement.

The system has gone in and out of fashion since. Smith believed that it should be used more widely, and felt that the conservatism and caution of many English players was an obstacle. "Many of our defenders are very reluctant to try anything except what they've already been taught," he said. "They also like 4-4-2 because they have people around them. Full-backs want their winger to help them out and centre-backs don't want a sweeper behind them, they want him alongside. You need the right players who can deal with one-v-ones." 

Smith admitted that the system is harder to coach because defenders need to make more decisions for themselves on the pitch and are less reliant on a pre-set structure. Fluidity comes at a price and particularly in that most exposed position of all, the sweeper. He sometimes had to import players from abroad who were more familiar with the role, such as Taribo West at Derby. For Smith, fluidity was key to any formation and, perhaps not surprisingly, he has admiration for Arsène Wenger's Arsenal and Louis Van Gaal's Ajax. 

When Jim Smith made the change from 4-4-2 to 3-5-2, he was not just replacing one set of lines with another, he was drawing his full-backs and sweeper away from any defensive or midfield line. Effectively, he was challenging the whole team to stop thinking in terms of lines and to improvise to a much greater degree. Those who didn't grasp this drifted into a straight back five. The extra emphasis on improvisation also demanded that players retain possession and pass accurately on the ground, rather than hit the ball hopefully into space. 

Although the 3-5-2 formation did not take a lasting hold on the English game, there is now far greater flexibility, far fewer teams reliant on a basic 4-4-2. Four at the back may have remained, but attacking formations have become more fluid. It is tempting to trace this trend back to the quiet revolution that Jim Smith began all those years ago, when he challenged his players to take the initiative and absorb new ideas.

I asked Smith if there were any reason he had been the first to take the risk, but he could offer no explanation. "You just study formations and systems, and try them out," he said. It was in his nature as a manager to look outward and not inward, forward and not backward. It was also always evident that Smith loved football management, and to love management you must relish the tricky decision. Smith's appetite for the bold stroke, the choice of player or formation that would give his team an unexpected advantage, was clear. In his career, he was always prepared to take risks.

There is usually a gap between the hopes of the dreamer who loves to see football at its most vibrant, and the view of the professional who must make things happen within the harsher realities of an imperfect world in which the sack is always waiting. Smith, though, managed better than most to reconcile his vision with the reality. When he cajoled his players at QPR into sharing his vision, they would have experienced him not as an unrealistic theoretician, but as a man who talked their language and who could handle their doubts. 

Smith was successful, but not lucky in his career. He turned several moderate sides into good ones, but was never given the chance to turn a good side into a great one. But even if fate denied him the major opportunities and the major prizes, perhaps we can at least offer this genial Yorkshireman the recognition he deserves as a major innovator in the English game.