We can’t transport Diego Maradona into the present or push Lionel Messi into the past. So how can we bring Argentina’s luminaries together? By taking the innovative methods we use to assess Messi now back in time to do the same with his predecessor. Deep data provides us with such a time machine. We’ve gone back and scored matches from the 1986 World Cup so we can better compare a 25-year-old Maradona in Mexico to a 26-year-old Messi in Brazil. Spoiler alert: Each player was great in his respective tournament and better than his compatriot in certain areas. So what made Maradona a World Cup champion while Messi fell a goal short?
Diego Maradona committed at least two divine acts in 1986, which also happened to be the year in which his second-coming was conceived. Despite what history has made of it, neither of those acts was winning the World Cup as a “one-man team.” Rather, those two acts resulted in goals. Measurement was essentially left at that – goals – because football data didn’t allow for much more, and the story and legend filled in around those goals as time eroded the less-memorable minutia that makes football tick.
Two goals. One famous. One infamous.
Years on, one of those acts is shown as a wordless argument in favour of Maradona being the greatest footballer of all-time, and it’s known as the Goal of the Century. What’s divine about it? If you’ve seen it once, it’s memorable to the point of not needing a video refresher in order to see it in your head in great touch-by-touch detail from there on. The other is shown to illustrate his controversial nature, and it’s known as the Hand of God – a finish on which he at worst cheated to progress in the World Cup or at best inadvertently broke the rules of the game. They occurred four minutes apart in a match that came to define the magnificent prime of Argentina’s sporting hero.
What’s been said about that tournament, those goals and the player who scored them has been said over and over for three decades with little forward momentum as the detail of the team around Maradona is gradually overlooked more and more. We can tell you now with far superior objectivity that Maradona was prolific but didn’t act alone, and we can numerically state just how effective he was throughout the tournament within more meaningful context.
But we can’t stop with him because Argentina hasn’t stopped producing exceptional talent. In the years since winning in Mexico, Maradona’s home country birthed the player who might most closely rival – or narrowly surpass – him for the title of the greatest footballer of all-time.
Lionel Messi came close four years ago to committing another divine act: leading Argentina to a World Cup title on Brazilian soil. He fell minutes short, and that unfulfilled act of winning a major trophy for his country is the only blemish on an otherwise otherworldly career. But is it right to call it a blemish if he didn’t have the team around him to finish the job?
The discussion of who was – or is – the greatest includes a few more names that breach the Argentine border, and it’s a discussion that’ll never be conclusively resolved. So rather than try to offer answers to the unanswerable, let’s at least celebrate what’s now possible and bring the past into the present as best we can. As best we can is achieved through deep data. Lots and lots of it. Advancements are making it quantifiable and contextualised in football in ways few imagined when Mexico hosted that 1986 World Cup that Maradona eventually hoisted.
The world watched exceptional talent then as it does now with his compatriot, but it was left as that. Now, the STATS operations team has gone back and – as best as that old, grainy video would allow – given Maradona credit for everything we now give Messi. At times, it has the reverse effect and, in roundabout fashion, assigns credit to Messi for what Messi is doing now because it gives him a player of comparable greatness to measure against.
In doing so, we’ll provide objective data to arm you with discussion points on each player and each Argentina side. How did they differ as teams and players stylistically in those tournament spans? Who converted more difficult chances, and who left better chances on the pitch? Who was more dangerous or wasteful with their ball movement? Who did more in their respective final, and which faced a more daunting final opponent? Who had more talent playing with him, and who was really more of a DIY player?
What we won’t tell you is who was better. Some things, after all, must remain subjective.
What’s changed most about football over the last 32 years? Easy: The shorts. We never needed that much thigh, Diego. But the ways we engage and analyse are a close second. Football in its essence hasn’t changed, but what we know now is not equal to what we knew then. There’s also been a change in the number of teams in the World Cup field, which has a way of diluting group stage competition.
We’re going to attempt to bridge those disparities here by analysing a five-match sample size from both the 1986 and 2014 World Cups for each Argentina side. Why not assess the entire tournament? Because wholly comparing competition from a 32-team tournament and a 24-team tournament probably doesn’t make for an even analytical playing field, particularly when considering group stage matches.
Accessing the data wasn’t an issue for 2014 because STATS scored the entire tournament at the highest level of event data available today. For 1986, we went back this spring and scored one group stage match and four knockout phase contests using STATS’ post-game tool to achieve that same level of specificity.
To make things as even as possible, we went to the rankings available before choosing the matches. FIFA didn’t have its current world ranking system in 1986, but it did publish a retrospective report that ranked all teams in each World Cup up to and including Mexico ’86. If we evaluated the entire tournament, the average ranking of Argentina’s ’86 opponents would have been 11 and 17.1 for Argentina ’14. This led us to settle on the following breakdown:
For 1986, this comes to a mean opponent ranking of 8.4 by eliminating group matches against South Korea (20) and Bulgaria (15). We could have gotten the average rankings fractionally closer by including Bulgaria over Uruguay in 1986, but we wanted to include the entire knockout phrase for each World Cup.
For 2014, it comes to 7.4. That eliminates group matches against Nigeria (33) and Iran (49). Yes, Messi scored three goals in those matches, which we’ll still discuss in some detail. But fair’s fair, and remember, part of the point of this is to get beyond goals. Before we get into those individual comparisons, it’s necessary to assess the quality and styles of the teams on which each star was playing.
Messi’s Argentina began the 2014 World Cup as a dominant side, and that briefly carried over into the knockout phase. Even in their 1-0 extra-time win over Switzerland in the round of 16 – a close call at a glance – they managed 30 shots, nine on target, a 2.9-1.0 expected goals advantage, and 60 percent possession while conceding four shots on goal in 120 minutes.
That changed as the tournament wore on and competition picked up. Argentina managed two shots on goal a round later in a 1-0 win over Belgium, and some would say they were fortunate to reach the final after the Netherlands out-passed Argentina 729-576 in a match that was light on scoring chances as Argentina managed a 0.4-0.3 expected goals advantage. The final, which we’ll assess in greater detail later, was indisputably in Germany’s control.
Maradona’s Argentina were steadier and probably played a bit more of a conservative style, which certainly doesn’t corroborate the lasting image of their superstar’s international pinnacle. They didn’t sweep through their group and dropped points against Italy, but they didn’t experience as drastic of a dip in form as they progressed, which we’ll show when we compare the two finals.
Nevertheless, there’s probably always going to be someone out there willing to assertively tell you Maradona is the greatest footballer of all-time because he was able to do it alone. And there’s probably someone out there willing to make a similar argument about Messi. So how can we substantiate or challenge these claims? We can start by considering how their teams stacked up against the five opponents in question. Here come the numbers.
Something to note before assessing these numbers: Argentina ’14 went to extra time in three matches, so within this five-match sample, they played an entire extra game. Because of that, per-90 rates are included parenthetically:
If we look at categories such as possession, ball events, successful passes and bad passes, goals against and expected goals against, it might seem Messi’s team was the best of the bunch. They were fantastic defensively. If we look at chances created and converted (shots, goals, expected goals), we’d assume Maradona’s group was the best of the bunch. It’s never going to be that simple, especially when we’re trying to determine which Argentina team gave their superstar a better chance to succeed.
We can begin to dig deeper with location. What’s interesting to consider with this specific level of data is where a team’s average touch occurred. It might seem strange to see a team that reached the final with a possession advantage to have an average touch position behind the centre line, but that’s exactly what the 2014 team did with an average touch 1.3 metres withdrawn from midfield against the five included opponents. For the sake of comparison, Germany’s average touch for the tournament came 1.7 metres advanced of midfield.
But Argentina ’86 played withdrawn to far more of an extreme in a 3-5-2 formation. For ’86, it was a matter of style instilled by manager Carlos Salvador Bilardo’s very un-Argentine, un-South American and late-80s, early-90s European nil tactics of first focusing on not conceding. Their opponents often did the same with Italy taking their average touch 8.0 metres behind midfield in their group stage draw with Argentina. The beauty of “conservative” play for this Argentina team was it produced plenty of chances with an expected goal value that led the four groups considered in the table above. What’s more is they – and in particular Maradona – outperformed that expected goal value.
West Germany were the only ’86 opponent of the five considered to have an average touch advanced of the centre line (1.6 metres) against Argentina. It’s interesting to note how the data corroborates the stylistic stereotypes typically assigned to Italy and Germany through the years.
This is a start, but it doesn’t really lead us to much in the way of answers. To get there, we can assess the quality of ball movement and look at the particular styles in which these teams played. This was impossible to do on a meaningful and objective level with traditional football data, but that’s changed.
To do this, we’ll start with Ball Movement Points. The concept of Ball Movement Points captures players’ contributions in terms of their ball distribution and possession regains with objective values based on past results for the relevant pitch locations from six years of data. As a calculation basis, a match is split into ball moves made by a player from a start zone to an end zone. Each ball move is then assessed regarding the danger it poses to the opposition. It goes beyond expected assists by looking at the full chain of passes, weighing the probability of that pass leading to a shot later in the play. It’s broken down into categories of offensive and defensive as well as positive and negative (oBMP+, oBMP-, dBMP+, dBMP-) with net values telling the more conclusive story.
Offensive BMP is something of a creativity index, and because we’re ultimately assessing Messi and Maradona, it’s what we’ll focus on here. Positive oBMP is our way of differentiating simple successful passes lacking ambition from those that move the ball to more dangerous locations of the pitch. Negative oBMP relates to the value of possessions lost or wasted by players. These scores accumulate during a match or across a competition. Large net values imply that the player makes more successful passes in areas closer to the opposition’s goal and/or that the passing opportunities that are realised outweigh those that are wasted. It’s what football minds could always see but never objectively calculate, and it is imperative in assessing dynamic attackers such as Maradona and Messi, and in turn, the systems in which they operated.
Here, we’re charting Ball Movement Points as well as the specific styles in which the four groups operated. We’ll explain specific styles as we go, so don’t feel the need to attempt to make sense of all this immediately:
The first thing that stands out here is the lack of threat and ambition for the 1986 opponents in terms of BMP. Recall above that this group had the highest level of possession among the four considered, yet they have the lowest net oBMP. This makes sense because they had the lowest rate of successful passes and highest rate of bad passes, but it goes deeper than that.
How was it they maintained the highest possession percentage among the four with Argentina ’14 close behind? It likely has something to do with their maintenance involvements. Maintenance is a possession-based playing style that captures possessions in which a team looks to maintain and secure possession within the defensive area of the pitch. The build up playing style then occurs in the attacking half from midfield to the edge of the 18-yard box, and the sustained threat playing style is possession in the final third.
Maintenance devours time and pads possession percentage because it’s typically not as contested as play up the pitch, but it’s the least dangerous of the possession-based styles because it occurs in the back. There wasn’t a massive gap in maintenance involvements between those 1986 opponents and Maradona’s Argentina, but what’s telling here is how the advantage flips as we go up the pitch. The ’86 opponents averaged about 11 more maintenance involvements per 90 minutes than Maradona’s team. But Argentina averaged about 33 more build up and sustained threat involvements than their opponents, as well as 11 more fast tempo possessions (occurring in the attacking half when a player releases the ball to a teammate in under two seconds or a player dribbles at high tempo).
This is where we’re going to get into considering the efficacy of each group’s midfield play. Building off what we just said about the 1986 matches and what we stated in reference to the first table with regard to touch location, note in this one that Argentina ’14 averaged more than 30 more maintenance involvements than their opponents. But that didn’t progress up the pitch, and their opponents averaged 21 more build up and sustained threat involvements with an additional edge in fast tempo possessions. So Argentina ’14 had more possession in the back than their opponents, while their opponents had more of the ball up the pitch. The increased number of build up involvements for the ’14 opponents v Messi’s side might lead us question the quality of the midfield ’14 Argentina were working with – which we’ll expand on later – whereas Maradona’s team accounted for far more build up, sustained threat and fast tempo than their opponents.
Nevertheless, Argentina ’14, despite playing more withdrawn than the five opponents considered here, managed the highest oBMP among the four groups we’re considering. So when they got the ball forward, they were ambitious with their ball movement. To tease what we’re building up to, two ’14 players in particular had a great deal to do with that, despite limited opportunity, and one was injured at a crucial stage in the tournament.
As for other styles, Argentina ’86 and their opponents focused more on crossing than their 2014 counterparts, while Argentina ’14 and their opponents played more directly. In terms of counter attack, ’14 opponents did the most of this while Argentina ’86 also covered plenty of ground on the counter and, interestingly, did substantially more of it on the dribble than on the pass. The ultimate embodiment of this was the Goal of the Century, which we’ll get back to. Given the number of high press regains (15.0 per 90 minutes in 2014 v 6.0 in 1986), the 1986 teams seem to have been applying less pressure up top, which is supported by the higher number of maintenance involvements in these matches (321.4 in 1986 v 313.9 in 2014) with far fewer build up and sustained threat involvements.
Overall, the 1986 tournament might have displayed slightly more conservative football than 2014 in terms of style, which certainly corroborates the stigma of the conservative football of the time, but that of course wasn’t reflected on the score sheet for Argentina, with a combined 2.6 goals per match in these five games vs. 1.2 in 2014.
Maradona certainly had a say in that by making the most of his scoring opportunities, but which Argentine did more with the ball?
Enough team context. Time for a comparison of the two players, but the homework we did above was necessary to properly grasp what’s coming. So here we go: Maradona v Messi – one on one, playing for the same team, 28 years apart.
Messi scored three of his four goals in the games we scratched here, while Maradona scored against Italy in group followed a few games later by consecutive braces in the England quarterfinal and Belgium semi. Both players scored two match-winning goals, but one of Messi’s has been eliminated here because it came against Iran. His other came against Bosnia and Herzegovina in Argentina’s 2-1 win to open the tournament.
Messi played more minutes in those five matches than Maradona because of extra time, so we’ve again got per-90 rates listed parenthetically. First, the basics before we dig deeper into ball movement and style:
Let’s not bury the lede here: Against top competition, Maradona scored goals when he was supposed to, and sometimes he scored goals when he wasn’t supposed to. Of his top-five scoring chances relative to expected goals, he scored four times. The Goal of the Century was of course a beauty, but the least likely goal he scored based on goal expectancy from a shot location was actually his 35th-minute equaliser against Italy. It was a one-time volley after a beautiful bending chip by Jorge Valdano, and came 10.7 metres from the goal line and 10.9 left of centre on the sideline-to-sideline axis with a .087 xG (8.7 percent likelihood of a goal) for the least likely goal scored by either Maradona or Messi in the matches considered here.
As for the least likely goals scored by Maradona or Messi in either tournament as a whole, the top two belong to Messi. Least likely was his stoppage-time winner against Iran, which came from 22.3 metres off the goal line and 8.7 right of centre with a 2.4 percent chance of scoring. His second goal against Nigeria, a free kick from 26.2 metres from goal and 4.0 right, came with a 6.1 percent chance:
As for the best chances missed in the five matches considered, Messi was guilty of the top five while Maradona only wasted one opportunity on which he had greater than a 10 percent chance of scoring:
Goals are great, and goal expectancy is the metric that correlates to winning more than anything else that’s been developed in football, but we’ve got to get beyond it to assess the tournaments these players had on a deeper level. This is supported by the fact that the Goal of the Century was the second-most likely finish among the Maradona and Messi goals based on the shot alone. The finish was fine; the work he put into it was majestic.
What’s interesting to consider from the first table in this section is the number of possessions per 90 minutes of each compared to the number of ball events. Messi was working with roughly 14 fewer possessions per 90, yet that resulted in about nine more events. And among those events, it resulted in nearly 20 more touches per match. Touches are differentiated from passes, shots and other ball events in that the player controls the ball and looks to retain individual possession. Among Messi’s events, 74.0 percent were touches. Among Maradona’s, 66.8 were touches. And among Messi’s 843 touches, the opposition won the ball 32 times (3.8 percent). Among Maradona’s 565 touches, the opposition won the ball 26 times (4.6 percent).
This is at least interesting to consider given that our lasting image of Maradona from the 1986 World Cup is of him dribbling through England’s squad. We’re not saying one or the other is preferable. But Messi held the ball longer and theoretically tried to do more on his own than even Maradona, and he also had a better ratio of successful passes to bad passes (2.96 v 2.76). If nothing else, it’s impressive. Whether it was effective or wasteful is another story, so we’ll build on this with oBMP:
What we see here straight away is that while Maradona had a better per-90 oBMP+ than Messi in the five-match sample, he was nearly twice as wasteful with the ball as Messi on a per-90 basis, despite, as we noted before, Messi holding the ball longer.
That certainly wasn’t the case on the Goal of the Century as Maradona covered 52.4 metres in 9.5 seconds with 12 touches after receiving the ball from Héctor Enrique, 3.0 metres behind midfield and 10.4 right, in a tight spot with his back to the goal among two English players. It seemed doomed from the start. He spun out of it and turned it into a 1 v 11 goal.
But as for oBMP across the five-match sample, Maradona’s net value was lower than teammate Jorge Burruchaga, despite the midfielder playing fewer minutes. Maradona’s per-90 rate was also lower than Enrique, who partnered with Burruchaga in Argentina’s midfield for the majority of the minutes in the five matches we’re discussing.
It’s extremely common for midfielders to compile higher oBMPs than even the best forwards because net oBMP in particular can be a playmaking midfielder’s dream metric. Forwards tend to give the ball away more in more advanced positions and see unfavourable oBMP- totals. Nevertheless, no 2014 Argentina player had a higher net oBMP than Messi, and no teammate who played at least half the minutes in the five 2014 matches in question had a higher per-90 rate. That includes Javier Mascherano, who had 122 more possessions than Messi.
This is information that absolutely adds to the discussion of which player was asked to be more of a one-man team. It feels preposterous to say as the Goal of the Century replays in our heads, but it seems Messi was. His method of getting Argentina to the World Cup final might not have been as apparent or announced as Maradona’s with finishes that’ll go down in history, but BMP shows Messi did things with the ball at his feet that transcend his position in attempt to make up for what was missing in his midfield.
That void was never more apparent than in the final, and it likely had something to do with the absence of a teammate history will overlook.
We’ll begin this section by backtracking slightly: It’s irresponsible to call Maradona’s midfield conclusively better than Messi’s throughout the tournament. That’s particularly true if we go back to one of the early tables comparing the two teams, in which it was Messi’s team with the higher oBMP+ and net oBMP. The crucial difference, rather, lies in how those midfields performed in their respective finals.
Both Argentina teams were probably outplayed in the final, at least in the traditional sense considering possession and scoring threat, though one instance was more drastic than the other.
In 1986, West Germany pushed Maradona’s side’s average touch to 8.8 metres behind midfield, which resulted in a menial oBMP+ of 0.79 but came with an expected goals of 1.2 for Argentina. West Germany were 1.6 metres advanced with a 1.34 oBMP+ and 1.9 expected goals. Argentina were out-possessed 54-46, but ball events were about equal with West Germany holding a 942-934 advantage.
In the 2014 final, Argentina’s average touch came an innocuous 1.5 metres behind midfield. The Ángel Di María-less Argentina side posted a 1.07 oBMP+ (0.76 per 90) and 0.6 xG (0.4 per 90). Germany were 2.1 metres advanced of midfield with a considerable 2.15 oBMP+ (1.54 per 90) and 1.0 xG (0.7 per 90). Argentina were out-possessed 60-40, and Germany had a substantial 2,058-1,281 (1,470-915 per 90) advantage in ball events.
Given the on-ball advantages Argentina displayed earlier in the tournament, Di María deserves further discussion. He was named to the 10-man shortlist for the tournament’s Golden Ball – which Messi eventually won – but missed the semifinal and final due to a thigh muscle tear suffered against Belgium. As we noted earlier, things changed for Argentina a match later against the Netherlands and, as we see here, they changed more drastically against Germany.
Forty-eight players compiled more minutes than Di María in the World Cup, but only six of those players managed a higher net oBMP throughout the entire competition – Toni Kroos, Mesut Özil, Philipp Lahm, Thomas Müller, Arjen Robben and Messi. Among that group, on a per-90 basis, only Kroos (0.21), Özil (0.16), Lahm (0.15) and Messi (0.14) surpassed Di María (0.13). So in terms of ball movement, the five best players in the tournament on a per-90 level were on the two sides that reached the final. One of Argentina’s didn’t touch the ball.
As for the one who did, Messi was limited to 40.7 possessions per 90 minutes in the final, which resulted in very little attacking production (0.1 xG, 0.06 oBMP+). Maradona was involved more with 63 possessions and did more in the attack (0.2 xG, 0.14 oBMP+), but the final doesn’t have much of a role in his tournament legacy. Their offensive contribution percentages in these matches were nearly identical with Messi (12.7 percent) narrowly edging Maradona (12.4), but that has a lot to do with the fact that Messi’s Argentina did so little offensively in the final.
The help Messi and Maradona received or didn’t receive doesn’t end there, and the truth is likely that neither player won or lost the World Cup for his country. It’s indisputable that Maradona outperformed what was expected of the chances he had throughout the tournament. But in the final, the highlights and the data will tell you his teammates filled that role. His offensive contribution was the fifth highest among Argentina players in the match behind Valdano, Julio Olarticoechea, Enrique and Burruchaga. Burruchaga, Valdano and José Brown scored a goal each with a combined xG not quite reaching 0.5, meaning they scored about 2.5 goals more than expected among the three of them. In other words, in one match, the trio nearly equaled the expected goal plus-minus that Maradona posted in the five-match sample (+2.9).
Meanwhile, jump ahead to 2014 and Gonzalo Higuaín accounted for well over half of the scoring opportunity the ’14 team had with a 0.4 xG. His xG total was hundredths of a goal from the goal-scoring ’86 trio, so nearly identical. But Higuaín had nothing to show for it aside from an early goal called back for offside and two near misses that, on their own, each had a higher conversion likelihood than any of the three goals scored by Argentina in the ’86 final. His xG was actually a shade higher than 2014 Germany’s leader, Benedikt Höwedes, who, you might recall, missed Germany’s best chance of the match when he hit the post with a header off a corner near the end of the first half.
So Maradona not only had a more active midfield than Messi in the final, but he also had teammates finishing when he wasn’t. Which brings us to consider one chance from each match, on which Maradona and Messi played supporting roles.
Jorge Burruchaga scored a World Cup-winning goal with the last touch he had in the tournament. Assists weren’t an official stat until the 1994 World Cup, but anyone who’s seen highlights of the 1986 final knows who played the outlet ball to Burruchaga on the 84th-minute goal. It was a comparatively deep Maradona, and with deep data, we know that position to be from 2.0 metres behind midfield and 9.9 right of centre. Burruchaga’s first touch came 16.9 metres advanced of midfield and 10.9 right. He took three touches before his shot from 10.3 metres from the goal line and 9.7 metres right, which came with a 10.7 percent chance of scoring.
Twenty-eight years later, Higuaín was in on Germany keeper Manuel Neuer in the 21st minute of a scoreless match and took a one-time shot 16.4 metres from goal and a shade left of centre, which came with a 17.3 percent chance of scoring. It was Argentina’s best opportunity of the match. It came after Kroos – ironically enough given his ball movement efficacy throughout the tournament – played an errant header behind Germany’s back line following a Messi air challenge on Mats Hummels.
If Higuaín had scored, we might be talking about Messi’s tournament on the same level as Maradona’s. If we left it at goals and wins as we did in 1986, we’d potentially be at four goals Messi, five goals Maradona and one World Cup trophy each. We might even compare the beauty of his 10-touch run starting 45.2 metres from goal on his improbable finish against Bosnia and Herzegovina to the Goal of the Century.
He didn’t get it done in the final, but given the difference in the two matchups, would Maradona have? The German midfield dictated the match in 2014 in a way Argentina didn’t allow West Germany to in 1986, despite their withdrawn position, and the data will tell you neither Messi nor Maradona had an overriding say in that failure or success.
So which player was better during their moment in history? We’ll leave that up to you to decide and discuss with a new level of objective information. What we – or the data, rather – will tell you is Maradona wasn’t a one-man team. No matter how intricate the analysis has become in the last 32 years or will become in the next 32, it will probably never tell us one-man teams win World Cups.