The Title Is Ours: Four Teams, Four Matchdays to Decide La Liga
It’s May 7, and we don’t know who will win La Liga. We don’t know who will finish second or third. We don’t know for certain Sevilla won’t finish higher than fourth. Let’s bring in some help: Take an expected points model, apply it to the remaining fixtures, and you’ve got a league predictor table. Here’s what the predictor has to say about where Atlético Madrid, Barcelona, Real Madrid and Sevilla will finish.
Area 51, Studio 54, MI6, Matchday 35. All shrouded in secrecy and drama with seemingly arbitrary numbers a part of their names. Turns out one is less restricted and exclusive than we thought.
We enter Matchday 35 in La Liga with three or four title winners still in play: Three for those confident rounding down to zero, four for those of us in southwest Spain spinning hope from fractions.
Since La Liga became a 20-team competition in 1987, this season was the first to see as few as three points separate the top four teams with only five matches remaining (based on three points for a win). That of course ended as Sevilla’s five-match winning streak came to an end on Monday.
Now with four matches remaining, we’re still in an interesting spot, and we’re where we are because the second half of the season has been a
perfect storm fairly well-organized storm of hot and cold streaks by the right teams at the right times.
We for a while had clubs like Real Betis, Getafe, Athletic Bilbao and Granada to thank for that as they stole points to keep the top four together. We now have Bilbao to blame for it. Their win Monday didn’t eliminate Sevilla from title contention, but this is one of few cases in which being part of something mathematically and being part of something realistically are very different things.
There you have it: We have three teams with about a one in four chance or better to win the league with just four matches remaining, all of which still have higher than a one in four chance of finishing third in a league of three giants.
This is the first season since 2006-07 that four teams are still technically a part of the title race this late in La Liga’s season:
La Liga May 18, 2007:
Those top four ended in exactly those positions with Real Madrid winning the title with 76 points on the head-to-head tiebreaker (four points to Barca’s one), despite Barcelona having the superior goal difference. The 83 we’re predicting as 2020-21’s winning number would be the lowest winning total since.
If you’re looking for a bit more pragmatism, you still have to look back five years to 2015-16 when three teams were all within a point of each other this late:
La Liga April 22, 2016:
Barcelona ended up winning the league on 91 points with Real a point back and Atléti third (88).
The top four this season are sharing pitches this week with Barcelona hosting Atlético Madrid on Saturday and Real Madrid hosting Sevilla on Sunday, fixtures from which we hope for a bit more clarity heading into Matchday 36.
Scratch that. If you’re rooting for clarity, go back and watch highlights of the 2017-18 season in which Barcelona won the league by 14 points without a particularly good team. We want chaos, and right now we have it. We have so much of it that maybe the following statement is the best way to sum up how peculiar La Liga’s title race is: Real Madrid will benefit from Barcelona winning this weekend.
So here’s what’s up with Spain’s top four, none of which are necessarily playing like a prize is at stake.
The Fantasy: Just about everyone who cares about Spanish football but doesn’t care about one of the three Spanish giants is hoping for this: A Barcelona-Atlético draw this weekend (or a Barca win). A Sevilla win over Real. Then three more Sevilla wins over Valencia, Villarreal and Alavés with utter chaos as the big clubs drop point after point, week after week. That’ll happen right after Manchester United fans forgive the Glazers.
The Reality: 99.9 is a much larger number than 0.1, and the predictor sees Sevilla finishing seven points back of Atlético and Barca and just a point closer to Real.
So let’s back up a moment and just work with this weekend. A win over Los Blancos would put Sevilla on 73 points, and the league predictor we show you next week would be a bit more favourable for Julen Lopetegui’s side. They’d be within a point of Real.
Sevilla have already secured fourth place and automatic qualification to the Champions League group stage, but they did that last season too. Boring. Third place? They’ve got a 7% chance of that, and they haven’t finished that high since 2008-09. Second? 1.3% there, and they haven’t done that since 1956-57. But let’s get through this weekend and see. As of now, their 91.7% chance of finishing fourth is the highest of any club in any position on the table.
The Expected Number: -3.15. The five-match winning streak that made everyone wonder for a week or two was great for making things even more interesting at the top, but Sevilla haven’t been finishing well lately. They have four non-penalty goals with a 7.15 expected goals tally in their last four.
Their -2.35 xG underperformance for the season is the eighth-worst differential in the division, which is a massive swing from league leaders Barcelona (+14.79) and Atlético Madrid (+14.68). There’s a bit of a what-could-have-been moment when considering where Sevilla might be if they were even in the range of Athletic Bilbao (+3.86) or Granada (+6.81) with their finishing. Six of their eight losses have been by a goal.
Real Madrid (24.3%)
The Fantasy: Three points are retroactively awarded to Real for their narrow 1.00-0.92 expected goals triumph over Real Betis on April 24, replacing the one point from the frustrating 0-0 home draw. The extra two points put Real even at the top. The expected goals haters follow the Super League haters and Glazer haters as the next football group to demand justice.
Or, slightly less fantastical: A win over Sevilla this weekend is met with city neighbours Atlético dropping points against Barcelona followed by three wins over Granada, Bilbao and Villarreal to close out their 4,394th La Liga title.
The Reality: Why does the idea of four straight wins seem difficult for this team right now, even if they are out of Champions League? Is it because of what we saw against Chelsea? Why do four straight wins feel difficult for a side that haven’t lost in La Liga in 14 matches since the start of February?
Because they’re just not scoring much. In all competitions, they’ve only won twice in their last seven matches. They’ve gone goalless in four of those with 0-0 results against Betis and Getafe. That said, they do have 10 points from a possible 12 against the four teams they have left.
The Expected Number: -5.25. This time a negative number is more of a complicated thing than a bad thing as it was with Sevilla. Real’s La Liga opponents have scored six goals in 14 matches with an xG of nearly twice that (11.25). That’s quite the pace of overperformance to keep up, but if they can, they may win the league.
The Fantasy: April 10’s El Clásico and April 29 v. Grenada go very differently and Barcelona’s unbeaten streak in La Liga never ends, now reaching 24 matches. A win this weekend over Atlético Madrid puts Barca seven points clear with three to play and a La Liga title celebration is just days away.
The Reality: There’s a lot of work to do, and some of it isn’t in their power. A win over Atléti this weekend still requires Real to drop points at some stage. If that happens, it puts Barcelona in the driver’s seat for their ninth title in 13 years while never going consecutive seasons without winning. Also, after winning Copa del Rey, it would be a nice little trophy haul after things were looking grim during the first half of the season.
More reality sets in when we recall their 1-0 loss to Atlético in Madrid on Nov. 21, in which Barca trailed the entire second half but still only managed 53.8% of the ball and their top individual xG figure came from a defender (Clément Lenglet 0.35).
The Expected Number: 21. This is an actual goals number, not an expected goals number. But that doesn’t mean it’s not expected. Messi’s scored 21 La Liga goals after the new year, which would be noteworthy if he had different DNA. Dating to 2009-10, he’s scored 20 or more league goals after the new year in nine of 12 seasons.
Only one has been from the spot, and seven have been from outside the box. He’s nearly doubled his xG, which is perhaps a passable last-ditch way of explaining expected goals to that friend who still doesn’t get it: Phil, Messi is almost twice as good as the average player.
And so, a fifth straight golden boot will be his, which has never happened in the Spanish top flight. Hugo Sánchez won four straight, Ferenc Puskás won four out of five and Alfredo Di Stéfano either won or tied for four straight . But that’s not to say you should go and forget Mallorca’s Dani Güiza won one in La Liga’s striker identity crisis season of 2007-08 leading into the prime Messi era, which apparently persists to this day.
Atlético Madrid (40.6%)
The Fantasy: No fantasy needed for the frontrunners. A win over Barcelona and smooth seas against Real Sociedad, Osasuna and Real Valladolid on their way to their 11th Spanish top-flight title. That doesn’t sound so fantastical given they’ve already beaten each of those sides with a combined one goal conceded, but…
The Reality: You can’t win a title in February*.
* Offer excludes Manchester City.
After 21 matchdays, Diego Simeone’s side led the league by 10 points on Feb. 1 with a game in hand, having played just 19 matches to Barca and Real’s 20 and most of the league’s 21. Looking back – at a table that’s meant to look ahead – this is what we’d have seen. The same ordered top-four finish as we have today, but different point totals and percentages:
We put Atlético on a projected 87 points back then. We gave Simeone a nearly four out of five likelihood of winning the title. If your team is doing the right things, such percentages are supposed to rise as the season draws to a close. Instead, that percentage has been nearly halved and the projected finishing point total has dropped because Atlético have dropped 19 points in 15 matches dating to the start of February. As we said, Real Madrid haven’t lost in the league in that time. And Barcelona went 19 straight without a defeat to get back in it.
The Expected Number: 12.67 v. 2.03. The first is Atlético’s xG overperformance through the halfway point of the season. The second is their overperformance since – a much more realistic number to maintain over time – which begs the question: Are the results we’re seeing much more realistic too? Get back on your horse, Luis Suárez (no shots on goal in three straight competitive matches for first time in his top five European league club career). Your old teammates are right behind you.
The (Bonus) Expected Number: 83. The number of points Atlético and Barcelona are expected to finish on. But that’ll drop again for Atléti Saturday if Barcelona win.
There’s been one title won by clubs not named Barcelona and Real Madrid in the last 16 La Liga seasons. Atlético did it in 2013-14. They’re still expected to do it again, albeit by the narrowest margin on a head-to-head tiebreak with the team they play this weekend.
Stay tuned. There’s a decent chance that after this weekend one-television households will have greater clarity on whether they’ll need to make a decision about owning a second before May 23.
Our Expected Points Model Explained:
- The model estimates the probability of each match outcome (win, draw or loss) given each team’s attacking and defensive quality.
- The team’s attacking and defensive qualities are based on four years of historic results, with more weighting given to their most recent results.
- The model will take into account the quality of the opposition that a team scores or concedes against and reward them accordingly. For example, scoring against Man City is worth more than scoring vs Newcastle.
- We can simulate the upcoming matches using goal predictions from the Poisson distribution with the two teams’ attacking and defending qualities as inputs.
- Finally, we simulate the outcome of the season 10,000 times in order to estimate the likelihood of each team finishing in each league position.
Enjoy this? Subscribe to our newsletter to receive five stories each Friday from The Analyst. It’s free.