Elche arrived in Bilbao second from bottom, with 1 point from 4 rounds, 12 goals conceded, a goalkeeper under fire for his recent mistakes and Real Madrid to host three days later. On the other side was an Athletic team that had just won at Celta and put 3 past Atlético Madrid.
Elche left with a point. Their only open-play shot was struck from inside their own half.
The verdict, before the numbers
Athletic should have won, and comfortably. 3.13 expected goals against 0.82. Expected points, which translate those chances into probabilities of a result, give Athletic 2.84 and Elche 0.10: one point in every ten games like this.
But saying only that would tell half the story. Because Athletic's dominance sits almost entirely in one half, and Elche prepared the game precisely to survive that half.
The number: 0.009
Elche shot 3 times in 90 minutes. A penalty, converted by Buonanotte, worth 0.74. A direct free kick from Chust, 0.06. And a shot by Rubén Sánchez from inside his own half, which the model values at 0.009: nine thousandths of a goal. That is all of Elche's attacking output from open play.
In the first half, zero shots. In the same half Athletic had 13, for 1.54 expected goals, and hit the crossbar through Navarro from a ball played in by Iñaki Williams.
How the two teams moved
Anselmi changed his line-up to defend better, bringing in Redondo and Revivo. The design was clear: three centre-backs, wing-backs dropping onto the line to form a back five, Morcillo and Villar in front to protect the middle. PPDA shows how much Elche gave up on pressing: they allowed Athletic 30.2 passes before going to disturb them. They almost never did. They gave up the ball, retreated into their own box and defended it by sheer numbers, with more bodies than spaces.
After the game Anselmi explained why. In previous matches Elche had always conceded before the 20th minute, and the aim was to compete in a different way, alert to every detail. Put in numbers: survive the first half.
Terzic's Athletic played their 4-2-3-1 with Ruiz de Galarreta and Gerenabarrena in the middle, Sancet between the lines, Nico Williams heavily involved on the left, Navarro on the right and Iñaki Williams up front. Iñaki created more than he finished: twice he set up Navarro, and the second went off the crossbar. For half an hour Athletic moved the ball patiently without finding a rhythm, then in the last quarter of an hour they raised the tempo and built three chances in a few minutes. None went in.
Then, early in the second half, the game turned on Elche's first real attack. A lapse in the Athletic defence, Areso tripping Valera in the box, Buonanotte from the spot.
From then on Athletic stopped being the Athletic of the first half. From the penalty to the 90th minute, excluding the goal, they produced 0.15 expected goals in total. Nico and Sancet faded out of the game, the build-up became hurried, and Elche even tried the odd counter-attack. The equaliser on 79 minutes came almost by accident: a cross-shot hit without much power by Berenguer, and Iñaki Williams, tussling with Chust, brushing the ball just enough. Late on Dituro, the goalkeeper under fire, saved the point with a big stop.
After the game Terzic asked for the result to be separated from the performance, and said the first half was perhaps a step forward. He is half right: the first half yes, the second no.
The season context
Elche are a strange team to read. They have 2 points from 5.81 expected, and no side in La Liga has conceded more expected goals: 11.16 in 5 games. But the points came in the wrong games. On the opening day at La Coruña they deserved to win and drew. A week ago against Real Sociedad they produced 2.87 expected goals to 0.88 and lost 2-3 in the 89th minute. On Saturday they took a point from a game in which the model gave them one in ten.
Athletic have 7 points from 9.45 expected. After two opening defeats, against Sevilla and Barcelona, they seemed to have relaunched, and on Saturday the first-half numbers confirmed it. The second half, instead, shows a team that, when behind, does not yet have a plan B.
The honest bit
First, the model was generous to Athletic in at least two moments. Iñaki Williams's goal is valued at 0.70, like a finish from a yard out, but it was a touch on a cross-shot, more deflection than shot. And the chance the model credits to Iñaki on 5 minutes, 0.60, is described in match reports as a Sancet ball the striker just failed to reach. The 3.13 should be read with that in mind: the dominance was clear, but less clear than that.
Second, of Elche's 0.82 expected goals, 0.74 is the penalty. Without that incident Elche produced 0.07, and a free kick accounts for 0.06 of it.
Third, the reading opposite to ours, which is defensible. Elche's point can be called deserved without stretching the words: a team that had been conceding in every first half chose to give up attacking for an entire half, managed it, and struck at the first opportunity. The model penalises teams that shoot little by design, but it does not reward those who make the opponent's possession useless. In the second half, from 0-1 onwards, Athletic created almost nothing.
Our view is that both things coexist. On the numbers Athletic threw away two points, and the first half proves it. But Anselmi's plan was not luck: it was a precise choice, which turned 45 minutes of siege into Athletic's problem rather than Elche's.
On Tuesday 15 Elche host Real Madrid, on Wednesday 16 Athletic travel to Valencia to face Levante. The question is whether Elche have found a way to stay in games or just a lucky night, and whether they will try it again against Real Madrid.
T.L.
Want to see how much each team really deserves its league position? The Meritometro compares real points with expected points. For those who want the models rather than the headlines, there is the Nerd Zone.