What’s On It

The Two-Leg Picture

Saturday lunchtime at the Riverside is the first leg only. The return is at St Mary’s on Tuesday 12 May at 20:00 BST. Whichever side advances on aggregate goes to Wembley on Saturday 23 May. Lower-seeded Boro host first; second-leg home advantage sits with fourth-placed Southampton.

The Final Table

How the Championship finished after 46 games — Boro and Saints level on 80 points, separated only by goal difference.

Form & the Underlying Numbers

The form contrast is stark. Middlesbrough stuttered through the run-in, winning only three of their last ten league matches. Five draws in that sequence included a 0-1 home defeat to Portsmouth in which Boro recorded 1.31 xG to the visitors’ 0.14 — a humiliating outcome given the shot count. A 1-2 home loss to Millwall was even more galling: Boro generated 3.04 xG to Millwall’s 0.79 yet found themselves beaten. The 5-1 demolition of Watford and a 1-0 grind over Sheffield Wednesday were rare highlights. Across their final six games, Boro managed only eight points. Southampton, by contrast, surged. Eight wins from their final ten included a 5-1 thrashing of Wrexham (2.61 xG to 1.28), a 3-0 dismissal of Blackburn (2.52 xG to 0.23), and a 3-1 final-day victory at Preston. Across the last eight league fixtures, Saints scored 19 goals. Their 17-game unbeaten run before the final day was extended to 18 with that Preston win. The xG-vs-result paradox at Boro — outshooting opponents in defeats — suggests a team creating chances but lacking the cutting edge or luck to convert them, a vulnerability Saints will look to exploit. Boro’s home defensive record, however, remains the best among the four play-off teams, with only 18 goals conceded across 23 home games at the Riverside this season.

Last six matches — side by side

Rolling xG — 10-match window

xG scatter — for vs against

Each dot is a fixture: xG produced (x-axis) against xG conceded (y-axis). Below the diagonal = outperforming the opposition on chances.

Home vs away split

Boro have the joint-best home defensive record in the division — 18 conceded across 23 home games at the Riverside. Saints lost only three home matches all season and only one under Tonda Eckert (to Hull on 17 January).

When the goals come

Players to Watch

Two on each side. Boro’s midfield engine and their highest-rated defender against Saints’ creative engine and their most efficient finisher.

Hayden Hackney (Middlesbrough)

Hayden Hackney is the heartbeat of Middlesbrough’s midfield. The 23-year-old academy graduate started all 38 league appearances this season, accumulating 3,321 minutes — more than any other Boro outfielder. His five goals and seven assists from central midfield understate his influence: 58 shots, 24 on target, and an average rating of 7.24 make him the club’s highest-rated player behind only Luke Ayling. An England under-21 international, Hackney is the connector between defence and attack, the player who receives from the back four and distributes to the forwards. If Boro are to control the midfield battle on Saturday lunchtime, Hackney must impose himself on the game’s tempo from the first whistle, particularly against a Saints side that has been prolific in transition over their recent unbeaten run.

Luke Ayling (Middlesbrough)

Luke Ayling, at 33, is enjoying a quietly outstanding season. The right-back leads the Boro squad in average rating (7.26) and has contributed five assists from defence — including the pass that sent Samuel Silvera through for the second goal in January’s 4-0 win over Saints. Ayling’s 43 appearances (37 starts) underline his durability, and his 25 shots from full-back show a willingness to advance. In a team built on defensive solidity, Ayling is the most reliable performer, rarely spectacular but consistently effective. His experience of promotion battles — he previously won the Championship with Leeds in 2019-20 — could prove invaluable in the high-stakes environment of a play-off semi-final at the Riverside, especially against a Saints attack that has scored 82 league goals across the regular season.

Leo Scienza (Southampton)

Leo Scienza is Southampton’s creative fulcrum. The midfielder’s average rating of 7.39 is the highest in the Saints squad, and his 10 league assists are joint-most among the four play-off teams. With seven goals of his own, Scienza has been directly involved in 17 goals from 37 appearances. His shooting is remarkably efficient: 65 shots, 36 on target — a 55 per cent accuracy rate that speaks to excellent decision-making in the final third. Middlesbrough’s midfield, led by Hackney, must find a way to limit Scienza’s influence. If he is given time and space to pick passes or shoot from distance, Boro’s defensive record will face its sternest test of the season, as Saints have scored 19 goals in their last eight games.

Adam Armstrong (Southampton)

Adam Armstrong returns to the Riverside with a point to prove. The striker scored 11 goals in 29 appearances this season — a rate of one every 185 minutes, among the most efficient in the play-off pool. He scored Southampton’s equaliser in the 1-1 draw at St Mary’s in September, but his last visit to this ground ended in a 4-0 defeat. Armstrong has previous in Championship play-off promotion: his 21 goals in 2023-24 helped Saints secure an immediate return to the Premier League. At 29, he knows what it takes to deliver in high-pressure moments. Boro’s centre-backs will be wary of his movement and finishing, particularly on the counter-attack where Saints have been most dangerous across their recent run of form.

Head to Head

The two sides met twice this season, with contrasting outcomes. In September at St Mary’s, a tight 1-1 draw saw Adam Armstrong open the scoring before Mamadou Kaly Sene equalised from Luke Ayling’s pass. Southampton held a marginal xG edge (0.85 to 0.29) but Boro defended resolutely. The return fixture on 4 January was a different story. Middlesbrough thrashed Saints 4-0 at the Riverside, with Morgan Whittaker scoring a brace (54′, 66′), Samuel Silvera adding a third, and Alan Browne completing the rout. Boro registered 22 shots and 2.27 expected goals against Saints’ 1.12, dominating the ball at 56 per cent. That result, however, came four months ago, before Saints embarked on their 18-match unbeaten run. Over the past five years, Saints have not beaten Boro in four meetings (Boro two wins, two draws). The historical trend leans Boro’s way, but the recent form curve tilts towards Saints. The 4-0 scoreline remains a psychological marker, though both sides have evolved since that January afternoon. Notably, Boro’s 4-0 win was built on second-half dominance, with all four goals arriving after the 54th minute and three of them between the 54th and 66th — Saints’ defensive structure crumbled inside a 12-minute window. The earlier September fixture had ended honours-even but the underlying numbers across the two meetings still favour Boro 2.56 xG to 1.97.

Last six meetings

Goal-scoring trend — xG vs actual

How well each side’s goal output is tracking their expected-goals profile across 25/26. Above the line = scoring above xG; below = under-performing.

The Model’s Honest Reading

The statistical models offer a nuanced read. For the first leg, both the logistic regression and Poisson models favour Middlesbrough at home. The LR-cal model gives Boro a 54.7 per cent win probability, with Saints at 25.9 per cent and the draw at 19.4 per cent. The Poisson model is narrower: Boro 42.6 per cent, draw 25.5 per cent, Saints 31.9 per cent, with expected goals of 1.44 for the home side and 1.21 for the visitors. The conditional modal scoreline is 1-0 to Boro, with 2-1 and 2-0 as the next most likely outcomes. However, the picture flips for the second leg at St Mary’s. There, both models favour Saints — LR-cal gives them 49 per cent, Poisson 48.5 per cent — meaning the aggregate tie is genuinely open. The form gap is significant: Saints took 14 points from their last six games, Boro only eight. Yet Boro’s underlying numbers across the full season — an xG difference of +27.9 and just 47 goals conceded — sit behind only Coventry’s and Ipswich’s. BBC Sport’s statistical analysis notes that Saints have the most factors aligning with past play-off winners, but Boro’s defensive record and fourth-place finish also match historical trends. On balance, the first leg reads as a contest between form and structure: Saints’ momentum against Boro’s home advantage and defensive organisation. The 4-0 January result is a data point, but four months of subsequent form suggest it may not be the most relevant one. Boro’s home xG difference of +0.8 per game further underscores their Riverside solidity, while Saints’ away xG conceded of 1.3 per game hints at vulnerability on the road.

1X2 — model probabilities (leg 1)

The chart below shows the calibrated logistic-regression model’s read on Saturday’s leg 1. The Poisson + Dixon-Coles goals model agrees on direction but tighter — Boro 42.6%, Draw 25.5%, Saints 31.9% (λ_h 1.44, λ_a 1.21). Both models lean Boro; the goals model is narrower.

Conditional-modal scoreline

Taking the LR model’s most-likely outcome (Boro win) and selecting the most-likely scoreline from the Poisson + Dixon-Coles grid restricted to home-win cells, the conditional-modal prediction is 1-0 Boro. Lambdas: Boro 1.44, Saints 1.21. The next two most probable Boro-win scorelines are 2-1 and 2-0.

How has the model been doing this season?

Track record across the full 25/26 Championship campaign:

The Verdict

The first leg leans Boro: home advantage, the league’s joint-second-best defensive record, the third-best xG difference in the division, and a 4-0 Riverside thrashing of these opponents only four months ago. Saints’ counter is the form arc — eight wins in their last ten and an 18-match league unbeaten run that suggests a side now far better than the one that visited Teesside in January. The two-leg picture is genuinely open. Saints flip to favourites at St Mary’s per both models. Friday-night Hull-Millwall is the dramatic one; Saturday-lunchtime Boro-Saints is the evenly-matched one.

Models: goals_logreg_v1_cal (Platt-calibrated logistic regression with player-rating features) and goals_poisson_v1 (Poisson + Dixon-Coles, ρ = +0.0173). BeyondThePrem does not provide betting advice; probabilistic reads are presented for analytical interest only.