Leeds Predictions
AI-powered match predictions, accuracy tracking, and bookmaker consensus comparisons.
📊 Past Predictions (latest 10)
Leeds secured a 1-0 victory over Brighton on the road to a dramatic late winner from Dominic Calvert-Lewin in the 90+6th minute. The goal, arriving deep into stoppage time, proved to be the decisive moment in a tightly contested encounter that defied expectations on both sides. Brighton's inability to convert their chances and Leeds' clinical finishing in the closing stages ultimately separated the two mid-table sides, with the home team's resilience rewarded at the death.
Our model prediction of a 2-2 draw missed the mark on both scoreline and result direction, with Leeds given just a 37% win probability heading into the match. The prediction leaned heavily on historical head-to-head patterns and Premier League precedent suggesting a stalemate, supported by both teams' mid-table status and limited motivation. However, the actual match played out far more defensively than anticipated. While we'd flagged Brighton's key attacking injuries and Leeds' inconsistent form as relevant factors, the defensive solidity on display—particularly in limiting clear-cut chances until the late stages—proved stronger than our xG model of 3.52 suggested. The absence of both-teams-to-score, which we'd backed given Leeds' average of 1.77 goals at home, further dampened the anticipated goal flow. Calvert-Lewin's injury-time intervention provided a reminder that late-game drama can reshape matches that otherwise track along conservative defensive lines.
Tottenham and Leeds played out a 1-1 draw at the Tottenham Hotspur Stadium, with Micky Tel's 50th-minute opener cancelled out by Dan Calvert-Lewin's 74th-minute penalty. The result marked a missed opportunity for the home side to build momentum late in the season, while Leeds secured a valuable away point that underlined their capacity to remain competitive against top-six opposition.
Our pre-match prediction of a 1-2 Leeds victory did not materialise, and the model's assessment of the attacking dynamics proved only partially accurate. We anticipated Leeds would trouble Tottenham's defensive shape through pressing and transition play, and while that pattern emerged in patches, the hosts proved more resolute than our forecast suggested. Tel's clinical finish represented the kind of efficiency we'd flagged as a constraint for Tottenham, yet one goal proved sufficient to hold parity rather than concede the multiple chances we'd modelled. The decisive shift came from the penalty—a moment that reshaped the match's trajectory in the second half and prevented what could have been a comfortable home win.
The draw itself reflects a more balanced encounter than our directional call allowed for. Leeds' counter-attacking intent was present but not as devastating as the 1-2 prediction implied, while Tottenham's control, though unspectacular, was adequate enough to avoid defeat. The penalty awarded to Leeds suggests they maintained sufficient structure and discipline to stay in the contest, which validates part of the pre-match reasoning around their defensive shape. Where the model diverged from reality was in underestimating Tottenham's capacity to defend a narrow advantage without capitulating—a reminder that efficiency in both boxes, not just attacking ambition, determines outcomes at this level.
Leeds produced a commanding performance to dispatch already-relegated Burnley 3-1 at Elland Road, with the hosts establishing control early and never relinquishing it. Stach opened the scoring in the eighth minute from Bijol's assist, then Leeds compounded Burnley's misery with two goals in quick succession around the hour mark—Okafor added a second on 52 minutes via Bogle's setup before Calvert-Lewin made it three just four minutes later. Tchaouna's 71st-minute effort provided only consolation for the visitors, who offered minimal resistance throughout.
Our model predicted a 2-1 Leeds victory with 75% win probability, correctly identifying the outcome direction but underestimating Leeds' attacking output. The prediction landed on the result marker itself—we called it right on winner, wrong on scoreline. What we flagged held partially true: Burnley's severe defensive fragility (averaging 2.37 goals conceded) was evident in the capitulation, and their attacking impotence (0.82 per game) meant they couldn't threaten despite operating in a dead-rubber context. The quick-fire second and third goals weren't anticipated in our scoring projection, however. Leeds' xG profile and Burnley's porous backline suggested the hosts would dominate, which proved accurate, but the timing and clustering of those middle-period strikes reflected sharper finishing and possibly sharper intensity than the historical averages suggested.
Both sides' motivation questions—Leeds mid-table with little to chase, Burnley already down—were partially negated by Leeds' professionalism. The 3-1 scoreline reflects a team executing at full throttle against a side already mentally checked out, delivering a more emphatic statement than our conservative 2-1 projection anticipated.
Bournemouth and Leeds served up the exact opposite of what we anticipated on Saturday, cancelling each other out in a 2-2 draw that defied our pre-match model's 2-1 forecast for a home win. E.J. Kroupi's 60th-minute opener, set up by Mateo Senesi, initially looked to have put Bournemouth on course, but the narrative shifted dramatically when an own goal from J. Hill just eight minutes later handed Leeds an unlikely equalizer. Rayan restored Bournemouth's lead in the 85th minute with Thomas Adams providing the assist, seemingly securing three points—only for Sander Longstaff to snatch a dramatic 90th-minute leveller that left both sides frustrated on the final whistle.
Our prediction missed on two fronts: we underestimated draw probability despite flagging it as a genuine possibility, and we failed to anticipate both the volatility of the own goal and Leeds' capacity to salvage a result despite mid-table positioning. The data supported a Bournemouth edge given their home advantage and xG profile, but we didn't fully account for how both teams' apparent lack of stakes—neither fighting for anything significant—might translate to loose defending and clinical finishing from set pieces rather than the composed, lower-scoring affair we modeled. The both-teams-to-score outcome aligned with our H2H analysis showing these sides frequently find the back of the net against each other, and the total of four goals fell within the historical range we'd observed, but the sequence and timing proved unpredictable.
Leeds dismantled Wolves with a dominant first-half performance that bore little resemblance to the stalemate our model had anticipated. The hosts struck twice in quick succession through Joe Justin in the 18th minute and Noni Okafor in the 20th, both goals arriving from a fluid attacking sequence that suggested Wolves would have few answers. Darren Calvert-Lewin added a third from the penalty spot in the 90th minute to seal a comprehensive 3-0 victory.
Our prediction of a 1-1 draw missed the mark entirely, failing to anticipate the gulf in execution that emerged between the two sides. The model assigned zero win probability to both Leeds and Wolves, indicating genuine uncertainty about the outcome, yet what unfolded was a one-sided affair. The early goals from Justin and Okafor suggested Leeds' attacking transitions caught Wolves unprepared, a tactical vulnerability our pre-match analysis had not adequately weighted. The visitor's inability to generate threatening moments in response suggested deeper issues with their approach or personnel.
This result represents a notable miss for the prediction framework. While matches of this caliber often hinge on marginal moments and individual performances, the scale of Leeds' dominance warrants reflection on which underlying factors we undervalued. The quick succession of goals in the opening twenty minutes set the tone entirely, leaving Wolves to chase the game without ever truly threatening their hosts. For the model's ongoing refinement, this serves as a reminder that early momentum can rapidly compound in ways that pre-match statistical forecasting sometimes struggles to capture.
Leeds United pulled off a deserved upset at Old Trafford, claiming a 2-1 victory despite our model predicting a comprehensive 3-1 Manchester United win. Namibian forward Nanso Okafor set the tone early, opening the scoring in the fifth minute and doubling Leeds' lead in the 29th with an assist from Brandan Aaronson. The early two-goal cushion proved decisive even after Casemiro's 69th-minute response, which briefly suggested a Manchester United comeback. The match's trajectory shifted decisively in the 56th minute when Lisandro Martínez was sent off, leaving United to chase the game with ten men—a reality our pre-match assessment failed to anticipate.
Our prediction fundamentally misread the match dynamics. The model's confidence in a Manchester United victory and the specific 3-1 scoreline reflected an overestimation of the hosts' attacking threat and an underestimation of Leeds' clinical finishing. Okafor's early brace, particularly the second goal's execution, represented the kind of clinical conversion that separates comfortable victories from narrow ones. The red card to Martínez, arriving at a pivotal juncture when the match remained contested, compounded United's difficulties and eliminated any realistic path to a comeback.
This result serves as a reminder that dominant possession and home advantage don't guarantee outcomes against organized opposition. Leeds' direct approach and conversion efficiency exposed defensive vulnerabilities that our model weighted insufficiently. The prediction's failure to account for the margin of error in tactical execution—both positive for Leeds and problematic for United—represents a notable gap in this week's analysis.
Leeds and Brentford cancelled each other out in a match that ultimately reflected the deadlock our pre-match analysis suggested might occur. The sides played to a 0-0 draw at Elland Road, a result that deviated from our predicted 1-1 scoreline but aligned with the broader tactical narrative we'd outlined. Our model correctly identified this fixture as one likely to end level, though it overestimated the number of goals both teams would convert. The prediction hinged on Leeds' attacking intensity at home being sufficiently offset by Brentford's defensive organization and counter-attacking threat—and while that balance held true tactically, neither side managed to find the breakthrough their approach typically demands.
The match reinforced the statistical profile we'd flagged beforehand. Both clubs did generate opportunities at the rates we'd expect from teams of their quality, yet the execution that usually comes with those chances proved elusive. Leeds pressed with their customary urgency, seeking to impose rhythm through possession and territorial control, while Brentford maintained their compact defensive shape and looked to punish space on transition. The difference between our prediction and the actual outcome rested on finishing efficiency rather than the underlying pattern of play. In a fixture between evenly-matched sides with contrasting styles, the tactical setup we'd identified came to pass—it was simply the clinical moment that went missing.
For a model tracking accuracy across the season, this result sits comfortably in the "directionally correct but score-specific miss" category. The draw was called. The scoreline was not. That distinction matters less than whether the fundamental understanding of how these teams would interact proved sound, and here the evidence suggests it did.
Crystal Palace and Leeds played out a goalless stalemate at Selhurst Park, a result that departed significantly from what our pre-match model anticipated. The match remained tightly contested throughout, though the narrative shifted dramatically when Leeds' Gabriel Gudmundsson received a red card in the 45th minute, forcing the visitors to navigate the entire second half with ten men. Despite this numerical advantage, Palace proved unable to convert their pressure into goals, leaving both sides to settle for a draw that neither team may have desired.
Our prediction of a 1-0 Palace victory missed the mark on both the exact scoreline and the result direction. The model flagged Palace's home advantage and defensive credentials as likely to produce a narrow win against a Leeds side historically susceptible in the Premier League, but the actual match unfolded differently. While Palace's structured approach materialized as expected—the home side controlled possession and territory without generating an abundance of clear-cut chances—the expected breakthrough never arrived. The red card should theoretically have compounded Leeds' difficulties, yet Palace lacked the clinical finishing or tactical adjustments needed to capitalize on their extra-man advantage in the second half.
The 0-0 outcome represents a genuine miss for our model's directional accuracy. Despite having a man advantage for 45 minutes, Palace failed to find the net, while Leeds proved sufficiently organized and compact to hold firm. This serves as a reminder that numerical superiority doesn't guarantee goals in Premier League football, and that Palace's usual efficiency in front of goal cannot be taken as a certainty, particularly against a defensive unit able to absorb pressure effectively.