VAR Impact on Football Predictions: How Video Review Changes Match Outcomes

Introduction

VAR impact on football predictions is one of the most consequential analytical topics of the modern era. When the Video Assistant Referee system was introduced across elite football competitions — most notably in the Premier League in 2019, though earlier in other leagues and in FIFA competitions from 2018 — it changed not just how decisions are made on the pitch, but how the statistical patterns of football matches have evolved. For prediction analysts, VAR represents a structural shift that requires specific analytical adjustments across multiple markets: the frequency of goals disallowed, the rate of penalty awards, the time spent in video review, the psychological dynamics of matches, and the distribution of match outcomes. Understanding VAR's impact comprehensively is not an optional add-on for sophisticated analysts — it is a core analytical requirement for anyone making predictions in competitions where VAR is deployed.

This guide provides a thorough examination of VAR's impact on football prediction outcomes. It covers the statistical evidence of how VAR has changed goal rates, penalty frequencies, and match outcomes; the specific markets most affected by VAR's presence; the competition-by-competition variation in VAR implementation; the psychological effects on players, managers, and match flow; and the specific analytical adjustments that prediction models should incorporate in VAR competitions. The aim is to provide analysts with a complete, evidence-based understanding of how video review technology changes the game — and therefore how to account for it accurately in prediction work.

The Statistical Evidence: How VAR Changed Match Outcomes

Goal Disallowance Rates Before and After VAR Introduction

The most immediately apparent effect of VAR implementation was a statistically significant change in penalty award rates. Across the major leagues that adopted VAR in 2018-2020, penalty frequencies increased substantially — typically 40-70% above pre-VAR levels — in the seasons immediately following implementation. This increase reflected both the catching of genuine incidents that had previously been missed by assistant referees and on-field officials (a positive VAR effect) and the aggressive application of the handball law (a more controversial VAR effect that generated significant debate before the law was subsequently revised). The Premier League's introduction of VAR in the 2019-20 season produced a penalty rate of approximately 0.35 per match, compared with around 0.22 in the pre-VAR era — an increase of nearly 60%.

The second major statistical effect was a changed rate of goal disallowances. Pre-VAR football saw a relatively modest number of goals disallowed for offside, because assistant referees were instructed to give the benefit of the doubt to the attacker in marginal situations. VAR's ability to draw precise lines on frozen frames produced a spike in marginal offside disallowances — goals cancelled by centimetres or fractions of centimetres — that had not existed under the previous officiating framework. This increased disallowance rate had a downward pressure effect on actual goal totals relative to xG-predicted totals, particularly for goals scored from set pieces and from attacking moves involving players making late runs into the penalty area. The separate analysis of offside technology's specific effects is covered in our offside technology and prediction guide, while the penalty-specific effects receive dedicated treatment in our VAR and penalty decisions guide.

Result Reversals and Their Prediction Implications

The net effect of these two countervailing forces — increased penalties adding expected goals, increased offside disallowances reducing them — on total goal counts was competition-dependent. In the Premier League, the increased penalty rate approximately offset the increased disallowance rate, leaving average goals per match broadly stable in the early VAR era compared with the pre-VAR average. In some other competitions, the penalty increase was proportionally larger, producing a small net increase in goal averages under VAR. In competitions where semi-automated offside technology (SAOT) was subsequently introduced, the disallowance rate stabilised, changing the balance of these effects again. Analysts building goal-total models for VAR competitions must therefore work from VAR-era data rather than historical pre-VAR baselines.

VAR and Result Distribution: Which Outcomes Change

Draw Rate Changes Under VAR

Beyond the raw goal count effects, VAR has influenced the distribution of match outcomes in ways relevant to result prediction. The most analytically significant effect on result markets has been a reduction in what analysts call "gifted results" — matches where a clear and obvious officiating error in the pre-VAR era would have produced an incorrect result that VAR can now correct. When a genuine penalty incident that would previously have been missed is now awarded, the beneficiary team gains a scoring opportunity that they "deserved" in a probabilistic sense; the result becomes more consistent with the underlying quality of the match.

This means that VAR, at its best, pushes actual match results closer to the distribution that match quality data would predict — a development that should theoretically improve the accuracy of prediction models that are built on quality-based analytical foundations. Teams with higher quality but historically subject to occasional "robbery" results due to missed decisions should show a slightly smaller variance in outcomes under VAR. In practice, this quality-reinforcing effect is real but modest, because the VAR era has also introduced new types of controversy and inconsistency that partly offset the gains from corrected errors.

Home Advantage Reduction Through VAR Equalisation

The specific result outcomes most affected by VAR are matches where the margin of quality is relatively small — matches that could reasonably go either way — and where the type of incident that VAR reviews (penalties, offside goals, red cards) is most likely to be pivotal. High-quality matches between closely-matched teams generate more marginal incidents that VAR reviews and changes, compared with mismatches where the dominant team's quality is sufficient to win comfortably regardless of officiating. For analysts using Asian handicap analysis — which prices quality differentials precisely — this means VAR's effect is most analytically significant for markets involving evenly-matched teams, and less significant for obvious mismatches. The complete Asian handicap framework is covered in our Asian handicap guide.

VAR's Impact on In-Play Match Dynamics

Beyond the statistical effects on outcome distributions, VAR has fundamentally changed the in-play dynamics of football matches in ways that affect live analysis and in-play prediction. The most obvious impact is the introduction of review pauses — periods of 60 seconds to several minutes during which the match is effectively suspended while officials consult the video evidence. These pauses disrupt match flow, change team tactical momentum, and create psychological uncertainty for players and spectators alike.

Research on VAR review period effects on subsequent match dynamics has produced mixed findings, but several patterns are reasonably consistent. Matches that experience long VAR reviews — particularly for penalty incidents — tend to see a slight reduction in scoring pace in the 5-10 minutes following the review, as the match's rhythm has been broken and both teams are recalibrating to the changed situation. This "post-VAR disruption" effect is small but statistically detectable and has some relevance for in-play goal market analysis immediately following major VAR reviews.

The psychological effect of VAR on players has also been analytically noted. Defenders in VAR competitions have shown increased caution in the penalty area — being more conservative in their challenges and less willing to make sliding tackles — partly because the review process can catch incidents that on-field referees would not penalise but VAR can. This defensive conservatism has a small but real effect on the frequency of genuine penalty incidents relative to a world without VAR, as some challenges that would previously have been attempted (and possibly resulted in penalties) are now not made at all. For live analysis, this defensive VAR-awareness creates a slightly different challenge-frequency environment than pre-VAR football, and analysts should account for this when assessing in-play penalty probability. Our comprehensive live in-play strategy guide addresses how to manage these real-time analytical considerations.

Competition-Specific VAR Implementation Differences

Premier League VAR Usage Patterns

One of the most important analytical insights about VAR is that it does not operate uniformly across competitions. Different football associations and governing bodies apply VAR with meaningfully different philosophies, threshold levels, and consistency. The Premier League's "clear and obvious error" standard has been applied differently across different seasons, producing variability in VAR intervention rates that makes building stable Premier League VAR prediction models challenging. Serie A has had periods of relatively aggressive VAR application. La Liga has generally been more conservative. The Bundesliga's VAR use has been regarded by many analysts as among the more consistent implementations across the major leagues.

La Liga and Serie A Implementation Differences

At the international level, FIFA World Cup and UEFA Champions League VAR applications have their own specific patterns. Tournament VAR use has shown a tendency toward increased conservatism in knockout stages compared with group stages, as referees and VAR officials are more cautious about decisive interventions in the high-stakes elimination rounds. This produces measurably different incident rates at different stages of tournaments, a pattern directly relevant to prediction analysis for both penalty markets and result-based forecasting.

The competition-specific nature of VAR application means that analysts must maintain separate calibration for each competition's VAR environment, rather than applying a single "VAR adjustment factor" across all competitions. A model built primarily on Premier League VAR data will not transfer accurately to Bundesliga predictions without recalibration to the Bundesliga's specific VAR application pattern. For analysts covering the Scottish Premiership, the Turkish Super Lig, or other leagues with their own specific VAR implementations, competition-specific calibration is equally essential. Our dedicated league analysis guides — including the Scottish Premiership guide and Turkish Super Lig analysis guide — incorporate VAR context within their competition-specific analytical frameworks.

The Handball Law and VAR: A Special Case

The handball law and its application under VAR represents a unique case study in how officiating technology interacts with rule interpretation to produce dramatic statistical effects. The period from 2019 to 2021 — during which the IFAB handball law was applied in its strictest interpretation, with any ball touching the arm in the build-up to a goal triggering an automatic VAR review for potential penalty or goal disallowance — produced statistical anomalies that were unique to that era. Penalty rates in this period were measurably elevated above both pre-VAR levels and subsequent post-revision levels. Analysts using data from this period for baseline calculations need to treat it as a special era with its own statistical profile rather than representative of the ongoing VAR environment.

Since the law was revised — reinstating some element of referee discretion regarding "natural" versus "unnatural" arm positions — handball penalty rates have partially stabilised, though they remain above pre-VAR levels because VAR can still catch genuine handball incidents that would previously have been missed. The direction of further law evolution is uncertain, which introduces an ongoing source of calibration uncertainty into handball-related penalty modelling. The practical analytical response is to monitor handball penalty rates in each competition on a rolling basis and update calibration regularly as the law's application evolves.

VAR and Red Card Decisions: Impact on Match Predictions

Retrospective Red Card Frequency

Red card decisions represent another significant area where VAR has changed match dynamics. Under the pre-VAR system, serious foul play incidents in less prominent areas of the pitch — particularly off-the-ball challenges and tackles from behind in mid-range positions — were regularly missed by on-field referees. VAR can now review these incidents and recommend additional sanctions where the on-field official has made a clear and obvious error in not dismissing a player.

The practical effect has been an increase in red cards resulting from retrospective VAR review — situations where a yellow card or no card was initially given, and VAR intervention upgraded it to a red. This has changed the pattern of red cards from being concentrated in on-ball incidents visible to the referee to including off-ball incidents that were previously invisible to officiating. For prediction analysts, this means the red card event probability in VAR matches — particularly for teams known to defend physically and for players with histories of off-ball involvement — is higher than pre-VAR data would suggest. The analytical framework for incorporating red card effects into predictions is covered in our guide on red card impact on match predictions.

In-Play Probability Shifts from VAR Red Cards

Conversely, VAR has also produced some red card reversals — situations where a red card initially given in real time was overturned upon review because the incident did not warrant dismissal. These reversals are analytically significant in matches where a red card changes the predicted match outcome dramatically: a reversal restores the expected team strengths and substantially changes the remaining match probabilities. In live in-play analysis, the possibility of VAR red card review — in either direction — represents a source of uncertainty that models should acknowledge rather than assume away.

VAR's Effect on Correct Score Predictions

Late Goal Cancellations and Score Market Impacts

Correct score predictions are among the most directly affected by VAR's presence, because VAR changes both which goals stand and the probability of specific score milestones being reached through its penalty effects. The most analytically significant effects on correct score markets are: the increased probability of a match involving at least one penalty goal (due to the elevated penalty rate in VAR competitions), the decreased probability of goals from marginal offside positions standing (due to VAR's ability to catch these precisely), and the changed probability of specific scorelines that depend on penalty conversion.

Adjusted Correct Score Modelling for VAR Leagues

For correct score analysis in VAR competitions, the Poisson model — which remains the most widely used statistical tool for scoreline probability estimation — should be calibrated with VAR-era goal data rather than historical pre-VAR data. This is a basic but important calibration step that significantly affects the probability estimates for common scorelines. The methodology for Poisson-based score prediction is covered in detail in our Poisson method guide, and the integration of VAR-adjusted goal rate inputs into this model is a straightforward application of the calibration principles described there. Our broader correct score predictions guide also addresses the context-specific factors — including VAR environment — that should inform scoreline probability assessment.

Expert Insight: Analysts who have built models specifically calibrated for VAR competition environments observe a consistent finding that has significant practical implications: the predictive power of pre-match quality metrics — xG, xT, progressive passing statistics — is slightly higher in VAR competitions than in non-VAR competitions. The reason is intuitive: VAR reduces the frequency of results-distorting officiating errors, meaning the match outcome distribution is more aligned with underlying team quality than it was in the pre-VAR era. This quality-reinforcing effect means that prediction models built on solid analytical foundations — using quality metrics rather than just recent results — should, in theory, perform slightly better in VAR competitions than in non-VAR environments, because their quality-based predictions are now more likely to be vindicated by results that fewer officiating errors distort. In practice, this effect is modest but real, and it provides analytical support for investing in quality-metric-based prediction approaches in VAR competitions over simpler results-based methods. The analysts who have adapted their models most effectively to the VAR era are those who leaned into quality metrics precisely because they correctly understood that VAR makes football more "fair" — more aligned with underlying quality — than the pre-VAR system.

Analyst Note: For practical VAR-adjusted prediction analysis, the following guidelines should be maintained. First, always establish whether the specific competition you are analysing uses VAR — and if so, which generation of VAR (standard, or with semi-automated offside technology). Apply appropriate calibrations accordingly. Second, build VAR-era-specific statistical baselines for each competition: do not average pre-VAR and VAR-era data, as the structural shift between the two periods makes combined averages misleading. Third, track competition-specific penalty rates on a rolling basis, updating your baselines regularly as laws evolve and application patterns shift. Fourth, for in-play analysis, maintain awareness of the specific referee's VAR review tendencies — some referees initiate VAR checks more frequently than others, affecting the probability of review-driven match-changing events in any given fixture. Fifth, when using historical head-to-head data between clubs, check whether those historical matches were played under VAR conditions; pre-VAR H2H data provides less reliable prediction information for current VAR-competition matches than VAR-era H2H records. Sixth, monitor ongoing developments in VAR law interpretation — particularly regarding handball — as these are subject to periodic revision that can shift penalty rates measurably. For the specific referee profiles that affect VAR application frequency, our referee profiles guide provides the framework for tracking individual officiating tendencies across different competitions.

Case Studies: VAR Impact on High-Profile Match Predictions

The 2019-20 Premier League season — the first with VAR — provided an extraordinarily rich case study environment for documenting VAR's prediction impact. That season saw the highest penalty rate in Premier League history to that point, driven by the combination of VAR reviews and the strict handball interpretation. Liverpool, who won the title that season, benefited directly from multiple VAR-awarded penalties — including several handball awards that would not have been given under any previous officiating framework. For prediction analysts tracking Liverpool's goal expected vs actual performance in that season, the VAR-driven penalty contribution was a significant factor that properly calibrated models should have accounted for, and those who had adjusted their Liverpool goal projections upward to reflect the elevated penalty probability would have been better positioned throughout the season.

A second analytically instructive case: the 2021 Champions League knockout stages, which involved a series of VAR penalty reviews in tightly contested ties. In one particularly notable round, a long VAR review for a potential penalty completely changed the match dynamic — the extended pause disrupted the flow of a match that had been building toward a crucial goal, and the team that ultimately benefited from the awarded penalty won a tie they might otherwise have drawn. For analysts engaged in live in-play analysis during this match, the extended VAR review was a significant moment of analytical uncertainty that the live framework described in our in-play strategy guide helps to manage systematically.

A third case study involves VAR's effect on a specific team's full-season prediction performance. Brentford in their first Premier League season (2021-22) were a team whose analytical profile suggested competitive quality but whose penalty differential — both won and conceded — was shaped by VAR in ways that pre-season models could not fully anticipate. Their pressing, direct style and specific attacker profiles created higher-than-average penalty-area contact situations that VAR reviewed frequently. Analysts who tracked Brentford's penalty-winning and penalty-conceding rates in their first Premier League months and updated their models accordingly — rather than relying on Championship-era penalty data that predated VAR at that competition level — would have been better calibrated for the season's remaining fixtures.

Building VAR-Adjusted Prediction Models

Incorporating VAR Penalty Rates into Expected Goals

The practical task of building VAR-adjusted prediction models begins with recognising which specific model components require adjustment. The primary adjustments are: goal rate assumptions (use VAR-era data; separate penalty-derived goals from open-play goals in your model), expected goals calibration (account for the increased disallowance rate for marginal offside goals in your xG-to-actual-goals conversion), penalty probability (apply competition-specific VAR-era penalty rates), and result outcome probabilities (use VAR-era match result distributions).

League-Specific VAR Adjustment Factors

Secondary adjustments include: red card event probability (slightly elevated in VAR competitions due to retrospective serious foul play reviews), and goal timing patterns (the introduction of VAR review pauses slightly disrupts the Poisson assumption of uniformly distributed goals, though the effect on overall timing distributions is modest). The framework for building prediction models that incorporate these adjustments is explored in depth in our prediction model building guide, which provides the quantitative methodology for incorporating all of these factors into a coherent analytical framework.

Expert Insight: VAR's impact on prediction accuracy has been asymmetric across different market types. Match result predictions have been affected moderately, as VAR overturns roughly 3 to 5% of results in heavily reviewed leagues. Correct score predictions have been affected more substantially because marginal goal disallowances directly change scorelines. Goalscorer predictions have been affected least, as the player who would have scored the disallowed goal still receives credit for the chance that was taken. Understanding these market-specific VAR effects allows analysts to apply appropriate adjustments rather than treating VAR as a uniform disruption factor across all prediction markets.

Conclusion

VAR's impact on football prediction outcomes is real, measurable, and ongoing. The technology has changed penalty frequencies, goal disallowance rates, red card patterns, and the alignment between match quality and results in ways that require specific analytical calibration for prediction work in VAR competitions. Analysts who maintain VAR-era-specific statistical baselines, track competition-level VAR application patterns, understand the specific markets most affected by video review, and incorporate these considerations systematically into their prediction models will consistently produce more accurate forecasts than those working without this VAR awareness.

The key takeaway is that VAR has made football prediction simultaneously more challenging — introducing new sources of outcome variance through review-driven decisions — and potentially more rewarding for quality-focused analysts — because the quality-reinforcing effect of accurate VAR decisions means that strong quality metrics are more likely to translate into results than in the pre-VAR era. The analysts best positioned to benefit from the VAR environment are those who build their predictions on solid analytical foundations — using xG, xT, form analysis, and head-to-head data — and who maintain the competition-specific VAR calibration that distinguishes sophisticated from surface-level prediction work. Combined with the specific analysis provided in our guides on VAR and penalty decisions and offside technology impact, this guide provides the complete analytical framework for navigating the VAR era in football predictions with accuracy and confidence.