Under 2.5 goals settles if the two sides combined score two goals or fewer in normal time. 0-0, 1-0, 1-1 and 2-0 all win. Three goals or more loses. As with every half-goal line there is no draw on the bet itself.
It is the exact mirror of over 2.5, and the two prices are set against each other. If you believe a fixture is a poor over, you are by definition saying something about the under — the question is only whether the price rewards it.
A fixture reaches this page when the combined goal expectation sits clearly below the line. The inputs are the same as the over page — attack and defence ratings, recent form, home and away splits, head-to-head — read in the opposite direction.
Defensive records carry more weight here than attacking ones. Two organised sides that concede rarely will produce unders even when both have respectable attacks, because the limiting factor is how often either side gets a clean look at goal.
Under selections tend to carry steadier probabilities than overs, because a low-scoring fixture has fewer ways to break. That does not make them safer bets — it makes them bets where the price is usually correspondingly shorter.
One late goal settles the market against you regardless of how the first eighty-five minutes went. The probability shown is for the full ninety plus stoppage time, not for the balance of play.
Fixtures where both sides have more to lose than to gain — relegation six-pointers, first legs, matches between two sides already sure of their position — are the classic shape. So are matches on poor pitches, in bad weather, or with a congested fixture list behind them.
Derbies are worth a second look. They are often priced as though emotion produces goals, when in practice caution and cards produce fewer.
Backing unders purely on defensive reputation. A side that concedes little because it dominates possession against weak opposition is not the same as a side built to keep a match tight.
Ignoring a red card as a two-way event. A sending-off can close a game down or open it up, depending on who goes off and when; it is not automatically good for the under.
Treating the market as the safe option. Under 2.5 loses to a single deflection in the ninetieth minute exactly as easily as any other bet.
Under 2.5 is the mirror of over 2.5, and the two prices are set against each other. Saying a fixture is a poor over is by definition saying something about the under; the only question is whether the price pays for it. Under 3.5 asks for less and is correspondingly shorter, while under 1.5 asks for a great deal more and pays accordingly.
If your real view is that one side will be kept out rather than that the match will be quiet, both teams to score, No expresses it more precisely — and a 3-0 settles that bet while losing this one.
The methodology page explains the goal expectation that routes a fixture here rather than to the over page.
The limiting factor in a low-scoring match is rarely how bad the attacks are; it is how few clean looks at goal either side gets. Two organised sides with functioning midfields will produce unders even with respectable forwards, because the chances never arrive. That is why we weight defensive structure above goal-scoring record when routing a fixture here.
Context does much of the work. Relegation six-pointers, first legs, and matches between sides already certain of their position all suppress scoring, as do congested fixture lists, poor pitches and bad weather. Derbies deserve a second look: they are often priced as though emotion produces goals, when caution and cards usually produce fewer.
The league tables home and away splits are the quickest way to check whether a side's defensive record travels, which is where most under selections are won or lost.