When to Predict a Draw (and When to Avoid It)

Draws are the most under-predicted result in football. Roughly one in four Premier League matches ends level, but if you look at the average predictor's gameweek submissions, they will have predicted a draw in maybe one or two out of ten matches. That gap between reality and prediction is where points are hiding.

The problem is psychological. Predicting a draw feels like a cop-out. It feels like you are saying you do not know what will happen. But in reality, predicting a draw is often the smartest and most informed choice you can make.

The draw numbers

Across the Premier League era, draws account for roughly 25-27% of all results. That is consistent season after season. Some years it dips to 23%, some years it pushes towards 28%, but it always hovers around a quarter.

Within those draws, the split is roughly:

• 1-1 - the most common draw, making up about 40% of all drawn matches
• 0-0 - the second most common, around 25% of draws
• 2-2 - roughly 20% of draws
• 3-3 and above - rare, about 5% of draws combined

So when you predict a draw, 1-1 should be your default. It is by far the most likely drawn scoreline, and it lines up with what we know about the most common Premier League scores overall.

When draws are most likely

Evenly matched mid-table sides

This is the classic draw scenario. Two teams sitting between 8th and 15th, similar form, similar quality. Neither is good enough to dominate the other, and neither is bad enough to collapse. These matches regularly finish 1-1 or 0-0.

When you see a fixture like Bournemouth vs Brentford or Brighton vs Wolves, and neither team is in particularly good or bad form, a draw is probably the smartest prediction you can make.

Low-confidence away favourites

This is a situation most predictors get wrong. A top-half team travelling to a mid-table side at home. The away team might be slightly better on paper, but home advantage tilts the balance.

These matches are close to 50/50, and when a match is genuinely 50/50, a draw is the most efficient prediction. You will be wrong a lot, but you will also be right more often than the people who confidently picked one side.