Sports Predictions and Picks

What a prediction from us is

A prediction is three numbers and a decision. The market’s price, with the book that posted it and the time we read it. Our own figure for the game. The distance between them. Then the decision: bet, at a price we name, or no bet.

Nothing else on the page changes that decision. The form, the injuries, the history between the teams are there so you can see what the price may be reading that our number is not. The call itself rests on the gap.

What makes a best bet

A best bet is a game where the gap is big enough. Each sport has its own threshold, set out below, and a game that does not reach it is no bet, stated in the first line of its page. On any given day there may be several best bets across the leagues, or none.

That is the point of a threshold. A price is already a good forecast most of the time. Across the sport hubs we checked the market’s prices against a full season of results: in Europe’s top five soccer leagues the favorite was priced at 51.6% and won 53.2% of 1,087 games played from December to May; in the 2026 MLB regular season every price band landed within three points of its result. A small gap is not an edge. It is noise.

The betting year

The leagues we cover fill the calendar. Football runs from late August to February. Basketball from October to June, with the WNBA through the summer. Baseball from late March into the autumn. Hockey from October to June. Club soccer in Europe from August to May, and MLS from February to December.

So there is no off season for predictions, only a changing mix. In October almost all of them are playing at once.

Every sport is priced its own way

The main bet is not the same everywhere. In football and basketball it is the point spread: a handicap in points. In baseball and hockey it is the moneyline: pick the winner, at a price. In soccer it is the three-way price: home, draw or away. Each of our sport pages explains its own market, and our predictions use the one that sport is bet on.

The scoring explains much of it. Football and basketball produce enough points for a handicap to work across a wide range. Baseball and hockey produce few, so the handicap is fixed at a run or a goal and a half, and the moneyline carries the weight. Soccer produces fewer still, and a draw is common enough to be a result of its own.

Home advantage is not one number

How often does the home team win? It depends on the sport, and on the level.

College sports sit at the top: college basketball home teams won 67.8% of 5,592 games in 2025-26, college football 64.8% of 879 in 2025. Professional leagues sit lower: 55.5% in the NBA, 53.6% in the NFL, 52.9% in MLB, 52.1% in the NHL. Soccer is lowest at 44.0%, because a quarter of its games are drawn. A home edge carried from one sport to another is wrong in both.

How often favorites win

The same pattern holds for favorites. In college football the pre-game favorite won 76.5% of 907 games. In the NHL it won 55.5% of 1,275.

That gap is the difference between betting a sport where the better team usually wins and one where it often does not. A hockey or baseball favorite is a much thinner thing than a college football favorite.

How our numbers are built, sport by sport

Every method is stated on the predictions that use it. In short:

• NFL: last season’s scoring margin per game for each team; the difference is our spread; a bet at 2 points or more from the market’s.
• College football: each side’s results against FBS opponents over two seasons, the home side at home against the visitor away, the gap halved; a bet at 3 points or more.
• MLB: runs scored and allowed turned into a win rate for each side, the two rates matched up, then home field added; a bet at 5 points or more from the price’s chance.
• NHL: the home team’s home win rate and the visitor’s road loss rate, averaged; a bet at 5 points or more.
• Club soccer: each side’s expected goals from its scoring at the venue and its opponent’s conceding, against the league average, through a Poisson model; a bet at 5 points or more on the home win, the draw or the away win. National teams get no bet: they have no common league to measure against.
• Basketball: no fixed rule yet; each prediction states the number it uses and the gap it needs.

Each method is plain on purpose. The inputs are printed on the page, so the number can be rebuilt and argued with.

Free picks, and what they are worth

Free picks are easy to find and hard to judge. Three questions sort them. Does the pick name a price? A side without a price cannot be checked. Does it say how it was reached? A method can be tested; a feeling cannot. Does it ever say no bet? A source that bets every game is not choosing.

Ours are free to read. They name the price, print the method, and pass when there is nothing to bet.

Reading the price

American prices come in two forms. A minus number tells you the stake that makes 100 in profit; a plus number tells you the profit on a stake of 100. Turn either into a chance: -150 is 60%, +150 is 40%, before the book’s margin. Take the margin out of both sides and you have the market’s forecast, the number every one of our predictions is measured against.

The margin is why a coin flip is not a break-even bet. At -110 on both sides, staying level takes a 52.4% win rate.

Sports betting terms

• Price, or odds: what a bet pays, written as an American number such as -110 or +150.
• Margin, or vig: the share the book keeps, built into the prices on both sides.
• Implied chance: the chance a price stands for once the margin is taken out.
• Threshold: the size of gap between our number and the market’s that a bet needs.
• No bet: our call when the gap is smaller than the threshold.
• Unit: one standard stake, used for every bet so records can be compared. More in our guide to units.

Pick your sport

Each sport has its own page, with its markets, its numbers and every league we cover: football, basketball, baseball, hockey and soccer.

Questions and answers