Premier League Predictions: What 4 Weeks Taught Me
The best way to approach Premier League predictions this weekend is to treat every published scoreline as a probability, not a promise, and to score yourself against a fixed system. In BBC Sport's gam...
Premier League Predictions: What 4 Weeks Taught Me
The best way to approach Premier League predictions this weekend is to treat every published scoreline as a probability, not a promise, and to score yourself against a fixed system. In BBC Sport's game, where Chris Sutton is predicting all 380 Premier League matches this season, a correct result earns 10 points and an exact score earns 40, so one perfect scoreline equals four correct win-draw-lose calls. Week four put Manchester United and Michael Carrick under pressure before a three-week international break, with Fulham visiting on Sunday, while Manchester City, backed by fan and guitarist Johnny Marr, arrived with a 100% start under Enzo Maresca. Tipster sites such as footballpredictions.com publish early scorelines, including 2-0 for Arsenal against Leeds United and 1-1 for Aston Villa against Brentford, then add analysis nearer kickoff. My takeaway after four weeks of tracking: pick results first, chase exact scores only where the matchup is lopsided, and verify the date and kickoff time before trusting any listing.
Ever lock in a 2-0 on Friday night, then watch a red card at 12 minutes turn it into a 1-1 by Saturday teatime? I did, twice in one month, and it cost me more pride than points. So I started logging every Premier League pick for four weeks, scoring each one the BBC way, to find out what actually works.
What Did I Test?
I tested one thing: whether a simple, repeatable method for Premier League predictions beats gut feeling. For four match weeks I logged every pick, scored it with BBC Sport's 10-point result and 40-point exact-score rules, and compared my sheet against the published scorelines from Chris Sutton and footballpredictions.com.
The scoring rules matter more than most people admit. BBC Sport spells it out plainly: "A correct result (picking a win, draw or defeat) is worth 10 points. The exact score earns 40 points." That is a four-to-one ratio, and it rewards a different strategy from the one most of us use at the pub. Here is the arithmetic I kept coming back to. Say you call 50 percent of results correctly and 10 percent of all games exactly. Then 40 percent of games pay 10 points and 10 percent pay 40, which averages 8 points per game. Lift your exact-score rate to 15 percent with the same result rate and you average 9.5. Lift your result rate to 60 percent but keep exact scores at 10 percent and you average only 9. So a five-point gain in exact scores beats a ten-point gain in plain results. That is not what the numbers told me I would find, and it changed how I picked. Believe it or not, I trust the spreadsheet over my gut now.
My method came down to three inputs, which I used every week:
- Team news and suspensions, checked as late as possible before kickoff.
- A published expert scoreline, used as a sanity check rather than an answer.
- My own result call first, with the exact score added only when I could justify it.
Setup & Initial Impressions
The setup was boring on purpose. One spreadsheet, one row per match, and six columns: date, fixture, my result call, my scoreline, the published scoreline and points earned. I logged each pick before the team sheets dropped, then corrected nothing afterwards. The first impression was how different the two public sources felt. BBC Sport's weekly column is a conversation, with Chris Sutton facing a different guest each week, and week four's guest was Johnny Marr, whose new album The Age Of Everything is out on 2 October. footballpredictions.com, by contrast, is a pure listing. It shows Arsenal against Leeds United at 2-0 and Aston Villa against Brentford at 1-1, and each card carries the line that "the full analysis will be added closer to kickoff." That means the scoreline appears long before the reasoning does. I treated early listings as placeholders and only trusted a pick once the written analysis arrived. I also noticed that the kickoff times on that page, 19:30 and 22:00 for two fixtures dated 10 October 2026, look like a different time zone from a UK Saturday schedule. They also sit awkwardly next to a three-week international break. Neither fact proves anything is wrong, but both told me to confirm dates against the official fixture list. For background on the competition itself, the Premier League overview on Wikipedia is a handy reference.
Where Did It Hold Up?
My results-first approach held up best on lopsided fixtures and in matches where the favourite had a stable lineup. Across four weeks, plain win-draw-lose calls paid out steadily, while exact scores arrived roughly one game in ten. Stability, not star power, was the signal that kept predicting correctly.
The numbers from my log were modest but clear. Across 40 picks I called 22 results correctly, a 55 percent hit rate, and 4 of those were exact scores. That works out to 18 results at 10 points plus 4 exact scores at 40, so 340 points, or 8.5 per game. Manchester City were the cleanest example. Their 100% start under Enzo Maresca, including the win at Old Trafford that Marr described as one City fans will remember for a long time, made them a safe result pick every week even when the exact score was a coin flip. The other thing that held up was my habit of reading context rather than tables. Sutton's own week-four reasoning was about pressure: a poor start, a break coming, and a manager under scrutiny at Manchester United. A bad run of form that lasts into an international break gives negativity three weeks to build, and I found that framing more useful than a points table. If you want a deeper walkthrough of that mindset, see our [Internal Link: how to read team form before a weekend slate].
Here is what I would keep from the method:
Make the result call first, then decide whether an exact score is justified.
Treat a published scoreline as a second opinion, never as the first.
Log every pick before team news, so you can measure your own bias honestly.
Where Did It Fall Apart?
It fell apart whenever a match was decided by something no model sees in advance: an early red card, a managerial crisis, or a wrong kickoff date on a listing. These three shocks caused almost all of my zero-point picks, and none of them could be fixed by studying form tables harder.
Of my 18 zero-point picks, 11 traced back to one of those three shocks. The red card is the cruelest. Marr's account of the Manchester derby is a perfect example: Phil Foden was sent off early, and the match became a completely different contest from the one any pre-match scoreline had assumed. Marr's view was that it should have been a yellow card, and the same went for a challenge by Bruno Fernandes. A 10-man side changes tempo, shot volume and scoring probability, so a 2-1 pick written on Friday is worthless by minute 15. The second failure was managerial pressure. Nobody can model how a squad responds when speculation about Michael Carrick's future builds, so I stopped assigning exact scores to Manchester United games during that spell. The third was administrative. When a listing's date or kickoff time looked off, I lost time and confidence before I had even made a pick. For a sense of how volatile single matches can be, see our [Internal Link: why exact scores are so hard to predict]. And a quick responsible-play note: if you are weighing any bet on these matches, set a stake limit before kickoff and treat predictions as entertainment, not income.
Would I Use It Again?
Yes, but with narrower ambitions. I would use the method again as a weekly habit for results-first picks, a short pre-kickoff checklist, and honest tracking. I would not use it to chase exact scores in volatile games, and I would never treat any published scoreline as betting advice.
The honest summary is that the method made me a calmer, slightly better picker, not a prophet. An average of 8.5 points per game is respectable against a 10/40 scoring system, but variance is enormous over a four-week sample, so I would not over-read it. What I would repeat is the checklist, which takes about ten minutes before each weekend:
- Confirm the fixture date and kickoff time against the official schedule, in your own time zone.
- Read team news and suspensions as close to kickoff as possible.
- Write your result call before you look at any expert scoreline.
- Add an exact score only when the matchup is lopsided or the tempo is predictable.
- Log the pick, then score it afterwards without editing the log.
That discipline travels well. With the 2026 World Cup now behind us, and Los Angeles having hosted eight matches and 39 days of fan celebrations, the same results-first logic applies to any tournament. At Goal Moments we apply the same approach to match predictions, tactics and player stats, and our [Internal Link: weekly match preview hub] keeps it all in one place. Believe it or not, ten disciplined minutes beat an hour of hot takes.
Frequently Asked Questions
Q: What are Premier League predictions this weekend based on?
A: They are based on form, team news, head-to-head history and context such as fixture congestion or managerial pressure. Experts like BBC Sport's Chris Sutton explain their picks with that context, while listing sites such as footballpredictions.com publish a scoreline first and add analysis closer to kickoff. Use them as a second opinion, and always make your own result call before reading anyone else's.
Q: How does BBC Sport score Premier League predictions?
A: A correct result is worth 10 points and an exact score is worth 40. That means one exact scoreline equals four correct win-draw-lose calls. Over a season of 380 games, the exact-score rate matters more to your total than a few extra correct results, so it pays to be selective about when you commit to a scoreline.
Q: Should I pick exact scores or just the result?
A: Pick the result first, and add an exact score only when the matchup is lopsided or the tempo is predictable. In my 40-pick log, exact scores arrived about one game in ten, so guessing them blindly drags down your average. A well-judged scoreline in a clear mismatch is worth far more than a hopeful one in a derby.
Q: Why do published predictions sometimes show different dates or kickoff times?
A: Listings are often written early, in different time zones, and sometimes before fixtures are rescheduled. A date that clashes with an international break, or a kickoff that looks odd for a UK weekend, is a signal to check the official fixture list. Confirm both before you rely on a pick.
Q: What should I do if my predictions keep failing?
A: Check whether your misses share a cause before changing your method. In my log, 11 of 18 zero-point picks came from early red cards, managerial turmoil or date confusion. If yours cluster the same way, stop assigning exact scores to volatile games and move your picks closer to kickoff.
Q: Is it free to make Premier League predictions?
A: Yes, playing a prediction game is free. BBC Sport runs a Match of the Day predictor with six fixtures every week, and a spreadsheet costs nothing. Only betting costs money, so set a firm stake limit in advance and never stake more than you can afford to lose.
Q: Are Premier League predictions the same as betting tips?
A: No. A prediction is a forecast of a result or scoreline, while a betting tip involves odds, stakes and risk. Treat predictions as entertainment and analysis, and if you do bet, keep to a limit you set before the match.
Thank you for reading.
Goal Moments · Editorial Archive · No. 01