Where the FPL points went: attack died, defence got paid

Something changed in the Premier League over the last two seasons, and FPL changed its scoring at the same time. The two together moved a large chunk of the game's points from one end of the pitch to the other.

Most managers went the wrong way. This piece shows the numbers in plain terms, and what to do about them.

Everything below comes from our own per-match record of the last three seasons.

Part 1: attack is dying

Start with the simplest question. How many goals get scored in a Premier League match?

Read the two columns against each other, because that is the story.

The whole league scores 16% fewer goals than two years ago. The three clubs you were told to build a squad around have fallen 23%, noticeably faster than everyone else. Teams did not just start defending. They got specifically better at defending against the big sides.

Put another way: in 2023/24 a City, Liverpool or Arsenal fixture produced 1.46 times as many goals as an average match. Last season it was 1.34. The premium for owning a big-six attacker shrank, and the price tags did not.

The same shift in raw totals, plus what the defences got out of it:

Fewer goals, fewer assists, far more clean sheets. That is a league that decided defending was worth more than attacking.

Part 2: FPL started paying for it

In 2025/26 FPL added a new way to score, called defensive contribution, or DefCon. The rule is simple:

It is capped at 2 points. You cannot get 4 for doing it twice over.

Two points sounds small. Across a season it was not:

This is the single most surprising number in the piece. 2025/26 was the most generous season of the three, nearly 10% more points per appearance than the year before. Strip out the 2,834 points DefCon handed out and it lands on 2.75, identical to 2023/24.

So the attacking points that disappeared were replaced almost exactly by defensive ones. The pot did not shrink. It moved.

Where it moved to

Now the important comparison. For each position, what share of all FPL points did it produce, and what share of the game's collective money was being spent on it?

(Budget share means: of all the money tied up in FPL squads worldwide, how much of it sat in that position. Not how many players, how much money. Five cheap defenders and one expensive striker are five slots against one, and roughly the same spend.)

Defenders

Forwards

Defenders were always slightly underpriced. In 2025/26 that more than doubled: they delivered more than a third of every point in the game while holding about a quarter of the money.

Forwards did the reverse. Twelve per cent of the points, twenty-seven per cent of the money.

The part that actually cost people the season

Managers had all season to notice. The money did the opposite. Forwards' share of the game's budget, first six gameweeks against the last six:

Two seasons running, the market drifted out of forwards over the year. In the one season where forwards returned their smallest share in three years, it drifted back in.

That is why the overall winner scored 2,582, the lowest in six years, in the most generous season we can measure. The points were there. They were sitting in a position most people were not buying.

Part 3: the number everyone quotes is the wrong one

If you have read anything about DefCon picks, you have seen DefCon per 90, a player's average defensive actions per ninety minutes. It is quoted everywhere because FPL publishes it directly.

It is the wrong number, for one reason.

DefCon does not pay averages. It pays a threshold: ten actions in a match, or nothing. A defender who racks up 20 actions one week and 2 the next averages 11 per game and gets paid once. A defender who does 10 and 10 averages the same and gets paid twice.

So the number that matters is hit rate.

Hit rate is the share of a player's starts in which he actually cleared the threshold.

A defender who started 37 matches and hit 10+ actions in 22 of them has a hit rate of 59%. That figure tells you how often he will actually score you the two points.

Nobody publishes it, because it needs match-by-match data and FPL deletes that at the end of each season. We keep it.

Does the average get you close? Mostly. Across 63 genuine candidates last season, the average explains about 79% of hit rate. Four times in five it tells you roughly the right thing.

The fifth time is expensive:

Wan-Bissaka and Krejčí averaged exactly the same number of defensive actions. One of them paid out three times as often.

Fofana is the trap in its purest form. An average of 10.8 sits above the ten-action threshold, which makes him look like a lock. He cleared it in roughly one start in three.

Two more numbers worth knowing before you go shopping in the cheap defender aisle:

• Of the 79 defenders under £5.5m who started 15 or more matches, only 12 cleared the threshold in more than half their starts.
• Of the 15 defenders averaging 10+ actions per 90, only 3 cleared it in more than 65% of starts, and 2 cleared it in fewer than half.

"Buy cheap defenders" is not the lesson. Five sixths of them are noise.

Part 4: what FPL changed for 2026/27, and what it didn't

FPL saw the same numbers. What they did, and what it does to everything above:

FPL said the bonus-system change was meant to reduce the overlap with DefCon. We tested how much it actually does, by re-running every match of last season under the new rule and re-awarding the bonus points:

Fifty-three bonus points, taken off roughly ninety regular defenders, across a whole season. Less than one point each. Set that against the 2,834 points DefCon gave them.

The tweak trims. It does not correct. The 35%-of-points-on-26%-of-budget gap survives it.

What does bite is the price list. Comparing the same players' prices at the end of last season to their price now:

Defenders are the only line to move meaningfully. That is roughly a 7% tax, twice the size of the bonus change, and still not enough to close a gap that wide.

Part 5: where the value actually is

Last season's return on every million spent, by position and price bracket. Higher is better; think of it as points per £1m of squad budget.

Three things fall out of that list.

The £9m+ striker was the worst use of money in the game. Premium forwards did not play badly. Haaland was the highest scorer in FPL with 239 points. At £14.7m that is 16.3 points per million, while Jarrod Bowen at £7.8m returned 24.0 and Igor Thiago at £7.2m returned 25.1. You could buy both and have change.

Goalkeepers are the quiet best buy. They top the table, and the 2026/27 bonus changes push them further up.

The instinct is to take the cheapest starting keeper you can find and spend the saving up front. Look at what that saving actually buys. Going from a £5.0m keeper to a £4.5m one frees £0.5m, which at the going rate for a midfielder is worth about 9 points over a season. Last season the gap between the better £5.0m keepers and the bargain-bin options was two or three times that, before you count the bonus points the new save rules will add. You save half a million and pay for it twice over.

Cheap defenders are the engine, if you pick them properly. The best of them last season:

* DefCon hit rate is the share of a player's starts in which he cleared the threshold: 10 defensive actions for a defender, 12 for a midfielder or forward. A 70% hit rate means he earned the two points in seven starts out of ten.

Senesi returned more points per million than any player in the game, at any position. Every name on that list cleared the threshold in more than half his starts, which is what separates them from the other 67 cheap defenders who did not.

That table is also a warning about how fast this churns. Of those five, Senesi is now £6.0m at Tottenham, Lacroix is £6.0m at Chelsea, Mavropanos has left the league altogether, and Andersen starts the season suspended. The edge repeats. The names do not.

The shape this suggests

How to split £100m. Formations follow from it.

1. Pay for your goalkeeper. £5.0m to £5.5m rather than £4.5m. Best return per pound on the pitch, and the new bonus rules favour keepers further. The second keeper is a £4.0m bench filler and always will be.

What that buys at today's prices, using last season's return:

Notice what these have in common. Every one of them is a busy keeper at a club that concedes shots. Saves are the engine of a keeper's score, so the value sits at the mid-table clubs rather than behind the best defences.

2. Five real defenders, chosen on how often they clear the threshold. This is where the structural edge is. Three or four in the £4.5m to £5.5m band, picked on hit rate rather than on who has the biggest per-90 number.

Who currently sits in that band having actually cleared it last season:

Two words of care on this list. Kevin Danso at £5.0m posted the highest hit rate of anyone in the band at 71%, and only 63 points, because he barely played. Hit rate tells you what happens when a player starts, and nothing at all about whether he will. Check the minutes before the percentage.

The newly promoted names being pushed as DefCon bargains this August have no Premier League record at all. Their numbers come from Serie A, the Bundesliga's second tier and the Eredivisie, which historically deflate on arrival.

3. No £9m+ striker. Two forwards in the £7-9m band instead. That band returned 23.9 per million against 14.1 for the premiums, and it frees roughly £6m.

The top two returned more points per million last season than Haaland did, at roughly half the price, and you can own both for less than he costs alone.

4. Spend the freed money in midfield, though not at the top. Midfield still produces the most points of any line, 45% of them, while £9m+ midfielders returned 15.7 per million. One premium at most, then spread the rest.

Every price above is today's. They move weekly, injuries and suspensions land, and a name that fits in August may not fit in September. The bands are the advice. The players are an illustration of it.

What could break all of this

Three things, in order of how much they should worry you.

The football could turn back. This is the big one. The defensive edge exists because the league stopped scoring, and that is a tactical trend rather than a law. Teams adapt, managers change, one signing cracks a low block. City have already bounced from 1.89 goals a game to 2.03. If attacking returns recover, the whole argument weakens.

One season of evidence. DefCon has existed for exactly one season. Everything about hit rate, cheap defenders and where the value sits rests on 2025/26 and nothing else. That is a thin base, and we would rather say so than dress it up.

Everyone else reads it too. If the field piles into defenders, they will still score the points. They stop winning you rank. In a mini-league, an edge everybody has is not an edge.

One more caveat worth stating plainly: every "points per million" figure here is calculated on players who actually played 1,500 minutes. That is knowable only in hindsight. Rotation ruins a cheap-defender strategy faster than bad picks do, and hit rate measured over starts flatters exactly the players who are rotation risks.

The short version

• The Premier League scores 16% fewer goals than two seasons ago, and the big three have fallen 23%.
• FPL replaced those points almost exactly with DefCon. Last season was the highest-scoring of the three, and only because of defensive points.
• Defenders produced 35% of the points on 26% of the budget. Forwards produced 12% on 27%.
• Managers responded by putting more money into forwards as the season went on.
• FPL trimmed defenders' bonus for 2026/27 and raised their prices by £0.33m on average. Neither closes the gap.
• The value is in goalkeepers, £7-9m forwards, and cheap defenders selected on hit rate rather than on the per-90 average everybody quotes.

Every figure here covers the 2023/24 to 2025/26 Premier League seasons and comes from our own records. The bonus-point projection accounts for the change to clearances only. The tackled and goalkeeper-save changes are not included in it, because the match detail they need is not published anywhere. Rule changes for 2026/27 are as confirmed by the Premier League.