Design your own scoring system
A working method for building fantasy scoring that rewards what you actually want — the four rule types, position weighting, and the usual traps.
Designing scoring is the most interesting problem in fantasy football, and almost nobody gets to do it. Here is a method that works.
Start with a sentence
Before touching a rule, finish this: “In my league, the best player is the one who ___.”
“Scores goals” gives you the standard game. “Does the most work” gives you something else entirely. “Creates chances even when they are not converted” gives you a third thing. The sentence is the design; the rules are just implementation.
Leagues that skip this step end up as the default game plus noise.
The four shapes a rule can take
Per action. Flat points every time it happens. Three points a goal. Use it for discrete events you want to reward linearly.
Per N. Points once per N of something — one point per three saves, minus one per two goals conceded. Use it when a single instance is too small to be worth a point but the accumulation matters. Counts carry no remainder, so five saves pays one, not one and two thirds.
Bounded. Bands. Under 60 minutes pays one, 60-plus pays two. Also the only sensible way to score a percentage or a rating — “two points per percentage point of pass accuracy” is a league where one player scores ninety points a game.
Milestone. Flat points once a combined count is reached within a single match. This is the shape defensive contribution uses. Tiers stack.
Choosing the shape is most of the design. The numbers are easier to tune afterwards.
Weight by position, not by rule
The instinct when a rule feels wrong for one position is to add another rule. Usually the answer is to weight the one you have.
A clean sheet worth four to a keeper, four to a defender, one to a midfielder and nothing to a forward is one rule with four numbers. Set a position to zero and it simply does not score for those players.
This keeps the rule list short, which matters more than it sounds — a twelve-rule league is comprehensible to its players, and a forty-rule league is not.
Conditions: the underrated tool
A rule can carry a condition: score this only when some other statistic clears a bar.
The clean sheet is the canonical example. A clean sheet is a fact about the match. Whether your league pays for it — and whether the defender has to have played sixty minutes to deserve it — is a rule you own. In the official game that gate is baked in and invisible. Here it is a condition you can change or remove.
Once you see the pattern, uses multiply:
- Pay for pass completion only above thirty attempts, so a defender’s four sideways passes do not register as an elite passing display.
- Pay a milestone only if the player passed sixty minutes, so it cannot be farmed in stoppage time.
- Pay an appearance bonus only when the player started.
Conditions are evaluated per fixture, so double gameweeks resolve match by match rather than being averaged into something wrong.
The traps
Double-counting the same event. Goals, shots on target and big chances converted all fire on the same moment. Reward one layer, not three, or forwards run away with it.
Scoring everything. If every action pays, minutes become the dominant variable and the league is decided by rotation. A tight rule set with strong opinions beats a comprehensive one.
Ignoring the price consequences. Scoring is half the system; prices are the other half. Making tackles valuable re-prices every defensive midfielder, which is the system working — but you should look at the resulting market and check it still looks sane. Ours recomputes prices from your rules for exactly this reason.
Negative rules that are too punitive. A −3 for a yellow card sounds proportionate until you notice which players get booked: the ones doing the defensive work you just decided to reward. Check your negatives against your positives.
Test it against last season
The fastest sanity check is to ask what your rules would have done to a season you remember. Would the top ten look plausible? Is there a position nobody would ever buy? Is there one player who is obviously broken?
If a rule set produces a top ten of nine midfielders, that is information.
Then stop
The temptation is to keep tuning. Resist it — a league locks once people start picking squads, and that is a feature. Everyone builds a team around a known set of rules; a rulebook that keeps moving is not a rulebook.
Get it to “interesting and roughly balanced”, publish, and find out what happens.
Start building, or read the scoring rules guide for the mechanics.