Why your league table punishes people for showing up

Every league table starts the same way. Points for each round, added up, highest total wins. It is the obvious design, it is what every spreadsheet does, and it is wrong in a specific way that takes about four months to become visible.

The arithmetic#

Points by finishing position, 25 for a win down to 1 for completing a round. Two members:

Martins leads the season without having finished in the top five all year. Duarte cannot catch up by playing better, only by playing more.

And this is not a contrived case. It is the normal case for a society, because a mid-field finish in a field of 30 still scores something, and twelve somethings beat four excellent results whenever the schedule is long enough.

Why this is a real problem, not a philosophical one#

Three concrete consequences:

1. The season is decided by calendar availability. Members who travel for work, have small children, or play a second sport are structurally excluded from winning. They work this out around April and stop caring about the table.

2. Nobody can catch up. Once a leader has a round advantage, the only counter is to play more rounds, which the trailing member usually cannot. A table nobody can climb stops being watched.

3. It rewards the wrong thing. Attendance is a good thing and worth encouraging — but it should be encouraged with a participation prize, not by making it the dominant term in a competition supposedly about golf.

The fix#

Count only each member’s best N rounds.

Five is a good default for a season of 20 to 30 competitions. The rule is one line and its properties are exactly what you want:

• An extra round can replace a worse counting round. It can never simply add.
• A bad round is free once you have N counting rounds.
• Playing more still helps — more attempts at improving your worst counting result — but with diminishing returns rather than unbounded ones.
• A member who plays exactly N rounds and wins them all deserves to lead, and now does.

Re-running the example with best-five:

Choosing N#

The trade-off:

• N too high (say 10 of 25) and the irregular member still cannot fill it. You have moved the problem, not solved it.
• N too low (say 2) and one hot round wins the season. The table becomes a lottery.

A useful rule: set N so that a member attending at your median attendance rate can fill it comfortably. Your own attendance distribution is the input here, and it is one query. Look at last season’s attendance distribution, take the median number of rounds played, and set N a little below it.

For most societies that lands on 4 to 6. If you also run a form table, the same window question applies there with different answers.

The objections, answered#

“It punishes the people who turn up every week.” It stops rewarding them for turning up, which is different. If you want to reward attendance, do it explicitly — an attendance prize, or a small bonus for playing more than X rounds — rather than by letting it silently dominate the main competition.

“Now half the season doesn’t matter for the leaders.” True, and it is a feature: it means a leader who has five good rounds still has to defend them, because everyone below is still trying to improve their fifth. What it removes is the guaranteed accumulation, not the competition.

“It’s harder to explain.” Slightly. “Your best five rounds count” is one sentence, and it is easier to explain than the alternative once someone asks why they cannot catch the leader.

The related failure: counting positions that include guests#

If you award season points from the round’s overall finishing position, a guest finishing third pushes every member below them down a place — in a table the guest is not even in. Compute member-only positions for the season table. See handling guests.