Disc Golf Round Rating Calculator
Fit the propagator line for a round to get its scratch scoring average, what a throw was worth, and your rating — with the fit quality shown.
Propagators are the players in the round whose ratings are already established. Paste as many as you have — the fit needs at least two, and at least two different scores among them.
⚠️This is the core calculation, not the whole official one. The PDGA also decides who is eligible to propagate, throws out rounds that fall too far below a player's established rating, and re-runs the fit afterwards. Those are policy rules rather than arithmetic and are not all public in reproducible detail, so treat this as very close rather than authoritative.
What a round rating actually is
A 54 is not a fixed achievement. It is a different round into a headwind than on a still morning, and on a 10,000-foot layout than on a pitch-and-putt. So the round is calibrated against the people who played it whose ability is already known. Fit a straight line through their (score, rating) pairs and the line is the calibration: where it crosses 1000 is the score a scratch player would have shot, and its slope is what one throw was worth that day.
Two consequences follow immediately. A throw is worth the same to everyone in the round, so the gap between two players never depends on the SSA — that is why ratings compare across courses at all. And points per throw is a property of the round rather than a constant of the sport: a spread-out scoring layout makes each throw worth less, a tight one makes it worth more. Anyone quoting a flat "10 points a throw" is quoting an average of something that moves.
Which propagator actually decides your rating
Not the highest-rated one. Least squares gives propagator i an influence on the slope of (their score − the field's mean score) ÷ Σ(…)², so leverage is bought with distance from the average score and nothing else. Every row below was checked by perturbing that propagator's rating by 10 points and measuring how far the slope actually moved, against the formula:
| Propagator | Rating | Score | From mean | Influence on slope | |
|---|---|---|---|---|---|
| Propagator A | 1015 | 51 | -4 | -0.1 | furthest out, so the most leverage |
| Propagator B | 1002 | 53 | -2 | -0.05 | some leverage, in proportion to the distance |
| Propagator C | 990 | 55 | 0 | 0 | shot the field average — no influence at all |
| Propagator D | 975 | 57 | +2 | 0.05 | some leverage, in proportion to the distance |
| Propagator E | 960 | 59 | +4 | 0.1 | furthest out, so the most leverage |
⭐So the propagator who shot exactly the field average cannot change points per throw at all — not a little, exactly nothing, however wrong their rating happens to be. Their leverage is (score − mean) over something, their score is the mean, and the numerator is zero. A 1050-rated player and a 700-rated player who both shoot the average have identical influence on the slope: none. Checked by moving that propagator's rating by −200, −40, 0, +40 and +200 and confirming the slope did not budge.
They are not harmless, though — they still move the intercept, and so they still move the SSA and everyone's rating with it. The lesson for a tournament director is that a field bunched around one score calibrates badly no matter how many propagators are in it. Spread of score is what buys precision, not headcount.
What one bad propagator does — and it is not what I assumed
I first wrote this section expecting one sandbagged propagator to drag every rating down, and the check came back false: the rating for an average round moved by 2 points. The assumption was wrong in a way worth keeping. A least-squares line pivots about the centroid of its data, so an outlier rotates it rather than sliding it. Add one 1000-rated player who shot 68 to the sample field and:
| Round | Rating, clean field | With the outlier | Change | |
|---|---|---|---|---|
| Score 50 | 1023 | 996 | -27 | a hot round, robbed |
| Score 55 | 988 | 992 | +4 | near the field average, barely touched |
| Score 60 | 954 | 988 | +34 | a blow-up, flattered |
The damage is not a shift, it is a compression. Points per throw collapses from 6.85 to 0.82, so the gap between a hot round and a blow-up falls from 69 rating points to 8 — a 8.6× squeeze toward 1000. The player who actually separated themselves simply stops being paid for it, while the player who fell apart is handed most of it back. That is a more insidious failure than everyone losing a few points together, because the people it robs are the ones with the least reason to suspect the arithmetic.
It is also why this page prints r² instead of hiding it. The clean field fits at 0.9977; with the outlier in it falls to 0.061. The SSA moved from 53.31 to 45.38 and looked like a number the whole time. A rating you cannot sanity-check is a rating you should not trust, and r² is the cheapest check there is — anything much under about 0.9 means your propagators disagree with each other and the line through them is not describing a course, it is describing noise.
When it refuses to answer
Two cases, both returning nothing rather than a number. Fewer than two propagators, because one point does not make a line. And propagators who all shot the same score, because with no spread along the score axis there is no slope to find — the arithmetic would be a division by zero, and the honest output is "this round cannot be rated from these players" rather than a figure produced by a floating-point accident. Both are checked rather than hoped for.
How to use
- Paste the propagators from the round as rating, score — one per line.
- Enter your own score for the round.
- Read the rating, the scratch scoring average and what a throw was worth.
- Check r-squared before trusting it; a low value means the propagators disagree.
Frequently asked questions
How is a disc golf round rating calculated?
The round is calibrated against the players in it whose ratings are already known, called propagators. Fit a straight line through their score-and-rating pairs: where that line crosses 1000 is the score a scratch player would have shot, and its slope is what one throw was worth that day. Your rating is the line evaluated at your score.
What is SSA in disc golf?
Scratch scoring average — the score a 1000-rated player would be expected to shoot on that layout in those conditions. It is not a property of the course alone, because wind, pin positions and tee placements all move it, which is exactly why it is recalculated from the propagators every round rather than looked up.
What is a propagator?
A player in the round whose established rating is used to calibrate it. They propagate their known ability into an unknown round. The PDGA applies eligibility rules about who counts, which is one of several places the official process adds policy on top of the arithmetic here.
How many rating points is one throw worth?
It depends on the round, and that is the point. A layout where scores bunch together makes each throw expensive; a spread-out one makes each throw cheap. Values from roughly 8 to 12 are common. Anyone quoting a flat 10 points per throw is quoting the average of something that moves.
Which propagator affects my rating most?
The one whose score is furthest from the field average, not the one with the highest rating. Least squares gives each propagator influence on the slope in proportion to their distance from the mean score, so a propagator who shot exactly the field average has literally zero influence on points per throw, however wrong their rating is. They still move the SSA through the intercept.
What happens if one propagator has a bad rating?
Not what most people expect. A least-squares line pivots about the centre of its data, so an outlier rotates the line rather than sliding it: a round near the field average barely moves, while a hot round loses a lot and a blow-up gains a lot. The effect is a compression toward 1000, so the player who separated themselves stops being paid for it.
What does r-squared tell me here?
How much the propagators agree with each other. A clean field fits above 0.99. Anything much under 0.9 means the line is describing noise as much as a course, and is usually a sandbagged or mis-entered rating. It is the cheapest sanity check available, which is why this tool prints it rather than hiding it.
Why does the tool sometimes refuse to give a number?
Two cases. Fewer than two propagators, because one point does not make a line. And propagators who all shot the same score, because with no spread there is no slope to find and the arithmetic would be a division by zero. Returning nothing is more honest than returning a figure produced by a floating-point accident.
Is this the official PDGA calculation?
It is the core of it, not the whole of it. The PDGA also decides propagator eligibility, excludes rounds falling too far below an established rating, and re-runs the fit after those exclusions. Those rules are policy rather than arithmetic and are not all public in reproducible detail, so treat the output as very close rather than authoritative.
Why do ratings compare across different courses?
Because a throw is worth the same to everyone within a round, the gap between two players never depends on the SSA — it is just the throws between them times the points per throw. That cancellation is what makes a rating from a long wooded course comparable with one from an open park.
🔒 This tool runs entirely in your browser. Nothing you enter is uploaded, logged, or stored.