Best On Ground

Best on Ground · The season in data

How we score: raw points, normalized ratings and Dally M Sim

Learn how Best on Ground turns NRL fans’ 3–2–1 votes into player rankings, with worked examples and comparisons from the 2026 season.

2026 regular season · Rounds 1–27 · Independent fan voting

4,615ballots cast
142unique voters
198matches with votes
383players picked

Three ways to read the same ballots

Best on Ground asks each voter to select three players per match. The first pick receives three raw points, the second two and the third one. All three scoring modes below start from those same ballots; they differ in how results from different matches are combined.

In the 2026 dataset, Nathan Cleary leads raw points with 777, Nathan Cleary leads normalized scoring with 498.8, and Nathan Cleary leads the Dally M simulation with 77. The modes can agree on a leader while placing other players differently.

Raw points: add every 3, 2 and 1

For a player with 12 first-choice votes, two second-choice votes and no third-choice votes, the raw score is (12 × 3) + (2 × 2) = 40. This is direct support in points. A match with 100 ballots distributes 600 raw points, while one with 10 distributes 60, so season totals also reflect participation.

Normalized scoring: equal available weight per match

Divide the player's raw points by the match's total available points, then multiply by 100. If those 40 points came from a match with 20 ballots, the normalized score is 40 ÷ (20 × 6) × 100 = 33.3. Season normalized scores sum the match values; they are a cumulative rating, not a season-wide percentage or a per-appearance average.

Dally M Sim: two simulated judges

Players are ranked within each match by weighted raw points. The first simulated judge awards 3–2–1 to the top three. If the top player has more than 50% of first-choice votes, the second judge repeats that order, producing 6, 4 and 2 points. Otherwise, the second judge swaps first and second, producing 5, 5 and 2. Players below the top three receive zero simulation points.

This is Best on Ground's model, not official judging. It does not apply suspensions, eligibility deductions or other official award rules. The maximum is six per player per match, regardless of voter count. Equal weighted match totals retain the underlying scoring function's input order, so a tie is not evidence that one player was preferred by more fans.

The simulation top ten under all three scoring methods
PlayerRaw pointsNormalized ratingSim points
Nathan ClearyPanthers crestNathan ClearyPanthers777498.877
Kalyn PongaKnights crestKalyn PongaKnights516313.855
Sualauvi FaalogoStorm crestSualauvi FaalogoStorm428314.047
Herbie FarnworthDolphins crestHerbie FarnworthDolphins370288.347
Thomas JenkinsPanthers crestThomas JenkinsPanthers427257.644
James TedescoRoosters crestJames TedescoRoosters502267.242
Braydon TrindallSharks crestBraydon TrindallSharks454299.242
Dylan EdwardsPanthers crestDylan EdwardsPanthers479286.141
Matt BurtonBulldogs crestMatt BurtonBulldogs231197.839
Isaiya KatoaDolphins crestIsaiya KatoaDolphins308256.037

Which view should you use?

Use raw points to inspect the volume of support, normalized ratings to give each voted match the same available weight, and simulation points to follow a match-award race. None removes selection bias from voluntary fan voting. Always check the ballot count, and use the match and player pages to see where a total came from.

Keep exploring the season

Open the full player leaderboard →

Have your say in 2027

The 2026 season has finished. Get an email reminder when the 2027 season begins.