normalized_scores

Table actions
rigid
the simplest possible normalization: 5 points for 1st (equal), 4 for 2nd (equal), etc.; null for team tasks (only two ranks)
spread
Distribute points from a pool of {5, 4, 3, 2, 1}, with rank ties between N people sharing the next N available point values. (〃)
scale
Multiply every score by some common factor to achieve a sum of 15 for the task; where this causes any individual score to exceed 5, regrettably abandon the 15-point target and scale down to ensure a maximum of 5. Note that this value does exist for team tasks due to being point- rather than rank-based.
5+3, 3+2, 3½+2½
various ways to normalize team tasks based on rank

5 rows where task = 703

✎ View and edit SQL

This data as json, CSV (advanced)

id ▲ task contestant base ± points rank rigid spread scale 5+3 3+2 3½+2½
3403 Paint a self portrait with a selected brush on a selected canvas 703 Dara Ó Briain 88 2   2 4 2 2 2      
3404 Paint a self portrait with a selected brush on a selected canvas 703 Fern Brady 89 5   5 1 5 5 5      
3405 Paint a self portrait with a selected brush on a selected canvas 703 John Kearns 90 4   4 2 4 4 4      
3406 Paint a self portrait with a selected brush on a selected canvas 703 Munya Chawawa 91 3   3 3 3 3 3      
3407 Paint a self portrait with a selected brush on a selected canvas 703 Sarah Millican 92 1   1 5 1 1 1      

Advanced export

JSON shape: default, array, newline-delimited, object

CSV options: 

CREATE TABLE normalized_scores (
   id         INTEGER PRIMARY KEY,
   task       INTEGER REFERENCES tasks(id),
   contestant INTEGER REFERENCES people(id),
   base       INTEGER,
   adjustment INTEGER,
   points     INTEGER,
   rank       TEXT,
   rigid      REAL,
   spread     REAL,
   scale      REAL,
   '5+3'      REAL,
   '3+2'      REAL,
   '3½+2½'    REAL
);