South Africa at New Zealand

Club Level Predictions

The first set of predictions treats a club as the smallest object, as the club develops its members, organizes a gameplan, and deploys its players as needed for each match. This club model has a prediction of 0.607, which translates to predicting New Zealand to win by 3.5.

Each club has a rating and a rating deviation (simiar to a Glicko system), and expected performances can be generated. This allows for simulated matches and spreads like the ones below.

Projected Performances

Projected Spreads

Projected Results

Player Level Predictions

Treating teams instead as an entity made up of the currently active players, I have ratings for each player in an altogether different system. These can be combined to form team ratings once teamsheets are announced, weighting starters a bit higher than the reserves. After the match is played, players can be weighted by their minutes on the field, allowing for an accurate measure of the team’s composition. With these compiled team ratings, we can make predictions, measure inaccuracy, and update the individual player ratings.

Prediction without Player Minutes: New Zealand by 21.4

New Zealand by 17.4 on a neutral field

Away Player	Away elo	Away Percentile	Number	Home Percentile	Home elo	Home Player
Steven Kitshoff	106.88	94	1	77	90.62	Ethan de Groot
Bongi Mbonambi	118.92	96	2	80	95.22	Codie Taylor
Frans Malherbe	135.04	99	3	100	140.24	Tyrel Lomax
Jasper Wiese	102.04	87	8	98	125.47	Ardie Savea
Damian Willemse	100.21	80	10	100	145.55	Richie Mo’unga
Lukhanyo Am	109.06	91	13	64	85.86	Rieko Ioane
Willie Le Roux	102.44	84	15	99	146.16	Beauden Barrett
Thomas du Toit	97.16	83	17	87	97.72	Tamaiti Williams
Vincent Koch	63.77	17	18	93	105.38	Nepo Laulala
RG Snyman	107.93	89	19	36	72.66	Tupou Vaa’i
Pieter-Steph du Toit	86.22	69	20	87	102.26	Dalton Papali’i
Duane Vermeulen	121.03	97	21	79	98.52	Finlay Christie
Grant Williams	107.85	88	22	89	106.25	Braydon Ennor
Manie Libbok	105.62	87	23	84	99.85	Caleb Clarke