Shield to Baggy Green
Who has a case for a Test opportunity now — and who deserves a closer look for the future? Explore the evidence by role, with uncertainty beside every player.
| Player / last observed state | Role / age | Recent Shield evidence | This season | Adjusted runs / dismissal | 80% interval | P(meets Test reference) | Assessment |
|---|
Recent evidence covers two years. Higher adjusted ratios favour batters; lower ratios favour bowlers. These conditional rate ratios are not literal predictions of a future batting or bowling average. Click a player for their match evidence.
| Match ended | Opposition | Venue | Runs | Dismissals / wickets | Balls |
|---|
The comparison line
Each player is compared with the median estimated ability of recent Australian Test players in the same role. Reference membership requires 400 batting balls or 600 bowling balls in the last two years. The probability includes uncertainty in both the player and the reference. Reference members are labelled explicitly; their self-comparison probability is withheld.
Openers and middle-order batters are assessed separately. Pace and spin bowlers use separate references. Unknown styles and missing role references are shown explicitly. A batting comparison for a wicketkeeper says nothing about keeping readiness.
Evidence, assumptions and checks
Shield to Test
Stan estimates scoring and dismissal rates separately, with partially pooled player effects, recent time trends, competition, venue, opposition team, innings and batting role. Shared players connect Shield and Test evidence. Match effects absorb some conditions; opposition adjustment uses teams, not individual attacks.
Readiness and potential
Ready-now evidence requires four Shield matches and 500 faced balls / 600 legal bowling balls over two years. A provisional strong case means at least 70% model probability of meeting the role reference; 40–70% is in contention.
The development list filters confirmed age and at most five Australian Tests in the source. It uses current ability and uncertainty to identify players worth following. Future improvement, peak ability and eventual Test selection are unestimated.
Historical checks
Fixed cutoffs in July 2024, 2025 and 2026 train strictly on earlier released matches. They score subsequent observed Australian Test innings of players with prior Shield evidence. Positive log-score gains favour Stan. The latest fold is incomplete. These checks concern conditional performance among selected players; they cannot validate how unselected players would have performed.
| Discipline / target | Comparator | Observed innings | Log-score gain | 90% match-bootstrap interval |
|---|
Sampling checks
Sources and provenance
Cricsheet ball-by-ball archives · Cricsheet player IDs · public ESPN player metadata · Cricket Australia: current Shield series.
Source match end plus two days must precede the analysis date. Coverage can lag matches; current archives contain later revisions and incomplete early seasons. No missing performance is filled with zeros. Not-outs remain censored; run-outs are not bowler wickets. Zero-ball dismissal exclusions are retained in the input audit.
Each Stan fit uses five preceding years. Player evidence and reference samples use two years; candidates must have appeared in the Shield within the last year. The development screen does not include an age bonus in current ability.