Methodology & Analytical Framework
Comprehensive explanations of how SupeRR Intelligence measures match pressure, assesses player quality, accounts for sample uncertainty, and generates auction recommendations.
1. How Player Ratings Work
Conventional cricket statistics (such as raw batting average or total wickets) reward players who bat during easy game phases or bowl in one-sided encounters. SupeRR Intelligence calculates batting and bowling ratings completely independently based on situational performance:
- Independent Discipline Evaluation: Specialist batters do not receive an unearned bowling rating, and specialist bowlers do not receive an unearned batting rating.
- Runs Above Expected (Batting): Evaluates whether a batter scored more or fewer runs than average batters in identical match situations (taking into account balls remaining, wickets in hand, and required rate).
- Runs Saved (Bowling): Evaluates how effectively a bowler suppressed boundary scoring and captured key wickets relative to baseline expectations.
- Clean Cricket Scoring Compliance: Wides are excluded from balls faced; byes and leg-byes are excluded from bowler runs conceded; run-outs are excluded from bowler wickets.
2. How Match Pressure Is Measured
Every ball delivered in a T20 match carries a distinct situational challenge. The match pressure level represents the difficulty of that specific moment:
Comfortable situations with abundant batting wickets, manageable required rates, or routine opening overs.
Middle overs where required run rates climb steadily and partnerships must be maintained without sacrificing scoring momentum.
Critical death overs, tight second-innings run chases, or situations where top-order wickets have fallen rapidly.
3. How Player Reliability Is Assessed (Evidence Levels)
A player who hits two boundaries off six balls should not be rated higher than an established international who has faced 1,000 deliveries. Our system categorizes players into clear evidence tiers:
| Evidence Tier | Minimum Activity | Official Ranking Eligibility | Recruitment Meaning |
|---|---|---|---|
| Strong Evidence | 180+ balls, 12+ matches, 40+ pressure balls | Eligible | Proven frontline performers with statistically robust ratings. |
| Moderate Evidence | 60+ balls, 5+ matches, 15+ pressure balls | Eligible | Reliable sample size; ratings adjusted conservatively. |
| Limited Match Data | 10 to 59 balls | Display Only | Promising candidates requiring further scouting before official ranking. |
| Not Enough Match Data | Under 10 balls | Unrated (N/A) | Insufficient data to draw an official conclusion. |
4. How Auction Recommendations Are Generated
The platform combines player ratings with franchise squad constraints and market simulations:
- Positional Balance: Ensures the squad contains sufficient designated opening batters, middle-order anchors, finishers, wicketkeepers, spin bowlers, and pace bowlers.
- Budget Ceilings: Uses mathematical optimization to prevent overspending on single stars at the expense of roster depth.
- Simulated Bidding Scenarios: Tests 1,000 potential auction dynamics to calculate the realistic probability of securing a target within designated budget caps.
5. How Accurate Are the Predictions?
All predictive models are evaluated on unseen matches (testing on future chronological matches that the models were never trained on). Across all tests, the situational models significantly outperform simple league averages: