Separate a statistical change from a character judgment
What does a bounce-back number actually tell you?
The four rows contain authored rating changes, not dated game pairs. Filtering changes their mean; moving the sliders changes a separate demo score. Neither output establishes an athlete’s mental resilience.
Independent worksheet. No verified game dates, opponents, subsequent-game ratings, or quotation transcript support the fixture. No causal analysis, validated psychology scale, predictive calibration, or live statistics feed is provided.
Try a worked example
The four stored changes sum to 83.8; dividing by four gives 20.95, displayed as +20.9 because the code floors to one decimal rather than rounding. The ≥2-turnover filter removes the +12.0 row: 71.8 ÷ 3 gives about 23.933, displayed as +23.9. This selected subset does not establish that turnovers cause better recovery.
Use real passing statistics carefully
The Hall of Fame explains that NFL passer rating combines completion rate, yards per attempt, touchdown rate, and interception rate. It does not measure leadership or the other intangible qualities of a quarterback. Pro Football Hall of Fame: NFL passer rating
The Broncos’ Bo Nix profile provides season statistics and dated game summaries. To study a real recovery pattern, first identify each actual game and the next game, check the rating calculation, and retain the pairing dates. These four undated scenario rows cannot substitute for that work. Denver Broncos: Bo Nix profile
A rating change is a difference in points on that scale, not a percentage improvement or a probability of winning. Matchup strength, attempts, game conditions, and selection rules can also change a comparison.
Read this worksheet’s exact mechanics
The ordinary score is round(70 + (pressure − 50) × 0.4 + turnovers × 2.5 × multiplier ÷ 1.2), then limited to 50–99. Default inputs 75, 2, and 1.2 would produce 85 through that expression, but a retained special case forces 90. That exception is an authored fixture, not empirical evidence.
Increasing either turnovers or the multiplier can increase this score. That direction comes from chosen coefficients; it should not be interpreted as a benefit of turnovers or media attention. Bands 1, 2, and 3 use thresholds 88 and 78 solely for display.
The chart’s dashed line joins fixed endpoints (45,+30) and (75,+10). It is not a fitted regression. Dot radius is 8 + 2 × stored turnovers, so dot area does not scale directly with turnover count. Stored row indices are separate from the slider score.
Presets change the three sliders while preserving the current row filter. Reset restores the default inputs and all four rows. Download JSON or copy the summary to retain the selected rows, inputs, score, and caveats.
Sources and further reading