FIND ME MY QB

The research notebook

How to read quarterback statistics.

Six guides explain starts, measurements, comparison groups and dated evaluations.

A matching trait is not a hit rate

If a selected group shares a trait, that describes the group. To test prediction, count everyone who met the trait and apply one outcome definition over a stated window.

Read guide →

Count starts, not appearances

A game played does not establish a quarterback start.

Read guide →

Know which measurement you are comparing

Roster, Combine and pro-day values have different sources and contexts.

Read guide →

Match the time window

Current-season production and completed college careers answer different questions.

Read guide →

Build an evaluation that can be revisited

Keep facts, interpretation and personal judgment on separate lines.

Read guide →

Watch with one question

A focused question makes a game more useful than a vague impression.

Read guide →

A matching trait is not a hit rate

If a selected group shares a trait, that describes the group. To test prediction, count everyone who met the trait and apply one outcome definition over a stated window.

  1. Write down the starting population before looking at results.
  2. Separate known values, unknown values and event-specific records.
  3. Show selected players who missed the preference.
  4. Do not label everyone outside the selected group a failure. Recent careers may be unresolved.
Inspect the cohort

Count starts, not appearances

A game played does not establish a quarterback start.

  1. Use official starting lineups, season ledgers and the existing source notes.
  2. Reconcile every college and any junior-college component.
  3. Record a cutoff date. Future scheduled games are not completed starts.
  4. A verified lower bound may establish a minimum while the exact total stays unknown.
Filter the starts status

Know which measurement you are comparing

Roster, Combine and pro-day values have different sources and contexts.

  1. Read the event and date beside the number.
  2. Do not replace an absent Combine test with a roster value without labeling it.
  3. Treat unverified, did not test and not applicable as different states.
  4. Compare only available measurements; missing is never zero.
Inspect measurement records

Match the time window

Current-season production and completed college careers answer different questions.

  1. Choose current season, all-college career, NCAA or JUCO deliberately.
  2. Read the source cutoff for each quarterback.
  3. Counting totals reflect opportunities; rates still lack opponent and scheme adjustment.
  4. Freeze an evaluation before an NFL outcome if you intend to test what was knowable then.
Open grouped comparisons

Build an evaluation that can be revisited

Keep facts, interpretation and personal judgment on separate lines.

  1. Fact: record the measure, source, event and cutoff.
  2. Interpretation: explain what the evidence may mean and the strongest alternative explanation.
  3. Open question: name an observation that would help distinguish those explanations.
  4. Personal view: write your own position with a date. Do not attribute it to Mike or an analyst.
  5. Revisit: compare a later record with the dated snapshot; explain a revision rather than erasing it.
Open the first frozen snapshot

Watch with one question

A focused question makes a game more useful than a vague impression.

  1. Pick one aspect to observe: pressure response, timing, movement or decision-making.
  2. Record the game and play context; keep observation separate from a count or a grade.
  3. Include a counterexample and what you could not see from the broadcast.
  4. Check the official gamebook before updating starts or production.
Open the watch guide