Industry
The fields an actor awards record needs, and the totals it should refuse to store
Structure an actor awards record: the fields one body's decision needs, why nominated and won are separate results, and the totals to refuse storing.
One field ruins most awards data, and it is the one nobody argues about: a column called "awards" holding a number. It adds together decisions made by unrelated bodies, under different rules, from different fields, and presents the total as a quantity.
This page sets out the fields an awards record actually needs, and the derived numbers it should refuse to store. It names no winner, no nominee and no performer.
What to take away
- The atomic record is one decision by one body in one category in one of its years. Everything else is derived from that.
- Nominated and won are different results of the same record, not different records. Store the result as a value.
- Store no totals. A tally across bodies is a sum of different currencies, and once stored it gets quoted.
The record, field by field
| Field | What it holds | The error it prevents |
|---|---|---|
| Awarding body | The body's full name, as it names itself | Two bodies with similar names merged |
| Body's year label | The year the body itself uses for this cycle | A result filed under the ceremony's calendar year |
| Eligibility window | The dates the body defined, where published | An entry nobody can reconcile with another body's year |
| Category as submitted | The category the production entered | Losing the fact that placement is a submission decision |
| Category as decided | The category the body actually judged it in | A silent mismatch between entry and result |
| Award kind | Competitive, honorary, special, ensemble | Honorary results counted as competitive wins |
| Result | Nominated, won, shortlisted, withdrawn, rescinded | A nomination recorded as a win, the commonest transcription error |
| Production | The production as the body named it | A credit attached to the wrong title after a retitling |
| Source | The body's own published list for that year | An entry that can never be rechecked |
| Checked date | When the entry was last confirmed against the source | An error carried for years behind a confident display |
Category as submitted and category as decided need to be separate because they can differ, and because the difference is itself the interesting fact. The pillar on how award bodies decide explains why placement is usually chosen by whoever submits rather than settled by the role.
Awards record fields to check
- Awarding bodyfull name as it names itself
- Body's year labelthe year the body uses
- Eligibility windowdates the body defined
- Category as submitted vs as decided
- Award kindcompetitive, honorary, special, ensemble
- Resultnominated, won, shortlisted, withdrawn, rescinded
- Source and checked daterecheckable entry
Why the year field needs three parts
A body's year is a label the body chose. Its eligibility window has actual dates. The ceremony has a date of its own, frequently in the following calendar year.
Three parts of an awards year
- Body's year labelthe year the body chose
- Eligibility windowactual start and end dates
- Ceremony dateoften in the following calendar year
Store all three and every question becomes answerable. Store one and you have a number that two indexes will disagree about forever, with neither able to explain why. The mechanics of those windows, and how bodies change them, are set out in the awards calendar and how a body's rules change.
Derived numbers to refuse
- A total awards count.It adds peer academies, critics' groups, juries, audience votes and paid-entry competitions as though they were the same unit.
- A win rate.Its denominator is nominations, which reflects campaign spending and body composition rather than anything about the work.
- A weighted score across bodies.The weights are invented, and once published nobody ever asks where they came from.
- Any ranking derived from the above.The argument against it is the same one made in why a ranking of live work measures exposure.
If a count is genuinely needed, derive it at query time from a stated definition, and display the definition next to the number. A derived value stored in a column becomes a fact within two copies.
Should you store this derived number?
Do you need a total awards count?
derive at query time from a stated definition
do not store a total awards count
Identity, and the field that quietly breaks
An awards record links a body's decision to a person and a production, and the person link is the fragile one. Bodies publish names as they were billed, which changes across a career, and two people can share a name.
Authority control for person links
- Body publishes name as billed
- Record every observed name form
- Assign one identifier for the entity
- Attach source and confidence to each link
- Keep the body's printed spelling
Handle it the way libraries do, through authority control: one identifier for the entity, with every observed name form recorded against it. Each link carries its own source and confidence.
Do not normalize names into one preferred spelling and discard what the body printed, because the body's spelling is what lets anyone re-find the entry.
Fields that do not belong
A record of decisions by bodies does not need biographical material. Date of birth, age, nationality, family and location describe a person rather than a decision, and holding them turns a decision log into a profile with obligations attached.
Public reference projects that do carry such material argue their own case for restraint, and the encyclopedic policy on material about living people is worth reading before deciding you need any of it. For a record whose subject is what a body decided, none of it is required.
Common questions
Should a paid-entry competition be in the same table?
Only with the award kind recorded, so it is never silently pooled with peer decisions. The honest option is a separate field for how the field of entrants was assembled.
How do I record a withdrawn or rescinded award?
As a result value on the existing record, with its date and the body's statement as the source. Deleting the record destroys the history and guarantees the old value gets reinstated from somewhere else.
What about awards to an ensemble?
Award kind ensemble, linked to the production, with the people listed if the body listed them. Folding it into individual wins misrepresents what the body decided.
How do I check an entry I inherited?
Against the body's own published list for that year, matching body, category, kind and result exactly. The method, and what to do when the body cannot be found, is in the guide to verifying a claimed award.
