Make the TV Format's Scoreboard Explain the Game
Make the TV Format’s Scoreboard Explain the Game
Two names, two numbers, a bar between them. It is the page every pitch deck has, and it usually looks fine. Then someone asks the question it cannot answer: is the player on the left finished?
That is the whole test. A scoreboard is not a logo with digits attached. It is an explanation of the state of the game at one moment — what counts right now, who can still change the result, and when the result stops moving. When the display cannot answer those three, the rules underneath may be perfectly sound; the audience just cannot see them.
To keep this concrete, here is a small round you can hold in your head. It is invented for this piece, has not been played or tested, and exists so that the display questions have something specific to attach to.
Demonstration rules
- Two players, A and B.
- Each player gets two attempts.
- An attempt is accepted for 1 point, or not accepted for 0. There is no partial credit.
- Players alternate: A, then B, then A, then B. The fourth attempt ends the round.
- Each attempt is made against one of the round's two declared challenges. The player chooses privately and announces the choice when the attempt begins.
- The rules define no tiebreak.
Four points are available, two per player.
Ask what the viewer needs to know at this moment
Put yourself at the sharpest moment in the round. A's second attempt has just been rejected, so A is finished on 1 point. B is on 0 with a second attempt that has begun and is now under review.
What must a viewer understand to follow what happens next?
- A has 1 point banked and can never add to it.
- B has 0 points, and the only live thing in the round is the attempt being reviewed.
- If that attempt is accepted, the round ends 1–1. If it is not, A wins 1–0.
- Nobody can choose anything else. The round is waiting on an official decision, not on a player.
Notice the shape of that list. Three of the four facts are derivable from two columns of labeled numbers, which means the display can carry them. Notice too what is not on the list: which challenge B chose, how long the review has taken, what the adjudicator is looking at. None of it changes the consequence the viewer is trying to predict. The private choice belongs to the player until the attempt begins; the scoreboard shows the resource, not the content, and should not hint at either.
The fourth fact is the one displays get wrong most often, because "who can still act" has three different answers and most decks have one word for all of them.
- A player can still change the result. B, earlier in the round, with attempts in hand.
- An official decision can still change the result. Now, with both players done and a review open.
- Nothing can change the result. After the adjudication lands.
A frozen pair of digits looks identical in all three situations. If your display cannot say which one the viewer is looking at, it is not explaining the game, it is decorating it.
There is also a question the rules should answer and do not. This round has no tiebreak, and the pending attempt can produce 1–1. That is a hole in the format, not a graphics problem, and the correct move is to flag it: mark it as unresolved in the pitch and keep working. Do not draw an overtime bar to cover the gap.
Name the number's meaning before giving it a position
Before you place a digit on screen, write down what kind of number it is. Four kinds show up on competition displays and they behave differently.
Earned points. Cumulative, and fixed once the rule event that awards them has been adjudicated. The unit is points. In this round, one accepted attempt is 1 point.
Spendable resources. Something a player gives up to do something else. This round has none, and saying so out loud is a decision worth recording, because it is the first thing a collaborator will ask about. If your round has one — a token that buys an extra attempt, a banked pass — it needs its own label and its own group. Spending a resource is not the same as failing to score, and the two should not sit in the same column.
Attempts remaining. A count of opportunities not yet used. The unit is attempts, not points, and the two must never share a row.
Projections. Conditional statements about a future. "One accepted attempt ties the round" is a projection. It belongs in words beside the display, never in the point column where it can be mistaken for something that happened.
A minimal key for this round:
POINTS EARNED A 1 of 2 B 0 of 2 fixed once adjudicated
ATTEMPTS REMAINING A 1 of 2 B 2 of 2 falls when an attempt begins
Those two lines are doing different work, and they have to stay apart. Consider the pending state again: A sits on 1 point with no attempts left, B sits on 0 points with a live attempt. Read the point column alone and A looks like a leader. Read both columns and you see that A is finished, while the only thing that can still change the round is the decision on B's attempt. Same two digits, opposite meanings.
Denominators matter for the same reason, which is why the key carries them. "1" alone tells a viewer nothing until it is attached to a ceiling. If a player can bank at most 2 points, "1 of 2" says how much of that ceiling is reached, and a viewer who knows the ceiling can compute what is still possible without waiting for the commentator. Where two players have different ceilings, show the different basis rather than letting everyone assume a shared one.
Then there is the pending-versus-final distinction. Do not let color carry it alone. A dimmed digit, a pulsing ring, a warm hue against a cool one — each of those vanishes in a freeze-frame, in a press still, and for any viewer who cannot separate the hues. Put the word on the screen: "under review," "pending," "awaiting decision." Let color confirm what the label already states.
Last, hierarchy. A big number is not hierarchy unless the number has accountable meaning. If the largest element on the display is a projection, the screen is advertising a result that has not happened yet, and the size of the type is doing persuasive work instead of explanatory work.
Draw the display immediately before and after a rule event
Work on paper. Four states, each tied to an event the rules actually define.
These are text descriptions of states, not rendered graphics. Nothing here has been drawn, animated or shown to a viewer outside this page, and the interpretation check in the next section is a procedure rather than a result. What follows is a discussion tool for a pitch meeting.
State 1 — before play
POINTS EARNED A 0 of 2 B 0 of 2
ATTEMPTS REMAINING A 2 of 2 B 2 of 2
Event: A's first attempt begins, then is accepted.
POINTS EARNED A 1 of 2 B 0 of 2
ATTEMPTS REMAINING A 1 of 2 B 2 of 2
The attempts counter falls when the attempt starts. The point arrives when the adjudication lands. Those are two events with two changes, and if they are drawn as one beat, the viewer learns nothing about the difference between attempting and succeeding.
Event: B's first attempt begins, then is not accepted.
POINTS EARNED A 1 of 2 B 0 of 2
ATTEMPTS REMAINING A 1 of 2 B 1 of 2
Compare that with the state above it. The point column is identical. The state is not. This is the case that catches a score-only display: if your screen only repaints when points move, it sits frozen through the most informative stretch of the round.
Event: A's second attempt begins and is rejected; B's second attempt begins.
POINTS EARNED A 1 of 2 B 0 of 2
ATTEMPTS REMAINING A 0 of 2 B 0 of 2
B's second attempt is under review.
Follow what that zero in B's column means. The counter falls when an attempt begins, so B's remaining count reached zero the moment the reviewed attempt started. Both counters now read zero while the round is still open. That is a consequence of a defensible choice, and it means the counter can no longer say whether the game is live. Only the label can.
Now the misleading version of the same moment:
A 1 — 1 B
ROUND OVER
The right-hand digit counts an attempt that has not been adjudicated. The label declares the round finished while the decision is open. Declaring a result twice is worse than declaring it once, because a viewer who sees both will believe the more confident of the two. An animation that flies the pending point into B's column does the same thing slowly, and a viewer who misses the correction at the end never learns the point was rejected.
Two options fix the pending state. Either add the explicit line, as above, or keep the live attempt inside the attempts group where it is visible as a third state:
ATTEMPTS A — B 1 of 2 in review
Whichever you pick, the attempt counter and the point counter must never move on the same beat, and a freeze-frame from any of these states has to be readable on its own. Motion can explain a change. Motion cannot carry a state; if you delete the animation, the state must survive.
Event: the review resolves.
If accepted:
POINTS EARNED A 1 of 2 B 1 of 2
ROUND COMPLETE
Tiebreak: not defined by the round's rules.
That last line is not a broadcast graphic. It is a note to the format team, and it is the honest version of a state the display cannot repair. Nothing here has been tested, and this branch is exactly the kind of thing the test is for. Do not let a sudden-death ribbon appear because the display needed to say something. A tie branch is a format decision, and the scoreboard reports it rather than inventing one.
If not accepted:
POINTS EARNED A 1 of 2 B 0 of 2
ROUND COMPLETE
A wins.
Same layout, same positions, one branch different. That is what agreement between display and rule event looks like. The only things that moved are the things the rules said would move.
Check an ambiguous state before polishing the graphics
The pending state is your test case. A decision could end the round or level it, no player can act, and the rules stop short at the exact moment the numbers stop explaining.
Show it to someone who did not write the rules. Do not ask whether the layout is clear. Ask two questions and write down the answers:
- Who can still change the outcome of this round?
- Can the result still change at all?
Then compare what they said with what the rules say, and sort the findings into two piles with two different owners.
The label failed. The reviewer said B still gets to play, or that the round is over, and the rules say otherwise. This one belongs to whoever owns the display. A column headed "attempts remaining" that can read zero while an attempt is live is a naming problem, and it gets fixed with words.
The rule failed. The reviewer said "so it's a tie, then" and your format genuinely has no tiebreak. This one belongs to the format team, and no amount of typography will settle it.
The two piles are easy to confuse, because both arrive as a confused viewer. Keep them apart deliberately. A display revision cannot repair a missing rule, and a new rule cannot repair a label that is false in one of the states it has to cover.
If a graphics maker is on the pitch, the states they cannot build are the states your format has not decided. Ask which ones they would need, and treat the answer as a list of open questions rather than a list of design requests. That conversation has not happened here; the check above is a plan for one, and a real review would record what people actually said rather than what confusion is expected.
Leave one page that a collaborator can read back to you
Put the misleading state and the corrected one side by side. Above them, the rule event in one sentence. Below them, one question: B's second attempt is accepted. What is the score, and is the round over?
If your collaborator has to consult the rules to answer, the display did not do its job.
If they answer confidently and their answer disagrees with the rules, you have found the disagreement on paper, in a meeting, at the cost of a sentence instead of a graphics budget.
Polished graphics do not settle rules the format team has not decided. If the sample scoreboard in your deck implies a tiebreak you never wrote, the deck is making a format promise on your behalf — and the first person to notice will be the one you are pitching to.
Frequently asked questions
What must a scoreboard explain at a given moment?
It must explain the state of the game: what counts right now, who can still change the result, and when the result stops moving. A frozen pair of digits can look identical whether a player can act, an official review is open, or nothing can change.
Why keep points earned and attempts remaining on separate lines?
They are different kinds of number with different units and timing. Earned points are fixed once adjudicated; attempts remaining fall when an attempt begins. Read alone, the point column can make a finished player look like a leader. Denominators such as '1 of 2' show the ceiling, and projections belong in words beside the display, not in the point column.
How should a pending decision be represented?
Use words like 'under review', 'pending' or 'awaiting decision'; do not let colour carry the distinction alone. The attempt counter and point counter must not move on the same beat, and the state must remain readable in a freeze-frame without animation. A display that posts a pending point and says 'round over' declares a result twice while the decision is open.
What does the sample pending state show about counters?
It shows A finished on 1 point with no attempts left, while B is on 0 with a second attempt under review. Both attempts-remaining counters can read zero while the round is still open, because the counter falls when an attempt begins. Only an explicit label can then say whether the game is live or waiting on an official decision.
How should an ambiguous state be checked before graphics are polished?
Show it to someone who did not write the rules, without asking whether the layout is clear. Ask who can still change the outcome and whether the result can still change, then compare with the rules. A label failure belongs to the display owner; a missing tiebreak or other rule gap belongs to the format team and cannot be repaired by typography.