The best first guess in Classic mode, calculated

Why the first guess matters so much

On your first guess you know nothing. Every later guess benefits from the colours you have already seen; the first one is played blind, which makes it the only guess you can prepare in advance. A good opener does not need to be right. Its job is to split the list of possible answers into small groups, so that the colours you get back tell you which group to search next.

Think of the two extremes. A guess that always returned the same colours, whatever the answer, would teach you nothing. A guess that returned a different combination for every possible answer would hand you the solution on your second try. Real openers sit somewhere in between, and that is exactly what this page measures.

The method

For each of the 35 Classic-mode characters, we simulated a game against each of the 35 possible answers, using the very comparison function the game runs. Every simulation produces a combination of six colours (Category, Role, Gender, Affiliation, Combat, Zone). Answers that produce the same combination are then grouped together: after that guess, there is no way to tell them apart.

Three numbers sum up how good an opener is:

• Candidates left (average): how many answers are still possible, on average, after the guess. This is the number that matters. An answer is more likely to land in a big group than in a small one, hence the formula: the sum of the squared group sizes, divided by 35.
• Information: the same idea expressed in bits, as in the well-known Wordle analyses. Each bit halves the number of candidates on average.
• Worst case: the size of the largest group, in other words what you are left with when luck is against you.

The Classic-mode ranking

The best opener, Renoir, leaves 4.1 candidates out of 35 on average, which is 3.4 bits of information. In the worst case, 7 remain. At the other end of the table, Noco leaves 12.4 candidates on average, three times as many as the best opener.

Why two human bosses come out on top

The answer lies in what the Classic pool is made of. Without mobs or chromatic variants, it holds only 35 characters, and no value dominates: 13 Nevrons and 11 humans share most of the categories, story bosses and optional bosses are almost level, and gender splits into three similar shares, 40% of them "Unknown". In that landscape, the best opener is the one whose every cell cuts the mode into useful pieces.

That is exactly the profile of Renoir and Simon, who are tied at the top of the table. Both are human, both are male, and both fight with a sword and Chroma. Their first three cells each single out about a third of the mode: the humans, the men, and the story bosses for Renoir or the optional bosses for Simon. The Combat cell adds a valuable nuance: "Chroma" is the most common style in the mode (8 characters) and the sword is among the next ones, so the cell separates everyone who shares one of those styles from everyone else. Only their zone tells you nothing, since each of them is the only character to first appear in his.

Third in the table: Ultimate Sakapatate, with 5.6 candidates left on average.

Why the heroes make poor openers

This is the most counter-intuitive result. Opening with Gustave leaves you 9.6 candidates on average, which puts him at number 28 out of 35. Maelle is no better (11.2), and the best playable character, Lune, only comes in at number 22.

The reason is simple: heroes have rare values. Playable, Expedition 33, Lumière, a sword or a rapier… Against half of the possible answers, Gustave returns an all-red row, which only tells you the answer is not a hero from Lumière. Rare values are excellent for confirming a hunch and poor for exploring.

What a good opener looks like in practice

Here are three sample games opened with Renoir, computed on the current data:

These three cases show the three broad kinds of outcome:

• An all-red row is not a failure, quite the opposite. The answer is not human, not a story boss and not male, and uses neither a sword nor Chroma: only a handful of optional bosses remain, nearly all of them Nevrons.
• A green Category and nothing else points to a woman with no tie to the Dessendre family or Expedition Zero: only women of Lumière remain, heroines or NPCs.
• A green Role and an orange Combat point to a story boss who shares the sword or Chroma without being human: only two Axons fit.

What about Complete mode?

Complete mode adds the 34 mobs and the 45 chromatic variants, for a total of 114 characters. The ranking changes completely:

Chromatic Goblu becomes the best opener, leaving 10.7 candidates out of 114 on average. Here there are 87 Nevrons, chromatic variants make up 39% of the pool and the "Chroma" tag appears on 56 characters. A chromatic opener therefore tests the most common values all at once: the Role cell separates optional bosses from bosses, mobs and heroes, and the Combat cell tells you whether the answer shares "Chroma", another style, both or neither.

Among the chromatics, Chromatic Goblu stands out thanks to its two other clues. Its combat tags pair "Chroma" with "Brute Force", the second most common style in Complete mode (21 characters), and its zone, The Continent, is the most frequent one (15 characters). The best-ranked mob, Aberration, only reaches number 21, and Gustave slides down to number 108 out of 114.

After the first guess

The same reasoning applies to every guess, with one important nuance. While many candidates remain, a guess that splits the list well is worth more than a guess that fits every clue, even if it has no chance of being the answer. Once only a handful are left, the opposite is true: play a character that fits, because it can win on the spot.

A few habits for the rest of the game:

1. After an all-red row, look for a guess that splits the remaining candidates in two, for instance by zone or by fighting style.
2. After an orange on Affiliation or Combat, look for a character that shares exactly one of the values: the cell will turn green or red and clear up the ambiguity.
3. With two candidates left, stop thinking and play one of them. You have a fifty-fifty chance of winning now and a guaranteed win on the next guess.

For a column-by-column reading of the grid, see Reading the clues in Classic mode.

The limits of this calculation

This ranking measures an average: it guarantees nothing about today's puzzle. If you prefer safety, look at the "Worst case" column, which shows the largest group an opener can leave you with. The calculation also ignores what you know about the game: a player who knows every zone of first appearance by heart will get more out of a Zone cell than a newcomer.

Finally, these tables are not frozen. They are recomputed automatically from the site's data whenever the roster changes, so the numbers on this page stay accurate. To check any character's clues, see the character codex, then try the opener on today's puzzle.