Finding Every Mii's Favorite Food: A Testing Method

Every Mii in Tomodachi Life carries two hidden food assignments: a randomly generated favorite food and a randomly generated disliked food. Because the game never shows you either value, the only reliable way to know what a specific Mii loves is to feed it and watch the reaction — and the cheapest way to do that is to start with the foods its personality group favors. This guide turns that idea into a repeatable protocol: how the 48-food database behind our food chart is organized, how the five reaction levels work, what the community-estimate affinity model predicts for each of the four personality groups, and a seven-step testing loop that converts guesswork into a logged, reproducible search. Every number below comes from the site's own data files, so you can reproduce the whole method by hand.

Why favorite foods matter on your island

Feeding a Mii its favorite food produces the strongest positive reaction in the game's food system, and that reaction is worth planning around. Community wikis and long-running player reports describe the favorite-food scene as one of the most dramatic happy animations on the island: the Mii celebrates, its mood visibly jumps, and a satisfied resident tends to show up smiling in the social events that drive friendships, romance, and proposals. Feeding the disliked food does the opposite — a clearly negative reaction. Knowing both hidden assignments therefore does two jobs at once: it gives you a reliable daily happiness lever, and it keeps you from accidentally serving the one item that ruins the mood.

There is a second, less obvious reason to hunt favorites: food testing is one of the few island activities that produces clean, per-Mii information. Each feeding is a controlled experiment — one resident, one food, one observable reaction — and the reaction vocabulary is small enough to log in a single tap. Once a favorite is confirmed and written down, later decisions about that resident get easier: who gets the good stuff on your daily rounds, which reactions to expect when planning compatibility, and which foods to keep away from the dinner table. The rest of this guide is about making that experiment cheap — fewer feedings per confirmed favorite, and zero lost notes.

The search space: 8 categories, 48 foods, 16 common favorites

Your search space is exactly 48 foods organized into 8 categories — drink, dessert, candy, snack, main, fruit, vegetable, and other — and knowing its shape is what makes efficient testing possible. The database is the one behind the food chart. Each entry carries three properties that matter for the protocol: a stable id that the tracker keys on, a category, and a commonFavorite flag marking the 16 items the community most often sees land as favorites. The full breakdown:

Two structural details jump out of that table. First, the heavy categories are main (10 items) and drink (9 items), so a blind brute-force pass burns most of its feedings there — a protocol that can defer a heavy category saves real in-game days. Second, the 16 commonFavorite items cluster in the crowd-pleaser slots: Cola, Juice, and Soda among drinks; Cake, Ice Cream, Donut, and Cookie among desserts; Pizza, Hamburger, Sushi, and Curry among mains. Those flags make good early probes even before personality enters the picture. The two other items — Boba Tea and Sundae — are the niche tail of the list, and exactly where one personality group likes to look first.

How reactions work: five levels and one hidden pair

Every feeding resolves to one of five reaction levels — love, like, neutral, dislike, or hate — and each level is a distinct, observable animation in the game. The five-level scale is also the vocabulary our tracker uses: every food row carries five compact logging buttons, one per level (♥ for love, ▲ for like, – for neutral, ▼ for dislike, ✕ for hate), so recording a result takes a single tap. The scale is deliberately coarse — fine enough to rank two positive foods against each other, coarse enough that you are never unsure which level you just watched.

Under the hood, the site's model attaches a 0–100 affinity score to every food for every personality group, and a fixed banding rule converts a score into a predicted level: 90 and above maps to love, 75–89 to like, 40–74 to neutral, 25–39 to dislike, and anything below 25 to hate. Those bands only power prediction. Once a Mii's actual favorite is confirmed, the model short-circuits: the favorite food is forced to love and the disliked food to hate, no matter what the heuristic says. That override mirrors the game's real behavior — the hidden assignment always beats personality, which is precisely why testing exists.

What the model predicts for each personality group

The four personality groups — Outgoing, Confident, Independent, and Easygoing — each get a distinct ranking over the eight categories, and that ranking is your testing order. The matrix below is the complete community-estimate baseline from the site's data file; read a column top to bottom and you are reading that group's suggested menu:

Four clear menus emerge. Outgoing Miis — the Leader, Entertainer, Trendsetter, and Optimist — peak on candy (95), drinks (90), and desserts (85): an energetic, party-oriented sweet tooth. Confident Miis — Designer, Adventurer, Go-Getter, Charmer — peak on mains (90), the status-oriented restaurant end of the list. Independent Miis — Artist, Free Spirit, Thinker, Lone Wolf — are the interesting case: their top scores belong to the niche other items (85) and vegetables (80), which is why Boba Tea and Sundae earn a place in the database at all. Easygoing Miis — Dreamer, Sweetheart, Softie, Buddy — favor the home-style end: mains (90), snacks (85), and vegetables (85). The chart's reaction preview samples one resident per group — a Leader, a Go-Getter, a Thinker, and a Buddy — so you can compare all four menus side by side. For how a Mii lands in one of these groups in the first place, see our MBTI mapping guide.

One honest property of the matrix: every cell sits between 55 and 95, which means the heuristic alone can only ever predict love, like, or neutral. The two negative levels are never produced by personality affinity. In practice, negatives come from the second hidden assignment — the disliked food, which is exactly as random as the favorite. Personality tells you where to start looking for the upside; nothing tells you where the downside hides except testing.

The testing protocol: seven steps to a confirmed favorite

The protocol finds a favorite in as few feedings as possible by testing the group's strongest categories first, logging every result, and pruning categories as evidence arrives. You can run it entirely inside the food chart's tracker, which remembers your checklist between sessions.

1. 1Identify the Mii's personality group. Look the resident up on the personality chart or the MBTI mapping and note the group: Outgoing, Confident, Independent, or Easygoing. Only the group matters for food — the sub-type (Optimist versus Trendsetter, for example) does not change the affinity matrix.
2. 2Pull the group's top-5 starting candidates. The chart's recommendation panel sorts all 48 foods by the selected group's affinity and displays the five highest. For an Outgoing Mii that is all five candy items at 95; for an Easygoing Mii the panel starts with mains at 90. Those five foods are your first session.
3. 3Feed one candidate at a time, highest affinity first. One food per feeding keeps the evidence clean — rapid-fire gifts blur which item caused which reaction. Watch the animation, grade it on the five-level scale, and only then move to the next candidate.
4. 4Log every result immediately. Tap the matching reaction button on the food's row. The tracker persists the checklist in localStorage under the key lifesimgrid-food-tracker, so tested foods, their logged reactions, and the progress counter survive page reloads and browser restarts — no account, no sync, no lost notes.
5. 5Confirm the winner the moment you see the strongest reaction. Mark that food love in the tracker and set it as the Mii's favorite in the picker at the top of the chart. From then on the chart forces love for that food regardless of what the heuristic would predict, and the favorite is settled.
6. 6If a category comes back flat, eliminate it and move to the group's next-strongest. A run of neutral reactions across a category's items is evidence the favorite lives elsewhere. Keep any like results as fallbacks — a liked food is not the favorite, but it is still a dependable daily choice. Work down the affinity matrix category by category until the winner appears.
7. 7Log the disliked food opportunistically, and reset between Miis. The disliked food surfaces as the strongest negative reaction during ordinary testing — record it the same way and pin it in the picker so the chart forces hate for it. When you switch residents, clear the tracker with one button (or use a separate browser profile), because the checklist is stored per browser, not per Mii.

The absolute bound on the protocol is 48 feedings — the entire database. The group-first ordering exists so you almost never approach it. Complete coverage of a group's two leading categories costs at most 15 feedings (mains and snacks hold 15 items combined for Easygoing and Confident Miis) and as few as 6 for an Independent Mii, whose preferred other and vegetable categories hold only 6 items between them. Those are starting-point expectations, not guarantees — the hidden assignment is random, and an unlucky resident can push you deeper into the matrix.

Worked example: walking an Outgoing Mii through the loop

An Outgoing Mii — say a Trendsetter — should be probed with candy first, and the recommendation panel agrees: all five candy items at affinity 95, with the nine drinks waiting at 90. Here is a realistic session:

• Feedings 1–5 — Chocolate, Gum, Caramel, Licorice, Candy: Chocolate lands a clear like, Candy another like, the rest neutral. No top-level reaction, so candy is out — and because the first session happens to cover the entire candy category, five feedings eliminated it completely.
• Feedings 6–8 — Cola, Juice, Soda: all neutral. Three flat common-favorite drinks in a row is weak evidence against the whole 9-item drink category, so the player parks the six untested drinks (Coffee, Tea, Milk, Water, Beer, Sake) and jumps tiers rather than grinding them out.
• Feedings 9–10 — Cake, Ice Cream: Cake is neutral, and Ice Cream produces the unmistakable strongest positive reaction. That is the favorite.
• Confirmation — Ice Cream goes into the favorite picker, the chart now forces love for it, and the progress counter reads 10 of 48 foods tested — about 21% of the database for a confirmed winner.

Ten feedings produced a confirmed favorite, one fully eliminated category, one parked category, and two logged fallback foods the player can still serve on ordinary days. Notice what the protocol never touched: the 10-item main category and the 4-item vegetable category, an Outgoing Mii's two weakest affinities (65 and 60). The disliked food is still unknown — it is as random as the favorite — so during normal island days the player keeps an eye out for a strong negative reaction, and pins it as hate in the picker whenever it appears. Run the same loop on an Easygoing Mii and only the entry point moves: the panel would lead with mains and snacks instead of candy.

Why favorites are random — and why fixed-answer guides fail

Any guide that prints a fixed favorite food for a given personality is describing one save file, not the game. The data file behind our chart states the situation plainly: Tomodachi Life does not publish an official food-reaction matrix, and each Mii receives a randomly generated favorite food and a randomly generated disliked food. The randomization is per resident, not per personality — two Miis with the identical personality, name, and even identical editor sliders can carry different hidden favorites.

That single fact reshapes what a useful guide can be. A lookup table cannot work, because the answer is not a function of anything you can read off the Mii. What can work is a search strategy: an ordering that puts statistically likelier candidates first, a logging system that never loses a result, and an elimination rule that prunes categories as evidence arrives. That is exactly what the four-group affinity matrix provides — starting points, not answers — and it is why our food chart ships a persistent tracker instead of a table of claimed favorites. Treat any site promising fixed per-personality favorites the way you would treat a horoscope: entertaining, unfalsifiable, and no help with your actual island.

Limits of the model (read this part)

The affinity numbers in this guide are a community estimate, not extracted game data — Nintendo has never published the food system's internals. The presets are explicitly tuned so each group shows a distinct ranking across the categories, because a distinct ranking is what makes a 'which food to try first' recommendation possible at all. That tuning is a modeling choice. It approximates community observation closely enough to be useful for ordering your tests, and like all models it is wrong about something.

Three caveats worth keeping in mind. First, the banding rule means the model can only predict down to neutral — every preset value sits at 55 or higher — so negative reactions are always surprises from the hidden disliked assignment, never forecasts. Second, the 48-item database is a curated subset of foods observed across the Tomodachi Life titles, not a complete in-game food list; if a feeding produces a reaction you cannot find in the chart, log the nearest equivalent and keep going. Third, the tracker stores a single checklist per browser under one localStorage key, so it holds one active Mii at a time — clear it between residents, or keep one browser profile per islander if you are hunting several in parallel. None of this changes the core deliverable: a per-Mii, evidence-based log of what each resident actually loves. The model decides where you start; your feedings decide what you believe.

Try the protocol yourself

• Food Chart & Tracker — the full 48-food database, the per-group recommendation panel, and the persistent tested-food checklist.
• Personality Chart — the 16-type reference with color-coded groups, for step 1 of the protocol.
• MBTI Mapping — browse residents by four-letter code if that is how you keep your roster.
• Compatibility Calculator — pair two residents and decompose the romance and friendship scores once their favorites are logged.