Where to place the Connects you own

Find your shape, then its first and second placements

Find your shape in the left column and use the Board owner’s Leader, performer and singer conditions. The first and second candidates come from moving the same card among central, red, blue and yellow slots. This compares placements.

These are representative fully unlocked, bloom-5 comparisons. The search covers all 54 owners, four layouts and 216 slots. Both candidates share the same owner, team, outfit, song and unlocks. Yellow’s side and the song category identify the example conditions.

Only the owner’s Board is fully unlocked; other Boards are undeveloped and other Connect slots empty.

Scroll tables horizontally on narrow screens.

Songs the owner sings

Songs the owner does not sing

≈ means “about the same”: the first/second final-score difference is below 1%. An arrow marks a difference of at least 1% in that example. Other cards or teams can change the order; check the conditions before using a placement.

Rate and bloom can change the order even with identical shapes. This table uses the currently highest-rate card for each shape at bloom 5. Lower-rate and bloom 0–4 placements, named cards, songs and exact results are in the detailed study.

If the owner leads but does not sing the song, the central 10% Score Support does not activate. If the owner neither leads, performs nor sings, red, blue and owner-singer yellow effects are inactive; check unconditional green stats reached below central. Yellow also has other conditions such as all-member songs, so read the actual effect. Move the same card in the score simulator to check your own roster and unlocks.

Check the Board too: four layouts

Blue and yellow swap sides; red has two mirrored layouts. Green is always below central. Red mirroring changes both effect positions and some outer cell positions. Swapping blue and yellow alone preserves the Board’s complete set of occupied positions.

There is one slot each in central, red, blue and yellow. Green has no separate slot; reach it downward from central. Red requires the owner to lead, blue requires them to perform. Yellow depends on the actual song’s singer conditions; the owner need not lead or perform.

Read the shape directions

Numbers count affected cells, excluding the placement. Triangle follows its tip, cross its longer arm, and Ｌ its two arms. Convex follows the side with one protruding cell. Up 2 down 2 reaches two cells each above and below. ◎ is the placement, ■ affected, and · outside the range.

Left convex 6    Right convex 6
·■■        ■■·
■◎■        ■◎■
·■■        ■■·

Eight cells: reach two support nodes and stats together

Fuwawa's red slot ② is a useful example for right triangle 8. Travel directly upward from central ① to Connect ② and set Fuwawa as Leader. Bold entries below are within right triangle 8's reach. — means no effect node. The empty lower-right cell is also covered, so eight cells boost seven actual nodes.

Support increases by 8.1 percentage points in total. Each performer also gains an additional Sen +202.5, Per +337.5, and Tec +337.5. A 135% Connect multiplies the original node effect by 2.35; it does not increase final score by 135%. These seven nodes have no singer requirement and benefit all five performers, but the red effects require Fuwawa as Leader.

At the same slot, which matters: area or rate?

Fuwawa's red ②, surrounding nodes unlocked, every candidate bloom 5. Values are additional effects over having no Connect; stats are per performer.

Eight cells can suit teams with little support; twelve cells can win when support is already high. Iroha's red ② instead has 2% support on the upper left: with both sides unlocked, left triangle 8 was better in the earlier case. Do not assume every Board suits the same direction.

Is replacing your current card worthwhile?

These results replace an existing card with bloom-5 right triangle 8 / 135% on a Tec-focused team led by Fuwawa, using her everyone-Tec +120% outfit. Only Fuwawa's Board is fully unlocked; other Boards are undeveloped and other Connect slots empty. Every existing card below is also bloom 5. The denominator is the score with that existing card.

The replacement is Yukihana Lamy「火照りを鎮める癒しの一口」. Kobo above, Reine「おすまし孔雀と嗜む一杯」, Nene「一緒に歌お！イノセントリズム」, Ninomae Ina'nis「潮騒の記憶をたどって」, and Okayu「パラソル下のリバティキャット」share right triangle 8 with 85% before bloom and 135% at bloom 5: equal development gives equal effects. Japanese titles identify the exact research cards.

Adding right triangle 8 to an empty slot gave +4.15% / +4.11%. That is not the gain from replacing an existing card. The shape's advantage is not exclusive to Lamy.

The development turning point: unlock the 2% support node

One reachable route goes from Fuwawa's central ① to red ② and unlocks the seven effects above, above/below and to the right of ②. It costs 23 Board Pt, requires player Lv10, and uses 395 tier-1 plus 15 tier-2 red-region materials. Subtract existing unlocks. This is not a proven cheapest route or optimal account-wide allocation.

Here, another 2 Pt and 40 tier-1 materials reversed the order. A wide shape needs its important nodes unlocked.

Check card bloom separately. At bloom 0, right triangle 8 is 85%; it beat no best alternative shape in this all-Board comparison. Replacing bloom-5 surround 12 with bloom-0 right triangle 8 at Fuwawa red ② lost about 1.34% on both songs. Bloom-5 strength does not imply strength immediately after acquisition.

What the 54-Board study found

Full unlocks and bloom 5, matched comparisons across three stat-focused teams × two songs. Each holomem uses a different Leader and team, so these are not rankings between holomems.

At blue ③, narrow high-rate shapes targeting key activation nodes can be strong. The model multiplies original activation probability and caps it at 100%. An extra 13.5 in the modifier is not an extra 13.5 percentage points of actual probability.

For yellow nodes, singer and solo/group conditions come first. The model caps combined song bonuses at 10%; extra coverage after reaching the cap adds no score. These two songs cannot establish the best placement for every holomem's other songs.

Detailed conditions and research

The placement search covers 54 owners, four layouts, 216 slots, 17 geometries and 21 shape/rate groups, at bloom 0–4 and 5 separately. Four Leader/performer roles and singer/non-singer cases reuse saved owner-matched HARD solo/group selections: 1,984 matched cases and 168,640 score calculations. All owners use Tec-focused teams; four layout representatives also use Sen/Per teams. This does not optimize an account with all Boards developed or several Connects placed together.

Shape-specific first/second, rate and bloom details (Japanese) · All owner/slot/role scores and inputs

The right-triangle-8 replacement examples below use a separate saved comparison study.

Saved model study dated 2026-10-10. It reuses the recorded random selection of singer-matched 明日への境界線 HARD19 and BLUE CLAPPER HARD19. The study covers 54 holomems, 216 slots, and 22 distinct shape/rate pairs through 2026-10-07, including ★4. Each comparison holds team, outfit, unlocks, song, and bloom constant and gives other shapes equal placement-search opportunities. No rotation; other Connect slots are empty.

Performers use rarity-specific maximum levels, bloom 5, special training 4; AP, life 1000, zero Memory and Member Enhancement. Results are simulator expectations awaiting in-game verification. Two songs do not establish universal rankings, spending value, or optimal multi-slot assignments.

All 54 Board placements · Detailed study and reproduction (Japanese) · All results and inputs · Recorded random song selection

Training basics · Lamy's three roles