Open with Ghost bodies
Rattleghast, Nekoffin, and Ratacomb are the leading early signals. Use the cheapest pair that wins the next fight and keep Petrirex or Shikitsune when they arrive naturally.
A ghost Type line from the master late snapshot. The page keeps data signal, image evidence, likely editorial paths, and open review work separate.
Deep Guide / DATA + LIKELY
This is a data-led Ghost guide for the largest sample among the selected pages. Petrirex and Shikitsune appear in every final row, while Ratacomb, Nekoffin, Cawnushi, and Opalion form a replaceable support shell.
This guide is built from the current DATA row, official database records, and clearly labeled LIKELY route choices.
Final six-slot equivalent / DATA
Petrirex and Shikitsune appear in every sampled final board. The remaining slots are current DATA companions in this cluster.
Detailed runbook / DATA + LIKELY
This is a data-led Ghost guide for the largest sample among the selected pages. Petrirex and Shikitsune appear in every final row, while Ratacomb, Nekoffin, Cawnushi, and Opalion form a replaceable support shell.
Player decision
DATAConsider it when Petrirex or Shikitsune appears alongside a Ghost opener. The large sample makes the core easy to recognize, yet its lower win rate means you should still compare the lobby before forcing a weak support shell.
Player decision
VERIFYStart with the DATA row, use LIKELY steps as a practical default, and keep VERIFY decisions open when the lobby, item choice, or board position changes.
Route steps / early to late
Rattleghast, Nekoffin, and Ratacomb are the leading early signals. Use the cheapest pair that wins the next fight and keep Petrirex or Shikitsune when they arrive naturally.
Petrirex and Shikitsune are late anchors, so do not replace them for a temporary win streak. Let Opalion or Cawnushi absorb the early front-line duty while the core upgrades.
Ratacomb, Nekoffin, Cawnushi, and Opalion are common current support entries. Cairnage and Saberhorn are the clearest observed replacements when the late lobby wants a stronger body.
The row has several sub-60% companions. Use the last slot for the missing role in the fight rather than treating the most common support unit as mandatory.
Final core / DATA
Core names remain linked to the internal Batomon database. Frequency shown on each token comes from this comp row.
Substitutions / DATA
CairnageOpen databaseThe most observed replacement for the early Ghost slot when a high-tier late body is available.
SaberhornOpen databaseA high-tier flex option that changes the board's front-line profile.
DracanaOpen databaseA mid-tier replacement that can preserve the sixth-slot structure when the rare options miss.
Trinket plan / official effects linked

The leading item in this sample and a flexible way to add cooldown without committing to one lane.
Official effect: On Battle Start A random monster gains +5% Cooldown Speed permanently.
A strong economy option for the long upgrade path around two core units.
Official effect: On Victory Gain $8.
A direct income choice when the board needs to keep rolling and upgrading through the middle turns.
Official effect: Increase Income by $3.The item order follows the current frequency row. The reason text explains a use case, while each card opens the site's official Trinket record for the exact effect.
Upgrade priority
Petrirex41% observedThe first 100% current core and the safest carry checkpoint.
Shikitsune36% observedThe second 100% core unit. Protect its upgrade before investing in low-frequency support.
Ratacomb42% observedA common support target, though its lower final frequency means it should yield to a better late unit when needed.
Positioning / relative roles
This role map protects the two core units behind the first contact while leaving the support bodies movable. The dataset does not provide frame-level position proof, so adapt each lane to the opponent.
Player decision
LIKELYIf the Ghost core arrives late and the lobby is already demanding immediate damage, keep the strongest early board and pivot to a line with the units you actually upgraded. The high sample does not guarantee a good individual shop.
Common mistakes / VERIFY
Quick read
Published GuideRattleghast and Nekoffin are the clearest early signals in this snapshot. Petrirex and Shikitsune form the most consistently observed late core. The sample is useful for comparison, while exact timing, matchup choices, and placement remain open.
Entry signals / DATA
These early Batomon and percentages are signals from the source cluster. They are useful for recognition and comparison, not a mandatory purchase list.
Likely transition
These three stages are derived from the overlap between early and standard rows. Keep the wording probabilistic until a route video confirms it.
Observed early entries. They are entry signals, not a mandatory purchase list.
Likely keeps are early entries that also appear in the standard board. Bridge entries are early units that leave the late row.
Likely late targets are standard entries without a matching early observation in this cluster.
Core priorities / DATA
Core is a presentation rule for this page: standard frequency at 70% or higher. It describes observation frequency, not strength.
Flex choices / DATA
Each row says who replaces whom in the source cluster. It does not prove the replacement is optimal in every lobby.
Trinket direction / DATA
These are sample frequencies, not forced priority orders.
Pivot / caution
If a high frequency core unit does not appear, it is likely worth comparing another line before forcing this snapshot. That is an editorial heuristic, not a gold threshold.
The sample is useful for comparison, while exact timing, matchup choices, and placement remain open. Exact shop rank, reroll timing, gold thresholds, and matchup pivots stay VERIFY.
Positioning
Exact front and back placement, adjacency, and matchup positioning still need a direct video or frame review. Continue at the positioning guide.
Open positioning review guideEvidence footer
Deep guide boundary
The current frequency row is the DATA anchor. Purchase order, placement, pivot timing, and matchup choices are practical LIKELY or VERIFY guidance and should be adjusted to the lobby.
Open source method