Smart Pachislot A Certain Magical Index 2
Explore this model alongside other AT pachislot machines in our Japanese pachislot machine library. For the basics, read how pachislot works in Japan.
Smart Pachislot A Certain Magical Index 2 (スマスロ とある魔術の禁書目録２) is a FUJI smart-pachislot AT machine introduced in Japan on August 3, 2026. Its Fantasy Replay system temporarily upgrades the value of ordinary bells and replays, creating short high-opportunity windows during normal play and AT. AT progression combines battle wins, reward upgrades and premium game-addition routes such as Automatic Writing / John’s Pen.
Last verified: October 8, 2026.
Key terms: Smart Pachislot · AT · Upper AT · CZ · Setting · Setting hint · Payout rate · Initial hit · Small role · Rare role · Replay · Bell · Watermelon · Bonus · Push-order navigation · Penalty · Ceiling · Reset · Quit timing · Mode · Zone
Quick specs
Quick specs
| Item | Value |
|---|---|
| Official Japanese name | スマスロ とある魔術の禁書目録２ |
| Manufacturer | FUJI |
| Series | A Certain Magical Index |
| Machine type | AT |
| Japan rollout date | August 3, 2026 |
| Model | Lとある魔術の禁書目録2FA |
| Machine format | AT / Smart Pachislot |
| Key system | Fantasy Replay (幻想リプレイ) |
| Displayed-game protection | 800 displayed games + α for CZ or AT; 200G + α after setting change (Administrator research) |
| AT real-game ceiling | 1200 real games; 777 after setting change (Administrator research) |
| Main AT | Imagine Breaker RUSH — 40G + α; approx. 4.0 coins/G (Administrator research) |
Setting probabilities and payout rate
Setting probabilities and payout rate
FUJI publishes the setting-by-setting CZ and AT initial-hit figures used below. Payout rate also rises substantially toward the upper settings.
| Setting | CZ initial hit | AT initial hit | Payout rate |
|---|---|---|---|
| 1 | 1/235.6 | 1/398.8 | 97.9% |
| 2 | 1/233.4 | 1/394.5 | 98.7% |
| 3 | 1/230.8 | 1/389.6 | 100.0% |
| 4 | 1/222.3 | 1/369.5 | 104.5% |
| 5 | 1/215.8 | 1/358.0 | 108.2% |
| 6 | 1/207.2 | 1/338.4 | 113.3% |
CZ routes and success conditions
CZ routes and success conditions
Administrator research. The two normal-play CZs share the same basic ten-game success logic, but Accelerator Chance has substantially stronger role-based success draws.
| CZ | Length / expectation | Success condition | Success reward |
|---|---|---|---|
| Railgun Chance | 10G / about 52% | Role-based success draw. A stop-freeze or [Replay–Red 7–Replay] stopping on the middle reel confirms success. | Imagine Breaker RUSH AT |
| Accelerator Chance | 10G / about 80% | Same confirmation condition, with stronger success rates from push-order bell and rare roles. | Imagine Breaker RUSH AT |
In either CZ, the final-game rare role is success-concentrated. A full 10G with no bell navigation is also a strong success rule in current analysis.
Administrator research. Additional play and result interpretation:
Accelerator Chance is the stronger CZ even though its timer and visible success pattern match Railgun Chance. A final-game rare role and the complete 10G no-bell-navigation rule are phase-specific clear conditions. “No bell navigation” should be recorded from the actual displayed instructions; deliberately ignoring a shown instruction does not recreate that condition.
| CZ role | Railgun Chance conditional AT draw | Accelerator Chance conditional AT draw |
|---|---|---|
| Push-order bell | 2.0% | 10.2% |
| Fantasy Replay | Confirmed class | Confirmed class |
| Weak rare role | 33.2% | Confirmed class |
| Middle / strong / super rare role | Confirmed class | Confirmed class |
These are the success draws after the specified result inside the named CZ. They are not the occurrence frequencies of those roles. The transformed state can change role treatment, so keep the phase and transformation indication when recording the draw.
Small-role probabilities
Small-role probabilities
The small-role table describes appearance frequency. The chance to clear a feature after drawing a role is a separate conditional probability.
Administrator research. Additional play and result interpretation:
Use the right probability column for the transformed state
Administrator research. During Fantasy Transformation, the all-small-role combination is approximately 1/2.1, including replays and rare roles. Bell combined frequency is approximately 1/3.6 in both normal play and AT, but AT divides its outcomes between differently directed bell results. The transformed chance-role interpretation is not an extra independent role draw to add to these frequencies.
A probability table that excludes transformation and a table for an active transformation answer different questions. Count the eligible state games separately if measuring role appearance; otherwise a sample with many transformation periods can look artificially different.
Game flow and Fantasy Replay
Game flow and Fantasy Replay
Fantasy Replay can temporarily transform ordinary bells and replays into chance-role-equivalent opportunities. Because this changes the value of roles for a limited period, the current state matters more than the raw appearance of one bell or replay.
Normal progression reaches CZ or AT through the displayed-game system and role/state draws. AT uses battle wins and reward-upgrade routes, with Charge Game-style escalation and premium Automatic Writing / John’s Pen progression capable of producing large game-count additions.
Administrator research. Additional play and result interpretation:
A short transformation changes which role rule applies
- Fantasy Replay (幻想リプレイ) creates a 3G Fantasy Transformation (幻想変換) state. An ordinary bell / replay can receive chance-role treatment during this period; preserve whether the transformation was active.
- Reach Railgun Chance (超電磁砲（レールガン）CHANCE) or Accelerator Chance (一方通行（アクセラレータ）CHANCE). Both last 10G and use the documented stop-freeze / middle-reel pattern to confirm the clear.
- A successful route starts Imagine Breaker RUSH (幻想殺しRUSH) at 40G + α. Gabriel Battle (神の力（ガブリエル）BATTLE) and reward upgrades add games to that AT allowance.
- The special ending uses approximately 7 coins/G, distinct from ordinary AT’s 4 coins/G. Read the phase before applying a speed figure to remaining time.
Displayed-game ceiling, real-game AT ceiling and reset
Displayed-game ceiling, real-game AT ceiling and reset
Administrator research. This machine has two different counters that must not be conflated. At up to 800 displayed games + α, the machine awards CZ or AT; after a setting change this displayed-game protection is shortened to 200G + α. Separately, the real-game AT interval is protected at 1200 real games, shortened to 777 real games after a setting change, and this route awards AT.
The production budget estimates are about ¥29,000 to the 800-displayed-game CZ/AT protection and about ¥43,000 to the 1200-real-game AT ceiling from 0G under the site’s standard lending/base and normal-play/no-penalty assumptions. A displayed-game count and a real-game count are not interchangeable, and + α can require additional cash beyond the stated counter.
Quit timing should also consider the current Fantasy Replay/state condition and whether reset-related shortened protection remains relevant.
AT battles and reward-upgrade routes
AT battles and reward-upgrade routes
AT progression is built around battles and reward upgrades rather than a single fixed reward. Winning the relevant battle advances the reward route, while Charge Game-type progression can raise the resulting game addition.
Automatic Writing / John’s Pen is a premium route associated with large game-count additions. The appearance of a premium route is not a guarantee of recovering prior investment; it is an internal AT reward state.
Administrator research. Imagine Breaker RUSH is a game-count-managed AT starting at 40G + α, with approximately 4.0 coins/G net increase. Gabriel Battle is entered through prescribed game counts or rare-role draws; a win awards game additions. The ending phase uses a separate approximately 7.0 coins/G rate, so that higher figure should not be read as the ordinary AT speed.
Setting differences and identification
Setting differences and identification
The manufacturer-published CZ and AT initial-hit rates both improve toward higher settings, with the setting gap becoming clearer in a long sample. Use these totals together with direct setting-hint presentations rather than inferring a setting from a single CZ or AT hit.
Administrator research. A cleaner conditional indicator exists inside AT: when a watermelon is drawn while the AT is in its ordinary state, the lottery to enter Mikoto High uses the setting-dependent rates below. Count only watermelons drawn while not already in Mikoto High or Mugino High; mixing those states changes the trial definition.
| Setting | Watermelon → Mikoto High |
|---|---|
| 1 | 14.3% |
| 2 | 14.5% |
| 3 | 14.8% |
| 4 | 15.2% |
| 5 | 16.4% |
| 6 | 16.7% |
Administrator research. Additional play and result interpretation:
The odd/even ending-image tendencies are about 55:45 in their preferred parity, so either image remains common at the opposite parity. A single Touma/Index or Mikoto/Kuroko ending does not confirm odd or even. Preserve direct parity confirmations such as the specific Arisa/Shutaura image separately from weak tendencies.
Statistical setting ranges
Japanese Slot Guide calculation. The expected counts and central 95% count ranges below are calculated from the listed setting probabilities with a binomial model. They describe ordinary statistical variation under that probability; they are not confidence percentages for the actual setting.
Watermelon → Mikoto High (AT ordinary state)
Administrator-research conditional lottery. One watermelon drawn while AT is in its ordinary state is one trigger observation; do not mix watermelons drawn while already in Mikoto High or Mugino High.
25 trigger observations
| Setting | Listed rate | Expected count | Central 95% range | Δ vs S1 | Relative difference |
|---|---|---|---|---|---|
| 1 | 14.3% | 3.6 | 1–7 | — | — |
| 2 | 14.5% | 3.6 | 1–7 | +0.0 | +1.4% |
| 3 | 14.8% | 3.7 | 1–7 | +0.1 | +3.5% |
| 4 | 15.2% | 3.8 | 1–8 | +0.2 | +6.3% |
| 5 | 16.4% | 4.1 | 1–8 | +0.5 | +14.7% |
| 6 | 16.7% | 4.2 | 1–8 | +0.6 | +16.8% |
50 trigger observations
| Setting | Listed rate | Expected count | Central 95% range | Δ vs S1 | Relative difference |
|---|---|---|---|---|---|
| 1 | 14.3% | 7.2 | 3–12 | — | — |
| 2 | 14.5% | 7.3 | 3–12 | +0.1 | +1.4% |
| 3 | 14.8% | 7.4 | 3–13 | +0.3 | +3.5% |
| 4 | 15.2% | 7.6 | 3–13 | +0.4 | +6.3% |
| 5 | 16.4% | 8.2 | 3–14 | +1.1 | +14.7% |
| 6 | 16.7% | 8.4 | 4–14 | +1.2 | +16.8% |
100 trigger observations
| Setting | Listed rate | Expected count | Central 95% range | Δ vs S1 | Relative difference |
|---|---|---|---|---|---|
| 1 | 14.3% | 14.3 | 8–21 | — | — |
| 2 | 14.5% | 14.5 | 8–22 | +0.2 | +1.4% |
| 3 | 14.8% | 14.8 | 8–22 | +0.5 | +3.5% |
| 4 | 15.2% | 15.2 | 9–23 | +0.9 | +6.3% |
| 5 | 16.4% | 16.4 | 10–24 | +2.1 | +14.7% |
| 6 | 16.7% | 16.7 | 10–24 | +2.4 | +16.8% |
200 trigger observations
| Setting | Listed rate | Expected count | Central 95% range | Δ vs S1 | Relative difference |
|---|---|---|---|---|---|
| 1 | 14.3% | 28.6 | 19–39 | — | — |
| 2 | 14.5% | 29.0 | 20–39 | +0.4 | +1.4% |
| 3 | 14.8% | 29.6 | 20–40 | +1.0 | +3.5% |
| 4 | 15.2% | 30.4 | 21–41 | +1.8 | +6.3% |
| 5 | 16.4% | 32.8 | 23–43 | +4.2 | +14.7% |
| 6 | 16.7% | 33.4 | 23–44 | +4.8 | +16.8% |
How to read this: overlapping ranges are expected. A result inside one setting’s range can also be plausible under another setting. Do not use these ranges as a setting confirmation or as a promise of future results.
Normal-play internal modes
Normal-play internal modes
Administrator research. Five modes control the displayed-game CZ/AT target. The mode queue can continue across CZ/AT results, so current displayed games and the recent result both matter.
| Mode | Maximum / characteristic |
|---|---|
| Morning | 200 displayed G + α; setting-change-only mode |
| Normal A | 800 displayed G + α; 100G/even-hundred zones emphasized |
| Normal B | 700 displayed G + α; odd-hundred zones emphasized |
| Heaven Preparation | 800 displayed G + α; 100G zones and a 150G route; next Heaven favored |
| Heaven | 100 displayed G + α; 50G sub-zone also exists |
A setting change establishes the initial special mode and future mode sequence. After CZ/AT completion, the machine advances the queue and selects a new future mode. This internal mode is separate from the three-game Fantasy Transformation state after a Fantasy Replay.
Administrator research. Additional play and result interpretation:
Morning is a setting-change-only mode. Its 200-displayed-game target and the queued future modes are separate from a three-game transformation following Fantasy Replay. The short transformation can change the current draw value without changing which future displayed-game mode was already selected. A recent CZ or AT result advances the queue; it does not establish that all queued modes became Normal A.
Setting hints
Setting hints
Administrator research. AT-end presentation patterns include odd/even and higher-setting tendencies as well as stronger confirmation classes. Treat a directly confirmed pattern according to its published meaning; ordinary character variation without a confirmed classification should not be converted into a numeric setting probability.
Administrator research. Additional play and result interpretation:
| AT-ending image | Published implication |
|---|---|
| Arisa & Shutaura (アリサ＆シャットアウラ) | Even setting |
| Uiharu & Mikoto (初春＆美琴) | Setting 4+ |
| Touma harem (当麻ハーレム) | Setting 6 |
| ITEM (アイテム) | Within remaining 500G |
| SCHOOL (スクール) | Within remaining 300G |
| GROUP (グループ) | Within remaining 150G |
| Frog-faced doctor (カエル顔の医者) | AT revival |
The latter four are remaining-target / revival information rather than setting floors. The image must be the documented AT-ending presentation. A normal scene featuring the same faction or character does not inherit its ending law.
How to play
How to play
No navigation: stop the left reel first. If numbered stop instructions appear, press the stop buttons in the displayed order (1, then 2, then 3). A symbol-target instruction tells you which symbol to aim for; follow it together with any displayed stop order.
Use left-first play in ordinary unassisted normal play. Aim the blue BAR area on the left reel. If a watermelon stops, aim the blue BAR on the middle reel as a guide and stop the right reel freely; missing the watermelon can reduce the payout, so use the target procedure when that pattern appears. Other ordinary patterns can then be completed according to the displayed result.
Follow every push-order/navigation instruction during AT. Middle-first or reverse-order play without machine navigation can cause a penalty, so do not change the reel order during normal unassisted play.
Data policy and official references
Data policy and official references
First-party manufacturer information is labelled official. Administrator research identifies the administrator’s review of published information; it does not by itself mean original measurement or manufacturer confirmation. Calculated examples are labelled separately and explain their inputs. A page verification date records an editorial check, not a new experiment or a remeasurement of every value. See our editorial policy for the distinction.