Slolog - Pachislot Session Tracker & Decision Tools

• 

• 

  Home: this month's profit, days played, win rate, quick tools and recent sessions. Free, no ads.

• 

  Sample machine: your EV border becomes the minimum starting spin count, and Border Review splits your sessions by it (met vs. didn't).

• 

  Slolog Pro paywall: yearly pre-selected (¥3,800/yr) or monthly (¥480/mo); 7-day trial if eligible (checked via RevenueCat).

• 

  P/L calendar (v1.1.0): the month's total and every day's full amount on one card, with the day's sessions below.

• 

  One-tap share image (v1.1.0) for X, with the Slolog mark. Hall name and photos are off by default.

• 

  Stats: cumulative profit chart you can trace day by day, per-machine ranking and tag breakdown. Free.

• 

  Manual EV table entry by starting spin count (Pro on your own machines), with two rate columns for exchange-rate conversion.

• 

  Session record: parlor, machine, starting spin count, time, investment and payout. "Same as last time" fills it in one tap.

Entered for the HAMM Award (our main focus), the Grand Prize, the RevenueCat Design Award and the #BuildInPublic Award.

In pachislot, the most important decision happens before the first spin: which machine to sit down at. Pachislot is Japan's regulated slot game, played in parlors. Many machines guarantee a bonus after a set number of spins without one (a "ceiling"), so a machine someone left at a high spin count can have positive expected value (EV).

Strategy sites sell the numbers; Slolog doesn't. Logging, standard stats and export are free. Decision tools — EV tools on your own machines and Border Review — are Pro. Try them free on our fictional sample machine. Track for free. Pay only for decision tools. Slolog doesn't charge players to keep their data; Pro helps them test the decisions behind it.

Slolog runs the whole decision loop, not just record → analyze: find a machine → judge whether the seat is worth it → sit → time the session → play → record → analyze → check your border (your own sit-down rule) → next decision. Pro lives only on the two bold steps, and both run free on the sample machine, so you can see exactly what you would pay for before paying.

• Free, no ads: logging, timer, symbol counter, seat notes, standard stats, calendar, CSV/ZIP export and backup, app lock, plus the full EV workflow and Border Review on a fictional machine with 12 sample sessions.
• Pro sells only decision tools: EV tools on the machines you actually play (import your own table, the minimum starting spin count for your border, EV at any spin count, your parlor's exchange rate) and Border Review. ¥480/month or ¥3,800/year (US $2.99 / $22.99). The paywall appears when you tap "Import" for a table on your own machine, or tap a locked tool.
• We never hold your data hostage: cancel, and every table you imported stays on your phone and in your ZIP backups (delete it whenever you like); only the calculations pause, and they work again if you resubscribe. Your logs, charts and export stay free.

I built it because I kept my own sessions in the iPhone Notes app.

Why free can stay free

Our core infrastructure cost barely grows with each free user (no backend, no AI calls, data on the phone), so costs grow with revenue, not with users. The real ongoing cost is my time updating the hub, which is meant to be how players find the app.

What it does, step by step

• Find (free): our hub (https://respecthq.github.io/slolog/) has published 54 source-cited machine pages since Sept 14, 2026, plus a weekly new-machine calendar. On iPhone, the Add to Slolog (スロログに登録) link on a machine page opens the app and adds that machine after a confirmation (live since v1.0.2, Sept 25).
• Judge (Pro on your own machines; free on the sample machine): bring an EV table you trust (paste, file or a manual form), set your border (e.g. "sit only if estimated EV ≥ ¥2,000"), and Slolog shows the minimum starting spin count that meets it, adjusted to your parlor's exchange rate when the table has two rate columns.
• Sit, time, record (free): a one-tap timer turns sessions into an hourly rate; log investment, payout, starting spin count, tags and photos in seconds.
• Analyze (free): profit, win rate, hourly rate, calendar, per-machine / per-tag / weekday breakdowns, cumulative chart.
• Border Review (Pro on your own machines; free on the sample machine): your results, split by your own rule. For one machine, it runs every session logged with a starting spin count through your own EV table at your current border and exchange rate, then splits them into "met my current border" and "didn't", with count, total and average for each side. Change your border or exchange rate and every session re-sorts instantly, including ones logged free before subscribing; with tags you would re-tag each session by hand every time. Counts are shown, so a thin sample is visible: a mirror for your decisions, not a prediction.

Slolog itself takes no bets and moves no money. It is a log and self-review tool for adult players (18+), and it shows true totals, including losses and rules that didn't hold.

How we built it

• Flutter + drift (SQLite), local-first; the network is used only for RevenueCat and the public machine feed.
• RevenueCat (purchases_flutter, StoreKit 2): Current Offering on the paywall, the pro entitlement with Trusted Entitlements, a 7-day trial shown only when the user is eligible (checked via RevenueCat), and Restore.
• Live since v1.0.3: each paywall entry (Import on your own machine, locked EV memo, locked Border Review, My Page) requests its own RevenueCat Placement (ev_import, ev_locked, border_review, mypage) and falls back to the current offering.
• One gate for every decision tool: Pro, or the sample machine holding its bundled table. The sample machine accepts no other table, even with Pro, so the demo can't become a free calculator for a real machine.
• An Astro static site on GitHub Pages for the hub. Each machine lists its source URL and fetch date, and the build rejects strategy sites as sources. The deep link carries only a machine ID; the app re-fetches the data and asks before saving.
• A Monte Carlo simulator for our fictional demo machine, "Kaido Kinana", so every number in the video and screenshots is our own. Two AI-assisted adversarial security reviews and 320+ automated tests.

Challenges

• Staying neutral. Bundling third-party EV tables would have been the easy paid tier. Refusing meant Pro had to earn its price with tools, not data.
• Exchange-rate math. Parlors buy medals back at different prices. Converting EV correctly, and refusing tables that can't support it, took real math and many adversarial tests.
• Drawing the free line twice. At launch, EV tools were free on one machine at a time and imported tables kept working after cancelling: generous but blurry, since someone could subscribe once, set up every table and cancel. v1.0.3 moves the line to "free on our sample machine, Pro on your own machines" and still keeps every imported table when Pro ends: you pay for the tool, never for your data.

Accomplishments

• Live on the App Store since Sept 19, 2026 (Japan, the US and Canada); "Add to Slolog" links from the hub since Sept 25.
• v1.0.3 live since Sept 28, 2026: the new free/Pro line and per-entry RevenueCat Placements (served by a Targeting rule), shipped nine days after launch.
• Four updates in eleven days (v1.0.1; v1.0.2 on Sept 25; v1.0.3 on Sept 28; v1.1.0 on Sept 30), and the hub grew to 54 source-cited machine pages.
• v1.1.0 live since Sept 30: a one-tap share image of the P/L calendar (hall name hidden by default) and free CSV migration from other trackers (10 years / 3,131 rows in about one second in our test).
• Results so far (as of Oct 1, 2026 JST; production only; excludes Sandbox, TestFlight, the developer's own trial and judges' codes): 6 first-time downloads (App Store Connect); 1 trial start (a real user, on Oct 1) and 0 paid conversions yet (RevenueCat). No ads or paid acquisition yet; it's early and we make no conversion claim.

What we learned

• Another chart wasn't the missing piece. The more valuable question was whether the player's own rule held up, and that became the paid feature.
• RevenueCat's entitlement model let a one-person team wire up the paywall, the pro entitlement and Restore in a single commit (Aug 24, 2026); trial eligibility (Sept 5) and the Pro-state paywall (Sept 11) were added later.

What's next (planned, not built yet)

• Pro on the web through RevenueCat Web Billing on the hub, unlocked in the app by the same pro entitlement.
• A value-triggered paywall experiment: an insight_ready Placement shown when a player's own history raises a question worth checking (was my rule actually working?). A conversion hypothesis to test with the Placements above, not a built feature; no fixed session count is claimed.
• Per-condition EV tables for each machine.
• Also, smaller: a one-time supporter purchase for our Kinana character project (unlocks no feature).

Built With

• astro
• dart
• drift
• flutter
• node.js
• python
• revenuecat
• sqlite
• storekit