Data queried September 8, 2026. Recounted when a fifth table reaches the same depth of play.
This page is for the GM about to start Lost Mine of Phandelver, or Phandelver & Below, and trying to plan the real thing: how many months it runs, how many players fit, how deadly it is, and how many names and open threads you'll be holding by mid-module.
The numbers come from four tables that record their sessions here, given per table so you can find the one that looks like yours, not added up into totals that mean nothing. Four tables is a small sample, none of them is finished, and every figure is a running count that the next session could move. It is still four real campaigns' worth more than the back of the book tells you.
A good default: plan Phandelver as a six-month commitment at a session every three to four weeks, and treat session 8 as the middle, not the end.
The four tables have played 8, 8, 5 and 3 sessions, and none is finished. What matters more for planning is the calendar: all four run a session every three to four weeks, which is the cadence most adult tables actually manage once work and childcare get a vote. At that pace the two tables at session 8 have been playing for five months, and both are still adding sessions.
So the number to put in front of your players isn't "about eight sessions." It's six months as a floor, and a start date for whatever comes next that you don't promise yet. A table that meets weekly would compress that to two or three months, but none of these four does, and yours probably won't either.
What tends to work: having a place for NPC names before session three, because by then the list is already too long to hold in your head.
This is the number the book can't tell you. By session 8, the two tables that far along hold 25 and 64 named NPCs in their records, plus 19 and 43 locations. The two younger tables are already at 27 and 34 NPCs after 3 and 5 sessions. Call it three to nine new names a session, depending on how much your table talks to people rather than fights them.
The practical point: a GM who is still keeping names in their head at session three is about to lose one, and the player who asks "who was that guy in the tavern" will notice. Whatever you use, a campaign tracker sheet, session notes with a margin tally, or an app, have it open by then.
A good default: expect four to ten open quests at any moment, and build your quest list to grow, not to be ticked off.
The four tables have logged 6 to 13 quests each, and at every table most of them are still open: 5 of 11, 4 of 6, 7 of 13, and 10 of 10 at the youngest table. This module hands out hooks faster than a party can close them, so the open list grows through the early sessions rather than shrinking.
That's not a problem to fix, it's what running this module feels like. Plan for it: a quest list that shows what's open at a glance, and a habit of asking the table which thread they're pulling next, beats trying to steer them toward closure.
What tends to work: five to seven players, and telling the table up front that a character can die here, because at half of these tables one already has.
The four parties are 7, 7, 6 and 5 characters. Three of those 25 have died so far: two at one table, one at another, none at the remaining two. Read that as "survivable, not safe": half the tables have lost someone, and half have lost nobody.
If your table is new to the game, that's worth saying at session zero. A death in the first act lands very differently on a player who was told it could happen than on one who assumed it couldn't.
A good default: plan three to four hours a session; that's what the timed sessions here ran.
Only 7 of the 21 live-recorded sessions have a recorded duration, so this is a spot check, not a total. Where the clock ran, sessions averaged 4.0, 3.0, 3.7 and 2.1 hours at the four tables, 3.4 hours overall. That's the ordinary evening-session length, and the module doesn't seem to need marathon sessions to move.
Everything above, per table, in one printable block.
| Per table | Table A | Table B | Table C | Table D |
|---|---|---|---|---|
| Sessions played (none finished) | 8 | 8 | 5 | 3 |
| Months of play so far | 5 | 4 | 4 | 1.5 |
| Sessions per month | 1.3 | 1.5 | 1.1 | 1.4 |
| Players | 7 | 7 | 6 | 5 |
| Characters dead | 1 | 0 | 2 | 0 |
| Named NPCs in the record | 64 | 25 | 34 | 27 |
| Locations in the record | 43 | 19 | 32 | 22 |
| Quests logged | 11 | 6 | 13 | 10 |
| Quests still open | 5 | 4 | 7 | 10 |
| Session length, where timed | 4.0 h (2 timed) | 3.0 h (2 timed) | 3.7 h (2 timed) | 2.1 h (1 timed) |
Sessions, party rosters, codex entries and quest states come from each campaign's own record as it was played, not a survey. Session zeros, prep sessions and test transcripts are excluded everywhere. 21 of the 24 sessions were recorded live; 3 were logged by hand into the same record. None of the four campaigns was imported, so the session dates and the cadence they give are measured, not recalled. Session length exists for only 7 of the 21 live sessions. All four tables are still playing, so every figure is a running count. Ask for a cut we didn't run: hello@thread-fall.com. The data on this page is licensed CC BY 4.0: reuse it freely, credit Threadfall with a link.
These four tables recorded their sessions in Threadfall, which is where the NPC, location and quest counts come from: the codex builds itself from what gets said at the table, so the names are tracked whether or not the GM wrote them down.
Longer than eight sessions, at least at these four tables: the two furthest along are at session 8 after five months of play, and neither is finished. At a session every three to four weeks, which is what all four tables actually manage, budget six months as a floor and don't plan the next campaign's start date yet.
Survivable, not safe. Across the four tables, 3 of 25 player characters have died so far, at two of the four tables. The other two have lost nobody through sessions 3 and 8. Parties of five to seven are what all four run.
By session 8, the two furthest tables hold 25 and 64 named NPCs in their records, plus 19 and 43 locations. That's roughly three to nine new names a session. Whatever you use to track names, set it up before session three, because by then it's already a list you can't hold in your head.
About the tables. Nothing here quotes or summarizes the adventure's text, plot or NPC descriptions; it's what four real campaigns logged as they played: session dates, party rosters, codex entries and quest states. If you want the module's story, read the book.
Both, in effect. Lost Mine of Phandelver (2014) and Phandelver & Below (2023) open with the same chapters, so the first eight or so sessions look the same whichever book is on the table. All four tables here are tagged as the newer book or its Shattered Obelisk edition.
More: the campaign tracker template · the session notes template · the session zero checklist · how long D&D campaigns run · all Threadfall guides