How often should PCs level up?

For games with levels like D&D, how frequently do you think PCs should advance, given typical play? HOw do you use the XP system to facilitate that, if at all (group XP, individual XP, level by GM fiat)?

I was just musing about how my 5e DMG suggests PCs should level up every 2-3 sessions, which seems very fast to me. I ran a Pathfinder AP (Curse of the Crimson Throne, converted) using Pathfinder Medium Track XP (party xp), the campaign went from level 2 to level 14, and for most of it the PCs were levelling up every 2 sessions on average, about 8 hours of play. This meant their power was doubling about every 4 sessions; I felt this rate really harmed the game, pushing the PCs up into the double digit levels where the system really breaks down. With my 5e tabletop 'Runelords of the Shattered Star' game (mashing up the Paizo APs Shattered Star & Rise of the Runelords) I'm aiming for a level-up rate about once per 5 sessions, or about 20 hours of play, about half the advertised rate.  I think this should work well for long term play; I would like this campaign to run maybe 4 years, about 90 sessions or so, at 5 sessions/level that should take the PCs to 20th, though I'm fine if they cap out lower or we play awhile using the Epic Boon rules in the DMG. The default 5e system seems to support this sort of progression rate ok just by using mostly lower level monsters, using individual xp, and not being too generous on bonus XP; currently the PCs are level 5 after 15 sessions, with rapid progression to 3rd then slow thereafter.

Both those campaigns (Crimson Throne & Shattered Star) run/ran fortnightly. My recently resumed 4e D&D Loudwater campaign runs fortnightly, evening sessions so shorter, 3 hours or so, and the PCs (using party xp) have levelled up about every 4 sessions/12 hours for a long time - currently just hit level 27 after 96 sessions. 4e is so slow that we only get 1 fight per session and I have to give a good deal of bonus XP to hit that rate; it's still slower than the recommended 1 level per 10 hours of play, or 2.5 typical sessions - same recommendation as 5e.

I also have a couple weekly games:
The Ghinarian Hills is an online text-chat 5e sword & sorcery themed game with individual xp, I use standard monster xp and a fair bit of xp from other sources. After 74 sessions the highest level PC is 15th, so a bit over 5 sessions to level. The other PCs are in the 12-13 range. This rate works pretty well for online game, maybe a bit fast.
Finally there's my weekly tabletop Classic D&D Karameikos campaign. After around 13 months of weekly play the PCs are in the 8-11 level range, though the highest couple had been played previously in an earlier campaign and came in higher. A recently retired Thief PC who'd played from the start at 1st level hit 10th level. Typical advancement rate is about 1 level per 5 sessions, which is the recommended rate in the Rules Cyclopedia. I find this works well; I tend to get this through bonus XP rather than huge piles of treasure, though.

Overall I'm finding in my games that about 1 level per 5 sessions seems to work best, which fits with the norms in older games (eg Gygax recommended that a year of weekly 1e AD&D play should get a successful PC from 1st to 9th), a bit quicker with 4e. But this is about half the default rate recommended by the GM guidance in 3e/PF, 4e and 5e, which all seem to recommend 2-3 sessions to level and 20 sessions in a year of weekly play.   What do you find? What works best for you?

New DM Starting a 5e Campaign

For a new GM I would recommend starting with an episodic feel with a series of episodic adventures - which can be changed/amended according to player activity in earlier adventures, stated preference etc - growing over time towards an epic confrontation. I would therefore tend not to recommend an AP, which in my experience take a lot of practice to use well. The big risk with an AP is that players feel shuttled down a pre-set path with no real input. They can also place a heavy burden on the GM in understanding how it all fits together, though more episodic APs like my current Shattered Star one mitigate this.

The Primeval Thule campaign setting book is excellent IMO and would be a good basis for such an open campaign. It has adventures in it, but you can also drop in many published dungeon adventures as-is; Goodman Games stuff fits the swords & sorcery flavour but I wasn't impressed by the one GG 5e adventure I bought. Lost Mine of Phandelver would need some conversion to give a Thule feel - goblinoids become Beastmen or savages, use a Thulean dragon not the regular green dragon etc. I would generally think it easiest to use the adventures in the Thule setting book plus the NPCs, monsters etc, give the players choice in what to do, riff off that, and prep a few extra encounters you bring each week that you can always throw in if unsure what to do - "Bandits Attack!" Think about what excites you most, think of a villain or group with long term plans, and start introducing them over time. As your skills develop you can create your own adventures, following the DMG guidelines on monsters and treasure. You can draw your own maps or get free ones online; Dyson Logos' site is particularly great - https://rpgcharacters.wordpress.com/maps/. He has adventures too - something like Challenge of the Frog Idol would convert easy to 5e (has Classic D&D stats) & work great with Primeval Thule - and if you like his stuff then Dyson's Delves I & II are available for purchase.

Alternately you could use a WoTC Adventure Path as your basis, but keep it as sandboxy as possible. I think Princes of the Apocalypse might be the easiest for this but I'm not familiar enough with them to say for sure.

The main trick is to present a living world and give the players plenty of freedom of action. A good campaign setting book like the 5e Primeval Thule one supports this; APs rarely do.

Read more: http://www.enworld.org/forum/showthread.php?484731-Thoughts-on-New-Campaign-for-New-Group-for-Returning-DM#ixzz47DPV7YKT

Scarier 5e?

 Originally Posted by Libramarian 

@S'mon based on your experience, what do you think of 0 HP = dead in the 2nd tier, with no revivify and new PCs coming in at level 5?

I'm a big fan of taking PCs out without killing them; 0 hp = dead would have deprived me of several cool scenes where Hakeem the barbarian survived being KO'd by sleeping gas and chewed on by a wight, or the time he held off the orc horde & their sorcerer leader long enough for his friends to escape, was KO'd by a cone of cold and put in a wagon for transport back to their dark tower, only to wake up, break his bonds and escape, returning to his beloved... 

What I do in my online Wilderlands game is use negative hp, you die at hp = negative max. This makes death rare (rarer than 3 death save fails = dead) but also makes popping dying PCs back up again during the fight is unlikely. After an hour if a PC makes a death save they can short rest, spend hd and potentially wake up & be active again.

Not using revivify seems fine, it's never been used IMC. I would probably rather keep revivify and lose raise dead/resurrection/true res though, I think that would keep the fear of death more strongly while reducing in-game lethality. There was no raising IMC until late in the heroic tier, and only 
a couple NPCs (cute, female companion NPCs)  have ever been raised, never any of the dead PCs. 
The 5e game would work fine with no raising (including revifify) at all - rare, permanent 
death works well. 

Start at level 5 - I do that, new PCs start at the bottom of the Tier or with half the XP of the dead PC, whichever is better. Sometimes I have different starter points, eg in my 5e Varisia game 
the current start level is 4th. 5 works well for PCs in the 5-8 range, for level 8-10 PCs I start them at 8th. Currently in my Wilderlands game the active PCs are 12-14 and start level is 11. 
Eventually I'll raise the start level to 14, then 17.

Read more: http://www.enworld.org/forum/showthread.php?482389-Finding-5th-edition-too-quot-safe-quot/page13#ixzz44kPPLmP0

Open World, Tailored Encounters

Sandboxing does not necessarily mean Status Quo encounters.

My 5e online game is an open-world sandbox, but it's not entirely a 'status quo' sandbox - I fairly often have encounters tailored to party level, partly because that's what the available wandering monster tables are designed for. So eg when the party decided to go on a quest to destroy the Black Sun Gate I was using a lot of ca EL 10-12 encounters, matching party level.

You can have sandboxes that are entirely open world and entirely tailored, like Elder Scrolls III, and this is a way to get around the level-up issue. Generally the 'zoned' approach of low and high CR/level zones works better for immersion - eg the area around the Black Sun Gate was one of powerful magical energies.

Currently my party are 12th-14th level and command their own fleet; with 3 warships and 200 warriors (down from 280 before the last battle)  - significant encounters are with enemy armies! It's still entirely open-world; they can take their forces whereever they want and decide who to fight, who to seek to recruit, etc.

Read more: http://www.enworld.org/forum/showthread.php?482109-D-amp-d-is-not-a-good-sandbox/page5#ixzz44aSEdih4