Is Forum Logic overly in love with prepared casters? (Baldur's Gate Experiment)

[Endarire]

Intro
Greetings, all!

I've played D&D 3.x since the 3.0 books were hot off the presses at GenCon 2000, but I've studied the rules far more.  Much or most of this was from forums - handbooks, discussions, etc.  Now that I've played D&D more since then (including lots of D&D video games of varying editions), and noticed how people playing Pathfinder 1e seemed to prefer to play spontaneous casters over prepared casters despite the 1 level tax/penalty/delay on all spell levels of 2-9, I became curiouser and curiouser:  In practice, what matters more (spontaneous or prepared) and why?

The Arcane Experiment
I conducted an experiment by playing Baldur's Gate II with a Fighter (Kensai) dual classed to Wizard (who effectively acted as a Wizard for the BG2 portion of my run, complete with prepared casting) and another run done as a pure Sorcerer.  No spell swapping was allowed in this case, meaning whatever I chose, I had to use.  (One notable exception was using a save editor to change a spell known since the Enhanced Editions of the games 'fixed' the sunfire spell and made it subject to spell resistance, but that was all.)

In each case, I played to the EXP cap which put me as a level 31 Sorcerer and a Fighter (Kensai)13/Wizard28.  (I played a Sor first.)  I discovered just how burdensome spell preparation felt when I beat Baldur's Gate II: Enhanced Edition for the first time as this Kensai/Wizard.  Sure, I could know every spell in the game a Wizard could learn, and I at least came close to doing so, but smart spell selection and use were key to victory in each case.  Scrolls, potions, and wands were common enough that I could use them to supplement whatever spells I knew or had prepared.  Nevermind that as a Sorcerer Sorcerer I felt like a poser when I was awarded my Planar Sphere stronghold at the end of a quest chain, and didn't feel that way when I finished this quest chain as a Fighter/Wizard.

The Baldur's Gate games are normally long with tens or hundreds of hours of content in each game if you do everything.  (In each case, I did all the optional quests except for some NPC-specific quests in chapters 2 and 3 before progressing to chapter 4.)  In theory, my Kensai/Wizard was better because it didn't need to cast spells as much due to having 13 Fighter levels, and there were plenty of times I didn't cast spells because it was faster to auto-attack stuff, but guessing what spells would be most useful that day as well as long term (since I didn't like changing my spell preparations often) was nowhere near as satisfying in the long-term as playing "Superman" as one person put it and just spontaneously casting what I wanted.

A significant concern for me was that time spent swapping spells at first made me feel smart because I was adapting, then more burdensome since I missed the spontaneity of a Sorcerer.  I only realized how useful consumables were when I started using them as my Fighter/Wizard because I had horded these things, realized I was pretty close to the end of the game, and started using my stash.  I never got close to depleting it.)

Once the Enhanced Editions were available, I also ran characters through Baldur's Gate: Enhanced Edition since it included the Sorcerer class.  (Wizards were included since the original BG1's 1998 release.)  At lowish levels (1-9), Wizards were more favorable.  They had fewer spell slots than Sors, but their day-by-day repertoire flexibility mattered more.  It was quite possible to accidentally be a one-trick pony or feel useless as a Sor if you didn't do the research on the best spells, and you had to seriously consider your character's end level since learning spells with saves could work well in BG1 but not in BG2 since foes there very reliably passed their saves.  With a level cap for Wizards and Sors of 9, Wizards also got access to level 5 spells in BG1 whereas Sors only got level 4 spells.  (Oddly, Wizards were able to cast level 6, 7, 8, and 9 spells starting at the same level as Sors, reducing their relative advantage, though that's a minor point since I've only rarely played tabletop characters in any D&D edition past level 6.  This was also primarily a Baldur's Gate II thing.)

Sorcerers had another advantage in the Baldur's Gate series:  They could learn any spell of any level they could cast.  Wizards got 0 free spells per level past character creation (with the exception of 'epic' or high-level abilities starting at 3 million EXP or level 18, just like Sors), forcing Wizards to find all scrolls of spells they could cast.

Now, I understand this experiment was done with a small sample size of just me in a fixed environment of Baldur's Gate II.  This game series focuses on exploration and dungeon crawling, not intrigue, camping, base building, nor certain other things that are available in the D&D system.  However, the BG series was intentionally made to feel as close as possible to a real tabletop campaign, and just like in these video games, tabletop has certain top-tier spells and abilities that seem like obvious picks to pick, or that are generally useful overall when you don't know what to expect.

D&D 3.x Experience
Baldur's Gate since the 90s influenced me heavily to be a Wizard in D&D 3.0 when I played in my first non-convention campaign.  I didn't like needing to call every shot (that is, prepare every spell) then or now, but I kept playing Wizards instead of Sors with rare exceptions because Wizards got higher level spells sooner, and because Wizards Abrupt Jaunt, bonus feats, and generally better class features than Sors.  Considering that various campaigns I played started at level 1 and ended at level 3 or 4 (and, again, rarely went beyond level 6), having higher spell levels and probably more spell slots overall was mechanically a great boon, but I eventually felt like I was forcing myself to play Wizards without enjoying them as much as I should have because I didn't like feeling penalized.  If the people on the forums thought prepared casters were more worth playing than spontaneous casters due to higher tiering and generally more praise, I should just listen, right?  I only learned later that forum logic often focused on the theoretical view of acting perfectly instead of focusing on my preferences.

Now, I understand why forum logic exists:  We need a common basis for evaluating and suggesting things.  It's also a forum, and we want to talk about something:  That's why we're here!  Dwelling on the possibility of power is often more appealing to theorycrafting and dreaming than the likely reality of what people face.  Being able to change what spells you have available to you every time you prepare spells (normally daily) can be a huge boon, but, at least in my experience, people don't play that way.  They pick a certain number of favorite spell preps on a near-daily basis (like grease and enlarge person every day on a Wizard) and might swap out a spell or two per spell level to do something different (like preparing color spray one day and magic missile the next).  In short, prepared casting focuses on what you can be while spontaneous casting tells you more of what you are.

In 3.x, I've played or thoroughly made Druids, Clerics, Wizards, Sorcerers, Psions, and other character classes.  Only after playing a Druid from level 1-11 did I realize the power of spontaneously-cast spells.  Being able to transform a spell slot into something that would likely be useful in most circumstances was wonderful, and I used it occasionally since I liked buffs.  Yes, Druids are spiffy for reasons in addition to their spells.  Likewise with Clerics.  Still, Wizards and Sors are primarily what spells they know and have available to cast at any moment, and a bad spell pick for today for a Wizard or for much longer for a Sor is painful.  Yes, in D&D 3.x and Pathfinder 1e and probably many other games, there are generally useful, generally optimal spells that people are normally assumed to get soon after they become available:  Research matters in-game as it does elsewhere.

3.x consumables are still useful if done properly, and, in short, my Red Hand of Doom experience varied greatly between a high-powered group that generally plowed through content with little difficulty and a purposefully low-powered group that barely made it through the module, having spent an uncomfortably high amount of in-game currency in my opinion on consumables.  (Perhaps 10,000G+ for the group.)

Conclusion
There's more than one way to play any edition of D&D, and I felt that spontaneous casting is the better, more fun, and more convenient way to handle casting.  WotC eventually implemented spontaneous casting of sorts as standard for every casting class in D&D 5e, at least in the 5e PHB.  (Prepared casters chose a certain number of spells/spell levels to prepare and cast them spontaneously.  Fully spontaneous casters remained fully spontaneous with limited numbers of spells known.)  Actively using forums and heeding their advice drastically influenced how I played D&D, and I sometimes listened to others more than myself even when it hurt for the sake of making or playing something better.

Disclaimers
Explaining Baldur's Gate mechanics or D&D 5e in great detail also aren't main the points of this thread beyond what I've already shared.

[Eldritch_Lord]

Apologies in advance for the wall o' text, but the topic of Vancian casting vs. alternate systems is one that's near and dear to my heart.

Considering that various campaigns I played started at level 1 and ended at level 3 or 4 (and, again, rarely went beyond level 6)
[...]
At lowish levels (1-9), Wizards were more favorable.  They had fewer spell slots than Sors, but their day-by-day repertoire flexibility mattered more.
[...]
Only after playing a Druid from level 1-11 did I realize the power of spontaneously-cast spells.

While you toss in the topic of level ranges as an aside, I think it's actually key to why you might feel spontaneous casting is better and more fun overall while the forum consensus is the reverse.

At low levels (1-5 or so), prepared is nearly strictly superior to spontaneous, because (as you noted) with only a handful of spells available spontaneous casters get pigeonholed pretty hard and getting access to the key boosts of 2nd- and 3rd-level spells a whole level ahead of time is huge.  Then at low-mid levels (5-10 or so) spontaneous casting suddenly starts to look a lot more appealing, because both prepared and spontaneous casters have "enough" spell diversity and spell slots to get by most days, the relative combat power boost of 4th-level spells over 3rds is nowhere near that of 3rds over 2nds, and you start having enough gold to spend on wands and scrolls to cover a lot of the utility gap.  For folks who mostly end campaigns in the low-mid levels, I can definitely see why a lot of people gravitate to spontaneous casters over time.

But the thing I think you're missing is that at the high-mid and high levels, things flip back to favoring prepared casting.  5th-level spells shine for their utility, not their combat power, so a Wizard 9 having a single 5th-level spell where a Sorcerer 9 has none isn't really a big deal (e.g. you can always get a scroll if your 9th-level party needs to teleport somewhere) but a Wizard 14 having a bunch of 5th-level utility slots while a Sorcerer 14 has just 3 5th-level spells known which are likely to be more combat-focused actually is a pretty huge deal.  Reliable divinations start to come online, so the ability to predict challenges gets much easier and the effectiveness gap between a spontaneous caster's spells known and a prepared caster's generic daily loadout narrows considerably.  High-level play starts to revolve around having layers of buffs and access to broad utility, something spontaneous casters have difficulty keeping up with because they can't just dedicate 1/3 of their daily slots to that and have the rest of them for combat--and consumables for mid- and high-level spells are increasingly expensive and aren't sustainable to use every day, so you can't just buy lots of scrolls and wands to cover gaps anymore.  And so on.

You're definitely right that there's an issue with "forum logic," but it's not that it's all handwavey theorycrafting, it's that people just repeat "wizards > sorcerers" as a shorthand because they're talking in generalities, without getting into the fact that the degree to which that's true varies by level range, type of campaign, and so on.

It's much like how the Tier system aggregates many different optimization levels and people who just rely on that and don't get into the nuances often argue it's all theorycrafting and doesn't match their experiences.  At newb optimization levels where new players don't have a good grasp on what spells are good, what kinds of challenges they'll face, and the like, wizards are bad and sorcerers are utter shit because a sorcerer's bad choices haunt them forever but a wizard can fix their issues with a bit of research.  At low optimization levels where blasty wizards and healy clerics rule the roost, sorcerers are great and wizards are meh because the party casters are only doing one or two things anyway so spamming spells is king.  At moderate optimization levels where players start fiddling with minionmancy and seriously using divinations, wizards and sorcerers are mostly on par because, again, there are only one or two tricks in play but the wizard can get a bit deeper and broader with each.  At high optimization levels, wizards leave sorcerers in the dust for obvious reasons.  Yet if a new player just looks at the Tier system that says wizards are T1 and sorcerers are T2 so obviously wizards are strictly superior, they'll miss that detail and not get why their 5th-level evoker is falling behind their buddy's 5th-level summoning-focused sorcerer.

I have a lot of personal experience with the prepared-vs.-spontaneous-at-various-levels issue, since nearly all of my campaigns run the full gamut from low to high levels instead of running into the (apparently common) phenomenon of dying out around mid levels; in the past decade, 80% of the campaigns I ran made it to 18th level or higher, and the lowest level any of them ended was 12th.  Several of my players started off always playing prepared casters because they were "stronger" or always playing spontaneous ones because they were "easier," but as they got a feel for the playstyles at different level ranges they've all gotten more comfortable mixing and matching based on what the starting level will be, what their concept is, and so on.

Dwelling on the possibility of power is often more appealing to theorycrafting and dreaming than the likely reality of what people face.  Being able to change what spells you have available to you every time you prepare spells (normally daily) can be a huge boon, but, at least in my experience, people don't play that way.  They pick a certain number of favorite spell preps on a near-daily basis (like grease and enlarge person every day on a Wizard) and might swap out a spell or two per spell level to do something different (like preparing color spray one day and magic missile the next).  In short, prepared casting focuses on what you can be while spontaneous casting tells you more of what you are.

This is a bit of a misunderstanding about the benefits of prepared casting.  Having a standardized spell loadout is totally fine and fairly common; a wizard who completely shuffles his spells every day is likely ditching effective spells for no reason, and the benefit of being able to change just one or two spells each day is overrated.  It's the ability to have one standard loadout for adventuring and a completely different one for downtime that really matters.

You mentioned Red Hand of Doom, which is a good case study because it starts and ends in the "wizards > sorcerers" range but plays through the "sorcerers are more fun than wizards" range.  If you know that you're heading into a Red Hand of Doom game, your party wizard and party sorcerer are probably going to have very similar spells available most of the time: there are lots of large groups of goblinoids so load up on AoE blasting, there are lots of dragons so pack energy resistance buffs and anti-flyer spells, and so on.  But where the sorcerer is stuck with those spells every day, the wizard can take a downtime day to completely swap those out for scrying to try to figure out where the Red Hand forces are, animate dead to get a bunch of hobgoblin skeletons as meat shields (er, lack-of-meat shields), sending to be able to update allies at long distance rather than flying back and forth all the time, and so on, and then go right back to a combat loadout the next day--to say nothing of scribing a scroll or two or crafting a wand to allow more divining or reanimation or whatever on combat days.

A single day of the wizard's downtime casting every in-game week or so can make an incredibly dramatic difference at the campaign level even if at the tactical level the wizard and sorcerer play identically and even if the sorcerer largely edges out the wizard on the firepower front.  And of course if the party successfully finishes Red Hand of Doom and the DM says "Congrats, you saved the country from Tiamat's armies of goblinoids and dragons, but now she's made a deal with Bel and there are armies of devils invading your world!", it's easy for the wizard to pivot from the fireballs and energy resistance buffs that are good against dragons to the lightning bolts and
anti-teleportation/-scrying wards that are needed against high-CR devils, while the sorcerer is stuck with what he already has.

There's more than one way to play any edition of D&D, and I felt that spontaneous casting is the better, more fun, and more convenient way to handle casting.

Don't get me wrong, spontaneous casting is generally more fun if you don't like the bookkeeping, and on the rare occasions I get to play instead of DM I, too, try to get some amount of spontaneous casting on my prepared casters so I always have fallback options when something unexpected comes up.  But I think by singing the praises of spontaneous casting over prepared you're swinging the pendulum too far in the other direction.

WotC eventually implemented spontaneous casting of sorts as standard for every casting class in D&D 5e, at least in the 5e PHB.  (Prepared casters chose a certain number of spells/spell levels to prepare and cast them spontaneously.  Fully spontaneous casters remained fully spontaneous with limited numbers of spells known.)

I see this comparison a lot--"5e went more spontaneous and everything got better!" and the like--but it's flawed for a few reasons.  First, it's no longer comparing prepared to spontaneous, it's comparing a best-of-both-worlds pseudo-Vancian setup to plain ol' spontaneous, so obviously that version ends up looking better.

Second, the change was necessary due to how casting in general changed in 5e.  If you're going to require people to heighten spells to be relevant instead of auto-scaling by CL, you obviously can't require people to prepare everything at the beginning of the day, just like how 3e psionics has condensed powers and augmentation because of the differences between power points and spell slots, and just making psions prepare powers or letting wizards use spell points would be a total non-starter.

And third, most of the comparative benefit of prepared casting goes away in an environment where spell selections are incredibly limited, good utility spells are heavily nerfed, removed, or nonfunctional, and all the downtime spells you care about can be cast as rituals anyway.  It's much like how a wizard is going to end up a lot closer to a sorcerer both power-wise and spell-selection-wise in a 3e core-only environment where the DM bans polymorphing, teleportation, and permanent minions and there's basically no downtime between adventuring days: if there's very little actual benefit to being able to swap out your spell loadout from day to day, then obviously picking that benefit over in-combat flexibility is a sucker's bet.


As a side note, 5e pseudo-Vancian is much like the 3e spirit shaman's spellcasting mechanic, and one would think that a class that's "druid, but with more flexible spellcasting" would be obviously better than the druid, yet we know that the druid ends up being far better because while its casting method is indeed better the advantage of Wild Shape and an animal companion over a grab-bag of spirit-related class features outweighs it in practice.  Likewise, while spontaneous casting is basically "all the good spells the wizard wants to prepare anyway, but with more flexibility" in a bunch of circumstances, you need to take the full context into account when determining which one is better on the whole.

[Power]

You're missing a few other differences between prepared and spontaneous casters.

• First and foremost, prepared spellcasters (in 3.X at least) have a 1 level advantage over spontaneous casters. Having access to a higher tier of spells than your spontaneous counterpart is massive.
• Second, prepared spellcasters should not prepare once for the day and be done with it. Prepared spellcasters are supposed to leave some of their spell slots open and then flexibly prepare them for the upcoming encounter's needs based on scouting. This kind of tailor-made solution is extremely powerful.
• Third, prepared casters have their strengths rise considerably with clever people who know how to take advantage of circumstantial spells and take a toolbox approach to solve challenges. If your player just gravitates towards "all the strongest spells" then prepared casters are often simply more nuisance than spontaneous casters since they are using the same spells every day only the prepared caster has less flexibility using them. However, there are a lot of potent spells that only prepared spellcasters tend to have because they are generally too niche even though they can be extremely valuable to have access to (most healing spells, for instance, but also spells that offer unusual movement types, allow you to interact with limited types of targets, or let you fish for certain forms of information).
• Fourth, there are ways for your prepared caster to end up casting (semi)-spontaneously. In Pathfinder a major example of this is the Universalist Wizard with an Arcane Bond prestiged into Magaambyan Arcanist with an Amulet of Magecraft as his arcane bond with a Planned Spontaneity feat on top, who has more than enough spontaneous spellcasting at his disposal even without getting into Extra Spontaneous Spell Mastery, Greater Spell Specialization, and Preferred Spell feats.

Baldur's Gate is also a game with a more limited spell selection than D&D proper where a lot of utility spells are simply gone or ineffectual, so it's no surprise that spontaneous casters feel better there. Especially since as a videogame the number of daily encounters is so high that having more spells per day makes a significant difference.

[Eldritch_Lord]

You're missing a few other differences between prepared and spontaneous casters.

Oh, definitely, but I figured two screens' worth of rambling was long enough.

• Second, prepared spellcasters should not prepare once for the day and be done with it. Prepared spellcasters are supposed to leave some of their spell slots open and then flexibly prepare them for the upcoming encounter's needs based on scouting. This kind of tailor-made solution is extremely powerful.

In theory, yes, but in practice there's a surprising number of long-time caster players who don't do this.  I've run into quite a few players who didn't know that was a thing you could do at all, and quite a few more who thought it was a wizard-only thing because divine casters have a fixed time to pray for spells each day and the rule about them being able to prep spells later in the day is under a different header in the Magic chapter than it is for wizards.

So even though that's technically a baseline advantage of prepared casters, I generally count that among the more advanced player tricks, like the tidbit about multiple long-duration Abjuration spells clustered together being easier to Search for.

• However, there are a lot of potent spells that only prepared spellcasters tend to have because they are generally too niche even though they can be extremely valuable to have access to (most healing spells, for instance, but also spells that offer unusual movement types, allow you to interact with limited types of targets, or let you fish for certain forms of information).

Indeed.  Prepared casting is absolutely core to D&D's idiosyncratic magic toolbox, and without it casters tend to regress toward the mean of preparing as many generically-useful spells as possible, making casters less interesting and making things more railroady on both sides of the screen.  The various proposals to make all casters spontaneous/replace Vancian with spell points/etc. that come up as a fad every so often all have, as a hidden design goal, effectively deleting speak with plants and hallucinatory terrain and similar spells because no spontaneous caster can afford to take them.

• Fourth, there are ways for your prepared caster to end up casting (semi)-spontaneously.

Yep, even just sticking to 3e wizards, there's the Spontaneous Divination and Spontaneous Summoning ACFs, the Alacritous Cogitation and Uncanny Forethought feats, and the Mage of the Arcane Order and Magelord PrCs for giving a ton of spontaneous casting for very little investment.  And one shouldn't forget that non-evil clerics can spontaneously cast all Sanctified spells (and even though BoVD doesn't say the same, most folks do the same for non-good clerics and Corrupt spells by analogy).

[Endarire]

everyone, thank you for your well-reasoned arguments.  Part of this post was discovering that i had to learn a bunch of stuff for myself, and that others' guidance can't guide me beyond a certain point.  (In short, if I'm playing a character, it's ultimately my character.)  Part of it was discovering that prepared vs. spontaneous is somewhat a matter of preference as well as a matter of power.  I also learned my tendency to prepare only 1 of each spell I prepared unless I had a strong reason to prepare more, whereas spontaneous casters just pulled spells from a pool and are encouraged to repeatedly cast the same spells on a daily or near-daily basis.

I have never played any caster who left spell slots open to prepare later.  I didn't realize it was a rule until someone showed it to me, and even then, I wasn't in a position to significantly capitalize upon it.  i suspect this is useful in the right contexts, but I've also played in games where we were expected to rush, rush, rush in-game for 2-3 years of real time straight.|

I know about various means to make prepped casters spontaneous, like Hathran, and have even used Hathran + an acorn of far travel in a separate Red Hand of Doom campaign to access this spontaneity, but felt like it didn't matter much when I got it around level 8.  The game ended around level 10, meaning there was only a small window when I could and did use it.

[Power]

everyone, thank you for your well-reasoned arguments.  Part of this post was discovering that i had to learn a bunch of stuff for myself, and that others' guidance can't guide me beyond a certain point.  (In short, if I'm playing a character, it's ultimately my character.)  Part of it was discovering that prepared vs. spontaneous is somewhat a matter of preference as well as a matter of power.  I also learned my tendency to prepare only 1 of each spell I prepared unless I had a strong reason to prepare more, whereas spontaneous casters just pulled spells from a pool and are encouraged to repeatedly cast the same spells on a daily or near-daily basis.

That's not a hard rule. Preparing the same spell more than once is more than fine. Sometimes it's exactly what's called for.

I have never played any caster who left spell slots open to prepare later.  I didn't realize it was a rule until someone showed it to me, and even then, I wasn't in a position to significantly capitalize upon it.  i suspect this is useful in the right contexts,

Useful is an understatement. You can prepare spells that invalidate projectile attacks, render enemies unable to close the distance to engage your party in melee, target their weakest save, ignore spell resistance, render them unable to cast without even giving them a saving throw, give your party energy resistance against the type of damage they will do, bypass unpleasant DR types or incorporeal defenses, target their vulnerabilities, conjure up some friends specially suited to the encounter, use various ridiculously niche effects that can demonstrate dramatic effects under the particular circumstances, etc. Turning your spell selection into a kit geared towards perfectly countering your opposition can completely crush an encounter. It gets even sillier if you can use prep time to pre-buff the party before an encounter. The power of not instantly engaging is severely underestimated.

And that's for combat. In a non-combat circumstance, you can just prepare and cast whatever will solve the issue at hand, unless there is imminent time pressure.

but I've also played in games where we were expected to rush, rush, rush in-game for 2-3 years of real time straight.

There are also ways to prepare faster than 15 minutes, whether in 3.5E or PF.

I know about various means to make prepped casters spontaneous, like Hathran, and have even used Hathran + an acorn of far travel in a separate Red Hand of Doom campaign to access this spontaneity, but felt like it didn't matter much when I got it around level 8.  The game ended around level 10, meaning there was only a small window when I could and did use it.

There are ways to cast spontaneously at lower levels too. And then there's the usual "Scribe Scroll -> Keep copies on hand of any circumstantial spells you might need" stunt that comes naturally to Wizards but can honestly be done by any prepared caster (especially divine ones).

[Endarire]

Do prepared casters generally do better in campaigns with significant amounts of downtime?  The 2-3 year campaign we played had us on a strict timer for about 2 years of real time, meaning taking even one day off to craft something (a Healing Belt for my Druid in this case) was a significant investment.  We also were rarely in a position to ambush foes or pre-buff.

[Eldritch_Lord]

Do prepared casters generally do better in campaigns with significant amounts of downtime?  The 2-3 year campaign we played had us on a strict timer for about 2 years of real time, meaning taking even one day off to craft something (a Healing Belt for my Druid in this case) was a significant investment.  We also were rarely in a position to ambush foes or pre-buff.

Better in the sense that the more downtime they get for spell scribing, magic item creation, divinations, and so forth, the better off they'll be, yes.  But it's not the case that a lack of downtime penalizes them or anything--they take the same amount of time to refresh spells each day that spontaneous casters do, after all, and prepared casters (or, more specifically, the stereotypical fighter/wizard/rogue/cleric party) are the baseline against which published adventures (which often have strict timers or other railroad-y elements preventing lots of downtime) are designed, so by definition they're not disadvantaged there.

Pretty much all forum advice assumes zero downtime, you'll notice.  If you look up any wizard/cleric/etc. handbooks, they'll generally either not cover item creation at all or mark it down for being situational because you can't assume a given DM will give you any downtime, in the same way that item-reliant classes like fighters are marked down because you can't assume a given DM will give you any particular items.

As far as ambushing and pre-buffing not being an option for you, I find it pretty unlikely that you could never fit in a round or two to pop some 1 minute/level or 10 minutes/level buffs when combat was expected, or that timetables were so tight that taking an extra 15 minutes a few times a day would completely screw you over.

When I've heard stories about similar parties, it's generally been the case that either (A) the party hasn't done a good job of scouting, ignored stealth, and so on and thus rushed into encounters when they could indeed have slowed down, made plans, and pre-buffed, or (B) the DM is a fairly railroad-y one who gives guards and other stationary foes unreasonably high Spot and Listen modifiers (and mobile/skirmishing/ambushing enemies unreasonably high Hide and Move Silently) to prevent scouting and avoiding encounters, conjures up not-previously-existent reinforcements and wandering monsters the moment the party tries to take a breather, is overly stingy with Knowledge check results and the like to prevent strategizing, and so forth--or the DM is fine, but is running a module where things always happen at the speed of plot and everything is supposed to be unnecessarily mysterious.

Which is not to say that either of those is necessarily the case for your group(s), just that one should be careful not to conflate "prepared casters are worse overall" with "prepared casters are hit hard by X, Y, and Z party-/DM-/campaign-dependent factor(s) and so appear worse in those circumstances."

[Power]

Do prepared casters generally do better in campaigns with significant amounts of downtime?

They don't need downtime, but they can certainly take advantage of downtime better than almost anyone. Divinations will generally destroy a campaign if a prepared caster spends their downtime preparing divinations into every slot to answer all kinds of questions about the campaign, dungeon crawls, future enemies, where those handy artifacts are, etc. Crafting feats are always useful if you have the time to craft. Stuff like wall of stone, wall of iron, fabricate, skill-boosting buffs, and really any spells you can use to reshape your surroundings, call up help, investigate things, and build stuff also tend to be of a lot of use during downtime.

The 2-3 year campaign we played had us on a strict timer for about 2 years of real time, meaning taking even one day off to craft something (a Healing Belt for my Druid in this case) was a significant investment.  We also were rarely in a position to ambush foes or pre-buff.

Get a Ring of Sustenance, carry tools and materials for a whole crafting workshop, use rope trick, set up your crafting lab in there, and squeeze out an extra 6-8 hours of unimpeded crafting while the rest of the party is mostly sleeping to craft shit. There are also ways to increase your crafting speed. There are also rules for crafting while adventuring (ie. making do on the road while being generally distracted) with significantly reduced progress, and you can also generally scribe scrolls and brew potions in 2 hours so you really don't need downtime to do those.

[Endarire]

Eldritch_Lord and Power, thankee for this discussion!

Everyone:  What other points do you have regarding spontaneous and prepared casting?

[Power]

Being 1 level behind in spell progression is really the big penalty that you get. Other than that, with a stronger body of spells known (Pathfinder decided to add extra spells known as a favored class bonus to spontaneous casters, drastically boosting their versatility), spontaneous casters tend to do better. The better spontaneous casters all use tricks that let them flexibly cast any spells outside their regular spells known list. With those tricks you tend to close much of the gap, but you still tend to run into the issue that you can't just spam utility spells like crazy on downtime days or when you have to fortify a defensive position and so forth. Depending on the trick though you can expand your spells known more for a limited duration of time instead of a specific number of uses, which means you can spam them as much as you want within the time limit.

Obviously arcane prepared casters tend to suffer a lot more when they don't have easy access to ways to expand their spellbooks, at which point they become slightly more awkward Sorcerers. This is even worse when the DM uses setting rules to limit the types of spells that can be found (although usually they do a Core-only restriction which is not painful since most of the best spells come from Core anyway). Usually though they can copy spells from other Wizards. PF also has the Blood Transcription spell letting you transcribe spells from any spellcaster, even if it's a divine caster, even if they only have it as a SLA, but if your GM is enforcing setting rules like that you probably won't have access to the spell. Spontaneous casters usually tend to do better in these settings when they can get more exotic spells as part of leveling while Wizards cannot. That said, prepared arcane casters also tend to rise in strength when there are multiple prepared arcane casters to teach each other their spells. Depending on the ruleset there may also be downtime rules for learning new spells for spontaneous casters. PF has these rules, which will cost you (Spell Level)²x7x100 gp total, assuming you never fail a day of progress, which is not as easy when you cannot take 10 on the two checks (DC 10+2x(Spell Level) Spellcraft and Knowledge (Arcane or Divine, depending on spellcasting type)) involved and failing by 5 or more means you subtract a day of progress. At least you don't auto-fail skillchecks on a natural 1.

Prepared divine casters generally have it best because they always have full access to their entire spell list. It gets crazier when they have methods to cast spells they haven't prepared spontaneously, since they have their entire spell list at their fingertips that way.

On a basic level in both PF and 3.5 you can go a long ways towards obtaining some flexibility with the right spells. Shadow Conjuration/Evocation/Etc spells tend to give a wide assortment of spells to both Wizards (where it is akin to a limited form of spontaneous casting) and Sorcerers (who can use it to cast all kinds of spells that are not on their spells known). To use the spells offensively you need to pump your spell DCs quite a bit though. Summon Monster / Summon Nature's Ally spells usually give you access to a wide assortment of SLAs and spells known through your minions. Most obviously Planar Binding / Planar Ally spells can get you just about anything. Simulacrum is also highly abusable. And Dominate Person might be able to net you an arcane caster as minion. Necromancy spells as well. Polymorph spells tend to be highly versatile. Obviously spells like Limited Wish, Wish, and Miracle can get you an extremely broad selection of magic at your fingertips. So if your Sorcerer happens to invest his spells known into Summon Monster spells, Shadow Evocation/Conjuration/etc (with some DC-boosting and save-penalizing stunts), Planar Binding, and Limited Wish later on, he's not going to feel like a very limited spellcaster.

In addition, while spontaneous casters tend to categorically avoid situational and utility spells, picking up one or two of them can make a solid difference in their performance on downtime days and the like. And just having Suggestion and Fabricate at your disposal will do a lot to help with non-combat encounters and downtime.

Creativity with the spells at your disposal is also valuable in determining how versatile you are as a spellcaster. Fireball might not be a good spell to cast normally, but if you ready a Fireball to strike an enemy once he begins casting, that's a different story. Dispel Magic can be used to not only strip buffs and counterspell spells, but also to temporarily disable magic items from functioning, which can result in all sorts of penalties (although you may wish to maximize it if you are doing this) when someone's staff, magic sword, magic armor, magic belt, cloak, flying broom, etc. suddenly stops being magical. There are usually ways to use spells that make you versatile beyond the norm, especially if you get in the habit of readying actions, casting spells on objects, and generally interacting with or modifying the environment (or bringing your own objects to create a favorable environment).

The right items change a lot in terms of capability as well. A Bard is usually worthless at construction-related tasks, but a Bard with a Lyre of Building is a very different story. Even non-magical items can contribute a lot to utility. Sometimes there are some good stunts to pull when combining them with spells, and sometimes the items are good on their own. Just showing up to a dungeon with a pickaxe, shovel, and grappling hook (where you made sure to tie knots in the grappling hook's ropes in order to make the climbing checks effortless) will do a lot to change how well your party can navigate a dungeon (Mage Hand also makes it possible to maneuver grappling hooks in tricky ways).

[Eldritch_Lord]

Obviously arcane prepared casters tend to suffer a lot more when they don't have easy access to ways to expand their spellbooks, at which point they become slightly more awkward Sorcerers. This is even worse when the DM uses setting rules to limit the types of spells that can be found (although usually they do a Core-only restriction which is not painful since most of the best spells come from Core anyway). Usually though they can copy spells from other Wizards.
[...]
That said, prepared arcane casters also tend to rise in strength when there are multiple prepared arcane casters to teach each other their spells.

There are two corollaries to this:

1) Prepared casters vary in effectiveness on a per-campaign basis much more than spontaneous ones do, so when playing a prepared caster it's beneficial to sit down with the DM beforehand and ask a bit about the campaign (and if playing with a new DM, about their DMing style).

Will there be plenty of of downtime for scroll scribing and spell copying, or will it be a rollercoaster from one plot point to another?  Are there wizards' guilds in major cities where one can find rare scrolls and swap spells with the guild members, or do wizards camp out in towers on the fringes of society and refuse to work together?  Are the whole wizard list and all the various wizard ACFs available, or are there limitations based on sourcebook (e.g. "The game is in Eberron, so no FR-specific spells allowed") or setting flavor (e.g. "Only members of the Tenebrous Academy can specialize in Enchantment or Illusion and take those ACFs").  And so on.

Even if the answers to those questions are "No downtime, no friendly wizards, no setting-specific spells, Final Destination" you can run a wizard just fine, but it's much better to know that stuff up front and plan for it than have it sneak up on you mid-game.  If you don't do that and do end up getting surprised with highly unfavorable circumstances, that can contribute to the "Wait, why would I ever want to play a wizard?" effect.

2) If it turns out downtime will be rare and copying spells will be difficult, there are bunches of ways to expand on the number of spells you can get for free on level-up.  The Generalist Wizardry ACF and the Collegiate Wizard and Aerenal Arcanist feats are the easiest and most well-known ways to add to your free spells, but there are more obscure ones out there (like the second affiliation benefit for The One and the Five in PHB2) as well.  So it's certainly the case that a low-downtime no-scribing campaign makes a wizard lose some comparative advantage over a sorcerer, but if you know that will be the case going into things you can easily compensate for it.

-----

Power and I have both hammered a lot on why prepared casters are actually better, but it's worth noting that there actually are some times when playing a spontaneous caster is (or at least can be) better.  When exactly those might be will, again, be very campaign-dependent, but the three major scenarios where that applies are thus:

1) If you're building a tightly-themed caster.  A generalist wizard trying to cover as many bases as possible is strictly superior to a generalist sorcerer trying to do the same, but if you specifically want to build "a pyromancer" (or "a shadow mage" or whatever other themed caster) it can be better to play a sorcerer than a wizard because having several dozen fire spells at your fingertips at all times can feel more like a master of all things incendiary than a wizard who has to pick and choose a bunch of different fire spells, even taking things like Uncanny Forethought into account.

And of course if the theme you pick is "tricky caster" or "spooky caster" then the beguiler and dread necromancer are even better choices for that than an Enchantment-/Illusion-focused or Necromancy-focused sorcerer.

2) If you're building a dual-progression caster.  When combining magic systems (or even two classes of the same type like Ultimate Magus), the two things to keep in mind if you want to get the best out of the build and not lose caster levels for nothing is (A) both sides using the same key ability score(s) so you can double-dip for DCs and bonus spells/PP/etc. and (B) trying to avoid redundancy so each side fills in gaps where the other side is weak rather than having lots of overlap (unless you're trying for a tightly-themed dual-progression caster, which is its own can of worms).

In such a scenario, even if e.g. Wizard/Cleric is theoretically stronger than Sorcerer/Cleric in terms of spell breadth and versatility, the latter may be better because (A) a cleric/sorcerer shares Cha in common between the two and so needs two high stats whereas a wizard/cleric would want to keep all three mental stats high (or at least Int and Wis high and Cha moderate), which would be difficult and/or expensive, and (B) a sorcerer/cleric has all the benefits of both a prepared caster (awesome downtime utility and grabbing niche spells) and a spontaneous one (lots of in-combat choices and the ability to spam spells) and can split their spells known and -prepared foci accordingly, while a wizard/cleric doubles up on the prepared caster strengths and can end up less flexible day-to-day and a bookkeeping nightmare in general.

3) If you have an unusually large party or one with an unusual composition.  Fighter/wizard/rogue/cleric (and its variations like warblade/wu jen/scout/favored soul) is the classic baseline party for a reason, and most advice about building full casters assumes you'll be the only caster of your magic system in the party and thus need to cover everything yourself, but that advice doesn't necessarily apply if that's not the case.

If you're playing an arcane caster in a three- or four-person party with other arcane casters (e.g. a duskblade as the fighter-alike and a spellthief as the rogue-alike), then not only do you not need to cover all the arcane bases, you don't want to cover all the arcane bases because then you and the other arcanists will be stepping on each others' toes a lot.  Much better in that case to realize that the duskblade and spellthief have the debuffing, stealth, and short-range blasting spells covered and go with a dread necromancer for minions or a bard for buffing, or something like that.

If you're playing the only caster of your type in a party of five or six PCs, you're not going to be stepping on another PC's toes directly but you are probably going to overlap with another caster.  A party of fighter/wizard/rogue/cleric/psion/druid is going to see a bunch of overlap between the wizard and psion and between the cleric and druid if the former both focus on blasting and summoning and the latter both focus on healing and utility...and a party of psychic warrior/seer/lurk/ardent/wilder doesn't have a single non-psionicist in it but really isn't going to benefit from the breadth of a cleric with all of the self-buffing, self-healing, and utility already on display.

In those cases, playing a spontaneous caster to focus your choices and clearly signal to the other players what role(s) you're going to fill is definitely a benefit.  Of course adding a wizard or cleric to those parties isn't ever going to be a bad thing, and I have actually run for a party full of focused casters where the wizard took the opportunity to focus on all those crazy niche spells you rarely see cast in an actual game, but it's a lot easier to stay in your lane and avoid scope creep as a spontaneous caster than as a prepared one.

[Power]

Honestly, the easiest way to get around the issue of a campaign with limited spell selection issue is to just play a Cleric or Druid much like you would a Wizard, especially if you use tricks to allow them to cast Wizard spells. In Pathfinder, you can just play a Shaman (using Wandering Hex for Arcane Enlightenment off of the Lore mystery) as a divine Wizard Plus with ease (build is a bit MAD though). But there are many ways to gain more spells as a Wizard too.

When it comes to dual-progression casters, honestly, a Cleric/Wizard can afford to dump Charisma all the way. As long as you have 18 in both casting stats you're pretty much fine. However, most dual-progression casters suffer a lot for being behind 3 whole levels, so it's not particularly recommended. You can do it, though, but typically you will use early entry tricks to make it worthwhile, using a spellcasting prestige class that has accelerated progression. In Pathfinder Mystic Theurge is a lot more desirable thanks to the Lesser Spell Synthesis and Extra Spell Synthesis feats letting you break action economy. Being Cleric/Sorcerer is generally worse than being Cleric/Wizard since now you lose an extra level of progression on both classes. Moreover, for dual-progression, you usually want to use methods to ensure both prepared spellcasters cast using the same ability score, if possible. The 3.5 Geomancer prestige class is essentially a better Mystic Theurge that lets you use the same ability score to set the DCs of both classes. In Pathfinder it is possible to combine Feyspeaker Druid and Seducer Witch to be charisma-based on both sides of the Mystic Theurge prestige class if you want. Most of the class feature losses and jank in both archetypes are irrelevant if you are only advancing the class itself for 3 levels anyway. That said, there are usually ways to cast off of the spell lists of multiple classes without resorting to Mystic Theurge, so the appeal is somewhat limited.

If your party has an unusual party composition, that is not so much a reason to play a spontaneous caster as it tends to be a reason to simply differentiate yourself from the norm. It is true that if your party already has a spellcaster covering utility roles, being a more specialized spontaneous caster will not cost your party much (although you still won't be much use outside your specialty). Also if your party has two Wizards, as I noted above, they can both copy out of each other's spellbooks which will usually boost their spell access significantly, especially in campaigns where your Wizard doesn't have easy access to other Wizards and scrolls to copy spells out of. Honestly an all-Wizard party is extremely powerful, even in the low levels, so it's not like you have to differentiate yourself by picking another class either. Also, it was mentioned that a psionic party doesn't need a Cleric because it has no need of buffs/healing/utility, but Clerics bring so much more to a party than just support spells. With the right domains, spell selection (Clerics get access to their entire class list as divine casters too, so they can help themselves to any fitting spell on their entire spell list whenever they take a break to prepare spells), and maybe some other tricks and prestige classes if you want to go further, Clerics can pretty much be Wizards in plate with better will saves and perception checks if they want to be.

The biggest reasons to play spontaneous casters are when you have methods to gain access to your entire class's spell list or you're deliberately specializing into a more limited array of spells and would prefer flexibility in how many times you cast them to diversity in how many spells you can cast.

[Eldritch_Lord]

When it comes to dual-progression casters, honestly, a Cleric/Wizard can afford to dump Charisma all the way. As long as you have 18 in both casting stats you're pretty much fine. However, most dual-progression casters suffer a lot for being behind 3 whole levels, so it's not particularly recommended. You can do it, though, but typically you will use early entry tricks to make it worthwhile, using advancing a spellcasting prestige class that has accelerated progression.
[...]
Moreover, for dual-progression, you usually want to use methods to ensure both prepared spellcasters cast using the same ability score, if possible.

In my experience, Wizard/Archivist is much more common than Wizard/Cleric, though that will of course vary by group.  Almost all of the [Arcanist]/Cleric builds I've seen have leaned heavily on Divine Metamagic (either metamagicking blasting/debuffing spells and using the arcane side for utility, or DMM: Persisting a bunch of buffs and using the arcane side for offense), so they wanted high Cha for that.  Not to mention that if Dragonlance feats are on the table, a Sorcerer/Cleric going Cha-SAD with Dynamic Priest has more synergy potential (diplomancy and such) than a Wizard/Cleric going Int-SAD with Academic Priest.

Also if your party has two Wizards, as I noted above, they can both copy out of each other's spellbooks which will usually boost their spell access significantly, especially in campaigns where your Wizard doesn't have easy access to other Wizards and scrolls to copy spells out of. Honestly an all-Wizard party is extremely powerful, even in the low levels, so it's not like you have to differentiate yourself by picking another class either.

I've run an all-wizard party before that got a ton of mileage out of free spell swapping, but that game started at level 7 or thereabouts so they could already use PrCs to differentiate their builds; I've never seen a multi-wizard party in a low-level-start and/or limited-scribing game, as using each other as the primary source of spells means lower-level wizards tend to be more same-y unless they have essentially opposed specializations (in which case spell swapping is of limited benefit).  If two players are fine with having basically the same capabilities, though, then it's definitely a good option.

Also, it was mentioned that a psionic party doesn't need a Cleric because it has no need of buffs/healing/utility, but Clerics bring so much more to a party than just support spells. With the right domains, spell selection (Clerics get access to their entire class list as divine casters too, so they can help themselves to any fitting spell on their entire spell list whenever they take a break to prepare spells), and maybe some other tricks and Prestige Classes if you want to go further, Clerics can pretty much be Wizards in plate with better will saves and perception checks if they want to be.

It's not that they can't benefit from a cleric at all (because, again, adding a wizard or cleric to a party is never a bad thing), just that adding one to a party of full and partial manifesters with plenty of existing self-buffing gishiness is going to be much less of a comparative benefit than adding it to a party with no divine casters or less magic in general.  In that scenario, one might want to go for a bit more gap-filling variety and try out something like a face bard or an illusion-focused sorcerer (or of course something further afield like a stealth-focused swordsage).

The biggest reasons to play spontaneous casters is when you have methods to gain access to your entire class's spell list or you're deliberately specializing into a more limited array of spells and would prefer flexibility in how many times you cast them to diversity in how many spells you can cast.

The former is definitely true for high-op games, but not necessarily in low- or mid-op ones.  Playing a spontaneous caster for party dynamics reasons can still leave you with a perfectly functional character, just like a wizard can get a long in those games without a lot of spontaneous tricks.

[Power]

Missed one of the more basic reasons to play spontaneous casters, and that's just having more spells per day. If your campaign forces you to do a lot of encounters per day, the added spells per day of spontaneous casters can make a considerable difference in effectiveness. That said, prepared casters usually end up with Pearls of Power to give them more spells at high levels.

In my experience, Wizard/Archivist is much more common than Wizard/Cleric, though that will of course vary by group.  Almost all of the [Arcanist]/Cleric builds I've seen have leaned heavily on Divine Metamagic (either metamagicking blasting/debuffing spells and using the arcane side for utility, or DMM: Persisting a bunch of buffs and using the arcane side for offense), so they wanted high Cha for that.  Not to mention that if Dragonlance feats are on the table, a Sorcerer/Cleric going Cha-SAD with Dynamic Priest has more synergy potential (diplomancy and such) than a Wizard/Cleric going Int-SAD with Academic Priest.

Interesting, although an Archivist can pretty much double as a Wizard by himself with enough domain spells and druid spells scribed into his spellbook. Sorcerer/Cleric though (in PF you can do something similar with an Empyreal bloodline Sorcerer which is wis-based) will put you behind an extra level of spell progression on both the arcane caster and the Cleric (because there is 1 more level of non-Cleric) when Mystic Theurge is already giving you massively delayed spell progression. I wouldn't really recommend it.

I've run an all-wizard party before that got a ton of mileage out of free spell swapping, but that game started at level 7 or thereabouts so they could already use PrCs to differentiate their builds; I've never seen a multi-wizard party in a low-level-start and/or limited-scribing game, as using each other as the primary source of spells means lower-level wizards tend to be more same-y unless they have essentially opposed specializations (in which case spell swapping is of limited benefit).  If two players are fine with having basically the same capabilities, though, then it's definitely a good option.

Well, if you have generalist Wizards it's pretty easy. You do get fewer spells per day on both Wizards, but you also have two Wizards who can probably share a number of resources between them. In Pathfinder Wizards can cast their banned schools (they just have to pay 2 spells to prepare them, and they can burn their 5th level bonus feat to remove even that penalty from 1 of their schools) anyway so it's effortless to go down this road.

It's not that they can't benefit from a cleric at all (because, again, adding a wizard or cleric to a party is never a bad thing), just that adding one to a party of full and partial manifesters with plenty of existing self-buffing gishiness is going to be much less of a comparative benefit than adding it to a party with no divine casters or less magic in general.  In that scenario, one might want to go for a bit more gap-filling variety and try out something like a face bard or an illusion-focused sorcerer (or of course something further afield like a stealth-focused swordsage).

I suppose, but I feel like the lack of comparative benefit is compensated for by simply not preparing spells the party no longer needs and preparing other spells that would be of good use to the party.

The former is definitely true for high-op games, but not necessarily in low- or mid-op ones.  Playing a spontaneous caster for party dynamics reasons can still leave you with a perfectly functional character, just like a wizard can get a long in those games without a lot of spontaneous tricks.

Honestly in PF doing things like wearing Mnemonic Vestment or playing a Half-Elf so you can Paragon Surge into temporary extra spells known feats are more like mid-op at best nowadays. It's pretty much considered part and parcel of the spontaneous spellcaster kit.

[Eldritch_Lord]

Honestly in PF doing things like wearing Mnemonic Vestment or playing a Half-Elf so you can Paragon Surge into temporary extra spells known feats are more like mid-op at best nowadays. It's pretty much considered part and parcel of the spontaneous spellcaster kit.

I get the impression, from the things you've mentioned in this thread, that PF has done a lot to blur the prepared/spontaneous divide; I haven't touched it myself since pretty early in its run.  With pure 3e, stuff like Mage of the Arcane Order or Ancestral Relic cheese are the best a sorcerer is going to get for faking a bit of wizardry, so I can see where the divergence in advice and play experience comes from there.

[Power]

It has. Some of it was intentional (extra spells known favored class bonus, bonus bloodline spells for Sorcerers, Mnemonic Vestment, Page of Spell Knowledge, Arcanist class including the Blood Arcanist archetype which is essentially a Sorcerer Plus, Razmiran Priest Sorcerer, Expanded Arcana as a buffed Extra Spell feat, Skalds with their Spell Kenning class feature, etc.) and some of it was accidental (Paragon Surge spell, Mongrel Mage Sorcerer which can take a new Arcane bloodline every day, putting Ampoule of False Blood on and off to convert into an Arcane bloodline anew for different spells known, etc.). There are also stunts for prepared casters to cast a lot more spontaneously than the norm. Generally speaking Pathfinder design has been a love song to spellcasters and general disdain for non-casters. Or to be exact, it is a design that believes in cool powers for the cool magical classes, and realism taxes as well as half-baked notions of balance and "this is the way these things are done" to keep mundane classes down.

In 3.5 stunts also include dipping two levels of Chameleon PrC so you can get a different Extra Spell feat each day (but you don't progress Sorcerer for these 2 levels - still a decent way to spend levels 19 and 20 though), using a rechargeable advanced psionic tattoo of psychic reformation, and equipping a Ring of Theurgy (consider using summons to fill it up with spells you want).

[Endarire]

Thankee!

PF has been kind to spontaneous casters.

My experience in a low-level all-Wizard group in core-ish 3.5 was that we generally found the best way to win was to go first and spam magic missile until we could learn fireball, then spam that.  These spells do little on their own, but having 5 Wizards do these every turn got us through the Temple of Elemental Evil!  (We also crafted items and cast other spells, but these two were our mainstays.)

[Power]

Generally Sleep and Color Spray are preferred at the lowest levels.

[Power]

On the topic of making a versatile Sorcerer with careful spells known selection, here is a 3.5 Core only Sorcerer list:

Cantrips: Detect Magic, Read Magic, Prestidigitation, Message, Mage Hand, Light, Detect Poison, Mending, Arcane Mark

• Grease, Tenser's Floating Disk, Charm Person, Feather Fall, Mage Armor
• Rope Trick, Glitterdust, Alter Self (can be replaced once you have Polymorph), Resist Energy, Invisibility
• Haste, Dispel Magic->Stinking Cloud (once you have Greater Dispel Magic), Magic Circle Against Evil, Tongues
• Shadow Conjuration, Polymorph, Dimensional Anchor, Scrying
• Shadow Evocation, Dominate Person, Teleport, Fabricate
• Greater Dispel Magic, Planar Binding, Create Undead
• Limited Wish, Greater Shadow Conjuration, Simulacrum
• Greater Shadow Evocation, Greater Planar Binding, Create Greater Undead
• Shades, Gate, Shapechange

Shadow Conjuration/Evocation is probably going to be a popular pick in combat, but fortunately you can pretty much always make those spells do good work in combat if you are creative enough (especially if you aren't playing Core-only). This Sorcerer is clearly going to have to invest heavily into getting high spell DCs, so you should probably play a Gnome (if doing Core races) and get Greater Spell Focus (illusion), maybe go for old age too. Get a Heighten Spell feat and Grease, Glitterdust, and Stinking Cloud will also continue doing good work into the high levels, giving you a few more options to ruin enemies depending on their saves (if Will is weakest save, shadow spells will do very good work).

This really isn't a perfect list, and there are a number of wands that might be worth nabbing too (like Silent Image and Unseen Servant). At any rate, a Sorcerer with a list like this can do all kinds of stuff depending on application of shadow spells, binding spells, dominated minions, undead minions, simulacra, and limited wish while also retaining a basic scry & teleport option and really just using planar binding or limited wish to get a plane shift or whatever when needed. Movement and flight can be acquired through polymorph and shadow conjuration (phantom steed).