3. The Journey There are four general stages to the quest for 99: 3.1. Questing and Character Development 3.2. Early XP Grind (through Level 93) 3.3. Levels 94-98 3.4. Levels 98-99 The subsections below address the general strategies for each stage, with detailed data around XP and number of runs for different targets included for the last two stages. Variants and special considerations for untwinked play are noted. A more abbreviated version of this can be found in this post from @Pb_pal. Note: The focus is on v1.13+ and therefore assumes availability of respecs and the higher drop rates for high runes. 3.1. Questing and Character Development This consists of building the character (i.e., finishing Normal and Nightmare difficulties), pulling together core gear for the build, and then some early leveling at key targets in Hell (Pits/AT/Pindle). For twinked play and with respecs available, this can be pretty straightforward: pick your highest level character for the class you want, respec, and equip, using your most efficient runner for relevant targets (MF runs or LK/Trav for runes) if missing key gear. For early leveling, hit Pits/AT/Pindle as necessary. Eldritch (superunique just north of first WP in A5) is also a nice target into the early-80s. It’s possible to go with Baal runs right away. But it’s better to focus on other targets while they give efficient XP and save the Ancients quest XP bonus for later in the journey when the XP penalties ratchet up. Spoiler: Details for Untwinked Play For untwinked play, there are three general approaches to this stage: 1. Moving through Normal and Nightmare as quickly as possible by completing only what’s absolutely necessary on p1, leveling on Pindle or another target as needed before moving to the next difficulty. MF is generally ignored until Hell, when the character reaches either an area (AT/Pits) or target (LK/Meph) that can be taken on low p-setting and help the character gear up. 2. Questing straight through to Hell LK/Meph/Pindle on p7/p8. The character builds up MF and improves gear along the way, without re-running areas or targets. This guide for untwinked tourney playthroughs, which is based on another guide by @jiansonz, gives useful pointers for this approach. The player has the choice of clearing as much or as little of each Act along the way. The more direct the playthrough, the more time is likely to be spent on running targets to gather gear (as described in #1 above) before moving into the next stage. 3. Doing MF runs along the way, such as NM Meph. Whether proceeding under option 1 or 2, the character builds up an MF rig and pauses at various targets in the leveling process to improve gear along the way. There are pros and cons to each approach, especially depending on the build. For example, option 1 is probably only a good idea for characters that can run Pindle on higher player settings with limited gear . So It’s an especially good option for sorcs (Blizzard is ideal) or trappers; melee builds may have more trouble. Similarly, some builds struggle to take down act bosses without good gear (e.g., necro summoners), so gathering gear while playing through or from area runs may be the best bet. But either way, there will usually be a stop in LK for runes and charms and some dedicated MF runs to pull the necessary gear together before fully transitioning to XP grinding, and the boundary between this stage and the next one is blurry for sure. Additionally, there may be some frustration trying to pull the core gear together due to randomness of drops. For reference, @Pb_pal and @NanoMist went with option 1, as seen here and here. My untwinked 99ers have tried all three routes, and there was always at least one point where the character felt underpowered and questing/running targets was a struggle at times. But, in the grand scheme of things, as @T72on1 noted in this post (which also includes a good overview of early leveling), this stage represents a negligible amount of time in the total journey. So pick the approach that sounds most fun. Finally, a quick word on how LK fits into untwinked character development: The most efficient approach is probably questing straight through to Trav to take out the Council and destroy the compelling orb (this makes the zealot packs in LK run away from the character). Then run LK as many times as needed to gather runes needed for key runewords (for non-sorcs, this usually means Enigma; for almost all characters, it also includes HoTo and/or CtA). Having these runewords will lead to faster leveling, and LK is about the number of runs (whether early or late in the process). That said, the number of LK runs needed for Enigma, on average, is 3855 as @Gripphon explained in more detail here (that post is wealth of other great information). And it may take more than that (or less), depending on variance. So players may find breaks from LK welcome. And they may luck out with a needed high rune while leveling at Pits/AT/Pindle, even if efficiency of those runs isn’t optimal due to missing key runewords. As examples, @NanoMist and @Pb_pal have taken the more direct approach to LK, as seen here and here (RIP Stevesy!). My 99ers mixed things up, with a Jah from the Pits speeding up Mercy’s process quite a bit, but with Hiatus ultimately returning to LK at level 95 to get CoH because CS was too challenging and the runes hadn’t dropped leveling in AT, and Phaedrus finally going back to LK at 93 to complete Enigma after slogging through the early 90s. 3.2. Early XP Grind Once the basic gear is gathered for the build, the focus turns to running targets that give good XP at high p-settings, generally as high as possible. But the balance of speed versus p-setting is important, as underpowered characters may gain XP faster on lower p-settings due to faster run times and less risk of death, which slows things down quite a bit (the XP penalties aren’t as harsh in 80s and even early 90s, but it’s still there, and dying simply slows down the average run time). It’s generally worthwhile to test out whether lower p-setting might be better, especially if p8 “feels” slow. Here are the most common targets: - Pindle is a great target into the mid-to-late 80s and even up to 90 or so. - Most builds, however, will do better to transition to A85 areas in the 80s, most commonly Pits or AT. The approach is to hit boss packs (boss + minions), because they give more XP (in absolute terms and because they are level eight-eight, while normal monsters are level 85, making the level-difference penalty less later in this stage). A good map has dense concentrations of boss packs, similar to a good MF map. Characters with cold damage will do well in AT, while Pits is generally the choice for others. As noted, p8 will give the best XP return in theory, but lower p-setting might work better. - Beginning with level 90 through level 93, the A85 areas are the general choice for leveling. Powerful characters may transition to CS at some point as well (see the next stage for more detail on these runs). - Others transition to Baal runs in earnest for this stage (and the next one). As confirmed by @Gripphon here, focusing on the waves gives better XP/min at this stage. It may be more efficient to end the runs with Ventar if Lister is a challenge. But, as noted above, other options are generally more efficient, if for no other reason than allowing the Ancients XP reward to be saved for later when the XP penalties are higher. Tips for efficient Baal runs are included in the next section. - Trav runs also give decent XP, with the added benefit of potential nice rune and jewelry drops. Spoiler: Details for Untwinked Play For untwinked, the only difference may involve LK runs, as noted in the prior section. Characters may also be missing key pieces of gear (e.g., Arachnid’s or +skill rings), so players may choose to mix in some MF runs to round out the kit. That said, the general advice is to focus on the XP runs because those items may well drop along the way and time spent on MF runs doesn’t move the XP bar. 3.3. Levels 94-98 At this point, there are generally two paths: CS/Diablo or Baal. They are discussed in detail below, with some notes on other options following that. Based on current knowledge, one type of CS run is generally agreed to be the most efficient option for the most popular classes at this stage. But it’s more accurate to say that there is consensus that players should test out which option is best for their particular character. This section concludes with detailed tables, created by @NanoMist and explained in this post, that show XP per run for various targets, and examples of how forum members have used the information captured in those tables to determine which approach to take. CS/Diablo: The general approach was to hit the seals at p8, taking out boss packs along the way, and then hitting Diablo at p7 (max experience from any monster is capped in the game code, and p8 exceeds that cap; so using p8 for Diablo only adds more HP). Some players just run on p7 to avoid switching p-settings over the course of thousands of runs. At p8/p7, you are looking at ~3300 runs for 94-98. @Corrupted and others skipped Diablo at earlier levels for faster XP/min. More recently, @Gripphon did extensive testing to show that a particular type of CS run (now called a Gripphon run) is the most efficient way to gain XP for classes that are the most popular for 99ers (Sorc and Paladin) and potentially the most efficient for others (e.g., Necro). In short, these runs involve using the tele-trick (see Tidbits and Pro Tips for an explanation) to reach the Pentagram in the middle of CS, then hitting only two seals (Vizier and Infector) and taking out boss packs along the way. Further testing has confirmed that hitting the seal for de Seis and then Diablo gives less XP/min for some builds than just focusing on the other two seals and boss packs. The best maps for Gripphon runs have boss packs clustered together near the seals so there is less time taken to stop between the seals. It’s difficult to calculate the number of Gripphon runs exactly (or p8/p7 runs, for that matter) because the number of boss packs (and monster types) hit in addition to the seals changes the XP gain/run. But the number of Gripphon runs needed is much higher than p8/p7; assuming no boss packs are hit (which is very unlikely), it would take ~6750 runs. That number drops significantly when the average XP/run from boss packs are factored in and, because the runs take less than half as long for certain builds, you can see why this can be a faster than p8/p7 runs. (A quick note: Gripp is one of the best runners in these forums and his characters were equipped for max killing power (i.e., no MF). So the actual numbers will be lower for most mortals . But the findings hold true.) A couple thoughts on running CS, whether p8/p7, Gripp runs, or hitting all three seals and skipping Diablo: (1) be sure to max fire resists (for character and merc), as Infector's pack is tricky; (2) +max fire resist and/or absorb equipment is helpful for the same reason; and (3) convicted Infector can be nasty, so a little bit of caution to confirm he *isn't* convicted can save you some deaths (less important if you are overstacked with +fire resist or have absorb per (1) and (2) above). Baal: The general approach for Baal runs is to find a good WSK map. Good means a WP right by the stairs, and then stairs down from WSK3 again very close. With teleport, this can often mean going through a wall to reach the stairs down quickly on at least one of the WSK levels. Running to and then clearing the Throneroom is on p1, then p8 for the waves, then p3 for Baal (the XP cap applies to him also, so going above p3 only adds to his HPs). Depending on the character’s strength, the p-setting may be decreased to speed up Lister’s wave (this should only ever apply to untwinked characters though). @Corrupted pointed out that hitting boss packs at p8 along the way, particularly for WSK 3, can be worth it for XP (I found that to be true also with Mercy, who had steady boss pack spawns on one of her WSK maps, while also netting some nice drops as a result). But it involves an extra change to p-setting. As @ffs noted, the map seed can impact how many monsters spawn in the Throne Room. (I haven't seen definitive testing to confirm this, but I've had the same experience.) Seeds with fewer monsters can speed things up noticeably, especially when considered across 1000s of runs. If you find lots of monsters spawning in the Throne Room regularly, it might be worth re-rolling, even if you have good WSK maps. There are a couple strategies for speeding things up: - First, the time between Baal spawning the waves slows the runs down a lot. Delerium’s ctc confuse can speed this up, as when it triggers Baal treats the wave as complete as long as the superunique in the wave has been taken out. Mileage may vary on this, as it involves gear tradeoffs. But some people swear by it. - Second, Lister’s immunities can slow runs down, especially for elemental builds. Although unbreakable Lister spawns can be “parked” (retreating outside the throne room and then moving back into the throne room so he doesn’t follow), it often will increase the XP/hour to bail immediately and start the next run. - Finally, as noted in the previous section, focusing on the waves will probably be more efficient at the earlier levels. The tradeoff is losing the drops from Baal. I’ve seen different takes on the inflection point where hitting Baal becomes efficient, and that is likely due to differences among builds and run times. Miscellaneous: The other A85 areas are also options, although they are generally less efficient. Several forum members have noted that characters that have done extensive p1 AT runs to make progress toward a grail will be level 96. As noted in this post, Mercy gained decent XP at level 97 from p8 pits with a good map that wasn’t too much worse than her XP from Baal runs but had better drops. So staying in Pits/AT past level 93 is definitely an option. With a good map, it might not be less efficient than Baal runs, allowing the character to save Ancients for level 98. This idea was kicked around in the 99 theorycrafting thread, and @zgpmf hit 98 on p8 AT runs with his HC Sorc, getting around 4M/hour for 97-98. XP Tables: @NanoMist created tables that allow players to determine the best path for them, based on their goals and playing style (e.g., max efficiency, MF goals, safety, intensity). They are embedded below and this post includes an explanation for how they can be used. The first table reflects p8/p7 CS runs and p8/p3 Baal runs, the second p1/p7 CS runs and p1/p3 Baal. Spoiler: XP Tables Table 1 Table 2 The basic idea is to time a run captured in this table, which allows an estimate of the amount of time needed to complete a level for that approach. Timing another type of run and figuring out how long it will take to complete a level using that approach allows a comparison to that baseline to see which approach is faster/slower. This can also be applied to a type of run not captured here – say, Masoleum runs. By timing those runs and XP gain, I could get a rough gauge about how it compares to all of the other options in this table. This can also be used to guestimate how a type of run would compare to another option. For example, it’s not possible to test Baal runs without burning the Ancients quest reward. Instead, I can figure out the time/level for whatever approach I’m using and compare that to an estimated time/run for Baal to see if it would be worth it to move to Baal now. Examples of how players have used the info in these tables are shown here, here, and here. The second one focuses on 98-99, but collectively they still give the basic idea. 3.4. Levels 98-99 The options here are Nihl, Diablo, and Baal, listed in descending order of efficiency for most characters. They also are in descending order of dangerousness and risk of death, which is particularly painful at level 98. That said, for many (except HC characters ), the increased efficiency of Nihl more than offsets the chance of dying. Put differently, even with a death or two, Nihl will generally be faster (assuming successful corpse recovery). As one example, @ffs theorycrafted time difference among the three targets here for a light sorc here. But here’s a decent general ballpark using the 2.5M/hr benchmark for Nihl, 90s/run for Diablo, and 2.5 min/run for Baal (which is very solid): - Nihl; 116 hours; 100 hours with Ancients quest saved - CS/Diablo: 140 hours; 120 hours with Ancients quest saved - Baal: 160 hours Nihl: As noted earlier, Nihl gives ~22k/run. With characters that take out some monsters around him, it is ~23k, and closer to 23.5k for builds that take out everything on his platform (e.g., trappers). The basic benchmark of XP/hour is 2.5M, with many characters exceeding it. @ffs has recently shown that 4M/hr is possible with a lightning sorc (and on a Mac that avoids S&E lag). A couple key points for Nihl runs: - There are fixed maps for Nihl, most importantly for the Halls of Pain as explained here. Map 3 is the shortest of the options and easiest to spot because the room with the WP has pillars. It’s a point of personal preference about Halls of Vaught for where Nihl’s platform is, or more likely a function of shrines. - Having an XP shrine active when Nihl drops increases XP gained by 50% for that run. It is possible to roll maps with up to 3 shrines from WP to Nihl. There are different views about the value of picking up shrines. @NanoMist had a pretty thorough breakdown here. The bottom line seems to be the time taken to check shrines roughly balances out to time saved by more XP, but that means fewer runs total and lower risk of dying. But finding a 3-shrine map that doesn’t take the character off a straight teleport line seems highly unusual. So 2-shrine maps that are on a direct line may be optimal – more testing is necessary to confirm. - Vipers are dangerous with their not-bugged physical damage from poison clouds. It’s explained in this post by @ioupainmax (and the post linked in it by @onderduiker). The danger from vipers can be minimized by always stepping back (or sideways) from the merc after teleporting to the landing spot hear Nihl’s platform. -integer damage reduction gear also helps. - Equipping merc with Delerium is key for the confuse proc to minimize ranged attackers (which include the viper clouds and witches with red stars of death). This isn’t needed if the character has a crowd control skill (trappers with CoS or Necros with Dim Vision). - In terms of tactics, this post covers some of the key issues: finding a good landing spot (which may involve rerolling maps), the side-step technique discussed above, dangerous rolls that should lead to immediate S&E (e.g., witches with might Nihl), and not-ever-teleporting to pick up items (risk of viper clouds is very high, especially if amp’ed). - This post by @Pb_pal covers or reinforces a couple other points: don’t be shy about S&E’ing and make sure to have a solid landing spot. It also notes that resists aren’t critical, except FR to account for Nihl’s corpse explosion. - Finally, Nature’s Peace is a key peace of equipment unless the character relies on corpse explosion (trapper or Necro). CS/Baal: The general approach is p1 seals and p7 for Diablo, targeting only the seal bosses and skipping boss packs unless they’re collateral damage. Some people prefer straight p7 to minimize changes to p-settings and better chances of drops, but de Seis will be more dangerous. Otherwise, the only difference for 98-99 is shrine hunting: You’re looking at ~5600 runs for the last level on p1/p7, not accounting for shrines. Baal: There isn’t anything different from earlier levels here. Only another >3600 runs .