Optimal exalting levels & times
What is this guide?
I've been self-debating the "best" exalt fodder level for a long while, and I decided to do the formal math for it. This guide provides a spreadsheet that helps you calculate your exalt leveling options, introduces you to some tools that help to gather the data, and includes some comments about the factors that modify your exalting times.
Quick Links
[x] Template spreadsheet copy
[x] Section 0: Introduction
[x] Section 1: Spreadsheets and data
[x] Section 2: About dragon supply
[x] Section 3: Measuring and improving your exalt times
__________________________________________________________________________
Section 0: Introduction
0.1: Levels per time spent
0.2: Profit per time spent
0.3: XP boost events
0.4: Why true timing is significantly better than oversimplified math estimates
0: Introduction
Before you begin, ask yourself: are you trying to maximize your profit per time spent or levels per time spent? Unless fodder is abnormally cheap, these goals are usually mutually exclusive.
Of course, if you want a balance of levels and profit per time spent, you should compare your time data to the appropriate values to make an informed decision. How much profit per time spent are you willing to sacrifice in favor of levels per time spent, and vice versa? That is situational and therefore up to you to decide.
0.1: Levels per time spent
If your goal is to maximize level output, disregarding profit, level 4 will usually provide the maximum levels per time spent. This is because the ramp-up time for building breath makes exalting at level 2 or 3 wasteful; if you've already invested time in the ramp-up battle, doing an extra battle for level 4 is much better than stopping at level 2 or level 3. Due to the way that experience thresholds work, regardless of who is training, level 4 always trounces level 5 in levels per time spent.
Level 4 is still better than level 5 in terms of levels per time spent during a 50% XP boost event, but you'll have to decide for yourself whether that maximum level output is worth throwing away "free" boosted XP that you would get towards level 5 instead.
See Section 0.3 for more comments about 50% XP boost events.
0.2: Profit per time spent
When considering profit per time spent, the main question is, which stopping point should you choose? The answer: It changes depending on your average time to level a dragon compared to the amount of Treasure that you paid for it, and you must also keep in mind how many dragons are supplied to you at that price.
For reference, average exalt payouts were taken from this thread. Level 1 exalts go between 2000 and 3000 Treasure, with an average of 2500 Treasure. For higher levels, the average is (Level * 1500 + (Level + 1) * 1500) / 2.
0.3: XP boost events
Drakeharvest, Frigidfin Expedition, Springswarm, Sunparched Prowl, and Warrior's Way all boost experience gains by 50% while active. This reduces the amount of experience required without affecting the average time spent per battle, meaning that more experience can be gained for the same number of battles. During an XP boost event, the profitable levels tend to climb upwards. (For example, if the most profitable level for a certain fodder price range is level 6 on normal days, it may be bumped up to level 7 during an XP boost event).
For an exact reference, you should record your times during an XP boost event to compare them with regular training times. It's not accurate to just reduce your regular training times by 50% for the lower levels (4-6), because this fails to take into account the effect that the XP boost has on the number of battles that need to occur for a level. Sometimes there's no difference at all, and sometimes there's an important difference. The difference in training time per level is not distributed evenly. Please read the next section (0.4) for more comments about real timing vs. oversimplified math estimates.
Check this post for a performance comparison between regular times and XP-boosted times. In it, I share some numbers taken during Warrior's Way 2021.
0.4: Why true timing is significantly better than oversimplified math estimates
You may be asking yourself the following: "I have the total XP required for each level and the average XP per battle, and I can time how long 1 battle takes me. Why can't I just apply some simple multiplication and addition to estimate my times instead of recording and averaging my real times?"
There are several reasons. The short answer is "Average battle timing is more complicated than you might think, and simplistic mathematical estimates eliminate too many significant variables."
The first (minor issue) is that not all packs have an equal chance of appearing, even though most guides will knowingly make that incorrect assumption for the sake of easier maintainability. So, your calculated "average XP per battle" will be wrong from the get-go. It's a tremendous amount of work to find the true average and requires tens-to-hundreds-of-thousands-of-battles per venue, so we don't bother.
Next, a major reason: The effects of RNG, even for the same type of issue, are not evenly distributed. Example 1: An Eliminate miss in one situation can cost only a second or so of delay, whereas a different miss could cost nearly a minute. Example 2: Being forced to reset on the battle right after you've just finished building breath causes far more total delay than needing to reset after successfully completing many battles in a row with previously-built breath. The context of an issue matters greatly in determining the effect, and wild swings in RNG can cause a wide rift between the minimum and maximum times to train dragons to a particular level.
For the lower levels (4-6), the difference in training time for the same level can be almost 2x as large, and at higher levels, the differences can add up to cause a delay of several minutes. But good times and bad times resulting from RNG are not evenly distributed; your "best" time and your "worst" time do not occur with the same frequency, because there are so many factors that affect the amount of delay caused by RNG.
For example, with a GLR Meditate build, I can train a pair to level 4 in 16 seconds per dragon with good RNG, but bad RNG can drive the time up to 38 seconds per dragon. My average time to level 4 with GLR Meditate is 20 seconds per dragon. Note how the average is closer to the "good" time than the "bad" time. Extremely bad RNG doesn't happen often, but when it does, it can cause very large delays. Since RNG variance is so large, to properly gain a reliable "average time per battle" for an estimated average, you would need to record a lot of trials (hundreds at least, but preferably thousands). At this point, your real average times will be much more accurate than any theoretical ones derived from oversimplified math.
Even in the absence of "bad RNG" such as misses or forced resets, the sequence of the packs that you get can subtly alter the training times due to different animation timings and so on. Your luck with the exact starting pack, random CAPTCHAs, and other small things will add up to affect your times. Also, depending on which level you're training to, needing to reset at certain points can have either a small or large effect on your total time.
Shifting topic slightly: another problem with oversimplified math is that it doesn't reflect the true number of battles needed to train a batch of dragons. The leftover XP at the end cannot be ignored if you want realistic results. This is especially true for the lower levels.
For example, a 50% XP boost event does not mean that you'll be training 50% faster for each level. For level 4 with a GLR Meditate build, I take an average of about 20 seconds per dragon with or without an XP boost event, because in each case, level 4 for that build/venue requires 2 battles. With a Mire Meditate build, there's a small improvement due to the existence of 2-packs in the Mire, from ~21 seconds unboosted to ~19 seconds; the 50% XP boost contributes about a 9.5% improvement in this case. For level 5, without a boost, I take an average of about 39 seconds to train each dragon with Mire Meditate; with a 50% XP boost event, I instead take about 30 seconds. That's a ~23% improvement in speed from a 50% boost in XP.
There are multiple factors that affect average training times unevenly between one level and the next, so using simplistic multipliers in calculations can lead to significantly incorrect projections. I hope that these examples help to illustrate the importance of recording your actual times instead of making inaccurate theoretical estimates.
__________________________________________________________________________
Section 1: Spreadsheets and data
1.1: Summary of popular levels
1.1.1: Per-profit comparisons
1.1.2: Visuals
1.2: Spreadsheets
1.1.2: Visuals
1.2.1: Spreadsheet section overview
1.2.2: Modifying the spreadsheet for your needs
1.2.3: Example personal spreadsheets
1.2.2: Modifying the spreadsheet for your needs
1.2.3: Example personal spreadsheets
1.1: Summary of popular levels
I am including some visual charts, graphs, and analysis of the profit per minute comparisons for popular training venues with standard builds. I chose some common training levels and prices.
Disclaimer: The data used is taken from my individual times, and these values will vary per user. While profit per minute trends tend to translate similarly, exceptional cases of latency and other issues may throw off the patterns slightly. (I have a modest internet speed - no fiber - and I train either on my laptop or smart phone, so my connection and devices aren't particularly special for speed.)
See section 1.2: Spreadsheets below if you're interested in recording your own data!
1.1.1: Per-profit comparisons
When determining the highest profit you can make within a certain amount of time, you need to take into account how much you paid for the dragon vs. how much you're getting paid for the dragon. General "best" levels by price per adult are listed below, with the caveat that it can vary slightly if your exalt times have a very different level : time ratio from mine.
I take a lot of speed tests for a lot of builds, because I pay particular attention to which level is fastest in which venue. Currently, Ghostlight Ruins is the fastest venue for level 4s, and there are a variety of builds that will yield similar results. The fastest build for level 5 and up (when comparing every build being used optimally) is currently the 119 STR / 64 QCK Meditate Mire build. Other breath stone variations in the Mire and Ghostlight Ruins are reasonably close together in times, so you can choose a different one based on your preference if you don't mind being a little slower. If you train thousands of dragons, the difference per dragon is quite significant, though. Click here to see time comparisons.
Estimates for average exalt payout
Subtract 1250t from your price for each gene and breed bonus and subtract 2500t for a color bonus.
* Remember that these are estimates based on my times. Your times might affect the exact values if you play significantly differently.
_____________ |
Adult price (T)
Free - 5054
5055 - 5875 5876 - 6986 6987 - 8274 8275 - 9474 | ___ |
Most profitable Lv. *
Lv. 4
Lv. 5 Lv. 6 Lv. 7 Lv. 8 |
At custom exalt payouts
The spreadsheet linked here will help you easily calculate adjustments to pricing. This accounts for raffles, public buy threads, and personal bonuses from buyers. Please read Section 1.2: Spreadsheets if interested.
1.1.2: Visuals
Solo Ghostlight Ruins Meditate 115 STR / 65 QCK
This is a Nature dragon with 115 STR, 65 QCK, 39 VIT, 8 DEF. It is equipped with 3 Berserker and 2 Ambush, and the Ability Stones of choice are Meditate, Shred, Eliminate, Rally, and Haste.
2.0 turn order KO thresholds and alternate stats / stones (Scratch, Meditate, Anticipate) that all perform very similarly can be found here. Check speed test comparisons here.
Solo Mire Meditate 120 STR / 64 QCK
This is an Arcane dragon with 120 STR, 64 QCK, 26 VIT, 7 DEF. It is equipped with 3 Berserker and 2 Ambush, and the Ability Stones of choice are Meditate, Shred, Eliminate, Rally, and Haste.
2.0 turn order KO thresholds and alternate stats / stones (Scratch, Meditate, Anticipate) that all perform very similarly can be found here. Check speed test comparisons here.
1.2: Spreadsheets
I made a template spreadsheet copy for anyone to use, if you want. The formula section does not need to be edited unless you want to customize the spreadsheet further. The training time recording section is where you will be inputting your data.
You want to record your average times on the spreadsheet and consider your dragon cost and supply in order to find the right sweet spot for you. You also should keep track of different venue times separately, as each zone has different enemies, configurations, and experience averages.
1.2.1: Spreadsheet section overview
The spreadsheet is made up of a few main parts:
- Training time recording ("Time Recording" tab), where you should record the times for tracked levels; these times will be used in the level sections. You do not need to record times for every single level, and you should leave columns completely blank if you forgot to record the corresponding time. For example, if you level a dragon to 7 but you record levels 5 and 7 while leaving 4 and 6 blank, that is fine. Recording each level is good too. If you want accurate averages, don’t intentionally skip recording just because you got “exceedingly bad luck”; bad luck is part of the Coliseum.
Fill the times in HH:MM:SS (Hours:Minutes:Seconds) format, making sure to enter a 0 for hours if the time is less than an hour, etc. You're going to have to watch your back here, especially if you're doing some weird copy/pasting, because Google Spreadsheets likes to do strange things with automatic Number formats, and deviating from HH:MM:SS format will break the calculations. For example, if you slip up and input incorrectly (as I've done a few times), the sheet may try to count 4:01 as 4 hours 1 minute instead of 4 minutes 1 second, which will mess up your averages. You won't get an error, so you won't notice unless you're paying attention to the values. - Average profit per minute ("Average Profit" tab), which contains formulae that show you the profit-time efficiency based on the cost of exalt fodder. All of the cells in this table will automatically reference and update according to the values in the "Average profit per dragon" table.
- Average profit per dragon ("Average Profit" tab), which contains formulae that lets you see how much Treasure you can make from dragons of specified costs. It automatically updates based on the payout chart and bonuses in the References tab. It's directly below the Average profit per minute table.
- The exalt payout chart ("References" tab). Since the profit per minute and profit per dragon charts are hooked up to recalculate automatically, you can easily edit the exalt payout chart in the References tab to use custom payout values (such as when there are buy threads). The bonuses box in the same References tab will automatically update the profit sections as well, if you just place an "x" by the relevant bonuses in the Include column (and remember to erase the ones you don't want).
1.2.2: Modifying the spreadsheet for your needs
Feel free to modify your copy of the spreadsheet for your needs; it is meant to track information that you personally care about. Parts that you may typically want to modify are not limited to, but may include:
- The average profit per dragon section: Note that Dragon cost column can be modified with no issues, and if you change your levels in the level section, these should update the column labels and references automatically. Don't confuse this section with Average profit per minute, which automatically updates based on Average profit per dragon!
1.2.3: Example personal spreadsheets
Here is a list of sample personal spreadsheets that I've recorded. You will see the zones and builds that I used listed in the titles.
- Mire Meditate
Tab 1: Time Recording
Solo dragon: 120 STR/64 QCK/26 VIT/7 DEF Arcane (Meditate) (check my turn order thresholds here) - Mire Scratch
Tab 1: Time Recording
Solo dragon: 120 STR/64 QCK/26 VIT/7 DEF Arcane (Scratch) (check my turn order thresholds here) - Ghostlight Ruins Meditate
Tab 1: Time Recording
Solo dragon: 115 STR/65 QCK/39 VIT/8 DEF Nature (Meditate) (check my turn order thresholds here) - Ghostlight Ruins Scratch
Tab 1: Time Recording
Solo dragon: 115 STR/65 QCK/39 VIT/8 DEF Nature (Scratch) (check my turn order thresholds here) - Ghostlight Ruins Anticipate
Tab 1: Time Recording
Solo dragon: 115 STR/59 QCK/45 VIT/13 DEF Nature (Anticipate) (check my turn order thresholds here)
---
Change log:
- Aug 14, 2021: Added some comments about why you should record your real times instead of making inaccurate estimates based on average XP and time per battle.
- Aug 12, 2021: Split out some information into more sections and added some details about levels per time spent vs. profit per time spent, as well as some comments about 50% XP boost events.
- Feb 14, 2019: Updated the graphs and linked data for Coli 2.0 values.
- April 8, 2017: Added a section that lists the most profitable level per price range for average exalt payouts, and a note about finding cheap fodder in Auction House 2.0.
- August 18, 2015: Added some charts and graphs in a new section on the first post entitled "Visual analysis of popular levels". They represent a visual version of the profit per minute data on my personal spreadsheets.
- May 7, 2015: Updated the measuring/improving exalting times section. There were some out-of-date comments, and I changed the format to hopefully improve reading convenience, since there's a lot of text.
- May 5, 2015: Moved guide information into separate posts. Updated personal Kelp Beds spreadsheet to use Duration format.
- May 4, 2015: Updated template spreadsheet: Moved average payout chart to References tab and added new tab with Duration format data entry. Updated personal Mire spreadsheet to use duration format and consolidate data onto the same tab.
- Nov 21, 2014: Updated template spreadsheet notes and highlighted safe-to-edit sections in green. (Apologies to the colorblind, there are still notes with instructions.)
- Oct 25, 2014: Updated the template spreadsheet to make the dragon cost cells in the "average profit per minute" table reference the dragon cost cells in the "average profit per dragon" table.
- Oct 19, 2014: Added a new sample data sheet to the guide for training two exalts at once in the Mire.
- Oct. 5, 2014: Updated the guide to include more detailed commentary about dragon supply.
- Oct. 5, 2014: Added "average profit per dragon" section to the spreadsheet. Modified cell referencing in "average profit per minute" and "average profit per dragon" to make minor changes to levels easier.
- Oct. 1, 2014: Removed mentions of hotkey software because Xhaztol has said that such software should not be used, as it has the *potential* to be used for botting.
- Aug. 28, 2014: Swapped out some data and information because augment stones are now mostly working.
- Aug. 19, 2014: Added level 4 and level 7 formulae to template spreadsheet.
- Aug. 19, 2014: Added average levels per minute to template spreadsheet.
- Jan. 3, 2024: Updated super old template spreadsheet to hopefully more useful and user-friendly.