@Romaki Glad to help! Wishing you many future profits. :)
Guides
Community created guides, helpful strategies, and more.
TOPIC | [guide] Optimal exalting levels & times
Update 2024 January 3: Made a new post here with new data for the latest Mire update, plus data for lv7 with each compared venue + build combination.
---
It's been a few years, and my motivation to play FR has honestly waned... but I've rounded out my GLR + Mire QCK turn order comparison here for the different breath stones. Using what I consider to be the optimal stats and playstyle for each build / venue, I took a total of several thousand speed tests for levels 4-6. (In a perfect world, I'd have more trials for the higher levels too, but that'll have to wait for some other day. Higher levels need more trials to stabilize the average.)
This is the followup post to my now-somewhat-outdated post comparing speed tests here. Moving forward, consider this new post to be more complete and correct than the old one.
A few important notes:
All right, so here are the builds I'll be comparing.
A few QCK options are reasonable, but I chose very specific QCK levels for reasons outlined in the QCK turn order comparison post.
Average time per dragon
This is a comparison of the average time taken per dragon. Take note that dragons are trained in pairs, so the total time taken for a batch is divided by 2 to get the average time per dragon.
Variance
RNG can do some pretty nasty things to training times of individual batches. If you're experiencing burnout and frustration from bad luck, there's definitely a good reason; sometimes 2x of the average training time can be wasted, and it may not even be your fault! Some builds play more consistently than others and have less annoying options of recovery, though if you're having a really awful RNG streak, every build suffers.
IMPORTANT: Variance is not evenly distributed! You cannot add min time + max time and divide by 2 for the average. Do not use the min/max times in this section to decide which build to use. I only include this section to show how disruptive RNG can be. The average time is a much more accurate picture of what you should expect out of a build, so please check the average time comparison section (above) when making a decision about builds.
Here are the minimum and maximum times that I have on record for each build. It's probably not the absolute min and max that I've ever experienced, but it's what I've got for this recorded dataset.
Conclusions
GLR Meditate has the best lv4 average time (20 seconds per dragon), with Mire Meditate, GLR Scratch, and GLR Anticipate tied at 21 seconds per dragon.
Mire Meditate starts to get better than all of the other builds starting with level 5 (average 39 seconds per dragon). This is a significant difference from the runner up, which is GLR Meditate (42 seconds per dragon). Mire Scratch, GLR Scratch, and GLR Anticipate builds are pretty close together and lag behind both Meditate builds slightly (44-45 seconds per dragon). The discrepancy increases as the level increases. When used properly, both Meditate builds tend to outperform the other breath stone builds.
When the level pushes past 6 or more, Anticipate starts to lag behind the two Scratch builds, because higher health fodder begins to interfere with building Anticipate breath. It's pretty even for levels 4 and 5, though. Note that Anticipate has some of the wildest variance when slammed with bad RNG (with the biggest pain point being not getting hit for breath), but it manages to stay competitive in levels 4 and 5 because the RNG can also go the other way (getting perfect packs and getting hit at good times for breath).
I consider all of the aforementioned to be good builds, and the Mire and Ghostlight Ruins are both good venues. You'll have to decide for yourself whether the differences are small enough for you to keep your favorite build or large enough to convince you to change to a new one. I can only provide the data for a comparison in a situation where each build was used optimally. If the builds are not used to their full potential, then mileage can vary quite wildly.
Personally, I will use GLR Meditate for level 4 and Mire Meditate for levels 5+.
---
It's been a few years, and my motivation to play FR has honestly waned... but I've rounded out my GLR + Mire QCK turn order comparison here for the different breath stones. Using what I consider to be the optimal stats and playstyle for each build / venue, I took a total of several thousand speed tests for levels 4-6. (In a perfect world, I'd have more trials for the higher levels too, but that'll have to wait for some other day. Higher levels need more trials to stabilize the average.)
This is the followup post to my now-somewhat-outdated post comparing speed tests here. Moving forward, consider this new post to be more complete and correct than the old one.
A few important notes:
- Devs have hinted that the Mire is the next venue that's slated for revamp. We don't know what effect it will have on training times yet; it will depend on what they do to change up the packs. When that time comes, Mire speed tests will need to be retaken. [EDIT]: As of October 22 2021, that time is now! Here's a link to the revamp post.
- These times are without XP boosts. The speed comparison still mostly applies for XP-boosted events, except all average times will be faster. (Make sure to time yourself if you’re chasing profit during XP events, though, because the XP boost has an uneven effect on the profitable levels per venue, especially at lower levels.)
- RNG is highly variable and takes its toll more and more the higher the level, meaning that higher levels also involve a wider range of time extremes. I did my best to take a lot of trials so that I could post this summary, but generally, more trials is always better to stabilize the average times (if you can stomach the recording of that many).
- For Meditate, when starting on normal packs, I use Meditate x2 to build breath. This is not a waste of turns and ends up saving turns on average in the end, even for level 4s, because it is extremely common to miss Eliminates. Yes, even within 2 battles for level 4s, you can miss an Eliminate. If you only Meditate once, you then have to deal with wasted Meditate Exhaustion debuff turns to rebuild breath after missing only one time, whereas if you Meditated twice, you only used 1 extra turn and you have a second Eliminate as buffer. I use a similar concept to farm bosses with Meditate if I'm not speed leveling at the time (though with bosses, I Meditate more times in a row depending on the phase of the battle). The point of multiple Meditates in a row is to reduce the amount of time wasted on Meditate Exhaustion based on average expected breath consumption (from misses or from having to hit a boss). I strongly disagree with a Meditate x1 regular startup strategy. Meditate x2 on startup has been more efficient for me when averaged over tens of thousands of level 4s.
All right, so here are the builds I'll be comparing.
- Mire Meditate: 119 STR / 64 QCK / 29 VIT / 9 DEF (Arcane) [stones] [speed tests]
- Mire Scratch: 119 STR / 64 QCK / 29 VIT / 9 DEF (Arcane) [stones] [speed tests]
- GLR 115 STR Meditate: 115 STR / 65 QCK / 39 VIT / 8 DEF (Nature) [stones] [speed tests]
- GLR 115 STR Scratch: 115 STR / 65 QCK / 39 VIT / 8 DEF (Nature) [stones] [speed tests]
- GLR 115 STR Anticipate: 115 STR / 59 QCK / 45 VIT / 13 DEF (Nature) [stones] [speed tests]
A few QCK options are reasonable, but I chose very specific QCK levels for reasons outlined in the QCK turn order comparison post.
Average time per dragon
This is a comparison of the average time taken per dragon. Take note that dragons are trained in pairs, so the total time taken for a batch is divided by 2 to get the average time per dragon.
|
Build Mire: 119 STR Meditate Mire: 119 STR Scratch GLR: 115 STR Meditate GLR: 115 STR Scratch GLR: 115 STR Anticipate |
___ |
Lv.4 Avg 0:00:21 0:00:23 0:00:20 0:00:21 0:00:21 |
__ |
# 240 644 682 656 652 |
____ |
Lv.5 Avg 0:00:39 0:00:44 0:00:42 0:00:45 0:00:44 |
__ |
# 232 452 430 458 400 |
__ |
Lv.6 Avg 0:01:04 0:01:13 0:01:08 0:01:12 0:01:14 |
__ |
# 228 176 168 156 200 |
Variance
RNG can do some pretty nasty things to training times of individual batches. If you're experiencing burnout and frustration from bad luck, there's definitely a good reason; sometimes 2x of the average training time can be wasted, and it may not even be your fault! Some builds play more consistently than others and have less annoying options of recovery, though if you're having a really awful RNG streak, every build suffers.
IMPORTANT: Variance is not evenly distributed! You cannot add min time + max time and divide by 2 for the average. Do not use the min/max times in this section to decide which build to use. I only include this section to show how disruptive RNG can be. The average time is a much more accurate picture of what you should expect out of a build, so please check the average time comparison section (above) when making a decision about builds.
Here are the minimum and maximum times that I have on record for each build. It's probably not the absolute min and max that I've ever experienced, but it's what I've got for this recorded dataset.
|
Build Mire: 119 STR Meditate Mire: 119 STR Scratch GLR: 115 STR Meditate GLR: 115 STR Scratch GLR: 115 STR Anticipate |
__ |
Lv.4 Min 0:00:17 0:00:15 0:00:16 0:00:15 0:00:14 |
__ |
Lv.4 Max 0:00:38 0:00:46 0:00:38 0:00:45 0:00:54 |
__ |
Lv.5 Min 0:00:32 0:00:32 0:00:32 0:00:31 0:00:31 |
__ |
Lv.5 Max 0:01:10 0:01:19 0:01:14 0:01:43 0:01:15 |
__ |
Lv.6 Min 0:00:54 0:00:54 0:00:55 0:00:55 0:00:55 |
__ |
Lv.6 Max 0:01:39 0:01:59 0:01:47 0:01:43 0:02:23 |
Conclusions
GLR Meditate has the best lv4 average time (20 seconds per dragon), with Mire Meditate, GLR Scratch, and GLR Anticipate tied at 21 seconds per dragon.
Mire Meditate starts to get better than all of the other builds starting with level 5 (average 39 seconds per dragon). This is a significant difference from the runner up, which is GLR Meditate (42 seconds per dragon). Mire Scratch, GLR Scratch, and GLR Anticipate builds are pretty close together and lag behind both Meditate builds slightly (44-45 seconds per dragon). The discrepancy increases as the level increases. When used properly, both Meditate builds tend to outperform the other breath stone builds.
When the level pushes past 6 or more, Anticipate starts to lag behind the two Scratch builds, because higher health fodder begins to interfere with building Anticipate breath. It's pretty even for levels 4 and 5, though. Note that Anticipate has some of the wildest variance when slammed with bad RNG (with the biggest pain point being not getting hit for breath), but it manages to stay competitive in levels 4 and 5 because the RNG can also go the other way (getting perfect packs and getting hit at good times for breath).
I consider all of the aforementioned to be good builds, and the Mire and Ghostlight Ruins are both good venues. You'll have to decide for yourself whether the differences are small enough for you to keep your favorite build or large enough to convince you to change to a new one. I can only provide the data for a comparison in a situation where each build was used optimally. If the builds are not used to their full potential, then mileage can vary quite wildly.
Personally, I will use GLR Meditate for level 4 and Mire Meditate for levels 5+.
Update 2024 January 3: Made a new post here with new data for the latest Mire update, plus data for lv7 with each compared venue + build combination.
---
It's been a few years, and my motivation to play FR has honestly waned... but I've rounded out my GLR + Mire QCK turn order comparison here for the different breath stones. Using what I consider to be the optimal stats and playstyle for each build / venue, I took a total of several thousand speed tests for levels 4-6. (In a perfect world, I'd have more trials for the higher levels too, but that'll have to wait for some other day. Higher levels need more trials to stabilize the average.)
This is the followup post to my now-somewhat-outdated post comparing speed tests here. Moving forward, consider this new post to be more complete and correct than the old one.
A few important notes:
All right, so here are the builds I'll be comparing.
A few QCK options are reasonable, but I chose very specific QCK levels for reasons outlined in the QCK turn order comparison post.
Average time per dragon
This is a comparison of the average time taken per dragon. Take note that dragons are trained in pairs, so the total time taken for a batch is divided by 2 to get the average time per dragon.
Variance
RNG can do some pretty nasty things to training times of individual batches. If you're experiencing burnout and frustration from bad luck, there's definitely a good reason; sometimes 2x of the average training time can be wasted, and it may not even be your fault! Some builds play more consistently than others and have less annoying options of recovery, though if you're having a really awful RNG streak, every build suffers.
IMPORTANT: Variance is not evenly distributed! You cannot add min time + max time and divide by 2 for the average. Do not use the min/max times in this section to decide which build to use. I only include this section to show how disruptive RNG can be. The average time is a much more accurate picture of what you should expect out of a build, so please check the average time comparison section (above) when making a decision about builds.
Here are the minimum and maximum times that I have on record for each build. It's probably not the absolute min and max that I've ever experienced, but it's what I've got for this recorded dataset.
Conclusions
GLR Meditate has the best lv4 average time (20 seconds per dragon), with Mire Meditate, GLR Scratch, and GLR Anticipate tied at 21 seconds per dragon.
Mire Meditate starts to get better than all of the other builds starting with level 5 (average 39 seconds per dragon). This is a significant difference from the runner up, which is GLR Meditate (42 seconds per dragon). Mire Scratch, GLR Scratch, and GLR Anticipate builds are pretty close together and lag behind both Meditate builds slightly (44-45 seconds per dragon). The discrepancy increases as the level increases. When used properly, both Meditate builds tend to outperform the other breath stone builds.
When the level pushes past 6 or more, Anticipate starts to lag behind the two Scratch builds, because higher health fodder begins to interfere with building Anticipate breath. It's pretty even for levels 4 and 5, though. Note that Anticipate has some of the wildest variance when slammed with bad RNG (with the biggest pain point being not getting hit for breath), but it manages to stay competitive in levels 4 and 5 because the RNG can also go the other way (getting perfect packs and getting hit at good times for breath).
I consider all of the aforementioned to be good builds, and the Mire and Ghostlight Ruins are both good venues. You'll have to decide for yourself whether the differences are small enough for you to keep your favorite build or large enough to convince you to change to a new one. I can only provide the data for a comparison in a situation where each build was used optimally. If the builds are not used to their full potential, then mileage can vary quite wildly.
Personally, I will use GLR Meditate for level 4 and Mire Meditate for levels 5+.
---
It's been a few years, and my motivation to play FR has honestly waned... but I've rounded out my GLR + Mire QCK turn order comparison here for the different breath stones. Using what I consider to be the optimal stats and playstyle for each build / venue, I took a total of several thousand speed tests for levels 4-6. (In a perfect world, I'd have more trials for the higher levels too, but that'll have to wait for some other day. Higher levels need more trials to stabilize the average.)
This is the followup post to my now-somewhat-outdated post comparing speed tests here. Moving forward, consider this new post to be more complete and correct than the old one.
A few important notes:
- Devs have hinted that the Mire is the next venue that's slated for revamp. We don't know what effect it will have on training times yet; it will depend on what they do to change up the packs. When that time comes, Mire speed tests will need to be retaken. [EDIT]: As of October 22 2021, that time is now! Here's a link to the revamp post.
- These times are without XP boosts. The speed comparison still mostly applies for XP-boosted events, except all average times will be faster. (Make sure to time yourself if you’re chasing profit during XP events, though, because the XP boost has an uneven effect on the profitable levels per venue, especially at lower levels.)
- RNG is highly variable and takes its toll more and more the higher the level, meaning that higher levels also involve a wider range of time extremes. I did my best to take a lot of trials so that I could post this summary, but generally, more trials is always better to stabilize the average times (if you can stomach the recording of that many).
- For Meditate, when starting on normal packs, I use Meditate x2 to build breath. This is not a waste of turns and ends up saving turns on average in the end, even for level 4s, because it is extremely common to miss Eliminates. Yes, even within 2 battles for level 4s, you can miss an Eliminate. If you only Meditate once, you then have to deal with wasted Meditate Exhaustion debuff turns to rebuild breath after missing only one time, whereas if you Meditated twice, you only used 1 extra turn and you have a second Eliminate as buffer. I use a similar concept to farm bosses with Meditate if I'm not speed leveling at the time (though with bosses, I Meditate more times in a row depending on the phase of the battle). The point of multiple Meditates in a row is to reduce the amount of time wasted on Meditate Exhaustion based on average expected breath consumption (from misses or from having to hit a boss). I strongly disagree with a Meditate x1 regular startup strategy. Meditate x2 on startup has been more efficient for me when averaged over tens of thousands of level 4s.
All right, so here are the builds I'll be comparing.
- Mire Meditate: 119 STR / 64 QCK / 29 VIT / 9 DEF (Arcane) [stones] [speed tests]
- Mire Scratch: 119 STR / 64 QCK / 29 VIT / 9 DEF (Arcane) [stones] [speed tests]
- GLR 115 STR Meditate: 115 STR / 65 QCK / 39 VIT / 8 DEF (Nature) [stones] [speed tests]
- GLR 115 STR Scratch: 115 STR / 65 QCK / 39 VIT / 8 DEF (Nature) [stones] [speed tests]
- GLR 115 STR Anticipate: 115 STR / 59 QCK / 45 VIT / 13 DEF (Nature) [stones] [speed tests]
A few QCK options are reasonable, but I chose very specific QCK levels for reasons outlined in the QCK turn order comparison post.
Average time per dragon
This is a comparison of the average time taken per dragon. Take note that dragons are trained in pairs, so the total time taken for a batch is divided by 2 to get the average time per dragon.
|
Build Mire: 119 STR Meditate Mire: 119 STR Scratch GLR: 115 STR Meditate GLR: 115 STR Scratch GLR: 115 STR Anticipate |
___ |
Lv.4 Avg 0:00:21 0:00:23 0:00:20 0:00:21 0:00:21 |
__ |
# 240 644 682 656 652 |
____ |
Lv.5 Avg 0:00:39 0:00:44 0:00:42 0:00:45 0:00:44 |
__ |
# 232 452 430 458 400 |
__ |
Lv.6 Avg 0:01:04 0:01:13 0:01:08 0:01:12 0:01:14 |
__ |
# 228 176 168 156 200 |
Variance
RNG can do some pretty nasty things to training times of individual batches. If you're experiencing burnout and frustration from bad luck, there's definitely a good reason; sometimes 2x of the average training time can be wasted, and it may not even be your fault! Some builds play more consistently than others and have less annoying options of recovery, though if you're having a really awful RNG streak, every build suffers.
IMPORTANT: Variance is not evenly distributed! You cannot add min time + max time and divide by 2 for the average. Do not use the min/max times in this section to decide which build to use. I only include this section to show how disruptive RNG can be. The average time is a much more accurate picture of what you should expect out of a build, so please check the average time comparison section (above) when making a decision about builds.
Here are the minimum and maximum times that I have on record for each build. It's probably not the absolute min and max that I've ever experienced, but it's what I've got for this recorded dataset.
|
Build Mire: 119 STR Meditate Mire: 119 STR Scratch GLR: 115 STR Meditate GLR: 115 STR Scratch GLR: 115 STR Anticipate |
__ |
Lv.4 Min 0:00:17 0:00:15 0:00:16 0:00:15 0:00:14 |
__ |
Lv.4 Max 0:00:38 0:00:46 0:00:38 0:00:45 0:00:54 |
__ |
Lv.5 Min 0:00:32 0:00:32 0:00:32 0:00:31 0:00:31 |
__ |
Lv.5 Max 0:01:10 0:01:19 0:01:14 0:01:43 0:01:15 |
__ |
Lv.6 Min 0:00:54 0:00:54 0:00:55 0:00:55 0:00:55 |
__ |
Lv.6 Max 0:01:39 0:01:59 0:01:47 0:01:43 0:02:23 |
Conclusions
GLR Meditate has the best lv4 average time (20 seconds per dragon), with Mire Meditate, GLR Scratch, and GLR Anticipate tied at 21 seconds per dragon.
Mire Meditate starts to get better than all of the other builds starting with level 5 (average 39 seconds per dragon). This is a significant difference from the runner up, which is GLR Meditate (42 seconds per dragon). Mire Scratch, GLR Scratch, and GLR Anticipate builds are pretty close together and lag behind both Meditate builds slightly (44-45 seconds per dragon). The discrepancy increases as the level increases. When used properly, both Meditate builds tend to outperform the other breath stone builds.
When the level pushes past 6 or more, Anticipate starts to lag behind the two Scratch builds, because higher health fodder begins to interfere with building Anticipate breath. It's pretty even for levels 4 and 5, though. Note that Anticipate has some of the wildest variance when slammed with bad RNG (with the biggest pain point being not getting hit for breath), but it manages to stay competitive in levels 4 and 5 because the RNG can also go the other way (getting perfect packs and getting hit at good times for breath).
I consider all of the aforementioned to be good builds, and the Mire and Ghostlight Ruins are both good venues. You'll have to decide for yourself whether the differences are small enough for you to keep your favorite build or large enough to convince you to change to a new one. I can only provide the data for a comparison in a situation where each build was used optimally. If the builds are not used to their full potential, then mileage can vary quite wildly.
Personally, I will use GLR Meditate for level 4 and Mire Meditate for levels 5+.
So, since Warrior's Way gives a 50% XP boost and I've been working on adding info about XP boost effects recently, I decided to also take some speed tests. It's not much because I didn't want to dedicate a lot of time to it, but I did get times for a little over 100 lv5-7 dragons and 30 lv8 ones. I also did one pair of lv25s for the heck of it.
General comments:
50% XP boost events like Drakeharvest, Frigidfin Expedition, Springswarm, Sunparched Prowl, and Warrior's Way do not necessarily translate to a 50% improvement in training time. It's quite a bit less for lower level fodder, and you'll only approach the 50% improvement as you creep towards level 25.
This discrepancy is due to a combination of factors, including, but not limited to:
Now, this is not to say that you should only train level 25s during 50% XP boost events. It's perfectly valid to use the XP boost to train to lower levels. Just don't make the incorrect assumption that lower level fodder will be trained 50% faster when you receive 50% more XP.
Lv. 4 to Lv.8 boosted average times:
[Speed test sheet]
I used my 119 STR / 64 QCK Mire Meditate dragon and recorded boosted times that can be compared to the non-boosted times for the same build + venue. Levels 4 and 5 had times for 108 dragons (54 trials), levels 5 and 6 had times for 106 dragons (53 trials), and level 8 had times for 30 dragons (15 trials).
Here are the average boosted times per level, with the caveat that there aren't technically enough trials to fully stabilize the average. Once the Mire is revamped as hinted, I'll do a re-do during an XP event with more trials, but for now, I didn't want to commit that much time to a venue that is going to change.
Lv.4 avg. time per dragon (regular): 21 seconds
Lv.5 avg. time per dragon (regular): 39 seconds
Lv.6 avg. time per dragon (regular): 1 minute, 4 seconds
Lv.7 avg. time per dragon (regular): 1 minute, 38 seconds
Lv.8 avg. time per dragon (regular): 2 minutes, 30 seconds
Lv.4 avg. time per dragon (boosted): 19 seconds (9.5% faster) *
Lv.5 avg. time per dragon (boosted): 30 seconds (23% faster)
Lv.6 avg. time per dragon (boosted): 46 seconds (28% faster)
Lv.7 avg. time per dragon (boosted): 1 minute, 10 seconds (28.6% faster)
Lv.8 avg. time per dragon (boosted): 1 minute, 43 seconds (31.33% faster)
* If training level 4s specifically in the Ghostlight Ruins, the level 4 average times are the same regardless of whether or not there is a 50% XP boost, because in the Ghostlight Ruins you always only need 2 battles to get your pair to level 4. The level 4 average times for the Mire are slightly different because sometimes you get 2-packs. (This is subject to change if the Ghostlight Ruins / Mire packs are changed in the future.) Note that with the 50% XP boost, the Mire gains slight superiority over the Ghostlight Ruins for training level 4s because battles are slightly shorter if you're lucky enough to get at least one 2-pack in the Mire, since the battle will complete a bit faster when there are less enemies; without the XP boost, the opposite is true because 2-packs in the Mire extend your required battles to 3 instead of 2.
Lv. 25:
[Speed test sheet]
I trained a pair from lv1 to lv25 during the Warrior's Way event, so they received the full 50% XP boost. I used my 119 STR / 64 QCK Mire Meditate dragon.
Sadly, I do not have a regular lv25 time with this build yet, but eventually I'll take one. I expect it to be closer to a 50% improvement than the lower levels, but probably still a little less than 50% to account for reset times and misses and things.
Anyway, I did a per-level "bad RNG" record, so I'll just post those instead for this boosted lv25 speed test. The pair took me a total of 2 hours, 18 minutes, 13 seconds (1 hour, 9 minutes, 7 seconds per lv25 dragon). Getting closer to the true average would require training more lv25s, though I'm not really motivated to do that since I'd make more profit per time spent on lower levels, so I'm calling 1 trial "good enough" even though I know it's not. Sorry!
RNG notes
L6: 1 miss Elim
L7: 1 miss Elim
L10: 1 miss Elim
L11: 1 low health reset
L12: 9 miss Elim, 1 KO reset
L13: 3 miss Elim
L14: 5 miss Elim
L15: 2 miss Elim
L16: 2 miss Elim, 1 Boss reset, 1 low health reset
L17: 5 miss Elim, 1 Boss reset
L18: 4 miss Elim, 2 Boss reset
L19: 3 miss Elim, 3 Boss reset
L20: 5 miss Elim, 1 Boss reset
L21: 5 miss Elim, 1 Boss reset, 1 low health reset
L22: 16 miss Elim, 1 Boss reset, 1 KO reset, 1 low health reset
L23: 11 miss Elim, 2 miss (Shroud Blind) Elim, 1 Boss reset, 2 low health reset
L24: 11 miss Elim, 1 Shock Paralysis, 1 KO reset, 2 low health reset
L25: 11 miss Elim, 2 Boss reset, 1 low health reset
---
One separate note. I won't spend too much text space on this here since my main goal for this post was to add a 50% XP boost comparison.
Over time, I've seen a lot of comments that gloss over the importance of recording true times to work towards an accurate average with an appropriate number of trials (hundreds / thousands of trials for each level). A lot of people seem to think that estimates based on "average time per battle" and "XP to n level" are good enough. However, such estimates can be many seconds or minutes off per level, and I think it's very important to debunk this line of thought, so I added a section to my guide about it. (See Section 0.4: Why true timing is significantly better than oversimplified math estimates.)
General comments:
50% XP boost events like Drakeharvest, Frigidfin Expedition, Springswarm, Sunparched Prowl, and Warrior's Way do not necessarily translate to a 50% improvement in training time. It's quite a bit less for lower level fodder, and you'll only approach the 50% improvement as you creep towards level 25.
This discrepancy is due to a combination of factors, including, but not limited to:
- When stopping at lower fodder levels, the breath building overhead takes up a larger % of the overall training time.
- Without an XP boost, sometimes a lot of XP is "wasted" because you get more than you need during the last battle. In some cases, the 50% boost helps only a little. This "waste" is less noticeable when training to high fodder levels.
Now, this is not to say that you should only train level 25s during 50% XP boost events. It's perfectly valid to use the XP boost to train to lower levels. Just don't make the incorrect assumption that lower level fodder will be trained 50% faster when you receive 50% more XP.
Lv. 4 to Lv.8 boosted average times:
[Speed test sheet]
I used my 119 STR / 64 QCK Mire Meditate dragon and recorded boosted times that can be compared to the non-boosted times for the same build + venue. Levels 4 and 5 had times for 108 dragons (54 trials), levels 5 and 6 had times for 106 dragons (53 trials), and level 8 had times for 30 dragons (15 trials).
Here are the average boosted times per level, with the caveat that there aren't technically enough trials to fully stabilize the average. Once the Mire is revamped as hinted, I'll do a re-do during an XP event with more trials, but for now, I didn't want to commit that much time to a venue that is going to change.
Lv.4 avg. time per dragon (regular): 21 seconds
Lv.5 avg. time per dragon (regular): 39 seconds
Lv.6 avg. time per dragon (regular): 1 minute, 4 seconds
Lv.7 avg. time per dragon (regular): 1 minute, 38 seconds
Lv.8 avg. time per dragon (regular): 2 minutes, 30 seconds
Lv.4 avg. time per dragon (boosted): 19 seconds (9.5% faster) *
Lv.5 avg. time per dragon (boosted): 30 seconds (23% faster)
Lv.6 avg. time per dragon (boosted): 46 seconds (28% faster)
Lv.7 avg. time per dragon (boosted): 1 minute, 10 seconds (28.6% faster)
Lv.8 avg. time per dragon (boosted): 1 minute, 43 seconds (31.33% faster)
* If training level 4s specifically in the Ghostlight Ruins, the level 4 average times are the same regardless of whether or not there is a 50% XP boost, because in the Ghostlight Ruins you always only need 2 battles to get your pair to level 4. The level 4 average times for the Mire are slightly different because sometimes you get 2-packs. (This is subject to change if the Ghostlight Ruins / Mire packs are changed in the future.) Note that with the 50% XP boost, the Mire gains slight superiority over the Ghostlight Ruins for training level 4s because battles are slightly shorter if you're lucky enough to get at least one 2-pack in the Mire, since the battle will complete a bit faster when there are less enemies; without the XP boost, the opposite is true because 2-packs in the Mire extend your required battles to 3 instead of 2.
Lv. 25:
[Speed test sheet]
I trained a pair from lv1 to lv25 during the Warrior's Way event, so they received the full 50% XP boost. I used my 119 STR / 64 QCK Mire Meditate dragon.
Sadly, I do not have a regular lv25 time with this build yet, but eventually I'll take one. I expect it to be closer to a 50% improvement than the lower levels, but probably still a little less than 50% to account for reset times and misses and things.
Anyway, I did a per-level "bad RNG" record, so I'll just post those instead for this boosted lv25 speed test. The pair took me a total of 2 hours, 18 minutes, 13 seconds (1 hour, 9 minutes, 7 seconds per lv25 dragon). Getting closer to the true average would require training more lv25s, though I'm not really motivated to do that since I'd make more profit per time spent on lower levels, so I'm calling 1 trial "good enough" even though I know it's not. Sorry!
|
Lv. 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 | _ |
Avg. Time (per dragon) 0:00:20 0:00:31 0:00:47 0:01:13 0:01:44 0:02:23 0:03:16 0:04:43 0:07:21 0:09:09 0:11:21 0:13:48 0:16:57 0:20:26 0:24:38 0:29:14 0:34:00 0:39:20 0:46:09 0:53:32 1:01:25 1:09:07 | _ |
Avg. Time (per pair) 0:00:39 0:01:01 0:01:34 0:02:26 0:03:28 0:04:45 0:06:31 0:09:26 0:14:42 0:18:17 0:22:42 0:27:35 0:33:53 0:40:51 0:49:16 0:58:28 1:08:00 1:18:40 1:32:17 1:47:04 2:02:50 2:18:13 |
RNG notes
L6: 1 miss Elim
L7: 1 miss Elim
L10: 1 miss Elim
L11: 1 low health reset
L12: 9 miss Elim, 1 KO reset
L13: 3 miss Elim
L14: 5 miss Elim
L15: 2 miss Elim
L16: 2 miss Elim, 1 Boss reset, 1 low health reset
L17: 5 miss Elim, 1 Boss reset
L18: 4 miss Elim, 2 Boss reset
L19: 3 miss Elim, 3 Boss reset
L20: 5 miss Elim, 1 Boss reset
L21: 5 miss Elim, 1 Boss reset, 1 low health reset
L22: 16 miss Elim, 1 Boss reset, 1 KO reset, 1 low health reset
L23: 11 miss Elim, 2 miss (Shroud Blind) Elim, 1 Boss reset, 2 low health reset
L24: 11 miss Elim, 1 Shock Paralysis, 1 KO reset, 2 low health reset
L25: 11 miss Elim, 2 Boss reset, 1 low health reset
---
One separate note. I won't spend too much text space on this here since my main goal for this post was to add a 50% XP boost comparison.
Over time, I've seen a lot of comments that gloss over the importance of recording true times to work towards an accurate average with an appropriate number of trials (hundreds / thousands of trials for each level). A lot of people seem to think that estimates based on "average time per battle" and "XP to n level" are good enough. However, such estimates can be many seconds or minutes off per level, and I think it's very important to debunk this line of thought, so I added a section to my guide about it. (See Section 0.4: Why true timing is significantly better than oversimplified math estimates.)
So, since Warrior's Way gives a 50% XP boost and I've been working on adding info about XP boost effects recently, I decided to also take some speed tests. It's not much because I didn't want to dedicate a lot of time to it, but I did get times for a little over 100 lv5-7 dragons and 30 lv8 ones. I also did one pair of lv25s for the heck of it.
General comments:
50% XP boost events like Drakeharvest, Frigidfin Expedition, Springswarm, Sunparched Prowl, and Warrior's Way do not necessarily translate to a 50% improvement in training time. It's quite a bit less for lower level fodder, and you'll only approach the 50% improvement as you creep towards level 25.
This discrepancy is due to a combination of factors, including, but not limited to:
Now, this is not to say that you should only train level 25s during 50% XP boost events. It's perfectly valid to use the XP boost to train to lower levels. Just don't make the incorrect assumption that lower level fodder will be trained 50% faster when you receive 50% more XP.
Lv. 4 to Lv.8 boosted average times:
[Speed test sheet]
I used my 119 STR / 64 QCK Mire Meditate dragon and recorded boosted times that can be compared to the non-boosted times for the same build + venue. Levels 4 and 5 had times for 108 dragons (54 trials), levels 5 and 6 had times for 106 dragons (53 trials), and level 8 had times for 30 dragons (15 trials).
Here are the average boosted times per level, with the caveat that there aren't technically enough trials to fully stabilize the average. Once the Mire is revamped as hinted, I'll do a re-do during an XP event with more trials, but for now, I didn't want to commit that much time to a venue that is going to change.
Lv.4 avg. time per dragon (regular): 21 seconds
Lv.5 avg. time per dragon (regular): 39 seconds
Lv.6 avg. time per dragon (regular): 1 minute, 4 seconds
Lv.7 avg. time per dragon (regular): 1 minute, 38 seconds
Lv.8 avg. time per dragon (regular): 2 minutes, 30 seconds
Lv.4 avg. time per dragon (boosted): 19 seconds (9.5% faster) *
Lv.5 avg. time per dragon (boosted): 30 seconds (23% faster)
Lv.6 avg. time per dragon (boosted): 46 seconds (28% faster)
Lv.7 avg. time per dragon (boosted): 1 minute, 10 seconds (28.6% faster)
Lv.8 avg. time per dragon (boosted): 1 minute, 43 seconds (31.33% faster)
* If training level 4s specifically in the Ghostlight Ruins, the level 4 average times are the same regardless of whether or not there is a 50% XP boost, because in the Ghostlight Ruins you always only need 2 battles to get your pair to level 4. The level 4 average times for the Mire are slightly different because sometimes you get 2-packs. (This is subject to change if the Ghostlight Ruins / Mire packs are changed in the future.) Note that with the 50% XP boost, the Mire gains slight superiority over the Ghostlight Ruins for training level 4s because battles are slightly shorter if you're lucky enough to get at least one 2-pack in the Mire, since the battle will complete a bit faster when there are less enemies; without the XP boost, the opposite is true because 2-packs in the Mire extend your required battles to 3 instead of 2.
Lv. 25:
[Speed test sheet]
I trained a pair from lv1 to lv25 during the Warrior's Way event, so they received the full 50% XP boost. I used my 119 STR / 64 QCK Mire Meditate dragon.
Sadly, I do not have a regular lv25 time with this build yet, but eventually I'll take one. I expect it to be closer to a 50% improvement than the lower levels, but probably still a little less than 50% to account for reset times and misses and things.
Anyway, I did a per-level "bad RNG" record, so I'll just post those instead for this boosted lv25 speed test. The pair took me a total of 2 hours, 18 minutes, 13 seconds (1 hour, 9 minutes, 7 seconds per lv25 dragon). Getting closer to the true average would require training more lv25s, though I'm not really motivated to do that since I'd make more profit per time spent on lower levels, so I'm calling 1 trial "good enough" even though I know it's not. Sorry!
RNG notes
L6: 1 miss Elim
L7: 1 miss Elim
L10: 1 miss Elim
L11: 1 low health reset
L12: 9 miss Elim, 1 KO reset
L13: 3 miss Elim
L14: 5 miss Elim
L15: 2 miss Elim
L16: 2 miss Elim, 1 Boss reset, 1 low health reset
L17: 5 miss Elim, 1 Boss reset
L18: 4 miss Elim, 2 Boss reset
L19: 3 miss Elim, 3 Boss reset
L20: 5 miss Elim, 1 Boss reset
L21: 5 miss Elim, 1 Boss reset, 1 low health reset
L22: 16 miss Elim, 1 Boss reset, 1 KO reset, 1 low health reset
L23: 11 miss Elim, 2 miss (Shroud Blind) Elim, 1 Boss reset, 2 low health reset
L24: 11 miss Elim, 1 Shock Paralysis, 1 KO reset, 2 low health reset
L25: 11 miss Elim, 2 Boss reset, 1 low health reset
---
One separate note. I won't spend too much text space on this here since my main goal for this post was to add a 50% XP boost comparison.
Over time, I've seen a lot of comments that gloss over the importance of recording true times to work towards an accurate average with an appropriate number of trials (hundreds / thousands of trials for each level). A lot of people seem to think that estimates based on "average time per battle" and "XP to n level" are good enough. However, such estimates can be many seconds or minutes off per level, and I think it's very important to debunk this line of thought, so I added a section to my guide about it. (See Section 0.4: Why true timing is significantly better than oversimplified math estimates.)
General comments:
50% XP boost events like Drakeharvest, Frigidfin Expedition, Springswarm, Sunparched Prowl, and Warrior's Way do not necessarily translate to a 50% improvement in training time. It's quite a bit less for lower level fodder, and you'll only approach the 50% improvement as you creep towards level 25.
This discrepancy is due to a combination of factors, including, but not limited to:
- When stopping at lower fodder levels, the breath building overhead takes up a larger % of the overall training time.
- Without an XP boost, sometimes a lot of XP is "wasted" because you get more than you need during the last battle. In some cases, the 50% boost helps only a little. This "waste" is less noticeable when training to high fodder levels.
Now, this is not to say that you should only train level 25s during 50% XP boost events. It's perfectly valid to use the XP boost to train to lower levels. Just don't make the incorrect assumption that lower level fodder will be trained 50% faster when you receive 50% more XP.
Lv. 4 to Lv.8 boosted average times:
[Speed test sheet]
I used my 119 STR / 64 QCK Mire Meditate dragon and recorded boosted times that can be compared to the non-boosted times for the same build + venue. Levels 4 and 5 had times for 108 dragons (54 trials), levels 5 and 6 had times for 106 dragons (53 trials), and level 8 had times for 30 dragons (15 trials).
Here are the average boosted times per level, with the caveat that there aren't technically enough trials to fully stabilize the average. Once the Mire is revamped as hinted, I'll do a re-do during an XP event with more trials, but for now, I didn't want to commit that much time to a venue that is going to change.
Lv.4 avg. time per dragon (regular): 21 seconds
Lv.5 avg. time per dragon (regular): 39 seconds
Lv.6 avg. time per dragon (regular): 1 minute, 4 seconds
Lv.7 avg. time per dragon (regular): 1 minute, 38 seconds
Lv.8 avg. time per dragon (regular): 2 minutes, 30 seconds
Lv.4 avg. time per dragon (boosted): 19 seconds (9.5% faster) *
Lv.5 avg. time per dragon (boosted): 30 seconds (23% faster)
Lv.6 avg. time per dragon (boosted): 46 seconds (28% faster)
Lv.7 avg. time per dragon (boosted): 1 minute, 10 seconds (28.6% faster)
Lv.8 avg. time per dragon (boosted): 1 minute, 43 seconds (31.33% faster)
* If training level 4s specifically in the Ghostlight Ruins, the level 4 average times are the same regardless of whether or not there is a 50% XP boost, because in the Ghostlight Ruins you always only need 2 battles to get your pair to level 4. The level 4 average times for the Mire are slightly different because sometimes you get 2-packs. (This is subject to change if the Ghostlight Ruins / Mire packs are changed in the future.) Note that with the 50% XP boost, the Mire gains slight superiority over the Ghostlight Ruins for training level 4s because battles are slightly shorter if you're lucky enough to get at least one 2-pack in the Mire, since the battle will complete a bit faster when there are less enemies; without the XP boost, the opposite is true because 2-packs in the Mire extend your required battles to 3 instead of 2.
Lv. 25:
[Speed test sheet]
I trained a pair from lv1 to lv25 during the Warrior's Way event, so they received the full 50% XP boost. I used my 119 STR / 64 QCK Mire Meditate dragon.
Sadly, I do not have a regular lv25 time with this build yet, but eventually I'll take one. I expect it to be closer to a 50% improvement than the lower levels, but probably still a little less than 50% to account for reset times and misses and things.
Anyway, I did a per-level "bad RNG" record, so I'll just post those instead for this boosted lv25 speed test. The pair took me a total of 2 hours, 18 minutes, 13 seconds (1 hour, 9 minutes, 7 seconds per lv25 dragon). Getting closer to the true average would require training more lv25s, though I'm not really motivated to do that since I'd make more profit per time spent on lower levels, so I'm calling 1 trial "good enough" even though I know it's not. Sorry!
|
Lv. 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 | _ |
Avg. Time (per dragon) 0:00:20 0:00:31 0:00:47 0:01:13 0:01:44 0:02:23 0:03:16 0:04:43 0:07:21 0:09:09 0:11:21 0:13:48 0:16:57 0:20:26 0:24:38 0:29:14 0:34:00 0:39:20 0:46:09 0:53:32 1:01:25 1:09:07 | _ |
Avg. Time (per pair) 0:00:39 0:01:01 0:01:34 0:02:26 0:03:28 0:04:45 0:06:31 0:09:26 0:14:42 0:18:17 0:22:42 0:27:35 0:33:53 0:40:51 0:49:16 0:58:28 1:08:00 1:18:40 1:32:17 1:47:04 2:02:50 2:18:13 |
RNG notes
L6: 1 miss Elim
L7: 1 miss Elim
L10: 1 miss Elim
L11: 1 low health reset
L12: 9 miss Elim, 1 KO reset
L13: 3 miss Elim
L14: 5 miss Elim
L15: 2 miss Elim
L16: 2 miss Elim, 1 Boss reset, 1 low health reset
L17: 5 miss Elim, 1 Boss reset
L18: 4 miss Elim, 2 Boss reset
L19: 3 miss Elim, 3 Boss reset
L20: 5 miss Elim, 1 Boss reset
L21: 5 miss Elim, 1 Boss reset, 1 low health reset
L22: 16 miss Elim, 1 Boss reset, 1 KO reset, 1 low health reset
L23: 11 miss Elim, 2 miss (Shroud Blind) Elim, 1 Boss reset, 2 low health reset
L24: 11 miss Elim, 1 Shock Paralysis, 1 KO reset, 2 low health reset
L25: 11 miss Elim, 2 Boss reset, 1 low health reset
---
One separate note. I won't spend too much text space on this here since my main goal for this post was to add a 50% XP boost comparison.
Over time, I've seen a lot of comments that gloss over the importance of recording true times to work towards an accurate average with an appropriate number of trials (hundreds / thousands of trials for each level). A lot of people seem to think that estimates based on "average time per battle" and "XP to n level" are good enough. However, such estimates can be many seconds or minutes off per level, and I think it's very important to debunk this line of thought, so I added a section to my guide about it. (See Section 0.4: Why true timing is significantly better than oversimplified math estimates.)
*raises a hand* I'd like to state that for those who happen to have wrist pain when grinding in the Coli, and I believe it'd count here as a form of increasing speed for profits, if you are an artist, use your drawing tablet (or any tablet).
I personally use my XP-Pen to Coli grind, and muscle memory makes for super quick clicks without fear of screwing up my wrist. It's worked for most sites that have a click-heavy feature, so I'd highly advise it!
I personally haven't charted any times for my fodder grinds with that method, but I can do so and get back with the results.
I personally use my XP-Pen to Coli grind, and muscle memory makes for super quick clicks without fear of screwing up my wrist. It's worked for most sites that have a click-heavy feature, so I'd highly advise it!
I personally haven't charted any times for my fodder grinds with that method, but I can do so and get back with the results.
*raises a hand* I'd like to state that for those who happen to have wrist pain when grinding in the Coli, and I believe it'd count here as a form of increasing speed for profits, if you are an artist, use your drawing tablet (or any tablet).
I personally use my XP-Pen to Coli grind, and muscle memory makes for super quick clicks without fear of screwing up my wrist. It's worked for most sites that have a click-heavy feature, so I'd highly advise it!
I personally haven't charted any times for my fodder grinds with that method, but I can do so and get back with the results.
I personally use my XP-Pen to Coli grind, and muscle memory makes for super quick clicks without fear of screwing up my wrist. It's worked for most sites that have a click-heavy feature, so I'd highly advise it!
I personally haven't charted any times for my fodder grinds with that method, but I can do so and get back with the results.
so with my own recorded times and with the updates to mire, from what it can tell GLR should be defiantly the fastest zone fr most lvls atm correct? Also i see you record times via a spreadsheet, I personally use LiveSpilt, a time recording tool for speedruning and i think it works great and recommend it.[emoji=wildclaw love size=1]
so with my own recorded times and with the updates to mire, from what it can tell GLR should be defiantly the fastest zone fr most lvls atm correct? Also i see you record times via a spreadsheet, I personally use LiveSpilt, a time recording tool for speedruning and i think it works great and recommend it.
Whoa, I'm a few years late with this. I meant to make an updated time comparison after the new Mire changes released, but I forgot and never did. So, I just took some recent data and here's another comparison! GLR times haven't been re-polled yet; I suspect we'll get a GLR update in the future since it hasn't changed for a very long time, so I'm going to keep using the current data set until then. The main reason that Mire times had to be re-taken is that some of the added enemies require 120 STR instead of the old 119 from several years ago.
If the summary looks a bit familiar, it's because I reused some phrasing from my previous time comparison post that I still find relevant now. (The main things that changed since the previous post are: I took new Mire times + added some lv7 data.) The Meditate data set ended up having almost the same times; the Scratch data set's times improved since the previous set, but still not enough to be better than Meditate.
Using what I consider to be the optimal stats and playstyle for each build / venue, I'm going to compare speed tests for levels 4-7. (In a perfect world, I'd have more trials for the higher levels too, but that'll have to wait for some other day. Higher levels need more trials to stabilize the average.)
Notes:
Here are the builds I'll be comparing.
A few QCK options are reasonable, but I chose very specific QCK levels for reasons outlined in the QCK turn order comparison post.
Average time per dragon
This is a comparison of the average time taken per dragon. Take note that dragons are trained in pairs, so the total time taken for a batch has been divided by 2 to get the average time per dragon.
Variance
RNG can do some pretty nasty things to training times of individual batches. If you're experiencing burnout and frustration from bad luck, there's definitely a good reason; sometimes 2x of the average training time can be wasted, and it may not even be your fault! Some builds play more consistently than others and have less annoying options of recovery, though if you're having a really awful RNG streak, every build suffers.
IMPORTANT: Variance is not evenly distributed! You cannot add min time + max time and divide by 2 for the average. Do not use the min/max times in this section to decide which build to use. I only include this section to show how disruptive RNG can be. The average time is a much more accurate picture of what you should expect out of a build, so please check the average time comparison section (above) when making a decision about builds.
Here are the minimum and maximum times that I have on record for each build. It's probably not the absolute min and max that I've ever experienced, but it's what I've got for this recorded dataset.
Conclusions
GLR Meditate still has the best lv4 average time (20 seconds per dragon), with Mire Meditate, GLR Scratch, and GLR Anticipate tied at 21 seconds per dragon. This may change once GLR updates to have more packs.
Mire Meditate starts to get better than all of the other builds starting with level 5 (average 39 seconds per dragon). This is a significant difference from the runner up, which is GLR Meditate (42 seconds per dragon). Mire Scratch, GLR Scratch, and GLR Anticipate builds are pretty close together and lag behind both Meditate builds slightly (43-45 seconds per dragon). The discrepancy increases as the level increases. When used properly, both Meditate builds tend to outperform the other breath stone builds.
When the level pushes past 6 or more, Anticipate starts to lag behind the two Scratch builds, because higher health fodder begins to interfere with building Anticipate breath. It's pretty even for levels 4 and 5, though. Note that Anticipate has some of the wildest variance when slammed with bad RNG (with the biggest pain point being not getting hit for breath), but it manages to stay competitive in levels 4 and 5 because the RNG can also go the other way (getting perfect packs and getting hit at good times for breath).
I consider all of the aforementioned to be good builds, and the Mire and Ghostlight Ruins are both good venues. You'll have to decide for yourself whether the differences are small enough for you to keep your favorite build or large enough to convince you to change to a new one. I can only provide the data for a comparison in a situation where each build was used optimally. If the builds are not used to their full potential, then mileage can vary quite wildly.
Personally, I will use GLR Meditate for lv.4 and Mire Meditate for lv.5+. Funny how this hasn't changed for so many years even with significant pack updates.
If the summary looks a bit familiar, it's because I reused some phrasing from my previous time comparison post that I still find relevant now. (The main things that changed since the previous post are: I took new Mire times + added some lv7 data.) The Meditate data set ended up having almost the same times; the Scratch data set's times improved since the previous set, but still not enough to be better than Meditate.
Using what I consider to be the optimal stats and playstyle for each build / venue, I'm going to compare speed tests for levels 4-7. (In a perfect world, I'd have more trials for the higher levels too, but that'll have to wait for some other day. Higher levels need more trials to stabilize the average.)
Notes:
- These times are without XP boosts. The speed comparison still mostly applies for XP-boosted events, except all average times will be faster. (Make sure to time yourself if you’re chasing profit during XP events, though, because the XP boost has an uneven effect on the profitable levels per venue, especially at lower levels.)
- RNG is highly variable and takes its toll more and more the higher the level, meaning that higher levels also involve a wider range of time extremes. Generally, more trials will always be better to stabilize the average times. At least 200 per level is ideal; more is better.
- For Meditate, when starting on normal packs, I use Meditate x2 to build breath. The point of multiple Meditates in a row is to reduce the amount of time wasted on Meditate Exhaustion based on average expected breath consumption from misses. I strongly disagree with a Meditate x1 regular startup strategy. Meditate x2 on startup has been more efficient for me when averaged over tens of thousands of lv4s. Yes, you miss Eliminate often even when training lv4s.
Here are the builds I'll be comparing.
- Mire Meditate: 120 STR / 64 QCK / 26 VIT / 7 DEF (Arcane) [stones] [speed tests]
- Mire Scratch: 120 STR / 64 QCK / 26 VIT / 7 DEF (Arcane) [stones] [speed tests]
- GLR 115 STR Meditate: 115 STR / 65 QCK / 39 VIT / 8 DEF (Nature) [stones] [speed tests]
- GLR 115 STR Scratch: 115 STR / 65 QCK / 39 VIT / 8 DEF (Nature) [stones] [speed tests]
- GLR 115 STR Anticipate: 115 STR / 59 QCK / 45 VIT / 13 DEF (Nature) [stones] [speed tests]
A few QCK options are reasonable, but I chose very specific QCK levels for reasons outlined in the QCK turn order comparison post.
Average time per dragon
This is a comparison of the average time taken per dragon. Take note that dragons are trained in pairs, so the total time taken for a batch has been divided by 2 to get the average time per dragon.
|
Build
Mire: 119 STR MeditateMire: 119 STR Scratch GLR: 115 STR Meditate GLR: 115 STR Scratch GLR: 115 STR Anticipate | ____ |
Lv.4 Avg
0:00:210:00:23 0:00:20 0:00:21 0:00:21 | ____ |
#
366190 696 656 652 | ____ |
Lv.5 Avg
0:00:390:00:43 0:00:42 0:00:45 0:00:44 | ____ |
#
372190 446 458 400 | ____ |
Lv.6 Avg
0:01:030:01:08 0:01:08 0:01:12 0:01:14 | ____ |
#
368190 182 156 200 |
|
Build
Mire: 119 STR MeditateMire: 119 STR Scratch GLR: 115 STR Meditate GLR: 115 STR Scratch GLR: 115 STR Anticipate | ____ |
Lv.7 Avg
0:01:390:01:47 0:01:48 0:01:57 0:01:55 | ____ |
#
33070 102 12 40 |
Variance
RNG can do some pretty nasty things to training times of individual batches. If you're experiencing burnout and frustration from bad luck, there's definitely a good reason; sometimes 2x of the average training time can be wasted, and it may not even be your fault! Some builds play more consistently than others and have less annoying options of recovery, though if you're having a really awful RNG streak, every build suffers.
IMPORTANT: Variance is not evenly distributed! You cannot add min time + max time and divide by 2 for the average. Do not use the min/max times in this section to decide which build to use. I only include this section to show how disruptive RNG can be. The average time is a much more accurate picture of what you should expect out of a build, so please check the average time comparison section (above) when making a decision about builds.
Here are the minimum and maximum times that I have on record for each build. It's probably not the absolute min and max that I've ever experienced, but it's what I've got for this recorded dataset.
|
Build
Mire: 119 STR MeditateMire: 119 STR Scratch GLR: 115 STR Meditate GLR: 115 STR Scratch GLR: 115 STR Anticipate | ____ |
Lv.4 Min
0:00:160:00:15 0:00:16 0:00:15 0:00:14 | ____ |
Lv.4 Max
0:00:490:00:49 0:00:38 0:00:45 0:00:54 | ____ |
Lv.5 Min
0:00:300:00:30 0:00:32 0:00:31 0:00:31 | ____ |
Lv.5 Max
0:01:070:01:28 0:01:14 0:01:28 0:01:15 |
|
Build
Mire: 119 STR MeditateMire: 119 STR Scratch GLR: 115 STR Meditate GLR: 115 STR Scratch GLR: 115 STR Anticipate | ____ |
Lv.6 Min
0:00:500:00:50 0:00:55 0:00:55 0:00:55 | ____ |
Lv.6 Max
0:01:360:01:58 0:01:47 0:01:43 0:02:23 | ____ |
Lv.7 Min
0:01:210:01:25 0:01:31 0:01:31 0:01:33 | ____ |
Lv.7 Max
0:02:380:02:23 0:02:18 0:02:21 0:02:14 |
Conclusions
GLR Meditate still has the best lv4 average time (20 seconds per dragon), with Mire Meditate, GLR Scratch, and GLR Anticipate tied at 21 seconds per dragon. This may change once GLR updates to have more packs.
Mire Meditate starts to get better than all of the other builds starting with level 5 (average 39 seconds per dragon). This is a significant difference from the runner up, which is GLR Meditate (42 seconds per dragon). Mire Scratch, GLR Scratch, and GLR Anticipate builds are pretty close together and lag behind both Meditate builds slightly (43-45 seconds per dragon). The discrepancy increases as the level increases. When used properly, both Meditate builds tend to outperform the other breath stone builds.
When the level pushes past 6 or more, Anticipate starts to lag behind the two Scratch builds, because higher health fodder begins to interfere with building Anticipate breath. It's pretty even for levels 4 and 5, though. Note that Anticipate has some of the wildest variance when slammed with bad RNG (with the biggest pain point being not getting hit for breath), but it manages to stay competitive in levels 4 and 5 because the RNG can also go the other way (getting perfect packs and getting hit at good times for breath).
I consider all of the aforementioned to be good builds, and the Mire and Ghostlight Ruins are both good venues. You'll have to decide for yourself whether the differences are small enough for you to keep your favorite build or large enough to convince you to change to a new one. I can only provide the data for a comparison in a situation where each build was used optimally. If the builds are not used to their full potential, then mileage can vary quite wildly.
Personally, I will use GLR Meditate for lv.4 and Mire Meditate for lv.5+. Funny how this hasn't changed for so many years even with significant pack updates.
Whoa, I'm a few years late with this. I meant to make an updated time comparison after the new Mire changes released, but I forgot and never did. So, I just took some recent data and here's another comparison! GLR times haven't been re-polled yet; I suspect we'll get a GLR update in the future since it hasn't changed for a very long time, so I'm going to keep using the current data set until then. The main reason that Mire times had to be re-taken is that some of the added enemies require 120 STR instead of the old 119 from several years ago.
If the summary looks a bit familiar, it's because I reused some phrasing from my previous time comparison post that I still find relevant now. (The main things that changed since the previous post are: I took new Mire times + added some lv7 data.) The Meditate data set ended up having almost the same times; the Scratch data set's times improved since the previous set, but still not enough to be better than Meditate.
Using what I consider to be the optimal stats and playstyle for each build / venue, I'm going to compare speed tests for levels 4-7. (In a perfect world, I'd have more trials for the higher levels too, but that'll have to wait for some other day. Higher levels need more trials to stabilize the average.)
Notes:
Here are the builds I'll be comparing.
A few QCK options are reasonable, but I chose very specific QCK levels for reasons outlined in the QCK turn order comparison post.
Average time per dragon
This is a comparison of the average time taken per dragon. Take note that dragons are trained in pairs, so the total time taken for a batch has been divided by 2 to get the average time per dragon.
Variance
RNG can do some pretty nasty things to training times of individual batches. If you're experiencing burnout and frustration from bad luck, there's definitely a good reason; sometimes 2x of the average training time can be wasted, and it may not even be your fault! Some builds play more consistently than others and have less annoying options of recovery, though if you're having a really awful RNG streak, every build suffers.
IMPORTANT: Variance is not evenly distributed! You cannot add min time + max time and divide by 2 for the average. Do not use the min/max times in this section to decide which build to use. I only include this section to show how disruptive RNG can be. The average time is a much more accurate picture of what you should expect out of a build, so please check the average time comparison section (above) when making a decision about builds.
Here are the minimum and maximum times that I have on record for each build. It's probably not the absolute min and max that I've ever experienced, but it's what I've got for this recorded dataset.
Conclusions
GLR Meditate still has the best lv4 average time (20 seconds per dragon), with Mire Meditate, GLR Scratch, and GLR Anticipate tied at 21 seconds per dragon. This may change once GLR updates to have more packs.
Mire Meditate starts to get better than all of the other builds starting with level 5 (average 39 seconds per dragon). This is a significant difference from the runner up, which is GLR Meditate (42 seconds per dragon). Mire Scratch, GLR Scratch, and GLR Anticipate builds are pretty close together and lag behind both Meditate builds slightly (43-45 seconds per dragon). The discrepancy increases as the level increases. When used properly, both Meditate builds tend to outperform the other breath stone builds.
When the level pushes past 6 or more, Anticipate starts to lag behind the two Scratch builds, because higher health fodder begins to interfere with building Anticipate breath. It's pretty even for levels 4 and 5, though. Note that Anticipate has some of the wildest variance when slammed with bad RNG (with the biggest pain point being not getting hit for breath), but it manages to stay competitive in levels 4 and 5 because the RNG can also go the other way (getting perfect packs and getting hit at good times for breath).
I consider all of the aforementioned to be good builds, and the Mire and Ghostlight Ruins are both good venues. You'll have to decide for yourself whether the differences are small enough for you to keep your favorite build or large enough to convince you to change to a new one. I can only provide the data for a comparison in a situation where each build was used optimally. If the builds are not used to their full potential, then mileage can vary quite wildly.
Personally, I will use GLR Meditate for lv.4 and Mire Meditate for lv.5+. Funny how this hasn't changed for so many years even with significant pack updates.
If the summary looks a bit familiar, it's because I reused some phrasing from my previous time comparison post that I still find relevant now. (The main things that changed since the previous post are: I took new Mire times + added some lv7 data.) The Meditate data set ended up having almost the same times; the Scratch data set's times improved since the previous set, but still not enough to be better than Meditate.
Using what I consider to be the optimal stats and playstyle for each build / venue, I'm going to compare speed tests for levels 4-7. (In a perfect world, I'd have more trials for the higher levels too, but that'll have to wait for some other day. Higher levels need more trials to stabilize the average.)
Notes:
- These times are without XP boosts. The speed comparison still mostly applies for XP-boosted events, except all average times will be faster. (Make sure to time yourself if you’re chasing profit during XP events, though, because the XP boost has an uneven effect on the profitable levels per venue, especially at lower levels.)
- RNG is highly variable and takes its toll more and more the higher the level, meaning that higher levels also involve a wider range of time extremes. Generally, more trials will always be better to stabilize the average times. At least 200 per level is ideal; more is better.
- For Meditate, when starting on normal packs, I use Meditate x2 to build breath. The point of multiple Meditates in a row is to reduce the amount of time wasted on Meditate Exhaustion based on average expected breath consumption from misses. I strongly disagree with a Meditate x1 regular startup strategy. Meditate x2 on startup has been more efficient for me when averaged over tens of thousands of lv4s. Yes, you miss Eliminate often even when training lv4s.
Here are the builds I'll be comparing.
- Mire Meditate: 120 STR / 64 QCK / 26 VIT / 7 DEF (Arcane) [stones] [speed tests]
- Mire Scratch: 120 STR / 64 QCK / 26 VIT / 7 DEF (Arcane) [stones] [speed tests]
- GLR 115 STR Meditate: 115 STR / 65 QCK / 39 VIT / 8 DEF (Nature) [stones] [speed tests]
- GLR 115 STR Scratch: 115 STR / 65 QCK / 39 VIT / 8 DEF (Nature) [stones] [speed tests]
- GLR 115 STR Anticipate: 115 STR / 59 QCK / 45 VIT / 13 DEF (Nature) [stones] [speed tests]
A few QCK options are reasonable, but I chose very specific QCK levels for reasons outlined in the QCK turn order comparison post.
Average time per dragon
This is a comparison of the average time taken per dragon. Take note that dragons are trained in pairs, so the total time taken for a batch has been divided by 2 to get the average time per dragon.
|
Build
Mire: 119 STR MeditateMire: 119 STR Scratch GLR: 115 STR Meditate GLR: 115 STR Scratch GLR: 115 STR Anticipate | ____ |
Lv.4 Avg
0:00:210:00:23 0:00:20 0:00:21 0:00:21 | ____ |
#
366190 696 656 652 | ____ |
Lv.5 Avg
0:00:390:00:43 0:00:42 0:00:45 0:00:44 | ____ |
#
372190 446 458 400 | ____ |
Lv.6 Avg
0:01:030:01:08 0:01:08 0:01:12 0:01:14 | ____ |
#
368190 182 156 200 |
|
Build
Mire: 119 STR MeditateMire: 119 STR Scratch GLR: 115 STR Meditate GLR: 115 STR Scratch GLR: 115 STR Anticipate | ____ |
Lv.7 Avg
0:01:390:01:47 0:01:48 0:01:57 0:01:55 | ____ |
#
33070 102 12 40 |
Variance
RNG can do some pretty nasty things to training times of individual batches. If you're experiencing burnout and frustration from bad luck, there's definitely a good reason; sometimes 2x of the average training time can be wasted, and it may not even be your fault! Some builds play more consistently than others and have less annoying options of recovery, though if you're having a really awful RNG streak, every build suffers.
IMPORTANT: Variance is not evenly distributed! You cannot add min time + max time and divide by 2 for the average. Do not use the min/max times in this section to decide which build to use. I only include this section to show how disruptive RNG can be. The average time is a much more accurate picture of what you should expect out of a build, so please check the average time comparison section (above) when making a decision about builds.
Here are the minimum and maximum times that I have on record for each build. It's probably not the absolute min and max that I've ever experienced, but it's what I've got for this recorded dataset.
|
Build
Mire: 119 STR MeditateMire: 119 STR Scratch GLR: 115 STR Meditate GLR: 115 STR Scratch GLR: 115 STR Anticipate | ____ |
Lv.4 Min
0:00:160:00:15 0:00:16 0:00:15 0:00:14 | ____ |
Lv.4 Max
0:00:490:00:49 0:00:38 0:00:45 0:00:54 | ____ |
Lv.5 Min
0:00:300:00:30 0:00:32 0:00:31 0:00:31 | ____ |
Lv.5 Max
0:01:070:01:28 0:01:14 0:01:28 0:01:15 |
|
Build
Mire: 119 STR MeditateMire: 119 STR Scratch GLR: 115 STR Meditate GLR: 115 STR Scratch GLR: 115 STR Anticipate | ____ |
Lv.6 Min
0:00:500:00:50 0:00:55 0:00:55 0:00:55 | ____ |
Lv.6 Max
0:01:360:01:58 0:01:47 0:01:43 0:02:23 | ____ |
Lv.7 Min
0:01:210:01:25 0:01:31 0:01:31 0:01:33 | ____ |
Lv.7 Max
0:02:380:02:23 0:02:18 0:02:21 0:02:14 |
Conclusions
GLR Meditate still has the best lv4 average time (20 seconds per dragon), with Mire Meditate, GLR Scratch, and GLR Anticipate tied at 21 seconds per dragon. This may change once GLR updates to have more packs.
Mire Meditate starts to get better than all of the other builds starting with level 5 (average 39 seconds per dragon). This is a significant difference from the runner up, which is GLR Meditate (42 seconds per dragon). Mire Scratch, GLR Scratch, and GLR Anticipate builds are pretty close together and lag behind both Meditate builds slightly (43-45 seconds per dragon). The discrepancy increases as the level increases. When used properly, both Meditate builds tend to outperform the other breath stone builds.
When the level pushes past 6 or more, Anticipate starts to lag behind the two Scratch builds, because higher health fodder begins to interfere with building Anticipate breath. It's pretty even for levels 4 and 5, though. Note that Anticipate has some of the wildest variance when slammed with bad RNG (with the biggest pain point being not getting hit for breath), but it manages to stay competitive in levels 4 and 5 because the RNG can also go the other way (getting perfect packs and getting hit at good times for breath).
I consider all of the aforementioned to be good builds, and the Mire and Ghostlight Ruins are both good venues. You'll have to decide for yourself whether the differences are small enough for you to keep your favorite build or large enough to convince you to change to a new one. I can only provide the data for a comparison in a situation where each build was used optimally. If the builds are not used to their full potential, then mileage can vary quite wildly.
Personally, I will use GLR Meditate for lv.4 and Mire Meditate for lv.5+. Funny how this hasn't changed for so many years even with significant pack updates.