Ranking Dungeon Prep: Athena

Yet another early edition of Ranking Dungeon Prep (or at least I hope it is). This timing is more in line with what I’d usually like to do; the earlier I do the research, the more time I’ll have to prepare for it in-game. The downside is that things can change between now and when the next ranking dungeon hits. The highest impact changes I can think of are the dungeon order changing and new cards being released. If the circumstances do change, I’ll try to address them as soon as I can.

In many ways the Athena ranking dungeon is the antithesis of its predecessor, Takeminakata. Due to the greater floor count, this dungeon instead emphasizes minimal combo count meaning skilled players will have less of an advantage. In Takeminakata, players welcomed skyfall combos, but in Athena you want to avoid them for the most part. Then the dreaded PreDRA floor makes its ugly return. One positive is the ability to use actives again; while getting a good score will be about as difficult as always, the lowered variance will benefit players that want a decent score but don’t want to invest too much stamina. There are still plenty of ways to gain an advantage over the field and I hope to find some that go beyond “get your Shiva Dragon, Set(s) and ROdin ready”.

Dungeon at a Glance

I’m eager to jump right into team analysis, but I’ll be diligent this time and give the dungeon a quick look over first.

Data taken from PDX.

Athena Tournament
Floor	Monster	Types	HP / DEF	Notes
1			76,350 / 900	Random 3. This floor would normally be a button target, but skill-only kills aren’t good in a dungeon of this length.
			76,350 / 900
			76,350 / 900
			76,350 / 900
			76,350 / 900
2			357,358 / 210	Blind. 1,497 preemptive strike.
			215,025 / 210
3			565,673 / 980
			580,923 / 980
4			45 / 10,000,000	Random 2. They’re back!
			45 / 10,000,000
			45 / 10,000,000
			45 / 10,000,000
			45 / 10,000,000
5			1,608,819 / 1,700	4 combo shield. Not a huge deal since in most cases you’ll need to combo to kill him anyways.
6			31,618 / 1,540	Random 3. Another possible button target.
			32,635 / 1,540
			35,685 / 1,540
7			820,415 / 0	9,172 preemptive strike.
8			280,544 / 188,889	Moderate defense. Resists make it harder for light or dark teams; wood is obviously preferred.
9			772,137 / 23,667	Blind. Something like AA Lucifer should be able to button this floor.
10			2,226,875 / 1,675	5 combo shield.

The key points are:

• No restrictions – This is the first ranking dungeon since Izanami where there are no team restrictions.
• 10 Floors – This is the highest number of floors we’ve ever gotten in a ranking dungeon. This means that any combo — not just skyfall/cascade, but any type — will hurt your overall score. Technically an extra combo will help if the time moving orbs is lower than 0.1 seconds, but that’s unrealistic. It follows that you want to make the minimum amount of combos to clear each floor to reduce the time penalty on your score (outside your burst floor if you’re going for the maximum damage bonus). Refer to my table in the Takeminakata prep post for more info on where this data comes from.
• Active reuse – A characteristic of long dungeons is that many actives used early in the dungeon will be ready again near the end. Be sure to factor this into your sub selection and floor-by-floor plan.
• Low HP – This allows for the conservation of main-element orbs and possibly clearing trash with sub-elements.
• Possible button floors – Floors 1, 6 and even 9 can be buttoned, but due to the skill-only penalty weighing so heavily due to the nature of this dungeon, it’s not a realistic option.
• Blinds – As minimal combo is preferred, the blinds shouldn’t be more than a simple annoyance. The better memory you have, the less reliance you’ll have on revealing the board, saving time in the process.
• The PreDRAs are back – Hopefully you have your poison ready. I’ll go over the pros and cons of using poison vs. true damage later in the post.
• Combo shields on floors 5 and 10 – You’ll likely want to burst on one of these floors if you’re planning on going for the max damage bonus since you’re going to have to combo high anyways.
• Wood and dark have an easier time hitting maximum damage – If floors 5 and 10 are the most likely candidates to hit max damage, then wood and dark teams have a slight edge due to the elemental damage bonus.
• Neptune has 50% light and dark resist – Both light and dark teams will have to expend a tiny bit more effort to clear this floor. Dark teams with god killers are particularly tempting due to Athena on the last floor, but this floor ensures that advantage won’t come for free.

JP Top 10 Shiva Dragons

Since the dungeon rewards minimal combo count due to the amount of floors, it only makes sense that our old friend Shiva Dragon dominates the JP top 10:

JP Athena Cup Top 10

Looks like it’s going to be a fun event, yes? There are no restrictions for this dungeon, but it may as well read: Shiva Dragon Teams Only.

Using the Cookie Cutter Team

The strategy is demonstrated in this video:

This player got a score that was only about 300 points off the JP top 10 so it should be a pretty good example of how to clear the dungeon. Basically use orb changers, make TPAs and pray for no skyfalls. Verdandi is there to help against the few water monsters in the dungeon.

The big takeaway from this video is the final floor where he gets some generous skyfall combos to reach the maximum damage score. So, the question becomes: how much benefit is there going for maximum damage?

Is Reaching Maximum Damage Profitable?

Without even doing damage calculations, it’s fairly safe to assume reaching maximum damage is worth the time spent executing the turn. For this dungeon, Athena has the highest HP at 2.2 million which means the minimum amount of damage points you can get is about 550. As such, putting in the extra effort to reach the damage cap can reward up to 9,450 above the minimum. That amount of points equates to about 19 seconds in time score; in other words it’s almost guaranteed to be profitable.

I’m still going to take a closer look as there are other factors that need to be considered, namely the method taken to reach that number. Since the cookie cutter Shiva Dragon build can’t reach 40 million damage against Athena on-board — even factoring Set’s buff and enhanced orbs — there will need to be some amount of skyfall to boost damage to the maximum. As stated earlier, skyfall combos have a negative effect on your score. I still think it’s safe to assume there will be a profit, but I’ll run the numbers anyways.

* Avg. Verdandi + Set board, 20 fire orbs, 10 wood orbs; 7 fire row enhances; 8 enhanced fire orbs, 4 enhanced wood orbs; 1.5x fire damage (Set); on-board combos (6): 3 fire rows, 3 wood 3-orb match; all hypermax; solved with this board; damage calculated using PADSpike

Verdandi + Set Board – Rows
Damage By Sub
Main	4,391,300	4,359,350	4,359,350	5,265,675	209,700	4,391,300
Sub	0	77,800	77,800	0	1,305,475	0
Team Damage
Elem.						Total
Dmg.	22,976,675	0	1,461,075	0	0	24,437,750

Using an optimal two-element board setup, an average Set + Verdandi board will result in about 24 million on-board damage, well short of the 40 million target (surprisingly, going rows results in more damage than going full TPA or row-TPA hybrid). The remaining 16 million is up to skyfall.

The worst case skyfall combo is one that is totally off-element which contributes a 25% bonus to base damage (in this case, water, light and hearts). In the above case, an off-element combo will add about 2,715,326 damage which translates to 678.83 points. A cascade/skyfall combo costs 683 points due to time lost to animations, but for a 10-floor dungeon an extra combo also gives 500 points due to the average combo increase. In short, the worst skyfall type will still net a positive 495.83 points on your burst floor. It follows that an on-element skyfall combo (fire, wood or dark) would net even more points.

So, as assumed, going for the maximum damage bonus will indeed help your score. However, if you’re trying to get the most amount of points possible there’s a catch: every skyfall combo contributing damage past 40 million damage once again hurts your score since the damage points are capped. So, to my chagrin, not only do you have to hope for no skyfalls on floors 1-9, not only do you have to get the requisite amount of skyfalls to hit maximum damage on floor 10, but you also have to hope you get exactly enough skyfalls so they don’t start to hurt your score again.

It’s this obstacle that makes dark teams all the more attractive since they can theoretically reach 40 million on Athena without relying on skyfall (I guess wood teams also need to be considered for their burst on Deadjumper). Another consideration would be adding a skyfall buff to your team just to combat RNG on your burst floor. If you’re going to use the Shiva Dragon team, any method of increasing your team’s damage will lower the reliance on skyfall, leading to more consistent runs.

Do Killer Awakenings Change Anything?

Maybe. JP had Awoken Archdemon Lucifer and Machine Hera with their god killers at the time of this ranking dungeon, but they failed to make an impact on the top 10. Judging from a cursory search of YouTube, the plebs didn’t try to use them, either.

NA has access to a few god killer cards that JP didn’t: Ragnarok Dragon, Fenrir and the RGB Constellation 2s. There’s also a good chance that NA could receive the version 8.8 awakenings buffs soon which gives many old pantheon split uevos killer awakenings. So while killers didn’t make an impact on JP, there’s still the possibility of someone getting creative on NA.

Assuming killers are good, how much of an advantage is there? I’m really not sure, but at the very least I can present a very simple case to at least get a ballpark idea. Since the second Set isn’t absolutely critical to the cookie cutter Shiva Dragon build, I’ll replace one with Denebola.

* Avg. Verdandi + Set board, 20 fire orbs, 10 wood orbs; 7 fire row enhances; 8 enhanced fire orbs, 4 enhanced wood orbs; 1.5x fire damage (Set); on-board combos (6): 3 fire rows, 3 wood 3-orb match; all hypermax; solved with this board; damage calculated using PADSpike

Verdandi + Set Board – Rows – Testing Denebola’s God Killer
Damage By Sub
Main	4,117,425	4,087,575	9,477,750	4,937,125	203,513	4,117,425
Sub	0	75,513	175,089	0	1,224,000	0
Team Damage
Elem.						Total
Dmg.	26,940,813	0	1,474,602	0	0	28,415,415

You get about 4 million extra damage from using Denebola with a row board (going full TPA or row-TPA hybrid with a 20-10 board does significantly less damage on-board, even when factoring in Denebola’s TPAs stacking with her god killer). That’s about the same amount of damage as 1.5 off-element skyfall matches would do with the previous team. This basically means you’ll have to combat much less skyfall RNG on your burst turn with this team. However, this change doesn’t come absolutely free as you’ll have to adjust how you tackle the rest of the dungeon. It doesn’t appear to have a huge impact, but it’s a factor nonetheless. At the very least, this change doesn’t impact the timing of having ROdin up for the PreDRA floor.

This example is just food for thought. Whether killers help you or not, I’ll leave that to your own testing, but they’re definitely worth the consideration.

Other Tips

Poison vs. True Damage

I actually put some effort into researching whether poison or true damage is better for a given dungeon. The general adage of “true damage is better for high-combo ranking dungeons and poison is better for low-combo” generally holds true, but I wanted to see the numbers for myself. The answer varies depending on the number of floors, your average combo and, of course, player skill. An additional factor for poison usage is the number of combos made on that floor before the poison triggers. Being poor at math, I had a hard time articulating my thoughts so rather than rushing it, I’m going to save that content for a separate post (this post is already much longer than I expected). The main points I want to cover are under what circumstances true damage is better than poison and for where you want to use poison, how many combos is optimal (0? 1? More?). I hope to get it out before Athena hits NA.

I will say, however, that for this particular dungeon there’s no contest: poison is better. This mostly holds true even if you’re going for a higher combo count strategy.

Avoid Skill-Only Clears

This is mainly an extension of the poison vs. true damage comparison; the skill-only penalty is too much to overcome in such a long dungeon. I found a particularly interesting video from dacho where he used a ROdin tama with just enough levels leave the 6th floor at a sliver of life so he could finish it off through poison. I don’t know his actual reasoning behind this, but since I’m pretty sure he has access to multiple, full-sized ROdins, I can only assume this was done to avoid the skill-only penalty.

Minimizing Combo Count by Knowing Your Damage Requirement

Since minimizing combo count is the name of the game, knowing the minimum amount of combos you need to clear each floor is extremely important. I strongly suggest using using a damage calculator like PADSpike or Arkaether’s spreadsheet.

Also, if you’d like to know why minimizing your combo count is good and why skyfalls are bad for this particular dungeon, I cover that a bit more in my Takeminakata post.

Other Points Covered by dacho

In addition to his video covering the effect of skyfall combos on your score, dacho also made one to cover some of the smaller nuances of ranking dungeons. The main points are:

• Where that seemingly random number in the ones’ place in your score comes from
• The positive effect empty sub slots and missing sub-attributes can have on time remaining
• The negative effect egg drop animations can have on time remaining.

I won’t pretend to know enough Japanese to translate any of his findings and there may be some points that I missed, but while I was on the topic of min-maxing I thought it’d be nice to bring up for those that want to explore it further.

Alternative Shiva Dragon Subs and Strategies

There is a little flexibility in the subs you can use on the cookie cutter team. Even within the JP top 10, Urd and Ares made appearances.

Shiva Dragon w/Ares 103,238

A clear using two Ares. This demonstrates the weakness of not having Verdandi or Urd on the team to go for the maximum damage bonus.

Shiva Dragon w/Krishna 110,606

This video is the main reason I made this section. Krishna isn’t thought of as the greatest sub, but in this case his skyfall buff makes the burst turn more consistent. The trade-off is losing an orb change for floors 1-9, but for the most part that doesn’t seem like a critical loss. This could be a sub choice for players looking to lower stamina costs by mitigating RNG, but still want to score well.

Outside these two examples, is there any room for any other improvements? As Shiva Dragon’s sub pool is relatively limited, there aren’t too many options. One might think of using a larger damage enhance than Set, but the only other fire-god enhancers with a larger multiplier are the Shinrabansho silvers Apollo and Sai. You’re welcome to search for better options, though.

Alternatives to Shiva Dragon

What can players that don’t have both Shiva Dragon and ROdin use? It’s hard to say, but a flat multiplier would be preferable. A team that immediately comes to mind is Verdandi since a wood TPA team should have a pretty good matchup.

Verdandi 92,052

Verdandi 92,206

She’s not really a top 1% team, but she should be good enough to hit the top 10% with the right team.

The next most obvious teams would be dark-based since they can take advantage of Athena’s light attribute. There’s also the possibility of utilizing dark’s god killers (Lucifer, MHera, Fenrir). Here are a few non-MP dark teams I found videos for:

DValk 105,308

A bit to my surprise, there were no good AD Lucifer runs. DValk is the closest I found. She has some big advantages over Lucifer with a flat 3x ATK multiplier, a higher top-end and a better active. Granted, there was some huge luck with that final board, but on the other hand the team had very few +eggs. I think there’s some potential for a team like this.

Satsuki 100,604

Another video by the DValk guy that uses minimal +eggs. Satsuki provides a nice flat multiplier with plenty of TPAs.

DMeta 102,992

This dungeon has a couple small preemptive strikes which can prove to be annoying for a team like this.

DIza 89,054

Ah, my good friend, DIza. This run is pretty sloppy and could definitely be improved, but I couldn’t find videos of a better run. There’s definitely some top 10% potential here. If you use a poisoner instead of Ra and keep combos to minimum, you can easily improve on this score (which was top 3% at the time of the recording).

Finally, here are some alternative MP Dragon clears:

Neptune Dragon-BSonia 103,058

Not really a team I’d recommend, but it worked out surprisingly well. The player had a good overall strategy as all the actives lined up perfectly.

Ra Dragon 103,906

dacho proves you can top 1% with with a color match team. There aren’t many players on his level, though. I mentioned this earlier, but it should be noted that he uses ROdin Tama on the 6th floor on purpose to avoid the skill-only penalty by killing via poison; notice it’s not plussed nor max leveled (I can only imagine he get very lucky with skill ups or used Flampys). It also has the minimal advantage of not having a sub-attribute.

Yomi Dragon 104,216

The last MP Dragon to cover, this Yomi Dragon clear gives hope to those looking to match-five their way into the top 10% or higher. An interesting point to note is that the super garbage floors don’t even need a match-five to clear.

Sub Choices

Poison

I went over a few examples in my Noah post; not much has changed.

Killers

The killer situation for NA hasn’t really changed since Takeminakata ended; as of the initial revision of this post only Machine Zeus was added. The upcoming 8.8 awakening buffs could definitely shake things ups. Collabs also have some potential (Lightning, Ichigo, many Crows Worst cards), but that’s being really optimistic. There’s also the 4th anniversary uevos with Zuoh (dragon killer), Gadius and Typhon (both devil killer). It appears that god killer will still be the most relevant, though.

Here are the PDX pages for God Killer, Devil Killer and Dragon Killer.

Blind Resist

There’s no reason to force blind resists onto you team unlike jammers for Takeminakata, but it’s certainly a nice luxury to have if the sub fit into your team anyways.

Here’s the PDX page for blind resist.

Skyfall Increase

This might seem contrary to the overall dungeon strategy, but for many teams skyfall is necessary to hit the maximum damage bonus. You’d preferably pop the buff on the last floor as to keep the skyfalls isolated to your burst turn. Fenrir is one of the more intriguinging options as his god killer is coveted and maybe the (unlikely) jammer skyfall combo can enable the last damage you need to hit 40 million. The awoken Greeks could play a role if we get them quick enough.

To see a full list, visit the PDX Skill List and select the “Skyfall” filter.

Other Clears

There’s quite a few other clear videos, here’s the YouTube search link if you’re interested in looking yourself.

Conclusion

From what I could gather, it felt like Takeminakata was one of the most hated ranking dungeons so far. While skilled players had an advantage, the feeling of getting screwed over by skyfall RNG was unpleasant for everyone involved. While Athena might have a more pay-to-win feel to it, the most heavy RNG element is isolated to the very few players competing for top 10. This means players that don’t take the competition too seriously can get their top 10-20% in a handful of runs and call it a day. By nature, the ranking dungeons are unfair, favoring those that have access to the best teams, but having RNG play a reduced role will make the experience a less tilting for most players. Also, for those that don’t have access to Shiva Dragon, it looks to be more than possible to score well with other leads so don’t let that discourage you too much.