Tuesday 22 March 2016

Some background information of "dungeon data"

This post is not related to any game play experience or card / mechanics analyses. I just want to share with you something works in background, especially on the dungeons are you attempting.

Dungeon Monster Tables

The monster tables store the appearances of monsters with their respective actions in a dungeon. For example in Hera Sowilo it pulls a data of "Tornado Holy Dragon" on floor 5. Other dungeons like Nordis and Hel also pull the same data set of "Tornado Holy Dragon". That's why every time you encounter a Tornado Holy Dragon you are expected to see the same set of behavior (for instance, preempt shield → 3 turns count down followed by Disaster Air). The only differences are the encounters' stats (HP, ATK, DEF) and their drop levels.

Have you ever asked yourself "how do they determine the level of drops?" Why the same dragon in Hel Legend drops me a level 4 egg, and Hera-Sowilo drops me a level 5, and Nordis Mythical+ drops a weird level 11 dragon?

It is because the stats of the monster you are facing is tied to the drop level - the higher the drop levels, the harder the encounter. And from now on I'd like to call the drop level "Monster Level".

Monster Level

The Monster Level (MLv) determines the stats of the encounter. Take "Tornado Holy Dragon" as an example (does not include encounter of its previous evos), its stats can be determined by the following formulas:

ATK = 1560 + 5033.3 x (MLv-1)
DEF = 820 + 820 x (MLv-1)
HP = 100207 + 322882.6 x (MLv-1) 

For instance, the Tornado Holy Dragon in Nordis Legend+ has a MLv of 10 (since it drops as a level 10 egg). With the formula you can calculate the HP, DEF and the ATK of the encounter.

ATK*DEFHP
Fomular10457982003006150
In-game value10458082003006150

* the actual damage of "Disaster Air", which deals 300% damage of its ATK value.

Every encounter in this game is stored as a linear function of stats and the dungeon data stores the MLv of the encounter.

The internal MLv can go over their maximum level. For example the internal MLv of "Frost Devil" at MLv 10 in Aamir Legend, MLv 15 in Aamir Mythical. So "Frost Devil" you got from Aamir Mythical is dropped at Lv Max. If you have cleared the Alt dungeons in Co-op you should remember you have encounter the same "Frost Devil" that drops at Lv Max, but hits much harder - It is because they are at MLv 20 and 21 (in Alt Talos's Abyss).

You'll see same things applies to monsters that drops only at level 1, like High Emerald Dragons, it's MLv 30 in Legendary Earth, and MLv 1 at Super Emerald Dragon (the alert dungeon). There are also MLv 10 dub-lits at Sky Prison and MLv 12 Mystic Masks in Alt Temple of Trailokya.

The rank EXP (rXP) given upon defeating monsters is also tied directly to their MLv.
rXP = base x MLv
For example, Sapphire Carbuncles (and all carbuncles of the same evo stages) has a base rXP of 13.57, and non-evo Pengdras has a base rXP of 40, which is about 3x of a carbuncle.

Take Alt Shrine of Green Water as an example:

The carbuncles and pengdras are dropped at Lv Max, but their internal MLv are 31 and 21 respectively. For each carbuncles spawned and defeated, you gain 13.57 x 31 ~= 419 or 420 rXP depending on the round-off, and for each Pengdra killed you gain 40 x 21 = 840 rXP. And the High Emeralds gives 1500 rXP!

The last but the least, monsters that don't drop eggs also have an internal level to determine their stats. For example "Fire Serpent God, Hino Kagutsuchi" which appears in Izanami and Takemikazuchi has an internal level of 3, 5 for Izanami Legend and Mythical, and 3, 6 for Takemikazuchi Legend and Mythical respectively.

For your information, FD Hino's stats curve is:
ATK = 1564 + 5036 x (MLv-1)
DEF = 84 + 84 x (MLv-1)
HP = 100210 + 322899 x (MLv-1)

Encounter buff?

But wait fether, your formula is a bit off and is not consistence in eh.. some dungeons?

That's true because some dungeons also give an implicit buff to monsters! And this is how Gungho counter the buff/"conditions" of dungeons that enhance certain attr./types of monsters in our team. For example, Dark Knight! buff all stats of Physical and Devil cards by 1.5x. They also increases the monsters in the dungeon HP and ATK by 1.5x to counterbalance it. This way, they are not raising the MLv of the encounters directly so you won't get extra rXP (remember rXP is tied to MLv), and impose a "punishment" to players that use cards that does not fit the conditions (as oppose to the old restriction rule).

Below is a list of implicit monster buffs applied to the dungeons:

Player Buff Enemy Buff
x1.5 ATK Defense HP
Tengu!! Balanced 1.5 2.5
Draggie! Attacker 2.5 2.5
Dragon Zombie Dragon 1.3 1.3
RO ACE Healer 1.5 2.5
Gaia God 2.0 2.0
Zhang Fei Phy / Atk 1.3 1.3
Guan Yinping Phy / Healer 1.5 1.3
Dark Knight Phy / Devil 1.5 1.5
Takemikazuchi Balanced 1.3 1.3
Yamatsumi Wood 1.5 2.0 1.5
Gran Reverse Dragon 1.5 1.5
Alt Dungeons I * None 1.5
Alt Dungeons II ** None 1.5 1.5

* Alt Dungeons I : Alt Castle to Alt Trailokya. 1.1x ATK and HP instead to High Metals Dragons.
** Alt Dungeons II : Alt Hypno to Alt Talos's. 1.1x ATK and HP instead to High Metals Dragons.

Saturday 27 February 2016

Orb enhance vs Row enhance

YES ORB ENHANCE!

I'm talking the seemingly useless awakening here - why do they even exist? Are they solely there for the 5-orb-1-enhanced LS? Why the Time Goddess keep getting these "shitty" awakenings on UEvo?



Recently, after tinkering with my NepDra team, I think I got the answer.

Understanding the maths behind rows and OE would open a new world to team building. So let's start with the maths.

At the simplest form, the damage calculation formula looks like this:


Here we are interested in the Row bonus and the OE bonus. TPA bonus is more difficult to handle because it's a card-wise damage boost.

Row bonus are often considered as the primary damage multiplier for many row teams because it stacks additive per activation. The actual multiplier is:


For example, a team stacked 12 row enhances and matched 2 rows. The row bonus would be 1 + [(12/10) * 2] = 3.4. Simple enough.

OE bonus is rather complicated. Each OE awakening increases the chance of a skyfall enhanced orb by 20%. At 5 OE awakening, every skyfall orbs are enhanced. But here we are interested in the damage bonus given by enhanced orb.

The calculation of enhanced orb consists of two parts - one part is dependent only on the number of OE matched, and the other part is dependent only on the number of OE awakening in your team. The actual multiplier is:



For example, a team stacked 9 OE awakening and matched 8 orbs with only 3 enhanced. The OE bonus would be [1 + (0.06 x 3)] x [1 + (0.05 x 8)] = 1.652.

From the formula, there are few interesting things:

First, assuming everything else stays the same, for each extra OE awakening stacked in the team, that awakening gives a less than 5% increase in damage and is diminishing.

Second, if you can slip at least one OE into a orb match while having 5 OE awakening in your team, you are guaranteed to have a least multiplier of 1.06 x 1.25 = 1.325.

----

So what give?


Assuming you can freely trade one row enhance awakening to one OE awakening (which in fact right it's kind of officially 1:2), and in an extreme case assuming you are choosing between a card with 5 row enhances and a card with 5 OE. Reasonable, right?

Since people like stacking rows to an absurd amount, I assume on average each of your cards holds 2 row enhances, so you have 10 rows enhances and 0 OE by default.

Now let's see the amount of damage you just gained from the 5 row enhances.

Row matchedmultiplier @ 10 rowmultiplier @ 15 rowincreased in damage
122.525%
23433.3%
345.537.5%

The damage increase from 5 extra rows enhance at 10 rows enhances ranges from 25% to 37.5%. Remember, the higher the number your original RE, the less damage you will gain from additional RE.

Now let's see what can you get from having 5 OE. 

With 5 OE all skyfall orbs (of that color of course) are enhanced. Now you are making one row without using orb changers - that means all 6 orbs in the row are enhanced.

For one row, the multiplier you get from OE alone is [1 + (0.06 x 6)] x [1 + (0.05 x 5)] = 1.7. It's a 70% increased in damage you gained from taking 5 OE over 5 row enhances.

In reality it is not possible to make two rows without using orb changers. Assuming you are changing 6 heart to 6 attacking orbs. On your board you have 12 attacking orbs, enough for two rows, with 6 enhanced neutrally and 6 not enhanced.

Mathematically it doesn't matter how you split the 6 enhanced orb as long as you have at least one enhanced orb per match while matching all enhanced orbs, you average OE bonus will stay the same, and it's 1.475 in this case, or a 47.5% increase in damage. For 3 row matched (12 not enhance + 6 enhanced), it's a 40% increase in damage.

For least increase in damage in all cases, that you have 5 OE awakenings, but only 1 enhanced orb per row, it's still an 32.5% increase in damage.

It's all better than stacking 5 rows, in almost any case, not to mention it's often hard to stack 5 extra row even with two flex slot; but you can stack 5 OE relatively easy.

----

So... stacking OE or stacking RE?


The correct question is, "At how many row enhances stacked should I throw in 5 OE awakenings?"

It vaguely depends on your team compositions and orb matching nature since OE bonus varies largely depending on the amount of neutral enhanced orbs in on your board before bursting with an orb change.

For example, NepDra can usually do two row-match at maximum and usually save water orbs. It's beneficial to stack a few OE before stacking further row enhances (assuming your OE subs helps you of course, not just blindly put OE awakening in your team).

Deliberately enhancing orbs in a row team does seem to work the effort. Stacking enhanced orbs do higher damage than spiting them into more rows at high amount of OE awakenings. For example at 6 RE and 8 OE, stacking 12 orbs together net you 1.6x from RE and 2.41x from OE, or 3.85x total, somehow doing similar damage to matching 2 rows at 14 row enhances. However for some row teams that can often match 3 rows, you could get higher damage from row enhance rather than stacking orbs together and enhancing them.

Nonetheless, getting good amount of skyfall enhanced orb is far more important than
just mindlessly pumping the amount of row enhances in your team. After all, the math is here and it's you (and your calculator/spreadsheet) to build a team that suits your play style.

----

Below are some useful cards to help you stack OEs

Nut & Osiris: even if your team is row based, a card 4 OE awakenings with worth a lot more than a card with 2 row enhances awakenings

Hathor: Might as well consider putting her in your Saria/Thor team for some passive damage boost

Nephthys: Actual damage boost from her AS is, at least for row team, 2.53.

Urd/Verdandi/Skuld: 2~3 OE along with a good AS and stats, as well as extra heart enhance that helps your RCV and keep your attacking orbs from heart-to-X skills enhanced

Ares Nova/Goetia: 4 OE, quick enhance with orb spawn.

Constellation I: 2 OE and 2 heart OE. Self comboing with a heart-to-X skill that keep converted orbs enhanced.

Monday 10 November 2014

700 days in PAD - how my team changes the game and how the game changes my team


700 days.

It is the 3rd game that I spent nearly two continuous years in. The first one being WoW, which I started at beta and quited at version 5.0. The 2nd one is Tera Online which I am still actively playing. And PAD is the third game.

Looking back into the box sorted by chronicles dug some memories from the past.

(ps. In fact I played PAD on the first day since release, accidentally deleted the apps without backup at ~200 days and started a new account again.)

2 years ago there were no egypt gods, no awoken, not much dual element/typed monsters and no ultimate evolutions. That rerolling wasn't a consideration when you create an account. Getting Zeus is from descends in the single direct goal when you start the game. It worth a few stones for your first Zeus run because 3x at full HP was the most powerful leader skill in the game.

My first and second roll is a Chaos Dragon Knight and Piedra.

Starting with dark monsters was a no-brainer. There were no super gold and tri-colors super kings alerts. Dark monsters were the one that can easily leveled up, no weakness element, and most importantly Zeus and Valk are light.

CDK's 2x HP leader skill helps me a lot in early dungeons. I can easily reach 16000HP for weekday evo mat dungeons when pairing with other CDKs. The Piedra was the key heart maker in my team. Pairing with ADK so my dragon team get 2xHP & 2xATK.



My first few rolls were made in a god fest + dark fest. Loki strengthen my way as a dark team. The volcano dragon and CDD helps me a lot to my dragon team. The time was around Christmas, I quickly grabbed the Bahamut from CD collab and BAM! My dragon team is ready (except my team cost).

 

My first "real" god fest didn't went well. Rolling a Hades forgo my need to run Hera Descends with the CD crystal team (did it for stone anyway). Echidna is also nice. I aimed for Horus or Ra but no luck, also rest pull were pretty horrible (remember there were no awoken skills). Luckily I got a few light subs for my future Zeus team (which never existed, will talk about this later).


My team at that time without a doubt was

 

With help from first wave of ultimate evolution and other farmable dark subs like KotR, Hadars, Lilith etc, most Legend stuff won't cause any trouble to me.

My team slightly changed when Hera-Is descend comes around. I remembered I ran the Legend few times without luck. At the last minute I entered Mythical with a few stones and got Hera-Is. 2.5x ATK boost to dark team without HP requirement was arguably as strong as Zeus. Hera-Is also have a similar gravity and it was way easier to skill her up. Immediately Hera-Is was the leader in my team - partnered with Loki to form a 1/5/2 dark team with two gravity. HP was a problem, so KotR usually replaced CDK in my team. Nothing can't be beaten with 2 gravity, DJJD+Vamp and Loki's enhance.

For a month I was going with my strongest team for various descends until


TADA!! HORUS!! (let's pretend he is there)

16x without limitation. I gathered my previously rolled stuffs and immediately my Horus team was:


Every stages look weak from this point. All technicals, bi-weekly and descends. The team was composed towards defense and RCV. At the time Hera-Is looks like a pretty stable subs for multicolor team, not only she takes up two color, she also has very all-round stats and a gravity. I started skilling her up and pouring +eggs to Horus then Echidna.


For the next half a year I stuck with my Horus team. I have no hype with the Chinese gods because they didn't fit my multicolor team and I won't use them as lead (3.5x restricted vs 4x unrestricted). Time mage sometimes replaces my Valk in my Horus team. DL Hades sometimes get into my team for double gravity. My team composition didn't change much from doing regular descends to farming KotG.


The next couple of month wasn't exciting either. In fact I found no replacement for my Horus team. I kept rolling God Fest and hope for a new leader that is on par of my Horus team but I don't have a strong desire towards one specific monsters. Didn't succeed in finding a new lead. In fact I kept rolling more subs for my Horus team.

Yomi and Izanagi. Most sought multicolor team subs. Not very useful to me because I don't need the one second time extend. Putting Izanagi into my team took away either my shield or delay. That didn't make good sense to me.



Lakshmi was in my team for quite a while because she takes up two "bad" colors in multicolored team. High RCV with time extended and on demand heals fits my Horus team very well.

I thought the two greek gods can help me in forming a wood team and fire team but they didn't. Take a look on the previous pics again and you will realize that I didn't roll good wood/fire subs.

I hyped too much with Okuni and Meimei/Izanagi/Yomi combo build. With all skill out I merely do same damage output as my Horus team - minus a shield, a long delay and most RCV - that said, it didn't turn out well but eventually I put Okuni in my Horus team for an extra one turn delay and HP boost (and he has all my +eggs). I admit that feeding +297 to him is a mistake. I would have better team if I fed my plus to Susano.

** On the point of feeding +eggs - I tend to give +eggs to monsters that can be used on multiple teams. For example Echidna, Okuni and LMeta. This is also a reason that I don't want to invest into RSonia team.

About RSonia - I don't have good use of her as a lead except KotG farming even now. First I don't like glass cannon team build (No Lu Bu yet). Secondly none of the subs from my Horus team (especially hypermaxed one) are usable in RSonia team. I don't want to spend too much efforts bumping a new team to do less things of my original team.


Ultimately RSonia, Ares and Leilan are in my Anubis team. :/



I always have great luck in rolling REM. I was going to roll for the MH Collab but accidentally got a LMeta from the other machine. She immediately replaces Lakshmi for the water spot on my Horus team - unbindable, bind remove and even higher RCV. Hypermaxing her was the number one goal. I also started collecting healer subs because of her.


I seldom roll in Collab REM. Foreseeing the 2x attribute enhance are useful, lacking of GOdin, BOdin and they are cats, I rolled a few times in MH Collab REM. It turned out very well. I didn't get duplicated cards - they have 10 monsters in the REM and I got 9 of them in nine rolls. Stopped. Realized I won't get the remaining one which I really want. GOdin Neko and BOdin Neko often used as stat sticks and skill boost in my Horus team. Gargwa stabled on my healer team with skill boost and 2x Healer ATK boost. The quick six turn damage reduction from Great Baggi often interchange with Susano as a substitution of blue subs.

Yes. Nearly everything I roll are for multicolor team.


Introduction of Sandalphon pushes me to build a healer team. Lacking of orb changers and HP makes me feel nervous in descends. My healer team was mainly used for speed farming. At that time it was like:


Two 297 subs from my Horus team and the transition underwent very well. After a few runs I realized that I need more light orbs to do damage. Apollo was a good choice even he is not a healer type (also high HP).

My fire team were formed and thanks to Wangren, and my previously rolled RSonia, Ares, Uriel - I got an amazing tanky and high damage output fire team:


The OPness of this team depends on the skill level of Wangren. Try it yourself. Sadly until today he is still SLv 3 (annnd the water one is SLv 8, god damn it).

I still don't had a water team and wood team. My dark team also far less useful than my Horus team at this point.

The reason I started building different teams was that I found Horus team lacking of burst damage to the descends / dungeons they released. More jammer orbs and binds reduced the chance of doing an attack (espeically jammer sky-fall). The requirement of burst in descend is also a problem. I could do higher damage with similar team leading with Ra. Horus sits in the middle ground of damage and reliability and makes me feel he is falling behind. Luckily my plus eggs investment still brings him up back to the usable level.

Plus the release of the 3k god favor increased raw stats (HP/RCV) and damage and ability to build monocolor teams. This makes me feel more nervous about the future of my Horus team.






And then I rolled a DQXQ within 2 rolls during the introduction of Chinese God.




I wasn't expected I would abandon my Horus team this soon. DQXQ is a solid lead as my light-based multicolor team. She has straight up better stats, awokens and active compared to FL Horus. Don't directly compare 16x and 25x! DQXQ team can convert two orbs to light, making rows and still having 16x multiplier. Luckily Hera-Sowilo comes two weeks later and now she is a stabled sub in my DQXQ team. My initial team was:

Trade off delays and shield in exchange for a immense big burst compared to my Horus team (Actually four continuous bursts.), DQXQ is the new star in my box. With Hera-Sowilo descends I also max skilled Hera-Sowilo and Apollo. This team made me feel I started wrecking the game again.


With the new UEvo of Thor, he (or she) essential substituted Valk in my DQXQ team, totaling the number of row enhance to five, giving me a three turn enhance that suits my three orb changers. With that I successfully zero-stoned Fagan (yes, first time). I also started doing descends with DQXQ / LMeta team but not Horus.

And then I unexpectedly rolled a ROdin. Now my fail-safe Twinlit team is ready except I needed a max skill RSonia. I spent nearly a month worth of Thursday to skill max her.


I didn't plan to roll in Saiya Collab - I don't have interest in the anime, characters and the cards in PAD except one - Athena(saiya) for my healer team. Rolled once, gold egg but not the one I want. Tried with his 2 TPA and purely subpar than Athena(PAD).







UEvo Athena is the monster that changes my way to build team and select leads. The damage of 2 TPA is the catchy point for many players, but I mostly like the 1.25x HP multiplier and unconditioned 3x ATK. I put most of the members of my DQXQ into my Athena team and found out I got 30k HP, 4k heal, abilities to do burst damage and heal on demand. Immediately it solves most problem my previous teams have.



This was the point that I started thinking to feed my Horus to someone else but I didn't because I still need Horus to farm the twin dragons and sacred masks.

The descends and collabs started from here got more and more devastative at this point - they hit larger numbers, preempt with annoying mechanism (reduce match time, skill lock), actively binding monster and orb conversion etc, which drove me away from using mulitcolor team. I switch nearly 100% to Athena team because of the stability she gives me (via HP, mostly). With my previous 2/4/2 → 14x fire team, I started my team build towards high raw stats and defense. My idea was "you can't die = you win".



Hyped with the new Indian God because they looked great in theory. Also the increased skyfall mechanism used by boss fascinated me. I think this is the second pack of stones I bought, solely for Vishnu and Krishna. The PADZ Fest was up at the same time as the God Fest. Rolled ~13 times?



The PADZ dragons were so-so. There were always better replacement. I'm not a big fan of dual-purpose skills like "enhance orb and 1 turn delay" or "change X to dark and a shield".

My wood team suddenly "appeared" with gZGL, Parvati, Avalondrake and Vishnu. Tried Vishnu but the team didn't quite work in terms of damage and defense. Also not a fan of controlling HP. Put gZGL in my Athena team but feeling kinda meh. Tried Athena team with wood row enhance + light TPA with Meimei and Avalondrake, but the defense sunk to the bottom.

Also realized the two major problem of Athena team - bindable leader (lose of hp multiplier) and poison orbs.


Thanks for gZGL by the way, now I got a better 100% resist skill lock team for sacred masks without the need of Okuni. If I could get rid of Horus, I can free up two 297s. Those green subs also made me a better KotG farm team although I wasn't farming KoG that much. I switched EXP farming to the coin dungeons. Hopefully Michael wasn't fully awoken...
 

Since my healer team was ready and I need a replacement for Apollo, together with the new buff to healer girls, I decided to roll the pink REM for Fuu for my healer team and Mitsuki for my fire team. I rolled five times and I don't want to mention the result again.

Nothing special about the Bikkuriman Collab. Rolled once for memorial, gold egg, "wow"ed and moved on.




My DQXQ / Athena team was doing fine. And the next God Fest was Egypt / Greek 1 IIRC. Every gods looked nice to have (Ra/Bastet/Venus etc). Rolled once and got something really unexpected.


Finally I can transfer my plus on FL Horus to somebody else. Athena was the top choice, but also realized she has a few weakness that couldn't resolve by giving her plus eggs, I hesitated.

I tried my first God Rush with my Athena team and found no major problem. I further hesitated giving her pluses. Oh by the way that is the first Zeus I actually have. Tried God Rush again with the new HP buff of LMeta for the Tamadra. Totally fine and much better than Athena team because of high recovery and utilities. Further hesitated in giving Athena pluses. Still no Tamadra though.

Rolled twice during PC God Fest. Got a BSonia which still never get in any of my team. Also Dragon Shougun. Looks like Gungho is telling me to build dragon team.

Everybody rolls in DBZ collab and my luck still continues. I don't need another multicolor leads god damn it. Cell is nice with TPA and extend time. I though I will be saving him for wood demon team that might or might not exist. However judging at that time I don't really want to build new teams that has less than 25000HP and 4000heal, I doubt I would make good use with him.

My teams didn't change for a few months because I am settled to the sense that "I need higher HP and RCV". At the same time the descends start getting harder, often over 25k per hit, but at the same time less OHKO mechanism. The game wants me to build tankier team and I go with it.

Rolled twice during JP/JP2 God Fest because my friends do. With duplicated Izanagi's and the ヒカピィ from Challenge dungeon I decided to max skill my Izanagi for whatever reasons.

New end-game technicals featuring the 4th end-game dragon was a bliss. Did it with LMeta team without even looking at the monster table. Further stabled my sense to HP and RCV.

The pink REM comes around I still fucking want a Fuu. I rolled once and my brain goes blank for a few seconds. I realized my water healer team skill need offensive subs like BValk and Sun Quan to be as effective as my LMeta team. Water LMeta can't solve this problem. So she has become the leader for my LMeta team when no duplicated monsters are allowed (coin dungeons!!!).



With the introduction of Egypt 2 I think it's time to ignite my passion to multicolor/combo team. I am a fan of orb matching but that doesn't help me to use my DQXQ/Horus team because of the survivability problem. All Egypt 2 Gods are good asset to me because of the HP/RCV boost and they are multicolor/combo team. I rolled a few times without luck. Got GOdin, Bastet and a few Egypt 1 Gods. I started from zero good cards on wood to having nearly all wood Gods in the REM. I really start making a wood team (still none of my pluses are on wood monsters).

Final pull from my friends for me was a fucking Hathor and I immediately fed my Horus to her. No hesitation. I also has two back-up Horus so I lose nothing. My original plan was to fed my Horus to Set if I got one, and my Okuni to Osiris. Well, color lead to color lead and combo lead to combo lead. I don't need two combo team or two multicolor team.

And this is my 700(+1year) days adventure of PAD. A truly good game. I don't think I will fed up on PAD soon nor regret spending that much time on a mobile game. I still have many contents to clear and team building to do, at the same time they are pushing contents faster than I can digest.

Good god damn game.

These are the current teams in my 10 team slots.


-----

If you ever wonder when did I got that magician from, here it is. I totally missed this portion of my box. I finally completed the old MH Collab cats and also rolled the only cat that I want from the MH4G Collab with 2 rolls. Jealous much on my luck? :p



Monday 2 June 2014

Board manipulation - 3

Cascade 101

What is a cascade?

It's an idea that instead of clearing the orbs in one single step, you (deliberately) separate them into multiple steps.

Like this:

Is that important? Sort of.

There are two main reasons for me to build a cascade. The first one is that building a cascade can save me a few move. Sometimes the positions of orb are restricted, so instead of arranging every orb to a "nice" location, building a cascade can utilize spaces and have less unpredicted move on the board. Typical cascade would be like this:



The second is that cascade can give you more combo from skyfall - it's about probability. It's useful when you are running Kirin team missing one critical orb, or using a Anubis team but your board only have 7 set of combo.

Utilizing a cascade - orb restriction

If you are playing a multi-colored team - congratulations! You should be fucking familiar with a situation that the three orbs you need to activate your leader skill does not appear in the position you want.

Consider this case.


I want to  match this three yellow orb. The conventional way is the move the yellow up at the bottom up. In fact, I can also bring the yellow orb down without moving them. If I can clear the orbs below them, they will drop onto the floor level and match with the yellow orb on the right.


Put it in another way, I moved the floor of the two rightmost column UP by clearing the two vertical combos.


This is the movement in action:



This can be done in many occasion to save you many times to move around the orbs. Note that this only add an extra way when you consider your path - if you can join the orbs FASTER and EASIER you should try to avoid this because it is easy to mess up the cascade and is logically hard to image the board in your mind sometimes.




Building cascade improve the chance of getting skyfall combo because step clearing the board mean that it checks the board line by line whenever a combo is eliminated. If a skyfall has 1/6 (~17%) chance to give you an extra combo, checking that 4 times raise it to ~50%. Not to mention cascade can also cause another cascade!

To build a cascade, it is better to start from the bottom because that checks more orbs when the cascade starts eliminating orbs. To begin, you can start by convert a normal one step combo into two, like:



Or move your next combo so that it cuts through your built combo: (I wasn't aware of a yellow orb at the bottom when I start moving the orbs so I have to deliberately move that away just to have a cascade, that's a usual situation that have to deal with when building cascade.)


Flooring

Building a cascade from bottom require a good imagination in your mind and observation of the board. I have a little trick that might help you to "imagine" how the step works. Think in this way - every time you eliminate some orbs at the bottom, the floor moves upward.


Let's say the first combo you made is the green orb. Now the "floor" moves upward, and is denoted as light orbs.






Now every orb at the floor will align after the green orbs are eliminated. We can safely place orb on the new floor because they must match and eliminated.





The floor moves upward as we placed another combo. We can continue this process to build a high step cascade.






And that is! This is the link to the above orb pattern.
http://pad.dawnglare.com/?patt=HHLLHHHLGDLHHGDBDHLDBRRLGBGGGR

While building a cascade, pay attention to extra orb the in the mountain of combos. Take a look the rightmost dark orb - that orb is not eliminated because the dark combo matched the leftmost 3 at the step of the blue orb. In most case this will break the whole cascade on top of the remained orb. Be very careful!!

The another key while building a cascade is diagonal move.



Most of the time you need diagonal move to put orbs into corners without disturbing the made combo. Like the red orb at the lower right corner in the sample above of from the dawnglare simulator. Usually I will leave the corner orb untouched and move other matching orbs (2 red orbs in this case) to build the cascade. Sometimes it is not efficient to do so - you may lose time that you can make 2 extra combo just to move the two red orbs down. That's why building cascade is not a solution for any board.

I think that's all I can think of right now. Please leave a comment for anything I missed or suggestions in building cascade/combo!

Happy cascade'ing!

Thursday 1 May 2014

Me, as a long time player, trying to tell you how descends are changing and why you should stop hype with some team.

Hi.

Let's talk about descends with numbers.

Descends are ways to acquire power monsters, but the difficulty of descends are unimaginable. There are guides all over the internet talking about how to build a team for which descends - which I'm not going to repeat those here. What I am going to talk about is a history of descends in PAD, and how are they evolving.

Let's get started by this list of descends, from the old to new:

Non-technical
  1. Hera
  2. Two-heroes
Technical
  1. Zeus
  2. Valk
  3. Goemon
  4. Hera-Is
  5. Satan
  6. Zeus-Dios
  7. Hera-Ur
  8. Athena
  9. Takeminakata
  10. Beelzbub
  11. Izanami
  12. Heracles
  13. Hera-Beorc
  14. Sandalphon
  15. Hera-Sol
* Maked in red are descends that only have Mythical difficulty.


Before we go ahead I would like to introduce you some terms I am going to use:

DPS - The average damage you are doing per turn, or average damage you are taken per turn. DPS of boss also means the healing you need per turn to sustain you HP. This idea is very important on choosing your member of your team, although your healing varies by your skill, number of heart orbs appears or others, but this number should give you a rough idea on the difficulties of the descend. Stole this term from most MMO game because I don't want to make new abbreviation.

Largest Hit - The biggest number you will take on that floor. This only accounts for hits that you are suppose to take (not the 430k damage that are mean to one-shot you).

Enrage - Significantly increased in DPS when the HP of the boss is lowered to a level.

Mechanism - Different skill used by the boss including shields, orb change, binds etc.


Now let me categories the list.

Hera, Two-heroes, Zeus, Valk, Goemon are Type I descends

The common set up the Type I descends is that, they all have floors that let you stall for skills. The DPS of those floors is high, but you can lower the damage by killing some trash mob on the floor. Examples are the knights and demons in Hera, ninja's in Goemon. To effectively stall on these floors you need to kill some but not all of trash mobs, and carefully match orbs to get enough heal and not too high damage to kill them.

The other common set up is that there is a mid-boss floor that serve two purpose:
1. Make you use your skill so you have less tool when you face the boss, and
2. Let you stall.
Examples are Mitsuki in Goemon and MKIII in Valk.


Hera-Is, Satan, Zeus-Dios, Hera-Ur are Type II descends


These four are the worst descends in PAD history.

They are hard. Just plain hard. Everything tries to kill you in one single hit.

You got basically first floor to stall for everything you need and you got boss rushes from the 2nd to 5th. There are heavy preempt strikes that needed to take care. They have preempt shields that put Echidna out of your regular team member.

The conventional way to do these is a high multiplier team. Horus, Ra, Kirin, Ronia, DMeta. Just burn everything and hope you aren't getting orb trolled. This is the descend which makes people think DMeta is great. In fact she is not. fucking. viable. in late game! If you plan to do type I and II descend forever, go and feed 297 to your DMeta and roll like crazy for Hanzo.


Starting from Athena, up to Hera Sol (the most recent one as I am writing this) are Type III descends

They are boss rush, and in terms of stalling, you could do that on the most comfortable floor depending on your team. But each floor has a mechanism that tends remove, or bring difficulties to one type of build from the face-roll list. Type III descends focus on annoying skills rather than just OHKO you.

You can always find few of the following mechanism to stop teams that do not need to puzzle from face-rolling.
1. First floor oneshot to stop 414 team, especially SoD Lucifer.
2. Preempt heal+oneshot to stop team that require <100% HP, like Goemon and DMeta.
3. Heavy preempt strike+damage per turn to stop multiplier that require 100% HP, like Zeus or Izanagi.
4. There is always a mid-boss that oneshot you in few turns to test your team DPS.

And there are three new mechanism introduced in the newer dungeons:
5. Reduced orb matching and constant orb change to hinder combo team.
6. Preempt skill block to stop Sonia and DMeta from press-skill-and-win on the first turn.
7. Combo shield to stop heavy row enhance team, again targeting Sonia and DMeta.


Now let's talk about numbers.

In fact it is hard to get everything in a analytical way because of the random nature of this game. I will leave out the trash mob parts and mid-boss part, and focus only on the "number" of the bosses.

Take a look at the HP of different bosses.

Most bosses have HP  at around 1.8m except for Takeminakata (tri-color) and Valk (seriously why she has so high HP in Legend?). The enrage HP is the mark of the highest burst you should do because most boss during enrage oneshot you. A shield works to buy you one turn (kushinada) or 3 turns (golems) or 5 turns (susano).

But solely look at the boss HP does not reflect anything.

Let's say I use a spike team and decide to melt the boss before they can touch me. Surely I bring my lovely Echidna for a 3 turn delay so I have enough time to kill them. Now take a look on the DPS I need.

I cut the graph at 600k DPS. Why? Can I really output 600k per turn if my team is just ready for Legend? I'm not talking about oneshoting the boss, I am talking about steady DPS. Surprisingly you can't easily melt Valk down before she lays her sword on you.

For type III descends, some of them have preempt skill locks that you can't just pop Echidna and start killing. Luckily the skill locks are short and they don't recast. But you still need to survive without skill over that period of time.

Talking about survivability, let's also take a look at the DPS of boss (aka the healing you will need per turn).

What? "You sure there is no typo"? Surprised by the drop of DPS at Type III descends? Have you ever think of how to take one hit from Hera-Ur when your Echidna is not ready without dying with your burst team? Why Type III descends has so low DPS? Are they that fucking much easier?

This probably related to the history of available gods.

During Type I descend released, the popular team team we have is 2/4/2, or Zeus 9x but HP restricted. With a 2/4/2 team we can easily hit 25k HP and 5k RCV. Now take this in mind and suddenly everything within Type I descends make sense. They have high damage per turn because your team halves their damage (2hp&2rcv = 50%damage reduction). We don't have skill up material so there are floors that let us stall.

And now high multiplier team kicks in. The Egypt god, 3x multiplier on Type, Skilled up Zeus 9x with two inherit gravity plus descent light team member. Here's are the Type II descend begins. Since high multiplier is facerolling Type I descends, they take the descend to a even higher level. "Yeah you like oneshot? Fine."

Type II descend has gotten to a point where 2/4/2 didn't really work because you don't have enough RCV to survive. Zeus didn't work because of heave preempt strike and shields that stop you from using Echidna. All you need to do a make a 9x or 16x team, put a King Slime, and melt through everything. These are the worst days in PAD history because people without high multiplier can't do any of these dungeons (keep in mind there was NO awoken skill and ultimate evo that tackles binds etc).

Now the descends evolve into Type III, where boss has low damage because people like spike team - and they adjusted the damage so that spike team can actually heal up themselves and survive a few hits from bosses. Does that mean 2/4/2 can faceroll the descend? No. There is also a floor that check your burst DPS. Burst DPS is the weakness of 2/4/2 - it is hard to deal large amount of damage in short time frame. But hey row enhance helps, orb change helps, the twin dragons helps! But you need to prepare for this. Does that mean spike team (combo team and Egypt god) still faceroll Type III dungeons? Hmm, my team - yes. What can kill me with 4.5k RCV and 2 heal skill + 2 delays? jk. Jokes aside, the orb changing and damage absorb is frustrating but not lethal. Hey they can be solved by YOUR SKILL rather the monsters active skill.

Type III descends focus on mechanism to deal with rather than a very short frame of burst you needed. They just use every mechanism to stop you from bursting them down in the first turn. What you are doing is heal and kill. You will be annoyed with skill lock, damage absorb (especially dark), per turn binds, random orb change, jammer and poisons, combo shields etc. In fact if your team is not an extreme spike team that have low RCV (by low I mean ~2000), depend heavily on skill activation and row enhance and extremely susceptible to mechanism, these dungeon is a bliss for you because you can actually feel the puzzle part of the game, heal, and attack.


Also the largest hit of the boss also falls significantly from a oneshotting ~16k down to manageable 10k as a team in Legend level. There is also a grace period before the enrage phrase of Type III dungeons so you can prepare for a large hit. Back in time where I am doing Celestrial Pole and accidentally lowered Leilan's HP to 15% and she hits you for 63k and I say "wtf skyfall fucking please stop now", this is a very solid change to the game to remove obsolete and frustrating mechanics.



So what are you trying to tell me?

I'm not forcing anyone to abandon their team and use a Horus. The descends are more and more welcoming to different teams. You can use 2/4/2 team, or 1.5/7/1 team, spike team, kushinada, no problem, you will not locked from the descends because you don't have a certain type of team now. But you need to prepare lots of different monsters with different awakenings, skills that takes you through different situation. The old guides that give you template for descends are out of date and you don't need the "required monsters from REM" just to go through these new descends (disclaimer: Legend only). The game will evolve and so does the content. You will be seeing different descend in the future, but there will be something in common:
 
1. New mechanisms will be introduced from time to time. If you think they just annoy you then your team is good. If you can think up of a mechanism that specially kills your team, you should start prepare a new team, now.

2. Team that faceroll by pressing skill will be harshly blocked. If you still plan to roll that Hanzo for DMeta just for descend, prepare to get locked 3 out of 4 descend because apparently YamaP doesn't like this.
 
3. RCV > ATK
3a. HP is also important, as a buffer to heart orb trolling, and in extreme case, soak one hit during enrage

4. Roll a Horus and Learn to combo, seriously. You can't rely on 3 row to kill everything forever because of the combo shield.

5. Can I do X descend? Just ask yourself if you can pull up 5k heal consistently for a few turns without skill. 10k heal for Mythical.


Last of the least: Mythical data - for those planning to do Mythical.

Boss HP
Average DPS required
Average healing requied - How can I not say RCV is extremely important?
Largest hit

Sunday 13 April 2014

Board manipulation - 2

Part 1

Hi all.

Somebody asked an important questions on building combos.

"How to line up combinations so that there is more space without messing up other combos?"

I suddenly realized that I almost never mess up my board when I am doing combo. I watched again on my video and I came up with a few ideas on how I didn't mess up with my board.

Making combo from one side to another

That's said, you align combo from one side and align another one next to it, and go on. This may not be a efficient way to do combo, but this technique is really handy when you don't want to think a lot each turn. You don't need to memorize lots of move pattern, but it is a challenge to orb moving speed, and on-the-fly reactions.

Take a look at the following:




I want to align them one on each column, from left to right, so that I won't mess up with the previous one when I moved on to the next one.

The purple one is a no brainer - just let them stays on the leftmost of the board. On the next column there is a heart and red cluster. Which color stays on the 2nd column is not really important because I just want them align vertically, one on each column. It looks like it is easier for red the stay on the 2nd column and heart to stay on the 3rd. But that depends on how you move the orb. In both ways the disruption to other orbs are minimal. I will leave the 4th column to the green orb. Whatever it takes just settle it there, and the yellow on the right most. 

And here we go:
You could have noticed there is a few problem:

1. The heart orb I have highlighted didn't align in the block I designated. Instead I used the heart orb below the yellow for my combo. If there is only three highlighted heard orbs, I would not able to complete my heart combo.

2. The position of the green orb is bad for them to align vertically because I have to move the leftmost green orb to the right, which my path will be blocked by the heart combo. 

3. Vertical clears cannot cause cascade.

Make sure the orb you need are on the "other side" of the board

If I am making columns from left to right, I will make sure the orb I used for next column are on the right of the combo.


I could do that but there no effective ways to move the green orb at the bottom when the dark combo is set.  Try it here, start with the yellow orb.


It is not encouraged because there are too many random movements in this large area. Most of the time cascade will solve these type of problems.

Without planning for cascade, never aligned a horizontal combo in the middle of the column

I have no extra words on this, just no.


"But if I really need to combine that three red orbs, what should I do?"

Put them as far away of the middle line as possible so you could have more space to move orbs around, or use minimum movement space to move the orb to one side.



Use chunks of orb that utilize all the spaces, they will align themselves magically

The idea is simple - say you wanted fit 4x3orbs into 12 spaces. When you aligned 9 orbs, the remaining 3 orbs will also be set in place. All you have to do is the move along the orbs so you didn't break the combos, and continue the rest of the comboing.

Take a look on the following:


12 orbs in 12 spaces. All I need to do is align three of them. It doesn't matter which goes first because if the orb movement stays within this rectangle there will be no extra orbs moving into the area.


It's a pretty simple move.

This also works when the space utilized is high enough. For example, 12 orbs in 15 spaces (that is half of the screen). If you align the combo from bottom (which is better than from top because it allows you to build cascade easily, please do this as a practice), the higher you are building, the less you need to move. Think of it like, in the worst case, you need to move 3 orbs in 15 spaces, and then 3 orbs in 12 spaces, and 3 orbs in 9 spaces so on.

It's better to see that in action (or try that yourself! This really works well!). Take a look on the left side of the board in the following:


I think that's enough for this week! If you have other ways to avoid breaking combos, please let me know! Thanks.

Next time I will probably talk about cascade building