Guides

@Maki Thanks so much! I’ll fix those soon.

[EDIT]: Fixed!

@Maki I'm trying the reduced speed Mire trainer with Scratch and I like the turn order. (72 QCK is now wasteful thanks to the first turn bug being fixed.)

I was testing and didn't notice a difference in killability between 63 and 64 QCK, even for the higher QCK enemies. The important KO turns are the same. I think I'll tincture again and start checking the turn order from 60 QCK to see where the true cutoff is. It's probably either 62 or 63.

I recorded some tests on this spreadsheet.

63 QCK functions as desired for starting breath, up until the fastest Psywurm x3 pack. I want to know if 60-62 is enough. I'll test with 2x 5 QCK fodder and 2x 9 QCK fodder.

Update: I dropped all the way to 60, and although turn order changed a bit, the exact KO turns are the same for the Wetland Unicorn - Mossy Cerdae - Wetland Unicorn pack. (46 - 44 - 46). The cutoff for those QCK levels might even be lower than 60 but I haven't tested that far. Assuming no misses, Wetland Unicorn 1 KOs on turn 7, Wetland Unicorn 2 on turn 11, Mossy Cerdae on turn 15.

60 is slow-ish for the Psywurm - Salve Kamaitachi - Psywurm pack. When starting to build Breath, the important Salve Kamaitachi and Psywurm 1 KOs happen at the same time as 64 QCK, but the last Psywurm gets an extra turn before its KO. Also if not starting to build Breath (and just trying to Rally Eliminate), one Psywurm will get its turn before the dragon in the first group of turns because non-Hasted 60 QCK dragon starts with 2 turns before enemy instead of 3. I usually Eliminate the Salve Kamaitachi first anyway. (61 QCK is same as 60 for this pack, 62 QCK achieves the desired turn order on this pack like 64 QCK)

If people don't mind the slight turn order difference, 60 QCK is actually good enough even for the fastest enemies in the Mire. (Cutoff may be lower, not sure.) At this point I don't think I need so much VIT for shaving off that much more QCK, though, so I will opt for more QCK to not lose that last turn of the startup rotation.

Still searching for whether 62 or 63 is ideal. [EDIT]: 63 is better when starting up against Psywurms. You save some of the later turns for Psywurm x3 and Psywurm - Salve Kamaitachi - Psywurm. I prefer 63 QCK.

---

I noticed something interesting against bosses, which supports my suspicion that whatever determines whether the faster combatant gets multi-turns is affected by every participant on the field.

With Haste in the Psywurm - Salve Kamaitachi - Psywurm pack, my 64 QCK dragon gets her 4th turn before the first Salve Kamaitachi turn. However, against a boss (where there are 2 less enemies), my dragon only gets 2 turns after casting Haste. The frog moves after her 3rd turn.

Salve Kamaitachi is faster than boss (56 QCK vs 50 QCK). Also at 63 QCK I noticed turn order differences when I went into the battles without fodder (vs having the fodder present, though I was not at a computer to record those).

I did all the tests with 5 QCK fodder, but I'll have to do it again with 9 QCK. Very tedious.

@Sylvandyr

The difference between 63 QCK and 64 QCK should be the first Haste turn against Venomous Toridae x2 and Venomous Toridae x3.

62 QCK 56 QCK 52 QCK 50 QCK 47 QCK 46 QCK 44 QCK 43 QCK 42 QCK
1st round (extra Haste turn)

50x2-30=70 QCK
47x2-30=64 QCK
46x2-30=62 QCK
44x2-30=58 QCK
43x2-30=56 QCK
42x2-30=54 QCK

2nd round (no Haste, 2 turns)

44x1.5=66 QCK
43x1.5=64.5~65 QCK
42x1.5=63 QCK

@Maki For 63 and 64 QCK there may be slight differences in turn order, but the KO turns should be what’s important, though, right? If they both KO before enemy’s X turn, it’s all good.

I don’t Haste against 3x Venomous Toridae because I don’t start on them, and if I’m in Eliminate mode, I just Rally because either way you’re only getting out two Rallied Eliminates before the last one moves (on any 3-pack).

63 QCK is fast enough to Rally Eliminate two out of three against all 3-packs until the Psywurms (the fastest enemy). I think that the extra QCK isn’t valuable compared to the extra staying power.

I think it’s much more important to prioritize the KO turn rather than a Haste turn order. Whether I move before an enemy or after in the first round doesn’t matter much to me if the enemy doesn’t get KOed.

@Sylvandyr

Hm, I think I still prefer to be able to start on all physical 2-packs in the same way (there is no Scythe Kamaitachi x2 pack).

Venomous Toridae x2

64 QCK

Haste
Rally
Rally Scratch
Rally Scratch
Venomous Toridae 1
Venomous Toridae 2

Rally Scratch -> KO
Rally Scratch
Venomous Toridae 2

Rally Scratch
Eliminate -> KO


63 QCK

Haste
Rally
Rally Scratch
Venomous Toridae 1
Venomous Toridae 2

Rally Scratch
Rally Scratch -> KO
Venomous Toridae 2

Rally Scratch
Rally Scratch
Venomous Toridae 2

Eliminate -> KO

@Maki Ahhhh, I don't use Rally in my starting rotation for 2-packs, so that's where we differ. You lose turns by Rallying before you've built up Breath for Eliminate because Rally only gives 5 Breath, Scratch gives 6, which makes a huge difference on a startup rotation because you get to Eliminate earlier. (If I miss one Scratch at the start when starting on a 3-pack and the enemy is slow, I may use Rally right before double Eliminate to finish enemy 2 and 3, but not before then. Usually it's just Scratch KO 1 and Scratch Elim 2 and 3 due to availability + timing of Elim Breath.)

This is what mine would look like.

As you can see, VT 1 and VT 2 die "at the same time" as far as how many turns they get to go before KO. For both 63 QCK and 64 QCK, I get KOs on turn 7 and 10.

Venomous Toridae x2

64 QCK (mine takes 10 turns, Rallying takes 11)

1) Haste
2) Scratch VT 1
3) Scratch VT 1
4) Scratch VT 1

5) Venomous Toridae 1
6) Venomous Toridae 2

7) Scratch KO VT 1
8) Scratch VT 2

9) Venomous Toridae 2

10) Eliminate KO VT 2

[EDIT]: Actually, for 63 QCK, I observed a different turn order than was mentioned, see here. This was with 2x 5 QCK fodder and 2x 9 QCK fodder. Dragon gets Haste turn + 3 more turns in the beginning.

https://docs.google.com/spreadsheets/d/1-pV_AoR8i-uAPIecFShQ490OwKrfA4bJmDTd6XPgEgw/edit#gid=1314950190

63 QCK (mine takes 10 turns, Rallying takes 12 turns and lets VT2 attack extra)

1) Haste
2) Scratch VT 1
3) Scratch VT 1

4) Venomous Toridae 1
5) Venomous Toridae 2

6) Scratch VT 1
7) Scratch KO VT 1

8) Venomous Toridae 2

9) Scratch VT 2
10) Eliminate KO VT 2

@Sylvandyr

I don't think I can break the habit of Haste+Rally only for the 2-packs xD I refresh if I see the 4-packs, otherwise I start even on bosses, since I usually train to higher levels and I also like to switch venues.

I guess it depends on the play style, even at 54 QCK you can get the Haste turn against almost half of the monsters.

@Maki Hm, with more VIT, I can think of a few packs that I'd be happier to start on. I really like this build! It's a significant quality of life change since the first turn bug was fixed.

@Sylvandyr

Yes, it's also much easier to defeat the bosses.

By the way, I made this build for Boreal Wood and Crystalline Gala farming / training. I thought the low VIT would be an issue, but it works quite well and it's faster than other builds.

@Maki Interesting build, I'll need to check it out later.

I tested 63 QCK with Venomous Toridae x2 and got a different order than the one I previously listed. This time I used SD x2, 9 QCK each:

1) Haste
2) Scratch VT 1
3) Scratch VT 1
4) Scratch VT 1

5) Venomous Toridae 1 (Hit but not KO SD 2)
6) Venomous Toridae 2 (Hit trainer)

7) Scratch KO VT 1

8) Venomous Toridae 2 (Hit trainer)

9) Scratch VT 2
10) Eliminate KO VT 2

I'm going to retest that pack with 5 QCK fodder. It's possible that with this group, KOs of fodder might affect something too.

[EDIT]: Retested using 5 QCK Guardians and got the exact same order for 63 QCK.

Both 5 QCK x2 and 9 QCK x2 fodder resulted in 3 dragon turns right after casting Haste with 63 QCK for Venomous Toridae x2.

---

All these edge cases will give insight into the QCK algorithm. Wish I could look into it more this weekend, but I have a big work team meet from Fri - Sun. Probably gonna be a while before I sit and process the data to code algorithm alternatives.