Guides

@Sylvandyr

Awesome! Adding this post to the list- it's the same build I found in Maki's Battlestones Guide, but with context, an explanation, and strategy. Also, whoops- wrote 126 instead of 56 in the Maki build, but I see you knew what I meant. And why would I want to 'double up' in the Ruins? I have a 126/59 farming team planned for the Ruins through Beds, so my Ghostbuster can focus on training.

Any preference in the Ruins for a Scratch Ghostbuster vs. a Meditate one? What about training in the Workshop- do you think this DEF build with a Regen mage is worth testing?

Again, I'm trying to cut down on the amount of testing I need to do. I'll be doing testing like you did for a 117 vs 115 Mire Flyer here, except including loot sell value. Thinking I'll post the results in a new guide.

@WyvernVenom - For some reason, I was under the impression that you wanted to use the same solo trainer in both Ghostlight Ruins and Harpy's Roost, so that's what I meant when I was talking about doubling up. If you were going to do that, then I think that Ghostlight Ruins stats are more important than Harpy's Roost stats, though you'd be losing out on some Harpy's Roost boss kills by not having separate solo trainers.

Scratch vs. Meditate

I prefer Scratch for solo trainers in general, and I also like it over Meditate in the Ghostlight Ruins, for a few reasons. Note that my playstyle is focused on maximum speed efficiency for training; depending on personal preferences, other people might not mind using the Meditate build since it's interesting and different.

1) Due to the Eliminate breath refund bug, you can only store up to two Eliminates max, so if you miss twice in a row, you're slowed down a ton already with the Meditate build. Even if an enemy is 1 Scratch from dead, or if you are missing only 5 Breath for Eliminate due to a miss, you have to stall 3 turns and risk being KOed while also wasting time. Having Scratch, on the other hand, kills that low health enemy or builds the missing Breath in one turn. (Note that one Scratch from death wouldn't happen in Ghostlight unless you're fighting a boss, but the missing Breath scenario is quite realistic since you Meditate up Breath in multiples of 30.)

2) Defend is bugged and doesn't reduce damage even though you see the defend buff icon. This makes stalling less appealing to me, since you'll be taking full damage from things. Ghostlight Ruins has many 4-packs and no 2-packs. There is plenty of physical damage in the mix as well. The Plague and Lightning dogs would be dangerous ones to miss on with a Water trainer (weak to both) or Nature trainer (weak to Plague), because if you have to stall against them, they can get their elemental slashes up against you.

3) Scratch animation is faster for me than watching lots of buff animations. Scratch also kills while building Breath at the same time, which I like because I prefer if every one of my turns is used to advance the battle. Also, if you run into a boss, I don't like the idea of having to re-do the stall again; it just takes up more time especially if you had just finished building, and it really hurts if your first round Eliminate misses.

Golem Workshop build

I prefer Aid in the Workshop, since I'm mostly concerned with minimizing time spent. Using Regeneration with a defensive build means that you'll purposely be spending more turns. Taking longer in battles also means that the enemy builds up Breath for dangerous attacks.

In my opinion, the 126 STR physical and 127 INT Contuse works well for training in the Workshop. I personally use two Ambushes on the mage, but you could play with 1 Meditate to see if it suits your preference. I'm pretty aggressive with my turn order because I know what I can get away with depending on the situation, and double Ambush works best for me.

@Sylvandyr

I was already aware that the Meditate build only has one attack, but I hadn't considered the fact that the buff animations take forever. The Defend bug, which I didn't know about, is the clincher. For the Workshop, the main attraction to the Regen build was the fact that the healer had more opportunities to attack, since they only needed to heal once a battle. I was already leery of the purposeful delay for bonus healing, but I'd forgotten that a lot of enemies have Slashes that need to be built up, and so taking forever invites not only more attacks but more elemental attacks, which is doubly dangerous.

I already asked about the Discipline/Ambush choice a while ago, and settled on double-Ambush so the mage could both eat the turn bug and buff/debuff at the beginning. However, that was with two other 25's, not as a training team. I have a bunch of extra Disciplines lying around, I guess I could try it out.

Thanks for the stat recommendations for the mage. There's another mage slot, though, besides the trainer- one that farms alongside two other 25's. I've settled on a 133-STR build, so they can Scratch-Eliminate anything except bosses and Rally-Eliminate the weakest ones (golems, squirrels, one of the vultures). They'd be slow, but a healer could probably keep two of them on their feet. I ran the numbers and discovered that if the mage's INT is 132, its Contuse can substitute for a 133-STR Scratch, due to the +7 Contuse damage and the fact that every enemy in the Workshop has more DEF than MND. However, this would make the mage nearly as slow as the fighters (43 and 40 QCK, respectively). Also, Maki recommends 126 INT for a farmer, spikenard 127, and Culex 129. How much INT should a farming-oriented healer have for grinding the Workshop alongside two 133/40s?

Finally, a tangent. You've published Google sheets with your Coli data, and I think my comparison of different teams' effectiveness would be useful enough to also warrant publishing. Google Sheets seems like the best option for ease of sharing and editing, but I'd rather not accidentally publish my username and email with it. Normally I'd experiment, but that's clearly not an option in this case. How would I publish a Google Sheet to everyone with the link in such a way that nobody would be able to see my personal information?

@WyvernVenom - Ah, I thought you were only using two trainer dragons for the workshop. With 3, I prefer two physicals and a single mage. Will you be running two different parties, one with a trainer mage and trainer physical, and the second with a farmer mage and two farmer physicals? I want to make sure that I understand before giving recommendations.

I've unfortunately never tried to hide my email when sharing a Google sheet. I'm not sure if it's possible, since the editor name shows up in full, so my suggestion is to make a separate Google account without your name in it, and use a gaming alias. I have several accounts just for spreadsheet sharing or testing stuff. It might be the only workaround.

@Sylvandyr

Yes, I'm trying to get one team of a healer and a fighter to train in the Workshop, and a separate team of one healer and two fighters to farm it. The two physical farmers will also be part of a team of three identical 133/40 fighters that can farm the lower venues (Roost and down) without a healer. If the most optimal healer build is the same for training and farming, one dragon could perform both duties, but I doubt this will be the case.

@WyvernVenom - You're right, the optimal mage complement for those 133 STR farmers would have different stats from the trainer. That said, I'd prioritize the trainer mage first if you have time constraints; any decently statted / stoned mage makes the workshop a breeze!

Now, back to the farmer mage subject. You don't need to match the STR scratch level with its INT. Instead, I recommend the Contuse to be merely strong enough to do damage that lets the Eliminate KO the enemy after the Contuse. You just need 129 INT for that. :)

@Sylvandyr

Thanks! That was what I meant by Contuse being able to substitute- I thought that it took 132 INT to allow a Contuse-Eliminate to KO every enemy. I looked back at the Excel document, though, and it turned out that I'd clicked the wrong cell and my Eliminates were based on INT. A 132-STR Eliminate and a 132-INT Contuse aren't enough, but a 133-STR Eliminate and a 129-INT Contuse are.

@WyvernVenom - No problem! I just used calculator mode of my tool to input the 133 STR and 129 INT and used the pre-saved monster defensive stats. It separates physical and magic damage based on stat automatically.

@Sylvandyr

Cool! I copied the HP, DEF, and MND rows from the stats-mode Workshop table and then inputted the damage formulae from the thread, so I wouldn't have to do Contuse and Eliminate separately for each monster and each INT value.

@WyvernVenom

I (finally!) have a second Level 25 healer, so I changed the Golem Workshop farming team to 129 INT / 50 QCK / 13 VIT (x1) + 133 STR / 40 QCK / 6 VIT (x2), Battle Stones Guide. It seems to be slower than the 3 fighters team (2.3 battles/minute); I'll try to do the battle timing before the Fire festival to see if there is a significant difference.