The Final Fantasy Tactics Autobattle Tournament thread

Hey my team lost to the ninja, which I wasn't all that worried about, while the lancer I was afraid of did nothing. I appreciated the commentary on why you formed the team. Basically making it as unlikely as possible that you ninja would be dead at the same time as the lancer by making her tanky and slow(spending most of her time in the air).

Cool!

Anyway, I'm not making any changes for the loser bracket. Let it ride!

I think that's only going to be an issue when my Knight gets Stopped or Paralyzed; otherwise, my units are too fast for Jump to work most of the time.

I didn't get to finish my line of reasoning, but Setiemson for Always:Haste allows the Lancer to spend 2/3 of the battle airborne. I'm kind of thinking now that I should've put Speed Save on the Lancer instead of the Ninja, but that's a concern for later rounds at this point.

I have changelists from these people:
Torgo
reinoe (team)
Traumadore (no change)
PrivateJoker

Still need DeeMer, Solitayre, and Regulus. Well, I also need SPOILERS and REDACTED, but they don't know who they are yet.

Torgo did want to revise that Abandon unit; you may want to make sure he's done.

It's fine; we're not near cutoff or anything. I'm just trying to stay organized.

Well it's pretty clever because your strategy has relatively few moving parts and kind of exists in a blind spot of the game's AI. I think if someone is able to beat you it will be some really fun, improbable occurance.

These matches have been great fun so far!

I'll be surprised if his lancer jumps as much as he wants it to. I had that problem (not really much of a problem) last time.

Is there any possible explanation for why everybody else's All Math units just mess with faith but mine chooses the best possible spells with pinpoint accuracy?

I'd argue that Death is rarely a Best Possible Spell, and that you got stupidly lucky in landing it where Sol failed to connect with Raise 2.

But no, no reason aside from, again, stupidly lucky.

because it's impossible to defend against Faith/Innocence effects. Everyone else is protected from instant death, be it status protection or Escutcheon II + Abandon. Soli's monk has no such protection, other than from petrification. Your unit exploited an opportunity, because it was there + a little bit of luck.

Well, sure, but on two units with ~57% on each, it's a pretty dang solid option.

The Raise 2 was all stupid luck. That could very easily have gone the other way: damage split on Cersei -> missed Raise 2 -> Holy+Earth Slash, loss.

PrivateJoker's explanation makes sense. Those other guys weren't protected from Holy, though, were they?

Oh wow, Soli's team has zero M-Ev. Yeah, that makes Death a bit more attractive.

You'll also notice that units without much status protection (i.e. Batman) Also don't have very high Faith, or even very high HP. Stuff like status effects don't have a super high chance of hitting him, but his health is low enough that they can take him out with a handful of calculated holys, so they don't bother with status effects.

Soli's units have max Faith to prevent situations like the one that went ahead and happened to him anyway (i.e. Raise missing). His units are also well protected enough from damage that your guys opted for the non damage option that was available.

I went and amended my revision again. Much more modest changes this time. In retrospect, while the whole "role reversal" theme was fun, it would have been much better to have the Chemist to be the female with Chantage and the Knight to be male for the extra damage. Oh well!

My team changes:

Iris: Remove all cure spells

Change Robe of Lords to Chameleon Robe

Jepac: Remove Earth Slash

The Knight should probably have Abandon with that Escutcheon for 100% evasion from all things. Blade Grasp is good, but it's still less evasion, unless you're particularly worried about getting your shield broken.

Glen's Faith also doesn't need to be upped to 70. All that serves to do is make spells more effective against him. Punch Arts aren't dependent on Faith, so the chance for Opal to revive him isn't affected by Faith. The only thign that affects Punch Arts, to my knowledge, is the user's PA. I could be wrong on this point, so someone feel free to correct me.

Secret Fist is MA, but you're right otherwise.

Also Martial Arts, but that's implicit on a Monk.

I really look forward to a map change...

I'm more than a little late commenting on my own battle here, thanks for catching the lack of Math Skill in the first go guys!
You might be amused to know that I chose "run like a rabbit" for my Chemists AI as a direct result of that AI always sending the unit charging into the front lines during the previous tournament. Her job is to make herself the first target of any Throws, and Hamedo them.
The AI definitely favours Holy too much. No matter how low the odds were, Death would have landed long before the way the match ended via Holy spam.
I would be delighted if my secret Zombie strategy worked just one time, but it's a pretty unlikely combination of events.

This tournament is doing a great job of showing what the AI thinks, it's really interesting to watch.

Death wasn't an option, because that spell can be blocked, and Robin has a 100% block rate against magic. So no, no matter how low the odds were, it wouldn't have ended the fight any faster. The odds were 0%, she spammed Holy because it's the only spell that's unblockable. Robin seems like he's vulnerable to status effects, but he's really not, not from Math Skill anyway. Ironically, Batman is kind of the weak link in the Dynamic Duo. Robin is really gonna have to support the crap out of him if my team is gonna get anywhere.

It's not a secret if you tell everyone. And it almost worked for a little bit against my team. (I still wonder why my Knight didn't try to revive the mage.)

I wouldn't necessarily call what the AI does thinking.

So I bet I've played FFT like 10 times before, and I've never see the AI mistime a jump until now. Is that just because the speed values are so high with these levels/gear/haste that it starts making bad assumptions?

It's not that he missed a jump. The Lancer made the decision to jump because it was the best option at the time, and it was guaranteed. The only reason it missed is because the target died before the jump landed, which was just the result of the other unit making a decision that ended the target's life sooner. They don't plan 10 moves ahead like they're playing chess. Each unit just makes the best decision it can with the situation at hand. Things like it knowing how to time Jumps, and being really efficient with Math Skill may look like a well calculated master plan, but it's actually just a computer processing information faster than you, and making correct choices.

Right, I know the AI doesn't plan ahead, just missed that it was a valid jump at the time (I didn't realize the state of the target changed).

He did miss a jump at about 1:38 in Mogri v. Traumadore. Mogri's lancer jumps at Traumadore's wizard, who has time to Muramasa a dude in the face and move out from under the jump before it lands.

Right, that's the one I was thinking of, then.

What's up with that?

Huh, I guess I didn't notice that one. Just went back and saw it. Not sure what the deal is there, maybe the computer doesn't properly take Haste into account when it checks for that stuff?

That was my thought, but I think I'd have seen one missed jump at least if that were the case over the years, given that I love Time Magic (lancers are pretty uncommon in the campaign, though).

There's only like, one place in the storyline where I recall Lancers showing up. You really spend the entire game fighting Knights, and various mages, with the occasional Thief or Monk thrown in. Stuff like Samurai, Lancers and Ninjas only show up in a specific area, and then they never show up again.

The A.I. is completely bonkers when it comes to jump. You ever do an FFT run using just AUTOBATTLE? Yeah, you learn some stuff about the A.I.

Lancers will jump on their own their own teammates with if the fall asleep with no regard to the loss of life caused: Oh no my buddy with 200 HP fell asleep, better use a 550 damage jump to wake 'em up! A lancer will jump on a unit with a spell cast on it that will kill it. The A.I. will jump on an enemy unit that's going to die from another unit jumping on it. And of course if an enemy is about to die from jump the A.I. will kill a unit anyway, wasting ct.

The only way to get the A.I. to use Jump consistently is to go 1v1.