Theorycrafting! (Because I like competitive metagames too much)

[metroid composite]

Mmm...yeah, I'm a little stuck on this one; let me try a few measures.

Who wins in the 5v5 versus match?

Well...ok, for starters, what equipment are we assuming for the Thieves?  The numbers I had above were Air Knife, but they probably actually poach Zorlin Shapes (108 damage) and probably do take the time to steal a Chantage from Meliadoul.

At level 30 with Thief Hats, it's 10 speed to 9, so Thieves go first.  The males run into a corner singing Cheer Song.  Now, normally Chocobos could keep just about any opponent from running for more than one turn by going into the center of the map...so...the Chantage user goes to the center of the map to intercept their obvious path, and waits.  I'm pretty sure their best reaction is to charge through her (and not attack her at all because that would waste CT).  Cheer Song resolves, giving an average of +2 speed to all Thieves.  The Thieves get the next turn before the Chocobos despite not waiting: 180/9 = 200/10, so they would have been CT-tied, but if Cheer Song hit them at least once (94% chance) then they'll break the tie.  However, if the Chocobos play smart, the Thieves won't be in range to hit them anyway, while simultaneously having nowhere to escape Choco Meteor next turn.  The males just wait (which continues Cheer Song).  The Chantage user attacks for...120 damage if she's had an average boost from Cheer Song (2HKO threat).

The Chocobos move in, three Choco Meteors kill one male, and two more injure another.  Now Angel Ring kicks in and the Thieves kill three Chocobos (making sure their full health/Chantage users do any attack that risks being countered--counter only really has a 53% chance to land between brave and class evade, but it's still worth being cautious).

So...now the two remaining Chocobos get their turn, and there are five Thieves alive, four of whom still have reraise, four of whom are are about speed 14, the one who was waiting for reraise being speed 12 (compared to the Chocobo's speed 9).  Oh, and most of whom have 7 move.

The chocobos permakill one thief, and bring another to the point of needing reraise.  But unless they can escape up some high cliff they die next turn (and even if they can, Steal Heart ignores vertical tolerance).

Yeah, even if the person playing the Chocobos is a dick and tries to stall for crystalization at this point, I'm not sure they can.

[Dark Holy Elf]

they probably actually poach Zorlin Shapes

Argh. Plagues have a ~1/6 chance of appearing in Bervenia Volcano (and looks like a similar chance overall in Germinas Peak), and then you have a 15% chance of successfully poaching a Zorlin Shape. If getting into each battle long enough to see if a Plague exists takes 3 minutes, and then successfully winning the fight (with your sucky party) and checking the fur shop takes 20 (including the 3), it takes an average of 3 hours and 53 minutes to poach a single Zorlin Shape. Fuck no to them. You can quibble with the numbers (I forget if we now have a more accurate read on that 15% figure... if it's 25%, for instance, then we're only talking 2:20!) but the point stands, Zorlin Shapes are terrible and should not be assumed unless we're utterly convinced they're necessary.

In this case, bear in mind the chocobos could spend those hours of poaching just levelling up as efficiently as possible, since they benefit a lot from that.

(Bitter Thief SCC memories returning!)

[metroid composite]

Hmm...checking the battle list, two of the four north-entrance battles have a Plague which always spawns, but only one Plague (for some reason, I thought there was a fight there with five plagues).

There's a Coillery battle that guarantees three plagues, but you can only fight that once (and it's still only half a ZS on average).

And yeah, Chocobos benefitting more from levels than Thieves is definitely something I've been keeping in the back of my mind.  In particular Chapter 2: level 10, Chapter 3: level 18, Chapter 4: level 30 is...probably not a remotely accurate description of a Thief SCC.  (I don't suppose you still have level numbers recorded for your Thief SCC Elfboy?)

I am going to pause to note that Red Chocobo speed is not ass good as I kneejerked.  6.8 at level 1 to Thief's 6.6.  85 growth to Thief's 90 growth.  That's...nearly tied.  The speed gap without equipment goes from 0.2 at level 1, to 0.8 at level 99.  Naked thieves will have the same speed as Chocobos for the majority of the time below level 50, which means any bonus from equipment is usually a flat out advantage, and Thief Hat giving a +2 speed edge is actually the norm (just not at level 30).

In the mean time, I want to take a look at the 1v1, with the already calculated level 30 stats because it has a different dynamic.

The maximum unarmed damage a Thief can do with Power Sleeve, Twist Headband, Bracer at level 30 is...140.  Nowhere near a OHKO.  Swapping out for a Thief Hat gives 96 (side note: turns out the unarmed punch numbers I quoted were Archer without accessory, Archer having one more PA at level 30).  If they can get one PA song off, that's raised to 117.  And hmm...Bracered Zorlin shapes deal 132.  (If they have Air Knives, they might need to sing once to get into 2HKO range).

But basically it's the same scenario: 3HKO vs 2HKO (unless you're willing to allow something really silly like Rubber Costume poaches).  Which is a reasonably even fight--If Chocobo counters and hits (53% chance), Chocobo wins (excepting doubleturn scenarios), otherwise Thief wins.  Thief isn't really fast enough to force a double turn (unless they start so far away that Thief can double-wait first turn without getting hit).  Chocobo usually can't cliff abuse or otherwise escape Thief.  So...1v1 (where neither Angel Rings nor Chantages work, and Sing is bad) is a pretty even fight.

Although...hmm, I wonder if I should take Bard seriously for 1v1, because drain is good in 1v1.  Level 30 Bard has 7 MA and 2 PA, I'm assuming Thief Hat, Power Sleeve, Bracer, so 7 MA, 7 PA.  This means 91 damage with Bloody Strings, which almost completely heals Choco Meteor damage and doesn't trigger counter.  Let's see...if Thief had 95 base HP at level 30, Bard would have...58, so that's 228 with Thief Hat Power Sleeve (still beats Red Chocobos...and more to the point, still avoids 2HKO).

Yeah, ok, humans win the level 30 1v1.  Pretty convincingly too.

(Definitely I should adjust the level higher, though; the question is how much...).

[Dark Holy Elf]

In particular Chapter 2: level 10, Chapter 3: level 18, Chapter 4: level 30 is...probably not a remotely accurate description of a Thief SCC.  (I don't suppose you still have level numbers recorded for your Thief SCC Elfboy?)

I know I beat Wiegraf at what was then a record low level, which I think is somewhere in the low 30's. I think I beat the game around the high 40's (maybe even 50?)... in theory I probably still have that file so let me know if checking it is useful.

[metroid composite]

Hm, right.  Wouldn't surprise me if the level 30 was lower on this challenge because you can have Sing in that fight.

(As for exactly what level you were at for Wiegraf...35 maybe?  That's enough to get you the 9 speed point while still using Twist Headband Bracer.  It gives you 106 base HP, which combines with Twist Headband+Power Sleeve giving 126 HP for 232.  Combined with bad compatibility you could survive two Lightning Stabs.  Your PA just hit 8, 15 with equips.  With bad compatibility that would be 132--that would take three hits I think?  (270 HP needed to kill him?)  With good compatibility 198--two hits.)

Yeah, that seems about right--both of these require setting up a double turn (so not possible if you're less than speed 9, where he moves first and instantly Lightning Stabs you and kills you in the same number of hits as you kill him).  Both of these require luck in terms of landing all your hits and not getting countered.  I'd hazard a guess that you went neutral or good compatibility, ran away for the first two turns, putting yourself in a location  on the third turn where you could double wait without getting lightning stabbed, and then baited him into using an earth slash while you double-waited?  Hmm...no, that's still not enough CT, although...actually, wait you don't even need to survive a Lightning Stab in this plan because you just double turn him and kill.

Ok, backing up, that opens up the possibility of sub-level-35 strategies.  With a green Beret and good compatibility you can still deal 144 damage (enough for 2HKO).  But if you drop a PA point to 7, that doesn't work since you only deal 120.  The 8 PA point happens at level 32, which lines up better with your "low 30s" recollection.

Ok, mystery solved, you were level 32 with Green Beret, Power Sleeve, Bracer, ate a couple earth slashes to set up a doubleturn, then ran in and punched him twice (probably hitting the shield one of those two times, and needing both counter attacks to miss).

...But this is getting off-topic.

I propose these levels:

1. Roughly match enemy levels (which happens to line up approximately with ~30 at the end of Chapter 3, and ~50 at the end of Chapter 4)
2. Stop doing no-accessory calculations.  Always use Bracers.  Yes, Angel Rings and Feather Mantles and Germinas Boots might be better than Bracers (in a nebulous hard-to-define way).  But Bracers give us nice, straightforward comparisons.

[Dark Holy Elf]

Also, I misread the battle list. Huh, 50% Plague spawning sounds completely wrong... my guess is battle 4 unlocks later in the game than when I was trying for it (which was the start of chapter 4), as I would buy 1/3 and I was just having bad luck. Still, the real limit is the rare poaching, of course, and the fact that you only have a shot at one plague per encounter.

Colliery has its own set of problems, namely watching the Beowulf/Reis plot each time, and the fact that you have to do three extra unnecessary battles to even get there.

(And yes, for the record, Level 32-34 was what I was for Wiegraf. Strategy is as you described, except that I was able to lure him to one place where he couldn't counter due to a 3 height difference, for one of the hits. Still only had about a 14-15% chance of winning IIRC.)

[metroid composite]

So...

Chapter 2 (level 15) (Default Green Beret, Power Wrist)
HP:
Thief: 162
Red Chocobo: 126

Damage:
Thief: 35 (40 with Headgear)
Red Chocobo: 56

Speed:
Thief: 8
Red Chocobo: 7

Chapter 3 (level 30) (Default Twist Headband, Power Sleeve, Bracer)
HP:
Thief: 227
Red Chocobo: 225

Damage:
Thief: 140
Red Chocobo: 100

Speed:
Thief: 8
Red Chocobo: 9

Chapter 4 (level 50, default Thief Hat, Power Sleeve, Bracer)
HP:
Thief: 320
Red Chocobo: 358

Damage:
Thief: 140 (176 with Twist Headband.  156 with Zorlin Shape)
Red Chocobo: 152

Speed:
Thief: 12
Red Chocobo: 10


--------------------------------------------

All Chocobo damage values are for Choco Meteor.  Choco Attack from a 70-brave Chocobo actually deals almost identical damage numbers at all three levels. (54/96/150 compared to 56/100/152.  Not what I expected from a parabola).  Admittedly, this is assuming average PA; since Chocobo PA is random, if you get a perfect Chocobo, it's 70/117/187 (compared to your perfect stat's Chocobo's Choco Meteor of 64/108/168).  Choco Meteor is still going to be the attack of choice, but against charging/sleeping enemies there's also the option of Choco Attack to get the 1.5x mult.  (Choco Ball is consistently worse than both).

The biggest difference from the old numbers is that the HP gap has now evaporated; Chocobos are even slightly advantaged at high level.  Thieves can switch around their equipment, of course, pull out the Angel Rings and have more durability (but a lot less damage barring Zorlin Shape).

Damage and speed...the problem is they're back and forth now; sometimes Thieves are slower but deal more damage, sometimes Chocobos are slower but deal more damage.  Let's look at an aggregated offence measure:

Damage*Speed

Chapter 2:
Thief: 280
Red Chocobo: 392

Chapter 3:
Thief: 1120
Red Chocobo: 900

Chapter 4:
Thief: 1680 (technically 1760 with Twist Headband)
Red Chocobo: 1520

Actually, you know what, why stop at two?

Damage*Speed*HP

Chapter 2:
Thief: 45360
Red Chocobo: 49392

Chapter 3:
Thief: 254240
Red Chocobo: 20250

Chapter 4:
Thief: 537600
Red Chocobo: 544160



So...the way I read this...you probably want to keep your Thieves decently well-trained until Velius (which is their strongest point in the game--Bracers and Twist Headbands and Power Sleeves are brand new in the store, giving them a temporary leg up over Red Chocobo stat growth).  And you want a whole lot of physical damage against Velius, regardless of range.  And then you...probably ditch your Thieves; yeah, they still have a stat lead, but the next 15 or so fights are not Thief-friendly anyway (lots of evasion and bossses that generally spit on melee attackers such as Balk and Elmdor, and fights that curl up and cry to Ignore Height).  You might want to completely break the Dycedarg fight with Fly Bards, but Ramza probably has Bard unlocked, and Red Chocobos can support that plan just fine (picking off non-Dycedarg targets so that you only need to cliff abuse Dycedarg while singing).  Alternatively, it's red chocobos: just spread the hell out and blitz him.  After that you have a bunch more shielded enemies, and then the final dungeon, and by the final dungeon you're on paper nearing level 50, fast enough that Hashmallum can't hit everyone, and just generally matching Thieves as beaters while having range, so you don't really regret dropping them.

As for the 5v5, the roles have semi-reversed;  Chocobos are so close to 2HKO that they can often swing it if they find some decent compatibility.  If Thieves want Angel Rings, and Thief Hats, they 3HKO (although with Air Knife they really need a round of Cheer Song to reach 3HKO range).

Chantage user plays point once again while everyone else Cheer Songs once (at these speeds two Cheer Songs are hard).  Chocobos advance.  Chantage user stabs one, everyone else is out of range.  Chocobos kill two people and injure a third, standing in formation to protect the already stabbed Chocobo.  Thieves reraise, kill one Chocobo, damage another; they've attacked six times only one of which killed, so taken an average of 2.5 counterattacks (one of which was on their Chantage user).  If Chocobos can still find favourable compatibility to hit, they now need 8 moves to wipe the Thieves' party (three Thieves needed two hits. Two thieves needed one hit.  8 hits total, counterattacks deal with 2.5 of those).  But compatibility probably isn't that nice, let's say 10 hits needed; ignore the Chantage user, 8 hits.  Ignore the counterattacks on her, 1.5 counterattacks.  So...6.5 hits.  You have 4 Chocobos.  Make sure the two raisers are dead for good, and drop someone else.  I was going to suggest moving...but that's probably a bad idea as it guarantees a doubleturn fairly soon, whereas if you wait on spot and Thieves need to move to hit the right Chocobo, you can avoid the doubleturn.  So...the Chocobo that's taken two hits gets fried, and the Chocobo that took one hit on the way in stays meatshielded by that corpse, and another full health Chocobo got hit twice.  So...remember that we had one full-health or close Thief that was at 2.5 Choco Meteors of life, one Chantage user who had taken one counterattack on average, and one Thief that revived from Angel Ring.  The Chantage user takes another 0.5 counterattacks, and the 2.5 Thief takes a 0.5 counterattack too.  Pretty straightforward: kill the reraised Thief, and drop the Angel Ring Thief.  One Thief reraises, they finish off one chocobo, take a swing at the other, and then the two remaining Chocobos get a full party wipe.

Getting back to the ban list...

Yeah...you probably keep one Ramza because you have to; he probably mostly sings, one female because of Chantage (assuming you think it's worth going to the effort to steal--which it probably is; Chantage users are very tempting AI targets when everyone else is spamming attacks from long range) and three Red Chocobos for pretty much the entirety of Chapter 4.  And...Chocobos are notably better in Chapter 2, and most of Chapter 3 (the non-Bracer parts) too.

Ban List:

1. Calculator
2. Summoner
3. Wizard
4. Chemist
5. Squire
6. Ninja
7. Time Mage
8. Samurai
9. Lancer
10. Priest
11. Oracle
12. Monk
13. Geomancer
14. Knight
15. Mediator
16. Archer
17. Dancer
18. The Infernal Spawn of Boco (Red Chocobos)

[metroid composite]

Bard, Thief, and a bit of Mime spice is all that's left now.

First thing to note: Mimes are not a total joke.  Everyone's going to have Sing, and having one Mime is actually a nice party upgrade (60% more songs).  Mime deals somewhere around 105-112 damage per punch at level 50, which is...really sad, admitedly, but if they punch once and Mime punch four times, that's 500 damage.  Their level 50 HP is around 220--which is worse than Bard, but not that much worse.  Their speed is disappointing (10; Thief gets 12).  But still: a 60% upgrade to Sing and potentially 500 melee damage are just attractive.

Their path to Mime isn't really all that painful.  Grab sing first, because Bard is on the way to Mime, and then Singing unarmed Squire/Chemist is really not that much worse than singing unarmed Thief, because their PA is pretty good without being in ban range.  They can even both equip Knives, and use them slightly worse than Thieves so it's totally legal.  Monk and Geomancer are the ones that will kind of be annoying, because everything about those classes is pretty much banned (most likely including axes--they're a lot better than knives).  But still, you can get through those classes via sit in a corner and sing, or just do propositions.

As always, though, two Mimes?  Don't be crazy.

Bard kind-of dominates most of the skills.  Move+3 > Move+2.  Sing > Steal.

Thief has generally better stats.  More HP, more speed, more movement.  Damage is a bit more balanced: yes, Thieves deal more raw damage via Bracer and unarmed hits.  (Although granted: pre-Bracer, Bard deals about equal damage to Thief or better--up to 104 with Bloody Strings.  Even post-Bracer, if you decide to go Angel Ring over Bracer, Bloody Strings will tend to deal slightly more damage.  (Until Air Knife shows up, and all the sudden knives aren't laughable).

The big reasons not to go Bard are: you might be female (for Chantage and Steal Heart usage) or you might want to stick to good PA growth for slightly better blitzing.

The big reasons not to go Thief are: you're heading to Mime, or you're in Chapter 2/early Chapter 3 where harps are better than Thief weapons (but you'll still probably end up in Thief eventually).

Man, this is a mess too.  Most people are just plain going to use both: Bard for skills and Thief for stats.  The exception being one female and one Mime.

EDIT: although...if you are running a female for Chantage, she goes to Thief and just sits there forever (because she has nothing better to do) easily getting everyone spillover for Move+2 and Steal Heart, even if they beeline straight for Bard and ignore Thief at the start.  So...there's probably a section of the game where everyone has Move+2.  Certainly the Mime candidate will probably never stop and get Move+3 when he already has Move+2 from spillover.

[Dark Holy Elf]

I think you may be underrating Steal Heart a bit. Sure, Steal Heart is mediocre... when you have the ability to blitz safely and effectively. The Thief SCC doesn't, nor does the Bard SCC, and I doubt the intersection of the two (with Mimes lategame maaaaybe) is much better. On the other hand, Steal Heart is very dominantly useful when your slugging ability is poor. (As such I don't think you'd only have one female, for instance.)

My gut feeling on this whole comparison, is that:

(a) Thief > Bard for a chapter and a half by virtue of existing. Yes, you can grind to unlock Bard or whatever, but Thief doesn't have to do this, and chapter 1 thieves don't incur enough resets for bards to "catch up".

(b) Later on, thief still provides Steal Heart, and up 1-3 Chantages depending on taste. Now, Sing beats Steal Heart, sure, but I don't think it's a complete slaughter by any means... which one is more useful does vary battle to battle. I figure this allows Thieves' superior HP and move, as well as Chantage possibilities, to give them a clear enough victory overall.

[Grefter]

I think this is time for me to note that this is a thing of beauty once again.

You might want to completely break the Dycedarg fight with Fly Bards

So very very delicious.

[metroid composite]

Before I dive in, a few corrections on past matchups:

On Mediator vs Archer: I totally overlooked this at the time, but Mediator was literally the only class left that could use Robes.  Between White Robes and Chameleon Robes that's actually a pretty big deal.  Further cements Mediator as the correct choice there.

On Archer vs Dancer: I listed Archer as having the higher damage physical attack, but in reality I suspect that honour actually goes to Dancer (at least...post-Cachmere pre-Gastrifitis...although they pull into first again post-Gastrifitis if you grab the common-poach Ryozan Silk).  I don't think this is enough to tip the balance--everyone still uses Charge, Concentrate, and Arrow Guard all over the place.  Archer still has the best HP, making them an attractive carrier.  Archer still brings Shields, which break some fights.  Archer still has range on their damage.  Dancer still has the gender restriction.  But...it certainly makes things closer.

(a) Thief > Bard for a chapter and a half by virtue of existing.

Yeah, I'm a little hesitant to go along with this logic for the following reasons:

Suppose the two classes left were Thief and another high tech class--let's go with Samurai.  Would Thief win that just because they win almost half the game "by default" and are still somewhat relevant during the rest of the game?

Let's go with an example from another angle.  Let's say the two remaining classes were Archer and Bard.  You'd go to Archer first, grab a few charges, think you yourself "should I start unlocking Bard yet?" and then conclude "naaah, let me get Concentrate first--that's going to help regardless of class".  Then when you had Concentrate you'd go "Ok, should I start down the path to Bard now?" and then probably react "wait, wait, I'm going to want Arrow Guard anyway, and it's going to make a huge difference--let me get that first."  And then, only after you've gotten 1500-2000 JP worth of abilities, do you go "Ok, sure, I guess I might as well unlock Bard now."

By comparison, when it's Thief and Bard, you...grab Steal Heart, and then ask yourself "ok, should I go unlock Bard now?"  and...the answer is...probably yes.  What are you sticking around for?  Move+2?  You're only going to replace it.  Secret Hunt?  Not relevant until Chapter 4.  Maybe get it on one person just in case you run into an Ahriman at Fovoham Plains or Zigolis Swamp or Zerekile Falls or Barius Valley or Mandalia Plains (Ahriman have Air Knife as their rare poach).  But you don't need Secret Hunt on more than one person for the random Ahriman.  Caution?  Meh, unlocking Bard will probably help you more.

Add to this, what's the best time to unlock Bard?  Probably Chapter 1 when you can lean on guests.  At least if we're going by SCC-style guest rules, Chapter 1 has likely become the easiest Chapter.

I guess what I'm saying is this: the competition in Chapter 1 isn't between Thief and Bard, it's between Thief...and heading down the Bard unlock path.  And I'm actually feeling like the Bard unlock path is the better choice.

This doesn't exactly mean that Bard wins Chapter 1.  The best analogy I can think of...let's say in a normal game of FFT you learn Magic Attack Up pretty early in the game.  You're still going to use Gained JP Up in 75% of fights, but you might pull out MAU in the 25% of fights that are actually hard.  Similarly, Thieves with knives and Steal Heart are still better than Mediators with knives and Steal Heart, so you might pull out the Thieves in a hard battle.  But it's only a slight upgrade, so you'll try to squeeze by with Mediators whenever you can (much like Gained JP Up).

Chapter 1 feels like a wash to me.

I think you may be underrating Steal Heart a bit. Sure, Steal Heart is mediocre... when you have the ability to blitz safely and effectively. The Thief SCC doesn't, nor does the Bard SCC, and I doubt the intersection of the two (with Mimes lategame maaaaybe) is much better.

Actually, they are rather a lot better.  If you can run away for two turns (and you have 7 move: you almost always can) then Battle Song goes off three times per user.  15 Battle Songs is an average of +7.5 PA...but it's actually better than that in practice because it's random and fists are quadratic.

In Chapter 2, when your PA starts at 7, using Battle Song for two rounds (3 activations) raises you from 35 damage on average, to 147 damage on average.

In Chapter 3, when your PA starts at 14 (thank you Bracer), using Battle Song for two rounds (3 activations) raises you from 140 damage to 325 damage on average.

What if you only have one turn of prep?  If you're speed 6 you can still pull off two activations of Battle Song in a single turn.  Still enough, in Chapter 2, to raise your average damage from 35 to 99.  Still enough, in Chapter 3, to raise your average damage from 140 to 251.

You can't really do this on the Bard SCC.  Or rather, you can, but you need to spend a couple turns just to bring your PA to Thief levels, and THEN you can start making yourself powerful.  There's not much point to doing this when you have Red Chocobos--one Battle Song promises ~70 future damage (spread across five units) while keeping your distance from the enemy.  Choco Meteor deals ~70 damage right now, also while keeping your distance.

On the other hand, Steal Heart is very dominantly useful when your slugging ability is poor. (As such I don't think you'd only have one female, for instance.)

I'm actually entertaining the opposite possibility: whether 0 females is the correct party composition.  In any fight that isn't a total blitzfest where you can run away for at least one turn, I'm pretty sure Battle Song fists becomes the dominant strategy.  Females can't sing, are worse at punching, and don't get Move+3.  All this for a slightly better version of Steal Heart?

(b) Later on, thief still provides Steal Heart, and up 1-3 Chantages depending on taste.

Yeah, I'm not inclined to give a whole lot of credit to the 4-Chantage party.  It's something you only get by beating the deep dungeon, and then repeatedly fighting a battle that spawns only one Porky on average.  And then you get to use this party in...six whole story fights.  I mean, yes, it makes a decent case for being the strongest party you can field in those six fights (even though it can still lose).  But it's not really centralizing.  Now, what it DOES stop is anything else becoming too centralizing during the last six battles, because "grind for 6 hours and get Chantage Cheese instead" is always a viable counter-option.

[Dark Holy Elf]

Right, I was thinking that Uribos got poached for Chantages for some reason. My bad, that does lower my opinion of thief some here. Replace "1-3 Chantages" with "1".

I do think you've overestimating the number of battles where you can run away for two turns. I've done a Dancer SCC (which doesn't have the crazy move, but does have a faster-working gimmick) and watched a fair bit of the Bard SCC and I recall quite a few battles where the enemy starts in your face with nowhere to run. It's not a majority, but it certainly keeps this strategy from being dominant.

As for females vs. males, it really depends. Female steal heart is not "slightly" better, as enemies skew male to a ridiculous degree (there are 17 battles in which all charmable enemies are male but just 3 battles in which all charmable enemies are female, excluding optional fights and those before Sweegy). By far the worst of this skewing happens in chapter 4... which is also when Sing is at its worst.

On the thief SCC, I used three females and I think I ultimately decided this was fairly optimum. Fists only really supplant knives in late chapter 3, and I had enough firepower to kill Velius reliably anyway (Wiegraf was the sticking point), while charm makes so many fights winnable it's not funny. Of course, this isn't the Thief SCC; males have Sing and Power Song boosts fists much faster than knives. Three females is unlikely to be optimum. However, it's almost certainly 1 minimum. Chapter 4 is already a potential sticking point for the all-male team due to a lack of enemies to charm and Sing depreciating, tossing in one female not only ups your charming power, but it makes the whole Sing strategy far more viable due to Chantage. Think about it, the singers run away while the Chantage-user charges forward to distract some enemies from chasing the males, and even charms some of them if you're lucky. This is a pretty obvious improvement and I don't think "slightly better Power Song -> punch" strategy in chapters 2-3 is going to offset this.

[metroid composite]

Three females is unlikely to be optimum. However, it's almost certainly 1 minimum. Chapter 4 is already a potential sticking point for the all-male team due to a lack of enemies to charm and Sing depreciating, tossing in one female not only ups your charming power, but it makes the whole Sing strategy far more viable due to Chantage. Think about it, the singers run away while the Chantage-user charges forward to distract some enemies from chasing the males, and even charms some of them if you're lucky. This is a pretty obvious improvement and I don't think "slightly better Power Song -> punch" strategy in chapters 2-3 is going to offset this.

Yeah--you probably do want one Chantage user in Chapter 4 to distract the enemies.  (And I was thinking about this earlier: yes this character uses primarily Steal Heart because it does a better job of keeping enemies from targeting your males).

Food for thought, though: do you train a female the whole game, or do you recruit one after you've stolen the Chantage?  It's not like it requires levels to wear a Chantage and a Green Beret.

More food for thought: it is a surprisingly big drop in offence to do this.  One occurrence of Battle Song is kind-of like Elemental--70 long range damage that you can follow up with melee.  Except instead of being 70 damage now, it's really +14 damage to everyone's punches in the future.  Now, it's not like a Chantage wearing female can't make use of Battle Song at all, but they're like Bards--it's maybe half as effective.  Also, Chantage users are supposed to be suicidal and draw fire, so they spend a lot of time dead, missing songs.  So...the boost is +14 to four party members, and maybe +4 or so to the Chantage user.  A boost of 60 per Battle Song resolution.  Add to this, by using a female you have a lot fewer songs per turn, especially if there's a Mime or two in the mix.  With Mimes, removing even one singer potentially takes you from 9 songs per cycle to 6 songs per cycle (although 8->6 is a more probable setup).

So...in order to be break-even, a Chantage user has to basically buy the males 50% more time...which I'm pretty sure a Chantage user can do.  Hence I still lean on the side of advocating Chantage use.  But it's not actually a total no-brainer.

I do think you've overestimating the number of battles where you can run away for two turns. I've done a Dancer SCC (which doesn't have the crazy move, but does have a faster-working gimmick) and watched a fair bit of the Bard SCC and I recall quite a few battles where the enemy starts in your face with nowhere to run. It's not a majority, but it certainly keeps this strategy from being dominant.

Fights where enemies start in your face need to be handled completely differently, yes.  I haven't really covered those yet.

If it's early in the game (before PA boosting) you...probably still battle song, even if you're standing right next to the enemy.  It just deals so much more damage in the long run.

If you're in most of Chapter 3, you...equip Harps.  They flat-out deal more damage than Knives, and if an enemy attacked you before you could move, you get to heal the damage with Bloody Strings.

Not sure about Chapter 4.  Air Knives deal slightly more damage than harps, but don't drain.  I'm not sold on Bracer in situations where you get attacked before you can move (because you really want Angel Rings to keep your timers sane), which mostly rules out fists.  It should also be noted that Angel Rings have a fantastic interaction with Bloody Strings--revive and instantly heal half your health instead of immediately dieing again.  And yes: Steal Heart is also an option.  In the event that someone does become dead and the clock starts ticking, it's not a bad idea to Cheer Song if you get a chance--if your party is speed synced it will get roughly half your party an extra turn before crystallization.

Point is, I agree: in-your-face fights require often completely different strategies.  But...a decent amount of the time they're Bard strategies, not Thief strategies.

chapter 4... which is also when Sing is at its worst.

Ehh...as long as you're speed 8 (which you will try to be for a lot of Chapter 4--don't want to be faster than the enemies or you'll get charged on) you can do 3 Battle Songs in 2 turns.  Which is pretty much what you've been doing ever since Chapter 2.  The amount of bonus damage you get out of each Battle Song scales with PA (thanks to the quadratic formula) so it scales up too.

And...Mimes are possible in Chapter 4.  Wanna know a class that uses Battle Song-Punch arguably better than Thieves?  Mimes.  They nearly double the songs.  They have martial arts and concentrate, and therefore benefit more percentage-wise from each Battle Song.  They can sing and punch at the same time.*

*Let me give you a hillarious example of singing and punching.  Picture Poeskas Lake.  Now picture taking three Singers and two Mimes to Poeskas Lake.  The Mimes punch the two Revnants to death, while the singers huddle in the corner with the Mimes and sing.  Nine songs per turn, WHILE dealing with the enemies who start in your face.

**Ok, granted, that's not necessarily the ideal Poeskas Lake strategy, since you really want to poach the Revnants so that they don't revive, and super-frail Mimes potentially getting countered, even by Revnants, is a little risky.  And a Chantage user can get the mages to waste their MP.  But damn-it, the concept is hilarious.

--------------------

On a side note, I'm suddenly tempted to do some calculations on whether a couple level up-down cycles are worth considering as an option.  You have Bard.  You have Mime.  Three 1->10->1 cycles takes...2 hours maybe?  LESS THAN ZORLIN SHAPES.

You know what, let's do this!  Let's say you start as a level 40 Mime with...

HP: 184
PA: 11

Levelling down to 1 drops this to...

HP: 62
PA: 7

1->10

HP: 142
PA: 9

10->1

HP: 98
PA: 8

1->10

HP: 225
PA: 10

10->1

HP: 115
PA: 9

1->10

HP: 355
PA: 11

At such a low level, you'll probably gain 10 more levels before the game is over so...

10->20

HP: 577
PA: 14

By comparison, if you had used that time to just gain levels you'd gain...about 1/5 the exp because you're not close to level 1, although you'd at least be gaining exp in every fight, so gainin 10 levels could have gotten you...

HP: 224
PA: 12

Granted, you have a bit more speed in this form, but...you're a freaking Mime; you don't need speed that badly when you're miming stuff.

...And there you have it: level up-down.  PROVABLY BETTER THAN ZORLIN SHAPE.  Actually, just to add insult to injury--Zorlin Shape was closer to 3 hours; if you squeeze in one extra up-down cycle to match Zorlin Shape time...

Stats at level 10

HP: 559
PA: 13

Stats at level 20:

HP: 909
PA: 16

TAKE THAT, ZORLIN SHAPE!

-------------------

Anyway, umm...I'm not sure how serious I am about level-up-down as a strategy.  Well...I think it's more serious than Deep Dungeon Chantages or ZORLIN SHAPES.  I also think that if you feel the need to do heavy grinding, that it's likely better than pure level gain...but only on one or two people (because you need to keep some of your party high-level so that you can level up in two actions).

[metroid composite]

And for the curious--I asked myself "who should get more credit for level-up-down, Bard or Mime"?  Turns out, Bard.

4x 1->10 cycles down Bard, up Thief:

HP: 192
PA: 10.5
Speed: 7.4

4x 1->10 cycles, down Thief, up Mime

HP: 184
PA: 9.3
Speed: 6.8

Admittedly, down Thief, up Bard does gain slightly more PA per cycle, and significantly more HP per cycle, but it starts way behind since the initial 40->1 in Bard is a big head start over 40->1 in Thief.  PA will never catch up--down Bard, up Thief hits 99 before down Thief, up Mime.  Down Thief up Mime will win in HP after six cycles, and max out after 12 cycles (compared to 13 cycles for down Bard up Thief).  But...yeah, that's the only way in which down Bard up Thief isn't strictly better than down Thief up Mime.

(Not that it matters: down: Bard, up: Mime is blatantly the strongest up-down combo here, getting about twice the HP and PA of the other combos per cycle.  And which class gets the most credit for it doesn't really matter).

[Dark Holy Elf]

Late, but

Suppose the two classes left were Thief and another high tech class--let's go with Samurai.  Would Thief win that just because they win almost half the game "by default" and are still somewhat relevant during the rest of the game?

They very possibly would! The reason you kneejerk Samurai > Thief (and indeed, it is better in the vast majority of circumstances) has a fair bit to do with oodles of other classes doing the "easy to unlock, unimpressive fighter" better than thief, but a metagame that lacks all of those makes the comparison, at the very least, a far closer one. Samurai may still win that comparison, but it's far from a foregone conclusion and would merit some pretty serious analysis.

[metroid composite]

Late, but

Suppose the two classes left were Thief and another high tech class--let's go with Samurai.  Would Thief win that just because they win almost half the game "by default" and are still somewhat relevant during the rest of the game?

They very possibly would! The reason you kneejerk Samurai > Thief (and indeed, it is better in the vast majority of circumstances) has a fair bit to do with oodles of other classes doing the "easy to unlock, unimpressive fighter" better than thief, but a metagame that lacks all of those makes the comparison, at the very least, a far closer one. Samurai may still win that comparison, but it's far from a foregone conclusion and would merit some pretty serious analysis.

Hmm...Samurai may have been a bad example of what I was trying to demonstrate, because they like being female, and really appreciate Move+2, and might use Thief as a carrier, and will use Steal Heart for the vertical tolerance.

Different example: Suppose the last two were Ninja/Thief, and Ninja unlock requirements were changed to include job level 8 Squire (so that Ninja unlocked about half-way through the game).  Thief is now the only available class for 50% of the game, and then you almost completely ditch the class for the second 50%.  Does that mean Thief auto-wins?  The point I was trying to make was...here is a scenario where Thief is the class of choice for 50% of the game, but your party sucks for the first 50% of the game, and your party rules and smashes all the fights for the second 50% of the game.  I think that should be taken into account.

Furthermore, even in the first 50% of the game, if a fight is easy, you should be grinding in Squire to unlock Ninja.  If a fight is so hard that you literally can't beat it, your obvious grind path is also "unlock Ninja".  The only time you shouldn't be ignoring Thief and unlocking Ninja is when you really need juuuust a little bit of extra speed/mobility/damage to win a fight, or when you're learning Thief abilities.  (Which to me means that Thief isn't centralizing in the early chapters, because there is an obvious alternative--namely Squire/other Ninja unlock classes).



Anyway, current thoughts on how centralizing various stuff is.

1. Sing (It kills any fight where you can run, any fight pre-power-sleeve-bracer where it effectively deals more damage than your melee, and is the strategy of choice in several other key fights including Wiegraf/Velius, Dycedarg/Adramelk due to being split fights, and Altima/Altima due to being ridiculously long, and these are three of the traditionally nastiest fights).
2. Thief stats/equips (The best combo for sing for 80% of the game.  Stronger than Mediators before sing.  Sometimes the choice in assassination missions).
3. Move+3/(on some people Move+2) (It helps the primary strategy AND the blitzing strategy, and is a huge slap in the face for Mime usability)
4. Bard stats/equips (When you can't stop and sing and need to blitz now, Harps spend a lot of the game dealing better damage than Thieves, and there's the whole 3-range, draining, and vertical tolerance bonuses even after daggers pull slightly ahead).
5. Steal Heart (Even people going pretty much straight to Bard probably detour long enough to grab this...but past that initial 150JP rush I just don't feel it's centralizing inasmuch as it doesn't crowd anything out.  It doesn't dictate your equipment/class like Sing strategies.  If you're grinding to Mime in Squire/Chemist, Steal Heart gets crowded out by Sing.  If you're a Chantage user, you're only one character--you don't stop the rest of the party from using anything).
6. Catch (There's probably a few fights where you want this on everyone)
7. Level up-down (if you're ever having trouble, it's more efficient than just levelling up).
8. Chantage (See above, except only on one character which makes it hard to be centralizing, but there is a tendency to build party strategies around it, so it does dictate the rest of your party at least a little bit).
9. Mime (It's hard to be centralizing as a Mime since there's unlikely to be more than one or two of them, but you do dictate party structure/actions to a degree).
10. Fly (It completely breaks three or four fights fights, but it's very optional, while also being a lot of grinding, and just seems like it'll get drowned out by other grinding options that impact more than 3-4 fights).
11. Secret Hunt (So...miniscule chances of Air Knives early, showing Revnants who is boss, and LOL Zorlin Shapes.  Meh.  Revnants are no Ninjas (ala Catch) and you don't even need this on everyone)
12. Caution (It's technically an upgrade to everyone, and yet I'm finding it very hard to give a damn about it.  Especially with Sing sapping evade half the time).

[metroid composite]

Starjeweled!

So...starting to chat with some high level players and getting a better idea of strategies.

Notably the common opening for people with high energy production who've thought about these things is...

Strafe X

Get to near 1000 energy.  Produce one Ultra and freeze it.  Then get back up to 1000 energy and produce two more ultras as the previous one unfreezes.  Then spam ghosts.

Now, here's what I like about this strategy, especially if you're outproducing your opponents in energy.  (Other than the obvious "three ultras arriving at the same time".  If you're slightly ahead on energy, and just make offence units before they do, then you're going to end up fighting at their cannons.  And they're going to have only slightly fewer units than you.  And probably what happens is that everyone freezes everyone else's Ultras, and your Ultras die from cannon fire.  The advantage of Strafe X is that it slows down your initial push.  Now you're fighting back at your own cannons, and the roles are reversed.

On the choice of Ghosts over Hydras, I'm not entirely sure on the reasoning for that.  I'd hazard a guess that they're better against Zealots and Hydras, but worse against Ultras and cannons.  It does mean that you have to keep heal-wave energy ready at all times in case they psi-storm your ghosts.

In general, at high levels there is a tendency towards units that handle psi-storm, especially if you're going to put a lot of energy into them (psy storm is multi while warp cell is singletarget).  In practice this means Ultra, Colossus, and Immortal, being the only three units that don't need too much babysitting against storm.  Colossi are interesting--if someone told me 6 colossi was provably harder to deal with than 6 ultras, I would take such a claim seriously, because that sounds quite plausible.

Immortals are interesting because they're...actually decent fighting units, and you absolutely must respond.  You can't just let them come to your cannons.  And neither warp cell nor storm is terribly cost-efficient against them.

Zealots are also interesting against storm in a roundabout way--they're horribly vulnerable to it, yes, but they get next to the enemy.

[metroid composite]

Desert Strike SC2

Specifically Zerg, as that's what I've been playing.

Every zerg unit I've checked so far matches a 3/3 upgraded version of its SC2 incarnation in terms of stats.  Although they tend to have movement modifications, like Roach burrow move being super fast and Ultras having the Colossus ability to walk through your own army.

Zerglings:
5 / 230 minerals (46 minerals per ling)
8 / 300 minerals (37.5 minerals per ling)
(150% markup from in-game)

Roaches:
2 / 320 minerals (160 minerals per roach)
(160% markup from in-game)

Hydras
3 / 550 minerals (183 minerals per hydra)
(122% markup from in-game)

Hunter Killers (Hydras that deal 25 damage to armored, 10 to light, and have 100 HP)
2 / 600 minerals (300 minerals per hunter killer)

I'm not at all sold on Hunter Killers.  When you consider the price gap between them and Hydras, it's maybe a 15% increase in damage to armored, and 25% less total HP.  So...not necessarily better against armored (although they probably are just due to packing more densely and standing up to AoE better, I guess.  Oh hey, apparently they have range 7 too; that's something).  But woe unto you if they ever start attacking a light unit.

Ultralisk
2 / 1850 minerals (925 minerals per ultra)
(185% markup from in-game)


Note, for the air units...I haven't actually checked if their stats are the same....

Mutalisk
2 / 700 minerals (350 minerals per muta)
(175% markup from in-game)

Corruptor
2 / 1000 minerals (500 minerals per corruptor)
(200% markup from in-game)

Brood Lord
1 / 1400 (1400 minerals per brood lord)
(255% markup from in-game)

And all the spellcasters are changed, so a direct comparison isn't that useful.  And there's more new units than what I listed, but I mostly just know what they do, not their stats.  (Abbers tank with light armor and parasitic moves.  Slugalisks lay creep, and disable enemy abilities.  Brutalisks are high range ground to air.  Swarm Guardians are high range air to ground.  Lurkers sound great, but never seem to do much when I make them.  Like...hey, I've caught the enemy without detection and...they're just walking over the Lurker like it's nothing.  Scourge I actually haven't built >_>).

[OblivionKnight]

Oooohhhhh sexy Desert Strike update!

Mmmm....

Love the Zerg changes.  Aberrations were always good, and now they're...better.  50 more minerals, but they heal more with attacks, and are better supported by other units (cost drops on Ultras, Hunter Killers got a huge boost, etc.).  They stand up a bit better against Diamondbacks than before (DB cost more, deal less damage, Hunter Killers can support the Aberrations better, etc.), and even the Dark Stalkers, which should counter them, don't do a good job of actually countering them.  Hunter Killers...I actually built them!  I normally only built them vs. other Zergs who spammed Roaches and/or Ultralisks (and maybe to support vs. Corruptors).  Now...getting an extra Hunter Killer makes a big difference in terms of firepower - they were always decent in theory (extra range for another firing line, more HP, take up more space), but only getting 2 for 50 more minerals didn't give them a huge advantage over Hydralisks.  Now, I see myself making a line of each on average, as it may seems to be great for adding more firepower.  Cost drops on Ultras, Slugs, and Roaches is overall helpful - in particular, I now find myself building a line of Roaches vs. Protoss to add another firing line. 

But the unit I like the most for Zerg?  The Taint.  Ok, it needs a new name (I laugh every time I see the name then look at it >_>), but it is awesome.  Excellent answer to Protoss anti-air (Void Prism + High Templar wrecks the Corruptor shield for the Swarm Guardians).  One-on-one they murder...everything.  140 HP and 3 armour, and 1200 for one (Corruptors are 1000 for 2, with 200 HP and 5 armour each), but their DPS is pretty good (0.43 cooldown, 9 damage, 12 vs. armour), in fact, better than Corruptors...combined with their range, and it gives you 2 firing lines in the air.  Do note they are Light (so Void Prisms can do a number on them...though it's a bit closer than you think), and are slower than Corruptors, but...they really help win the air.  1:1 ratio of Corrupt Spires to Tainted Spires seems to be the magic number - even vs. Void Prisms and High Templar, the Corruptors tank long enough to eat up the Void Prisms enough that that Taints kill them all (and combined with a couple Brutalisks...).  GTA units (Fades) help, but 6:6 of each Spire was demolishing everything in the air.  Corruptors are just such excellent air tanks that they give the Taints plenty of time to demolish everything else, even under Psionic Storm.

For Terran...Predator damage nerf makes me sad, but the extra Marine from the Barracks I think balances this out, damage-wise.  Cheaper Goliaths and more expensive Diamondbacks (and less damage)...probably overall is a nerf (since Diamondbacks are so good, while Goliaths are more situational).  The 2 Barracks per Starport requirement...I think is to nerf the fast Banshee rush.  Granted...I don't think it needed it, but eh.  Probably their biggest change was the Spartan...200 less minerals, 33% faster attack speed makes them more useful for what they're to be used for, and hey, I might build them now!  They actually work pretty well vs. Swarms and Broods and Colossi and BCs now.  Their AI seems a little better too, so...I need to play with them more, but they seem much more relevant now .

Protoss...mmm...Colossi, Outcast, Immortal nerfs (cost, HP, cost, respectively) hurt, though the Immortals definitely needed it with their 6 range boost.  Uh...yeah.  Preservers got a big boost (they now...uh...heal!), but they were specialized anyway.  Dark Stalkers...I don't know.  I rarely if every made more than 1 Stalker (Archons, Immortals, Outcasts were always my base...I only ever made 1 Gateway too).  Dark Stalkers do more damage, but have less range.  They...mmm...they can't kill a Zergling in one hit, and while they're meant to help deal with Predators and Aberrations, I need to play with them a bit more.  A group of them supporting Outcasts and Archons probably helps, though I think Archons still have better DPS, especially vs. bio with their splash.  I'll probably play with them a bit more and add a line of them vs. Terran or Zerg if I see Predators or Aberrations...otherwise, probably won't touch them.

Overall, neat changes.

[metroid composite]

Yeah, I haven't been posting my theorycraft on the new version in this topic because I got roped into being a playtester somehow, and stuff is still changing.

3 Barracks per marine is a huge game-changer.  TvT now tends to be a mass marine opening facing another mass marine opening.

Dark Stalkers are fine on paper; basically the same anti-light DPS as the new Diamondback, and better DPS against other targets.  Less HP (it's something like 260 vs 160 I think) but blink helps with that, and they're cheaper.  But...yeah, every time I start building them, I feel like my wave gets rolled.

Taints...given that everything has 3 armor...

DPS vs non-armor: 13.9
DPS vs armor: 20.9

Compare to Corruptors:

DPS vs non-massive: 7.4
DPS vs massive: 12.1

Except Corruptors cost 500, and Taints cost 1200.  So...more specifically, DPS per 100 minerals

Taints:
DPSPM vs non-armor: 1.16
DPSPM vs armor: 1.74

Corruptors:
DPSPM vs non-massive: 1.48
DPSPM vs massive: 2.42

Which is to say...Taints are good against armored non-massive, which means Vikings and Corruptors.  Their fast firing rate also chews through PDDs, so they're especially effective against Terran (notably, being light is a definite bonus against Terran, as all Terran's air to air is armored).  I'm still not sure how many to get against Toss; you want some just to spread out, but Void Prisms are Massive and not Armored, so... >_>

[OblivionKnight]

Yeah, playing more games today, the +15 mineral cost for an extra Marine is huge.  Marine DPS is massive, and effectively boosting it by 50% for 15 minerals is crazy.  It's very safe to just build Barracks for a while and then get add-ons later.  Makes Predator openings vs. Protoss more effective, I've found - Predators lost damage, but they still tank amazingly well, and the point of them was always to shield for Marine DPS, which effectively got boosted.  Also, that +15 mineral cost doesn't reduce the amount of structures you can build between gasses, so...yeah, potent.

Dark Stalkers are a bit too expensive...I think.  I don't know.  I've been playing with them, but still am not sure when and why I'd build them.  I'd like to see them made a bit cheaper, maybe an upgrade from the Cybernetics Core (270 for the Core, +100 for the unit conversion).  They seem like they fill the same role as Stalkers - to amass behind a wave that comes to support yours (i.e., a fast Zerg wave) and provide some early AA support (usually for a supporting fast Zerg wave).  Also, no AA limits them a bit (I know it wasn't the role they were designed for, but...).  They don't seem like they make that big of a difference, although maybe I need more (or the standard Archon+Outcast build I use is good enough).  I've been building 4-8 of them when I see Predators/Aberrations, but I always feel like Archons/Outcasts are a better option.  Still not sure what to do with them. 

Void Prisms are armoured/massive, I'm damn near sure, or at least they were last version!  If that's changed, then the Taints aren't as good vs. them as I thought...although they've worked wonders in practice.  In my experience so far, 1:1 of each spire works (ends up being 6 Taints and 12 Corruptors) for smaller numbers of air units (like 4 Prisms and some High Templar), and provides Colossus-killing power.  Although even if Void Prisms aren't armoured (and I'm fairly certain they are), the Taints still work well - they don't get stormed to death by High Templar and provide long-range fire support with good DPS.  It's the AtA unit Zerg's needed - also, yeah, as you identify, they're excellent more-so vs. Terran, since only Thors deal extra damage to them (but they spread really well, which helps alleviate the splash effect, unlike Corruptors).

[Captain K.]

Was looking at some more Plants vs Zombies Endless Survival designs.  On the Xbox 360 version Cob Cannons are less attractive because there's no mouse control.  So this (or minor variants thereof) is considered to be a better design:

[metroid composite]

Mmmm...I'm definitely familiar with cobless builds, but I'm not sure I'd recommend them for people with low mouse control.  The problem is that you need to be on top of your instants, which actually takes more APM on PC.  I think it would be a little troublesome to take that to wave 100.

That said, looks like 360 version doesn't require you to individually click every sun that shows  up, so that helps cobless builds.  (11 fewer clicks every sun cycle).  Also looks like 360 has two cursors, one on the plant bar, and one on the lawn, so you don't need to go rapidly back and forth.  So...instants, which on the PC version are about the same as a cob cannon, are probably faster on the 360 version.

On setup specifics...

I approve of the cattails--if you want to limit APM, they save you from using Blowvers.

I'm skeptical about only having 6 Twin Suns.  You're going to be instant-spamming.  I've used as many as 4-sunflowers when I was using a medium cob build (4 cob).  I'd be tempted to replace some of the redundant inner wintermelons with more twins.

I suspect you'll also be replacing the outer row pretty often.  The problem being the wake-up time on the second inner gloom.  On the inside, using that spot for a freezie means you're a little more vulnerable to inner jacks.  An inner jack disaster is manageable, because you have redundant glooms both in the pool and on the inner rows.  But yeah, it'll be an expensive job.

[metroid composite]

If you've played Starjewelled against online people, you may have played against Plex once or twice, and thought "OMFG, is this person human?"  This is what Starjewelled looks like from Plex's perspective:

http://www.youtube.com/watch?v=vZ3nCthQvRQ&

[metroid composite]

While we're at it, more plex:

http://www.youtube.com/watch?v=NAicQ6yZ_VI&

Onto metagame stuff that I'm realizing with horror I haven't covered.

Ultralisks are still extremely centralizing, but the metagame has begun to adapt.

Mass immortals is actually a viable strategy that I've seen used successfully many times.  The thing is, if everyone always freezes Ultras, and you make Immortals...what's your opponent going to do about it?  Freeze the immortals?  Now they're falling behind because they need to freeze three or four units and you just need to freeze two.  Don't freeze the immortals?  Immos can kill Ultras at an acceptably fast pace.  Storm the Immortals?  They take three storms to kill, and get fully healed by one heal wave.  Mass immortals actually does seem to counter mass ultras (as long as you never miss a Warp Cell).

Another strategy that I haven't fully experimented with yet, but which was working fairly well against me, is...Whenever you get to 1000 energy, 1x Colossus, 2x Banshee.  Just...completely blind.  If the opponent goes Ultras, then you have two Banshees already, and it probably backfires on them.  If your opponent goes Ghosts or Hydras, then the Colossus rips them up (Colo being more effective against these units than Ultralisks), and if they warp cell the Colossus, then storm does a number on them.  And...just in general, one of the weak point of Banshees is that while they do deal a ton of damage, they often waste it on Marines.  Colo is the best unit in the game for clearing out Marines.  Follow up is three or so Hydras (the same way double-Ultra tends to be followed up by some hydras as you get energy).  The funny thing is how none of this can actually kill cannons yet; Immortal seems to be the unit of choice for that.  This is quite spell-heavy, though, as it requires you to heal wave in response to storms.

I've also seen people who just blindly go for 4x Banshee openings, but these seem less good.  The cool thing about 2x Banshee is that they cost 500 Energy, whereas killing them with two storms costs 600 Energy.  4x Banshee can't claim that, and also loses to not terribly rare stuff like mass hydra.

Zealots appear to be a widely accepted counter to Mass Immortals.  They are not standard, though.

Roaches.  I've actually been incorporating them back into my arsenal.  It's gotten to the point that warp-celling Banshees is pretty standard.  Freeze your ultras, build a Banshee won't do much against good players.  But...what do you do if your opponent freezes your two Ultras and builds 8 Roaches?  You can't stop them from dealing damage--warp cell is not an option.  You could double-storm, but now you've spent as much on spells as I did on the original Roaches, and they still got their damage in.  Roaches are more of a defensive thing than a "look at my unstoppable army", however, as double-storm does stop them.