Theorycrafting! (Because I like competitive metagames too much)

[metroid composite]

Pobes vs Zeelot

So...how to optimally blend HP and armor

Armor is weird, the cost for it goes up exponentially like everything else, but the values are...

9% reduction
18% reduction
27% reduction
36% reduction
45% reduction
54% reduction
63% reduction
72% reduction

Which looks linear at first glance, but it's actually harmonic (1/x) meaning it grows slower than exponential at first, and faster later.

So without further ado...

no armor: 0-800 defencive minerals (highly convenient number)
Lvl 1 armor: never optimal (o_O)
Lvl 2 armor: 900-1800 defencive minerals (again quite convenient; means 1600 HP 200 armor is optimal)
Lvl 3 Armor: 1900-3400 defencive minerals (meh)
Lvl 4 Armor: 3500-6200 defencive minerals (Well, this does mean that 3200 HP and 800 armor is optimal)
Lvl 5 Armor: 6300-11100 defencive minerals (Again, on the plus side spending 6400 HP and 1600 on armor is optimal)
Lvl 6 Armor: 11200-19300 defencive minerals (12800 HP and 3200 armor is optimal)
Lvl 7 Armor: 19400-32500 defencive minerals (stacked 2x 12800 HP and 6400 armor is optimal)

Although, realistically, you're probably not using three equipment slots on durability (and if you are, you might be going regen)

So...here's where it gets a little troubling: gas.  At gas it's:

1 gas: 92% reduction
2 gas: 96% reduction
8 gas: 98% reduction
32 gas: 99% reduction

A mineral-maxed zealot who isn't double-stacking HP dies in 3 hits to a T11.  Which is to say: usually game over.
A zealot with 1 gas on HP, and 12800 min armor dies in 14 hits to a T11.  This prevents feed, but doesn't kill.
A zealot with 1 gas on HP, and 1 gas on defence dies in 48 hits to a T11.  This doesn't prevent feed at all; in fact you will still feed pretty hard.

So...

0 gas defences = LOL dead to T11 (19x T9 to prevent feed, or something like that)
1 gas defences = 1x T11 to prevent feed
2 gas defences = 3x T11 to prevent feed
3 gas defences = 6x T11 to prevent feed
4 gas defences = 12x T11 to prevent feed

The problem is, going from 1 gas to 4 gas doesn't take very long particularly with gas-based damage.  The cost rises linearly in this section, but the durability doubles every time.  Now if you need more durability than this, then you need to go up to 10 gas next, then 16 gas, then 40 gas, and those take an appropriately long-ass time.

But realistically, a zealot who gets 3 gas total (1 weapon, 1 armor, 1 HP) should not lose ever.

Fun fact, while I'm at it: gas 2 regen + gas 2 defence only nulls out an extra four T11s.  (So for 6 gas, 16x T11 to prevent feed).  For all that regen is "good" once you hit gas, still not that good.

[metroid composite]

Pobes vs Zeelot:

So...some more discoveries.

First: I actively changed the metagame recently.  Lost a game as zealot to someone who blinked around the map, and started wishing there was a 4ish gas "give full map vision" item.  Then I realized: wait, treasure boxes give a tiny ring of vision (wide enough to cover base entrances).

So...if a probe pulls one of these stunts where they run and stall, pack your bags with a bunch of level 1 items, and drop them at the entrance to every base.  I usually keep a few serious items:

1. HP (so that if the probe does base up, you don't die)
2. Boots (it's a scouting mission; movement matters.  Also, they are pretty cheap and could be dropped as a treasure box)
3. attack (but not attack speed--nice to be effective if you actually do find something)


Another interesting discovery: pre-gas, if you expect a T11, doubling up on HP will let you survive for just long enough to warp out (so 2x of the 12800 HP, and 1x of the 12800 armor).



And...a few calculations I ran: apparently if you do a really awful durability build (namely: keep all three of regen, defence, and HP at the same level) you...actually do ok.  Like...your durability tends to be 80% of optimal, assuming medium opposition (by medium, I mean turrets that will kill in 40 seconds or so).  Basically, it plays like an HP build, and if you last for 40 seconds, then the health you get from the regen during that time is...about 60% of what you could have gotten with an HP setup, but not that bad.  And the ratio of amount spent on health stats to amount spent on armor stats is...reasonable.

[metroid composite]

Pobes vs Zeelot:

Feeding and travel time.

So...on the large map, the closest base takes 10 seconds to return home for a shopping trip.  A moderately far base (bottom left--still one of the closer ones) takes about 40 seconds round trip.

This has some interesting implications.  The amount of time you should spend feeding depends on how distant the base is.  But it also depends on the point of the game--early in the game you can feed enough to double your damage in about 70 seconds.  Later in the game, it takes 15 seconds to feed enough to double your damage.  This is purely a function of the attack speed upgrades.  (Attack power always costs the same amount per damage, but later in the game it's multiplied by 5, and your damage is your income).

What I found in my calculations was that doubling time played a very large part in how frequently you wanted to return to home base.  In extreme circumstances it can be worth spending more than half your time travelling, just to upgrade more frequently (though only slightly under given how things are balanced).  On the flip side, early in the game, when you feed slower, you want to stay a long time.  Feeding all the way to 400 minerals, for instance, can be optimal when it's early and the only known base is quite distant.  (90 second feed--although less in practice due to starting minerals and passive mineral generation on the way to the base).  Later in the game, when you can double in 15 seconds, you want more like 45 second feeds for distant bases.  By comparison, close bases (10 second round trips) are more along the lines of 45 second upgrade cycles, and then get faster (to like...35 or so) when you start doubling quickly.

Move speed boosters:

Cloak lasts for 10 seconds, and seems to be a +1.88 movement boost, with a 1 minute cooldown.
Void lasts for 5 seconds, and has a 45 second cooldown.  If void cooldown is about to end, that's a pretty good time to leave; worth taking a mental note of the timestamp of when you get voided.
Boots give +1 movement bonus.
So...here's the interesting thing: I think the zealot gets a +0.25 movement boost on the large map, so is normally 3.75 by default, but is 4.00 by default on the large map.

On the subject of boots, they are a 25% bonus to travel time, but early on you should be doing more feeding and less travelling (ideally) at maybe a 2:1 ratio for somewhat distant bases on the large map, so boots become good when the equivalent damage increase for the cost would be about 12%?  (so...around when you have the 800 cost weapons and 400 cost attack speed?)  Although...these aren't directly comparable--the comparison with boots I made there assumed staying to feed longer, but ideally with boots you'd have less travel time so you'd stay for shorter time periods to upgrade more frequently (which probably means boots expenditures should come from your durability fund, not your damage fund).

[metroid composite]

Pobes vs Zeelot:

So...the whole rigid 2 ultra -> 3 legendary -> perfects...except when you're pooling and want 0 ultra -> 3 legendary -> perfects...

Fundamentally this comes from you want to get your gas income above a certain level before you start saving for a certain threshold of expensiveness.

But what if we throw a wrench into this: you instead of pure automines get several T11s (which don't produce gas income).  Now the ideal values are often somewhat higher; instead of two ultra miners, it can be optimal to get three ultra miners.  Instead of three legendaries, it can be optimal to get four legendaries.

This also brings up another interesting question: let's say you don't plan to econ any time soon at all, and just want as many T11s as quickly as possible?  What level miner should you get?  Turns out: spam ultra miners.  Spamming Legendaries eventually catches up, of course, but not until you have 12 T11s or so.

But this value changes if you're pooling and going pure turret; with three people pooling, you might as well get Legendaries.  You won't delay your first T11 much (if at all).  Alternatively, pretend the maximum generator was lower (like 128 or so).  Now it can be worthwhile to go for master miners over Ultras.

Intuitively this makes sense.  If it takes sooooo long to get an ultra miner that two master miners could get you the 10,000 minerals you need for T11, then go for the Masters.  Mathematically it breaks down something like this...

dx/dt = floor(t/timecost)*income

x = sum(1/2*timecost*income * n) for n = 0 ... floor(t/timecost)

x = (n*(n-1)/2)*1/2*timecost*income

So...as n gets very large you end up with a formula that looks like : t^2 * income/timecost

And so here's the thing: income/timecost is always exactly the same--ultra miners always produce 36 minerals, and they always cost about 1/3 of what a legendary miner costs.  HOWEVER, the difference comes in the n*(n-1) term.  As you increase timecost, the n*(n-1) term gets smaller, and the timecost*income gets larger (assuming you have a target x).  But as the n*(n-1) term gets really small, it starts to shrink the other terms due to the (n-1) component.






....in a completely unrelated calculation:

T9 and T11 cost efficiency comparison.

It takes about 100 T9s to deal as much damage as one T11.
100 T9s cost 11200 minerals (and much, much more gas; T9s cost primarily gas).
one T11 costs 10240 minerals (plus the cost of a T10, which is like...8192 gas and 240 minerals--basically nothing)

...Yeah, T11 cost efficiency is kind-of insane.

[Niji]

You know...every once in a while I analyze the metagame for Starcraft or Magic the Gathering or Pokemon or Advance Wars.  I don't always have anywhere to put my thoughts (and notably, some of these communities will flame you off the board for theorymoning/theorycrafting/etc...and not without reason--it's easy to overlook stuff when you theorize).  But I just happen to find theory fun, so as long as I'm entertaining myself by theorizing, I may as well store my thoughts somewhere.  And the DL seems like the place that would most appreciate analysis just for the sake of analysis.


So...without further ado, I'm kicking this off with...Pokemon:

There is no ban's in pokemon what are you talking about. (yes I am aware of the existence of smogon universty and that only certain types of terrible people use their rule system...)

[Dark Holy Elf]

I'm pretty sure the vast majority of serious analysis of Pokemon assumes the existence of tiers (i.e., banning of ubers as at least a possibility) because for a long time there were so few ubers that they were completely overpowered and overcentralising. This trend long predates the existence of Smogon.

[Niji]

I'm pretty sure the vast majority of serious analysis of Pokemon assumes the existence of tiers (i.e., banning of ubers as at least a possibility) because for a long time there were so few ubers that they were completely overpowered and overcentralising. This trend long predates the existence of Smogon.

I was being snarky, the idea of the existence of "tiers" seems pointless when I can one shot a charizard @ lvl 100  or Hydragedon @ lvl 100 with my bellossom. Though If i had to ban stuff it would be flying gem acrobatics users

Even the legendary fire types pale to my mighty bellossom!

Banning "moves" would seem more logical in pokemon than baning the pokemon, even arceus kinda sucks despite his stat pool.

[Yoshiken]

Arceus is absolutely insane, actually. There was a time when people were considering banning him from the Ubers tier, just because of the sheer number of options and the crazy power he can get from Extremespeed.
Also, some moves/abilities are banned not because they're good, but because they make the game more an aspect of luck than skill. (i.e. Moody, OHKO moves)
Aaaand OHKOing stuff with a Bellossom outside of weather seems incredibly unlikely, unless you happen to count "sleep and then attacks" as killing, but then that would be why Breloom is one of the best revenge-killers in the game~

[Niji]

*leaf storm, it only works the one time and of course during clear skies since bellossom is poop slow otherwise.
And generally the legendaries and fire types dont max their special defense evs with positive nature for it.
Luck is an important factor in all battles, i have won and lost so many matches to lucky crits! And those are less than 20% chance, otk moves make more sense(not to mention mostt can learn  mindreader/lockon/ettc).

[Yoshiken]

Luck is an important aspect, but not something that should be relied upon, which is why OHKO moves are generally frowned upon - they're a bad strategy, worse than the actual top attacks, but because they're reliant on luck, they can beat a team that is technically better. The best strategy in Pokemon, and in a number of games that come to mind, will be one that can mitigate luck, not rely on it.
Think the only perfect accurate -> OHKO, other than Smeargle, is Articuno. Possibly Froslass too? Either way, not many.
And.. I'd be surprised if Leaf Storm OHKOs a Ho-oh, since they do tend to run HP EVs, so I guess you're excluding Ubers? (Hell, I'd be surprised if it OHKOed Lugia, since that thing's a tank and a half.) Not to mention things like Ferrothorn or Multiscale Dragonite...

[Niji]

Luck is an important aspect, but not something that should be relied upon, which is why OHKO moves are generally frowned upon - they're a bad strategy, worse than the actual top attacks, but because they're reliant on luck, they can beat a team that is technically better. The best strategy in Pokemon, and in a number of games that come to mind, will be one that can mitigate luck, not rely on it.
Think the only perfect accurate -> OHKO, other than Smeargle, is Articuno. Possibly Froslass too? Either way, not many.
And.. I'd be surprised if Leaf Storm OHKOs a Ho-oh, since they do tend to run HP EVs, so I guess you're excluding Ubers? (Hell, I'd be surprised if it OHKOed Lugia, since that thing's a tank and a half.) Not to mention things like Ferrothorn or Multiscale Dragonite...

I have OHKO a moltres but I never faught a Ho-Oh (for some reason they seem unpopular in battles being a physical fire attacker, but it does have boss special defense and x4 grass resistence)with my bellossom, the legendaries and dual types I've OHKO'd are: Deyxos (multiple forms but never defense form no one seems to run that), Lugia, Moltres, Articuno, Zapdos, Entei, Suicune(lol to easy that one though), Darkrai(got lucky since it was faster than me that darkvoid failed >.>), Dragonite, Charizard, Arceus(This time though it was on my phsyical attack bellossom and it was a double battle where I boosted its attack with the confusion atkx2 booster), Jirachi(that was surprising), Mewtwo(never fight mews), Hydragedon, Tyranitar. That is about it, the other legendaries I have fought are too fast still and usually running a choice item or choice Gyrados with fireblast >.>

Personally I find One Hit KO moves hit more often than 50-85% chance to hit moves(i've had fireblast and others miss all 7 times in a row....), and then there is the whole thing about Hail Storm and Sheer Cold and sandstorm and fissure(or fissure in certain games on certain map terrains), as well as the fact those moves can hit through protect under certain circumstances which make them really strong, They seem like they are lucky moves but i've seen computer trainers sweep my team with sheer cold before.

And a few more pokemon in Gen I/II could learn fissure from the TM as I recall, though the mechanics in Gen I were pretty different than onward :C
I MISS PRESS B at the right moment to make sandlash's slash always crit!

If its GEN I I will talk pokemon theory crafting all day, but GEN II onward and especially GEN V where everything seems finally balanced regardless of legendary usage in fights (baring of course the one trio that gets stab on acrobatics in addiction to being ludicrously fast >.> since I mostly use grass types I find acrobatics to be the supreme of over powered moves especially in triple battles since it counters all strategies and will one shot most pokemon on first use).

[Yoshiken]

The OHKO moves are inaccurate, even if it doesn't work out in practice. AI is rarely a good test, and as for things like Fire Blast/Blizzard/Thunder, I've had Thunder hit consecutively 5 times in sunny weather. Doesn't make that a reliable figure, unfortunately.

Just decided to run some calcs on Leaf Storm from a Bellossom. Positive nature, max EVs, figured Choice Specs was the best way to go for turn 1 damage. Running these on a Gen4 calc because it changes nothing else and I'm too lazy to find a Gen5 one.

Bellossom (Modest, 252 SpA) Leaf Storm vs. Ho-oh (Adamant, 4 SpD) - 14.2% - 16.7%
Bellossom (Modest, 252 SpA) Leaf Storm vs. Wobbuffet (Calm, 252 SpD) - 55.3% - 65.3%
Bellossom (Modest, 252 SpA) Leaf Storm vs. Dragonite (Mild, 0 SpD) - 22.6% - 26.6%
Bellossom (Modest, 252 SpA) Leaf Storm vs. Jirachi (Jolly, 4 SpD) - 42.5% - 50.1%
Bellossom (Modest, 252 SpA) Leaf Storm vs. Arbok (Impish, 0 SpD) - 54.9% - 64.8%

...I'm really not seeing it.

[metroid composite]

I have OHKO a moltres but I never faught a Ho-Oh (for some reason they seem unpopular in battles being a physical fire attacker, but it does have boss special defense and x4 grass resistence)with my bellossom, the legendaries and dual types I've OHKO'd are: Deyxos (multiple forms but never defense form no one seems to run that), Lugia, Moltres, Articuno, Zapdos, Entei, Suicune(lol to easy that one though), Darkrai(got lucky since it was faster than me that darkvoid failed >.>), Dragonite, Charizard, Arceus(This time though it was on my phsyical attack bellossom and it was a double battle where I boosted its attack with the confusion atkx2 booster), Jirachi(that was surprising), Mewtwo(never fight mews), Hydragedon, Tyranitar. That is about it, the other legendaries I have fought are too fast still and usually running a choice item or choice Gyrados with fireblast >.>

Most of those legendaries you listed are not considered "ubers" and are not banned from normal play.  In fact, most legendaries don't have especially great stats.  The ones that are banned to ubers....

Deoxys A (glass cannon)
Deoxys S (Mediocre durability--badly distributed stats for it)
Lugia (Unholy tank; and takes half-damage from grass, and most EV builds pour even more into defences.  Err...yeah, this is strange--you sure it was level 100 with EVs?)
Darkrai (Mediocre durability--kind of like Deoxys S but not as bad)
Arceus (Fantastic tank, but as you mentioned there were special circumstances here)
Mewtwo (Much more known for ridiculous speed and offence than durability, but his durability isn't bad)

But yeah, some of the ones you mentioned are actualy allowed in low-tier play (I believe Moltres is for instance).  Certainly being legendary doesn't mean you're an unstoppable killing machine.  Some legendaries have bad stats, bad moves, or bad typings.

[Niji]

^ This

Oh and critical hits lol probably should have mentioned that

I average 40-50% critical hit rate in this game regardless of a moves crit chance for some reason, But that is discounting even the 99% moves (90% boosted 10%) missing multiple times in a row. (aka only counting moves capable of a critical hit, sucessfully landing).

This is a trend in every RNG based game I have ever played though.

And No, only stab boosting items or defensive items get used.

I have never i think delt damage to a lugia of decent amount lol>.>

But I was just talking about legendaries and other beefying mon's with x4 resistance STILL getting OHKO'd.

UNKNOWN if these enemy pokemon were EV trained as it was wifi or battle frontiers/stadium (most of my bellssom OTKs took place in stadium games from Gen II-GV (i don't play her special atk in GenV )vs friends pokes so unknown how well their EVs/IVs looked I would assume focused on stats, My bellossoms IV was always in the 25-31 range with maxed speed and sp atk EVs for data purposes with sp atk postive nature in Gen IV and remake games. I won't discuss that those were usually the only kills I got against their pokemon but yeah >.> But the rage was pleasurable from them.

Like I said theorycrafting is pretty cool but vs actual RNG(or luck if you will) it falls short terribly when my 30% chance move connects more often more regularly than my 50-90% chance moves >.>

I think its banned because people get annoyed by lucky fellows that sweep OHKO with the OHKO moves their entire team without even a chance to defend themselves.
The chance to fail of 70% should be enough for them to be competitive for "lucky" folks. I think its just /rage.

And then...there is double and triple battles too >.>

[metroid composite]

We interrupt this discussion of games that people actually care about for...

Pobes vs Zeelot!

Quick durability calculations to demonstrate the different builds.  If the zealot has 12800 minerals, and the probe has two level 9 turretss...

12800 regen: OHKO'd.  (Survives 1 second...2 seconds if lucky and you regen between the two turret hits)
12800 8 Armor: 3 HKO'd
12800 8 HP: 41 HKO'd
6400 regen, 6400 Armor: 6  HKO'd (takes 296 damage, heals 256, net loss of 40)
6400 regen, 6400 HP: 30 HKO'd (takes 800, regens 256, net loss of 544)
6400 Armor, 6400 HP: 56 HKO'd

None of these setups is perfectly optimal, granted.

Not to mock regen too, too badly, though.  Technically, the longest surviving setup against 2 9s when you have 12800 minerals is...

6400 Regen, 3200 Regen, 3200 Armor: Survives indefinitely (takes 368 DPS, regens 384 HP per sec).

[metroid composite]

archiving this conversation

You: Turret the hell out at +8
Mitzu: Yea sure
what's the difference
o.o
You: Then salvage when the zealot leaves, and turret the hell out at +16
Mitzu: Ohhh
you meant at +8 income
You: yes
Mitzu: thought you meant level 8 turret
You: no.
Level 8 turrets suck
Mitzu: was like..
how is that going to help..
You: Mass level 4s into mass level 6s.
Mitzu: yea
just don't worry that much
about eco right?
You: Yeah, stopping the zealot's eco is higher priority than ecoing yourself
Mitzu: Mm hmmm
You: You do kind of eco just by existing; sitting on +16 gen is a decent amount of income.
Mitzu: mm hmm
should be better than zealot's
You: If you were really rushing you'd sell your generator.
(There are some builds that sell their generator; they're pretty hillarious).
Mitzu: Haha, I don't think I've seen any
at least not recently that I can remember
You: I wouldn't bother with them outside of 1v1.
Mitzu: Haha yea I thought not
You: Well...arguably it's worth it to sell your generator for your first legendary (as long as you immediately rebuild and reupgrade it).
Mitzu: mmm
You: But it only gets you about 10 seconds ahead anyhow.
Mitzu: that's true
actually
You: So...I usually don't bother.
Mitzu: ehh
You: (Honestly, there's a lot of stuff with econ as probe where I'm ok with losing 5-10 seconds.  Like...leaving my depot built on +32 gen; that's 8 wasted seconds, but I don't really care that much).
Mitzu: ahh yea
I dont normally salvage either
just because I might built more
later
Not worth taking probe out again
You: Yeah
Mitzu: For 256 to 512
I normally don't build workers though, I just end up buying minerals
in bulk half the time
You: But if I were more daring I would just head out with my probe every time.
Mitzu: As long as prices aren't sky-high
You: Ehh...256 to 512 is the one area where I almost always get extra miners.  Prices usually spike sometime around that time, and even if they don't, if you buy 200, you just raised the price by 100 all by yourself.
Mitzu: YOu do what, sell wall and buy 2 professionals?
You: 1 pro on gold will usually do it
Mitzu: Mmm, i feel like
4k for pro,
isn't really that much more then buying one set
You: 2 pros on gold generally means you plan on doing a lot of selling (1 pro implies buying if you just have 2 average + 1 advance)
Mitzu: in bulk
Does the pro, mine over 200 buying it at 256 to 512?
You: Hmm...it's about a 120 second upgrade time, so...
mines 240 I guess?
Mitzu: Ehhhhh
and buying 200 will purchase at the cost 50 higher then the current on average right?
You: More to the point, though, if it stays alive after 512, then you have some extra minerals; can sell them on the market as other people hit 512.
Mitzu: I feel like
after people hit 512s and even 256s
minerals just become dirt cheap
You: 100 higher, but 50 higher on average.
Mitzu: especially when everyone proxies the gold
You: 'Well yes, depends on the market
Mitzu: Yea, so as long as the price is 150 or less
You: Gradient certainly likes buying to 512.
Mitzu: isn't it just faster to buy in bulk?
Especially since
mineral income is pretty negligible after 512
since you rush ultra anyway
You: Mineral income is nonnegligible if you can sell at a decent price
Mitzu: After 512
why would people need to buy minerals?
You: And often enough it'll spike to 200 or so as other people reach 512.
Mitzu: That could be true
I guess it really just depends on the other plays
players*
You: (A lot of my builds are tuned for being ahead of other probes on econ >_>)
Mitzu: LOL
that could be why, I'm never as ahead of you
;)
You: Yeah, if I get to 512 first, the mineral price will drop.
Mitzu: Maybe that's why i buy in bulk
after you
haha
You: But anyway, as far as getting a miner between 256 and 512...usually even with one Pro I still have to do SOME buying on the market.
I just don't buy everything on the market
Mitzu: Yeaa, and you don't buy in bulk right?
You: Yeah, I'll buy like...50-80.
Mitzu: Buying in bulk isn't actually cheaper than buying in 10's is it?
o.o
You: Which doesn't spike the price too much.
I....don't think so?
Mitzu: I used to think it was at a discount, but then I realized it isn't
lol
XD
You: I've never actually checked.
Mitzu: if the way it calculates is based off individual sales, then it's equivalent
i Used to think it was 200 minerals at the starting price you buy
but that's definitely not true >.>
You: yeah, I remember looking at it, and thinking "that looks equivalent, or very close."
Mitzu: Haha yea
For +32 to +64
how many minerals do you normally end up selling?
you go 2 average 1 advanced
But I always seem to fall behind you in that step
You: I go 2 average for 32 to 64 usually, and sell about 40.
Mitzu: no advanced?
until 64?
You: I get the advanced by selling the wall at 64.
But...it all depends on the market
Mitzu: Ohh
That's what I've been doing wrong
I end up selling like 60-80 instead
and then i'm behind you
You: I've done things like getting all my miners on +32 when the market was doing crazy things (like hanging out at 350)
Mitzu: Yea
If it's at 250
you still wait for 64?
You: Buying all my miners on +32 is actually how I got my 14:38 solo max gen.
yes
250 isn't that much
Mitzu: Mm hmmm
You: If anyone else is selling, that can tank fast
Mitzu: Also, in games where the zealot is good, do you still bother trying to take gold with miners?
You: yeah
Mitzu: It seems like if the zealot is assuredly good, he'll just scan and kill
and then you get set behind a lot
You: If the zealot is really good, the zealot will usually ignore gold miners
They don't feed.
Mitzu: Doesn't it hurt players more than zealot?
That's basically stopping at least 50-60 percent of the mineral income
You: Ehh...as long as they last about 70 seconds before getting killed, it's probably still better than mining blue
You catapult ahead on gas income by upping your gen faster.
Mitzu: if they die
Do you just end up buying rest
or do you rebuild at blue?
Or does this all just depend on the price
You: Buy the rest to get to the next gen up, and then rebuild on gold.
Mitzu: Ahh I see
You: Depends how close you are to gen up, really
Also, you can do stuff like send one miner to each gold base; if the zealot wants to run around that much, sure, don't get feed, that's fine.
Mitzu: Yea, kind of what I did last game I guess
but I only split it off into 2 golds
Of course after you chased, I never retook the gold either
You: I only killed the one gold set because there were like...10 miners on there, and it was on my way home anyway.
Mitzu: haha
You: If you'd sent to the top gold, I wouldn't have bothered detouring that much.
Mitzu: yeaa
so far from my base though
T+T
You: Just drop a depot in there; you knew which base I was going to be returning to repeatedly; you can guess my path
(Good zealots are predictable; it's bad zealots who might kill you in transit)
Mitzu: I honestly thought
you were going to camp top right
so I figured safest base would be left gold
It's only when that other guy
went to left gold too
that I put one in the other one
You: Yeah, I knew there were miners on the left gold, but I never went in there; too much of a detour.
Mitzu: ahh
Then when zealots scan the golds
You: I was mostly scanning the golds to see if anyone would be dumb and actually move in.
Mitzu: do you run?
You: Sometimes
half the time running turns out to be wrong, and my scvs die on the way out
may as well get in the extra mining time
Mitzu: yea
You: Ideally split up the miners and send them different directions.
Mitzu: mm hmm
You: and hope he wastes time by chasing XD
Mitzu: I've tried doing that, but with only 3-4 miners
and zealot's speed and damage, they normally don't get very far
but I guess any time wasted is still good
You: Yeah, my splits are bad.
Mitzu: Ohhhh second point! What is your build for zealot?
You: I sometimes get like...1-2 out alive.
Mitzu: Damage, attack speed, health?
You: Most important thing with zealot is damage.
Mitzu: yea
You: overall DPS I mean.
Mitzu: Err
You: So...you'll notice that level 1 attack up is +2, and level 1 attack speed is +20%.
Mitzu: don't get level 2 attack
before you at least get attack speed
?
You: Thing is, +2 is actually a 40% bonus over 5 damage.
Mitzu: ohh yea
jk
XD
You: Yeah, it's an imbalanced square.
Mitzu: So damage over speed, until you get the 80%
thing right?
or after a certain point in damage, speed becomes worth it
>.>
You: You generally want to spend about 250 more on attack power than attack speed.
Mitzu: What about for armor life and regen
Life first?
I see a lot of people go regen
You: Basically, if you had 10 base power, then +2 would be a 20% bonus.
Regen sucks.
Mitzu: Really?
You: I generally never buy regen.
Mitzu: just pure life and armor?
You: Yeah.
And mostly life
With a cheap armor, because it's a percentage bonus
Mitzu: mm hmm
You: But like...better to double up on life than to get 9% more damage reduction.
Mitzu: Yea true
You: (The typical ratio I go for is to spend about 4x as much on life as on armor).
Mitzu: I see
You: (That breaks down a little towards the expensive armors, because it's harmonic).
(The gap between 63% reduction and 72% reduction is much larger than the gap between 9% and 18%).
Mitzu: So many things to think about =.= just being a probe is hard enough haha
You: But yeah: other stuff: never buy level 4 attack speed (but that's kind-of obvious).
Mitzu: err which one is the level 4 attack speed
100 percent?
You: yep
Mitzu: ahh yea
i normally just go from 80 to 200
You: 800 minerals for +100...when you can spend 400 minerals for +80
Yeah, 80 to 200 is right.
It's actually correct to stick on 80 for a really long time (like up around 50 damage).
Mitzu: mmm
You: But...yeah, past that, in the lategame I focus builds around dealing with T11s, because T11s are that good.
Mitzu: Mm hmm
so what, armor and life? XD
You: If you don't have gas, stacking two 128 HPs, and one 128 armor will let you survive an 11-void with a teleport
Mitzu: Really?
You: Barely; you survive 6 hits.
Err...the 6th hit kills you
But void only lasts 5 seconds.
Mitzu: mmm
That's not bad at all then
normally you'd have gas by 11 though right?
You: Well...yes, normally I have gas around the time people are still on +64 gen >_>
Mitzu: XD
what is that, 7 and 8 turrets?
You: You can support a level 9 turret on +64
Mitzu: mm
You: I mean, not without delaying yourself, but that's about what the Mega Wall 1 costs, which is what you need for the upgrade.
Mitzu: mm hmm
You: And yeah, other post-gas stuff: 1 gas HP, and 128 armor makes you pretty safe against 11s--they kill in 15 seconds (denies feed, but you'll live).
Mitzu: and not everyone should have 11's
You: 1 gas HP, 1 gas armor lets you live 45 seconds against 11s (feed as much as you want).
Mitzu: mmm
You: I usually stop getting durability around 1 gas HP 1 gas armor, unless I just have spare gas (in which case I'll get 2 gas armor, then 2 gas HP).
Or the probes are making multiple 11s, of course.
Mitzu: i thought you said life over armor
You: Gas armor is different
Mitzu: doesn't it just go
from 92 to 96?
You: 92% to 96% = double
Mitzu: ehh
does life double too?
You: Yes.
Mitzu: so then it doesn't matter
You: So...it's pretty much equivalent.
Mitzu: which one you get?
ahh kk
You: (Armor is still slightly better due to your innate 500 HP and innate passive regen >_>)
Mitzu: heh
XD
You: And 8 gas armor is just flat out better than 8 gas HP.
Mitzu: For sure
You: (armor doubles again, HP goes from 320000 to 490000)
Mitzu: yea
You: Oh, and Regen 11 is actually good; arguably better than 490000 HP.
Mitzu: it's the 20 k hp per sec right?
You: Yeah
Quadruple the previous one.
Mitzu: coupled with armor, I'D ASSUME SO
oops
caps
XD
You: Well....so here's the thing
IMO all the regens (Except regen 11) should be doubled in how much they regen
Here's why
for the most part, 100 minerals buys you 250 HP, or 4 regen.
In order to get more than 250 life regened in a single visit, you have to keep knocking on the probe's door for 62.5 seconds.
It's basically never worth it to stay that long without returning home to upgrade.
Mitzu: yea
You: Double the bonus from regen, however, and now you're looking at 30 seconds; that's quite attractive
BUT: regen is still vulnerable.  If you go all regen, and someone gets a T9, one void and you die.
Mitzu: mm hmm
life is just safer
You: Yeah; if regen bonus was doubled, I'd....still go life a lot of the time, but go regen when I feel I should take a risk.
Mitzu: so you never buy rege
until max?
You: Pretty much
Regen 11 is good; it also stacks
Mitzu: Mk, sounds good to me!
You: If you have to deal with multiple max turrets, for instance, you can have one damage sponge wearing 4x regen 11.
They can absorb 3-4 T13s.
Then load a kitty up with like...four max blades to kill the final wall.
Mitzu: wait
4 x regen 11
with what
armor and life?
You: Yeah
Mitzu: won't they just kill
your cats?
o.o
or damage is too high
You: You have to hope they don't focus fire
Mitzu: for them to focus in time?
ehh
without kitties you lose
as soon as they hit final
You: Well yes
Mitzu: right?
You: Well...ok, no not exactly
Double max weps can break it.
If there aren't too many repairs.
Mitzu: but then if they build turrets
in time
you die right?
You: Well...one max turret isn't too bad; regen 11 and 99% pretty much nullify it.
Two max turrets means you can only stay for 20 seconds, though.
Mitzu: ouch
so by 4
it's auto kill
with void?
or wait..
8 is auto kill
o.o
You: 4 kills in...7 seconds?  Something like that.
But yeah, you need a kitty if you want to break 4 max turrets and a final wall.
Mitzu: Mm hmm
You: One max turret and one final wall is breakable solo, at least if they only have one probe repairing.
But...more to the point, it's very very rare that you ever get to this point; either probes or zealot tend to pull ahead and kill/completely feed deny long before final wall.
Mitzu: haha yea
like that game where we had like 15 final walls and like 8 turrets
XD
You: Haha, not actually the best way to do the endgame though.
Just cover the map with final turrets and zealot detectors.
Mitzu: yea i figured
You: What's he going to do, attack one of the final turrets?
Ok, he dies.
Oh, yeah, the other thing you can do against final wall if you catch them just as they get it, but before they get a final turret, is 3x max blades.
Mitzu: =0 lol, just pure damage
XD
You: You still have HP and armor >_>
Mitzu: well yea >.>
You: (Although, if you want to play really risky, you can sell the armor too, and go for 4x max blades)
Mitzu: lol
You: (You'll be fine as long as they don't have a 13....)
(But you'd have to run from like...a 12 x_x)
Mitzu: haha
You: (Well...you could stay for like...20 seconds against a 12, which might be enough time to knock down the wall)
(Wait...20...no that isn't right is it?  Hmm...ok, no more like 8 seconds).
Mitzu: lolol
XD
not to mention, that'sonly 1?
Most people would have like 2 -3
if they already have final
at the minimum
You: Some people will sell all turrets when they realize they're not even hurting you.
(Which is actually not a bad strategy)
(Might as well try to econ faster)
Mitzu: mmm
Anyways, I'm pretty tired, gonna call it a night. Thanks for all the info and tips Let's play again sometime
You: ya later

[metroid composite]

Super Smash Bros Melee

So...interesting things have been happening to the metagame, particularly a lot of characters are surfacing that weren't surfacing five years ago.

Peach: arguably the top player in the world right now is a Peach player from Sweden.  Now, Peach is a character who was considered good on-paper for a very long time, but most of the people who played her were players who weren't good at doing all the physics exploits in melee (wavelanding, wavedashing, L-cancelling).

Jigglypuff: also a character that people knew was pretty good, has recently moved into being dominant, been moved to the top of the tier list.  Major tournaments have ended in Jiggly vs Jiggly grand finals (which incidentally, is an incredibly boring paring).

Pikachu: now here's where it gets interesting.  Pikachu wasn't really on anyone's radar in, say, 2005, but now one of the top oh...10 or so people in the world is a pikachu main.  Very strange to watch, too--Pikachu basically can't deal damage (even his downsmash is like...11%) so most of the kills revolve around gimping at low percents.

Mewtwo: There is in fact a Mewtwo player who is...not completely irrelevant.  Placed like...third at an international tournament.  Not dominant on the level of the above, but not irrelevant either.


Now, there are a couple of theories for why this shift might happen.  One would  be that there is less competition--for instance, back in 2005, the best SSBM player in northern California (that went to regional tournaments) was a Link player.  Draw from a smaller pool, and the results become less consistent.  I don't really buy that, though--several of the players that were good in 2005 are still good and still playing tournaments...and losing to these 17 year olds playing non-standard characters (I believe Armada is still in high school).

Part of it seems to come with a shift in the metagame--people playing more defensive, just focusing on not making mistakes.  Part of it comes with the game being around so long that people can learn multiple characters.  Every one of these players can play at least one of Fox/Falco/Marth fairly well in addition to their main character, so if they run into a particularly unfavourable matchup, they just switch.

The particularly interesting thing is how, at least it appears that Brawl strategies have influenced the Melee metagame.  Brawl is a game that has become a lot of "stay away from the enemy forever, occasionally throw a projectile or an attack in their direction in case they're stupid enough to try to get close".  Certain melee matchups end up kind of like that (Jigglypuff vs Young Link, for instance), and while people played them aggressive at first, as these matchups become more and more important, and people figured out what really matters, it's more common for melee matches to look kinda Brawl-esque (no combos, no attacking, lots of camping).

[Niji]

Gosh that sounds awefully boring to watch if i went to tournaments, even more so to play.... But I am very happy to hear about peach, I always knew her potential(she is the only melee character I have ever won a fight against fox/falco/marth that can do all the exploits, and the dude was one of those idiot savants so 1 win out of a 1000 games as my sole victory is EXTREMELY REWARDING lol, I have never played Melee since then, that is how satisfying it was(so about 1.8 years)). SO THIS PLEASES THE NIJI.

Pikachu is REALLY surprising to hear about, I guess cheap tactics like with kirby's throw off a cliff but press up the same moment your opponent dies because of priority delay?
Mewtwo, I can see it, that sounds like it would actually be really exciting to watch him battle in a tournament.

[metroid composite]

Pobes vs Zeelot:

So...a friend and I have started handicapping ourselves as Zealot.

To review, pobes vs zeelot is a game where you econ race the other team.  The probes spawn at 0:20 into the match, and the zealot spawns 1:00 into the match.  The challenge that we have been imposing on ourselves is that...the zealot sits at spawn for 5 minutes, gaining no income until the 6:00 mark.  The other player then suicides, chooses to help zealot, and spawns as a Hunter (spawning at 6:00) and the two move out together.  The question is, how much of a handicap is this really?

So...some quick calculations--you passively gain 1 mineral a second (starting from 0:20) giving you 340 minerals by the time the hunter spawns.  Obviously put everything into damage (usually divided between the hunter and the zealot) so...at the 6 minute mark this gives you 6 damage from passive mineral income, added to the 5 base for the zealot and the 4 base for the predator.  (15 damage, but slowest possible attack speed means 10 DPS).

An average zealot will generally have 2x that by the 6:00 mark (and a good zealot with a close feed might have 3x).  Which is to say: you're not initially as behind as you might think.  However, you tend to fall further behind for a few reasons.

Primarily among them is that we're splitting resources between two units.  This has several implications.

1. You flat out deal less damage than a single unit.  1600 attack power + 1600 attack speed gives you 74 DPS.  3200 attack power +3200 attack speed gives you 230 DPS.  That's more than triple for equal cost.

2. You don't share durability.  We've been in several situations where the pred stacks up damage, and hopes that the Zealot draws the turret aggro, but then the probe re-targets the turrets at the pred, forcing the pred to run (while the zealot continues to feed at lower damage).  Imagine if you crammed all the damage and all the durability into one unit instead, so you could both tank and deal damage...

So...basically, splitting resources between two players actually takes the somewhat behind state (15 damage at lowest speed) and puts it more behind.



Despite all this, 4/4 wins so far (with some very close calls).

[Scar]

I used to love going to Smash tournaments. Never won any, but I usually placed in the top 4 by the end...if that counts for anything! >_> (It doesn't, I know.)

My obscure character to use in tournaments was Luigi. His air game is a lot of fun, and I have a lot of control of his movements in the air. I love trying to predict an opponents movements when they are sailing and try to juggle them for mass-hysteria.

When I am serious I tend to use Sheik as my main, and while I was active in the Smashing scene, she was quite a site to behold. It was like my Luigi, except a lot faster. pin-pointing where someone is heading in the air takes a lot more precision since she is a freaking blur.

Another thing, I was never a wave dashing person. I figured that I wanted to test my skills against players that did wave dash, and I was not too far behind them. Sure, there are advantages to knowing how to do stuff like that, but I'd have to re-learn so much and I never wanted to get that focused. I was having fun with what I was doing, so I left it at that. Plus it was always fun to beat wave dashers.

Man I miss playing smash with people. =(

[metroid composite]

My obscure character to use in tournaments was Luigi. His air game is a lot of fun, and I have a lot of control of his movements in the air. I love trying to predict an opponents movements when they are sailing and try to juggle them for mass-hysteria.

Luigi's got a lot of strengths.  Grand finals of DLC2 (after the various Marths and whatnot had been knocked out or switched characters) was my Luigi vs Elfboy's Samus.

In fact, while we're on the subject of Luigi, here's a video of one of the best players in the world using Luigi:

http://www.youtube.com/watch?v=BGq-e_6FxVw

...which...also highlights why he's not higher tier: his recovery is garbage.  If he had normal recovery, my gut feeling is that he'd be a relative constant in tournaments (not top tier, but like...Captain Falcon level of "lots of people play him, he's good.")

Another thing, I was never a wave dashing person. I figured that I wanted to test my skills against players that did wave dash, and I was not too far behind them. Sure, there are advantages to knowing how to do stuff like that, but I'd have to re-learn so much and I never wanted to get that focused. I was having fun with what I was doing, so I left it at that. Plus it was always fun to beat wave dashers.

Wave dashing is actually somewhat balanced.  It's another tool in the toolbox, so going without it is like saying "hey, could you beat that person without using forward-B or down-B"--chances are you can if you're better than the other player.  But there are plenty of situations when wavedashing isn't even the right answer, and plenty of other situations where running or rolling are 90% as effective anyway.

L-Cancelling, on the other hand, is just flat out better and gives people a blatant advantage.  Of all the mechanics that SSBM has, I always found it curious that people who complained about SSBM tournament players would yell "those goddamn wavedashers" instead of "those goddamn L-Cancellers".

[TranceHime]

I feel the need to talk about Pokemon so I'm very happy that this came up because I theorycraft to shit about Gen5 meta

Most of those legendaries you listed are not considered "ubers" and are not banned from normal play.  In fact, most legendaries don't have especially great stats.  The ones that are banned to ubers....

Deoxys A (glass cannon)

Tears through most holes in the Ubers metagame, especially given the presence of Life Orb. However, it is correct that Deoxys-A is pretty much the epitome of Glass Cannon - shitty defensive typing in Ubers (lels pure Psychic?) and abysmal defensive stats (50/20/20 don't be foolin' no one) mean all strong users of Choice Scarf sets, Mach Punch and Vacuum Wave pose as good checks to the speedy menace. However, keep note that +1 90 base speed Pokemon can not outspeed maximum speed Deoxys-A. Packing Scizor or Metagross on a Deoxys-A that doesn't have HP Fire is also a good idea, and Lugia can also tank Deoxys-A and OHKO it back.

Deoxys S (Mediocre durability--badly distributed stats for it)

Deoxys-S's worst enemy is the fact that Team Preview exists - allowing people to easily counter or check Deoxys-S before the game even begins, since you would be downright stupid not to play Deoxys-S as a lead in ubers metagame; as a suicide lead or not. Mostly, people run Deoxys-S on the support since it's capable of using dual screens, Stealth Rock, Spikes, Taunt, Magic Coat, and a whole slew of other useful moves for support. However, it's got some mediocre offense, ironically its best options on the offense being ExtremeSpeed for other Deoxys (of A or S variety) and Fire Punch (for the ever annoying Forretress lead). Other than that it's not very overwhelming. I guess.

Lugia (Unholy tank; and takes half-damage from grass, and most EV builds pour even more into defences.  Err...yeah, this is strange--you sure it was level 100 with EVs?)

Crits are bullshit lelelelele

Did you know Lugia has Multiscale?! lelelelel (Though sadly Multiscale+Roost is illegal, so you need to rely on Recover instead)

Don't really need to say too much about Lugia that hasn't already been mentioned - it has ridiculous bulk; only a few things even scare it (Dark-types come to mind, as does Kyogre Water Spout, Zekrom Bolt Strike, and Kyurem Ice Burn all have the potential to ruin Lugia's day). It can phaze with Whirlwind/Dragon Tail, and has Toxic alongside its bread and butter move: Roost. For defensive purposes, Lugia can also pack a Sub or Reflect, which just further improves its massive bulk.

Darkrai (Mediocre durability--kind of like Deoxys S but not as bad)

Darkrai's major advantage is that it's Dark type in a tier where lots of things take super-effective or neutral damage to the Dark type. 90/135/125 offenses is really solid, especially considering that it has its signature Dark Void, which is the second most accurate Sleep-inducing move before Spore.

Dark Void
Dark Pulse
Focus Blast/Nasty Plot
Nasty Plot/Substitute

This is the set that Gen 4 players should recognize if they knew about Darkrai's potential as an incredibly powerful sweeper. It has access to a great way to set up and destroy the opposition, and its access to Focus Blast/Dark Pulse allows it to deal heavy damage to almost everything in the tier. Only two Pokemon resist the combination of attacks: Heracross and Toxicroak, with the former pretty much being one of the only reliable checks to Darkrai. Heracross takes the Dark Void happily and Sleep Talks, either 2HKOing or OHKOing back with its STAB moves thanks to Guts. Other than that, Darkrai's got great offenses.

Arceus (Fantastic tank, but as you mentioned there were special circumstances here)

Arceus having Multitype gives it a bazillion options, pretty much, and what Plate the player chooses usually dictates what set Arceus is going to run, and the sheer unpredictable factor of which Arceus is going to come out until it does come out makes Arceus very scary. Of course being able to OHKO an Arceus takes some considerable luck if you're running Pokemon who aren't: A) offensive behemoths or B) abusing the weakness of the type Arceus currently is, but eh.

Mewtwo (Much more known for ridiculous speed and offence than durability, but his durability isn't bad)

Mewtwo's 106/90/90 defenses are decidedly reasonable for Ubers.
Mewtwo gained a very powerful and new toy this generation - Psystrike. It's a 100 BP Psychic move that hits the Defense stat, but runs off of Mewtwo's Special. And only Mewtwo gets it (everyone else gets Psyshock, the 80 BP variant). In fact, this one single addition is what makes Mewtwo incredibly terrifying this time around. Combined with ridiculous speed, Mewtwo now effortlessly puts dents into Chansey and Blissey (especially with Calm Mind, a +1 Mewtwo pretty much OHKOs most things in general at this point) and has ridiculous coverage, Ice Beam/Fire Blast/Aura Sphere and more. Furthermore, it gets Unnerve as its Dream World ability, which allows an offensive Mewtwo to completely trivialize Chople Berry T-Tar, Yache Berry Garchomp and Wacan Berry Manaphy. Did someone mention Custap Berry Wobbuffet? Well now that ain't working anymore either.

Fun fact: A Choice Scarf Shadow Tag Chandelure can and will ruin Mewtwo's day with an Overheat under the Sun. Steel type Pokemon will give Mewtwo a hard time as well if Mewtwo doesn't have Fire Blast. If it does, welp. Psychic Arceus can do something about Mewtwo too, unless it has Shadow Ball. Then welp.

I've got some experience with Random WiFi battles on Pokemon BW, if anyone wants to hook up for some WiFi battles then just shoot me a message on the boards here or on IRC (I'm Mikan or some other stupidly girly nick)

[metroid composite]

Might as well copy-paste a Plants vs Zombies post I made:

would be better if there we more viable strats for endless other than cob cannons and winterpults everywhere

There are plenty of cobless builds for endless.

I wouldn't make a build with no winterpults, but I'm pretty sure it's been done successfully too.


In fact...let me provide examples:

Cobless:

http://www.youtube.com/watch?v=ZTOKFFAEP6A

This is actually a reasonably normal strategy.

http://www.youtube.com/watch?v=CdVIqqqeflQ

A build with no winterpults. I'd consider this fairly abnormal.

And...I guess both of the ones I've linked have had either quite a few winterpults or quite a few cobs, so let me link one that's light on both:

http://www.youtube.com/watch?v=KWyxchwyw7s

And yeah, it still has a couple of winterpults, but...only AoE status and all that: you generally want at least two. (And I'd consider this an extremely standard build, for the record; I play almost exactly this, swapping out a couple sunflowers for a couple cobs--sunflowers and cobs basically do the same thing anyway--flowers let you blow stuff up with cherries etc on a regular basis, cobs let you blow stuff up with corn).


That said, if there's one strategy I wish was more viable in endless, it'd be firepea:

http://www.youtube.com/watch?v=DlkoM2aT5pI

It's not bad, because fire gives peas AoE, and then they deal a ton of AoE DPS (more than a winterpult or fume, certainly. Similar to a Gloom--less AoE more range). And while they unfreeze, they only unfreeze singletarget--their AoE damage does not unfreeze. The problem is that they have an arbitrary weakness that if they're hitting a shield (screen door, ladder, zomboni) then...they don't get double fire damage to the shield (meh, whatevs) AND don't get their AoE effect (o_O). Still usable, and will roast their row 70% of the time, they just sometimes need babysitting.

So umm...sure, I can agree that they shouldn't lose their AoE effect against shields so that they would be more viable in survival endless.


I feel like I'm missing some viable plants...oh yeah! Spikerocks and Cattails:

http://www.youtube.com/watch?v=h8Tya-ZOs70

Rocks are obvious--arguably better than Glooms in that location because you don't need to replace pumpkins. Cattails, despite being entirely singletarget, tend to do a decent job at countering balloons. Can't really build a strategy or clear a lane with these, but they're an option.

And...while I'm on the subject of interesting support plants...

http://www.youtube.com/watch?v=g9IZii1-RjQ&feature=relmfu

He wants to get ladders set up on his back pumpkins, so...he lets his front plants get eaten, then uses...Garlic! and Tall Nuts! Granted, that's a pure short term fix, but I've also seen Tall Nuts used long-term in the pool to stop dolphin zombies.

Not to mention the obvious support stuff (Umbrella Leaf, Cherry Bomb, Squash, puff shroom, blowver, Freezie, Doom Shroom, pumpkin, sunflower etc).


But umm...sure, in general you need AoE DPS (which you can only really get from Cob Cannons, Glooms, one-time-use plants, and to a lesser extent from firepeas melons spikerocks and fumes). And unless your DPS is insane you'll generally also want some AoE status (which you can only really get from winter melons and Ice shrooms). So...your backbone will pretty much always be built out of some of these. But...I feel like damn near half the plants are viable in some setup or another, even if just in a support role.

[metroid composite]

SC2 Heart of the Swarm

So...most people playing the Beta seems to think that Warhounds are broken; let's take a look...

Cost:
150 minerals
75 gas
2 supply

Stats:
2.81 movement speed
7 range
23 damage, 1.3 cooldown (17.7 DPS)
220 HP

So...I've seen a lot of people comparing this to Marauder.  Stat-wise the unit this immediately reminds me of is actually the Roach.  About the same movement speed as upgraded roaches (slightly slower even).  Twice the damage and 50% more HP in exchange for twice the cost (so...3x overall stats for 2x overall cost--this should result in a dead even fight assuming no micro: good ol n*(n-1)/2 formula and all that).  If you were to double all the costs of the roach, I would expect basically this -exact- HP/DPS to keep the balance the same.


There are two stats that immediately jump out at me however:

7 range.  Combined with high speed this allows them to kite just about everything.  Not only that, but 7 range actually lets them siege some stuff--outrange Planetary Fortresses and bunkers, for instance.  Compare to roaches which are characterized by having super short range.
2 supply.  Remember beta roaches?  That cost 1 supply?  Well THEY'RE BACK!  And they're a Terran unit now.

[SnowFire]

Old post from Slashdot, but since this is mc's thread and this was something I had to check to see if mc had posted due to previous highest-finite-number shenanigans, the Magic The Gathering Turing machine:

http://www.toothycat.net/~hologram/Turing/HowItWorks.html