So, when FA was released, I think experimentals were intended to occupy a different role than they currently do.

When FA was closing in on a release, the devs had a serious problem...

They had T3 land, which was complex and diverse, and they didn't want experimentals to just outright replace T3 land on high mass maps.

However... they needed experimentals to be scary units. They needed them to be impactful, both for sales reasons and to justify their existence in the game. If their role wasn't to be some kind of hyper end game replacement for T3 land, then what should their role be?

And I think that what they came up with, was that experimentals should be a sucker punch.

Their niche was providing a means of avoiding steep production costs in engineers and factories (factories were much more expensive back then, so this was probably a major factor) you could always take a risk and go for an experimental... and going for an experimental was always a very risky proposition...

The upsides were:

• Fast to produce

• Required next to no infrastructure

The downsides were:

• Significantly less effective on a per mass basis than T3 land spam

Experimentals were always a major risk to build, because they were under-powered by design. If your experimental was delayed and your opponent gained enough production to match it, then it would die to significantly fewer units than it cost to build in the first place, and your opponent would get its mass as reclaim. They weren't units that you built on a whim, there was real risk in building them.

This gave T4 units a role that didn't step on the role of T3 land. They served two major roles that were related. First, they could be used as a sucker punch against opponents that weren't scouting effectively. Secondly, in the event of a significant influx of reclaim mass after a battle, they were an effective way to quickly put that reclaim to good use.

That is, I believe how experimental units were intended to work by the devs at launch. That was their top level design philosophy. They were risky units that could be built to give a player a large but temporary force advantage at the price of poor mass efficiency.

This thread takes the position that by veering away from this original vision for T4, we've inadvertently broken some aspects of the game...

So if that's how T3/experimentals used to work, if that's how the game used to define the role for experimentals, how are experimentals used today?

Today, Experimentals are much closer to being "T4" units.

I've avoided using the term "T4" thus far in this thread, because I want to separate them from being part of a tier system. They aren't really tier 4 units, they're experimentals, but we've moved them pretty far from their original roles.

Now their design philosphies justifying the existence of Experimentals more or less match the philosophies justifiying untis at other tiers.

• They are more powerful

• They have significant advantages over units at lower tiers

• They have a mass efficiency that is close to in line with that of other tiers.

• They require significant production and time to produce like units in other tiers.

And because they function more like T4 units than they used to, I think they are replacing T3 land in very late game engagements.

And I don't think that's really what we want. I don't think we want Experimentals to step on the role played by diverse late game T3 unit formations.

How did we get here?

So I admit that I don't really know how we got here, and that this is just a guess, but I bet its more or less correct.

The old system included the "sucker punch". That was the role of Experimentals. They were tough, but for the mass less tough than most other options. The thing that made Experimentals something that you feared was really the speed with which you could push them out rather than their innate combat effectiveness.

And that led to people losing games to units that they didn't scout.

And that led to people complaining about it without really understanding why Experimentals worked the way they did. They didn't understand why Experimentals had to be sucker punch units. They didn't get that this role was chosen for them to avoid having them step on the role of T3 land.

Over time this complaining became consensus. Its not just complaining, the build times really ARE too short. Its not a design choice that's working as intended to avoid stepping on the role of T3 land units, its a "bug" that needs to be fixed.

And so we fixed it, over and over again we fixed that "bug"...

And as we fixed that bug over and over again, we didn't notice that T4 units were starting to step on the role of T3 land.

And I wouldn't suggest that we just move Experimentals back to their release state. Perhaps the sucker punch really was too strong, perhaps they were too quick to build.

But I think that at some point, we've pushed these units too far away from their original role.

And I think that is having an impact on gameplay that is detrimental to the game and the community.

Ask yourself, is T3 land what you plan on fighting with in late game engagements, do you see it as the default late game force, or is it a stepping stone to keep you alive long enough to get out a Monkeylord or Chicken? Are experimentals T4 units, or are they niche units designed for very particular circumstances, either to surprise your opponent or capitalize on a reclaim field?

And if it really is a stepping stone, why the hell is that the way we think it should function? Are we really sure that we've made the right choices?

@Mod_Councillor

You clearly don't know what makes a good map, because you have some kind of elitist bias that prevents you from learning anything from maps that have become wildly popular.

It is blatant unveiled narcissism that allows you to claim that maps that have been enjoyed by thousands, a map which currently occupies the #2 and #4 slots on the "most played" page, is a bad map.

If this belief happens to be held by a large number of ladder players, that doesn't make it any less ridiculous.

I kind of figured this would be an issue. No objective criteria can ever be instituted because any such criteria would have to accept the obvious truth that popular maps... maps that thousands enjoy playing, maps that people pass over carefully crafted very pretty 1v1 style supposedly "good" maps in order to play... are good maps...

This absurd group think mindset runs so deeply that even the most played maps of all time, hell - particularly the most played maps of all time - are somehow considered bad.

At the end of the day, you have this notion that your opinion is superior to that of others, that no amount of "votes" by players means anything at all. You are the one that determines map quality, not the playerbase.

And that is both laughable and sad.

Any map which no one plays unless they're forced to play it by a matchmaking system is a failed map. If you think otherwise its because you are using a broken system for determining map value. Sitting around with a bunch of buddies and picking out maps might be fun for you, but good luck getting people to give a shit about your opinion and play those maps willingly.

If the map making community spent more time learning why some maps are so heavily selected by the playerbase, and applying those lessons to their maps rather than bitching about how those maps aren't any good, maybe people would rotate off of some of these "bad" maps once in a while.

I definitely find them useful as units for deterring a com push, although I worry that usefulness is probably mitigated by priority targeting mods.

If they were just a little faster, they'd be a good counter to T4. Not saying that we SHOULD do that, but... I wouldn't complain...

I think firebeetles have a limited role as eco damage drop units, although this isn't particularly sexy as T1 arty for other factions does it better.

Its not a terrible option though, as they don't leave reclaimable wreckage usually, so if you get a mex, the mex is gone and they don't even get the wreckage of your beetles to help rebuild with.

If Cybran transport capacity wasn't so low they might be seen as slightly less useless.

I think they're no longer in the "completely useless for any particular purpose" category, and are now in the "needs reasonable minor tweaks to really shine" category.

Dropping them to say 160 mass might be enough to change people's perceptions of them. I don't think they need anything like a major rework at this point.

In Leage of Legends, the devs have figured out a neat trick...

When they deploy a new character, the character is always moderately OP. This is to get people to take notice of them.

I think the firebeetle needs that. They need a moment where they're OP enough that people acknowledge that they exist and serve a purpose.

After that you can nerf them back into a reasonable position, and people remember how good they were and see how they can be useful and everyone will be happy.

@Mod_Councillor

You already read my other post, but for the sake of anyone that comes through this thread I'll say this.

At a very low level, team maps are mostly about providing safety. Players want a safe map.

As you get into higher and higher ranked games, it becomes more than that. Players want a "good game". They want a game between 20 and 45 minutes long. They don't want to sit in lobby for 30-45 minutes waiting for a match only to have it either end, or become overly one sided before the first 10 minutes is up.

Good team maps provide this. They ensure that the game doesn't immediately end. They are relatively forgiving of fuck ups like commander over-extension. They don't provide lots of contestable expansions that advertise that your team is behind.

If I designed a map for high rated players, or just tried to host one, I imagine it wouldn't fill. There aren't enough top 100 ladder players to justify many maps that cater to that crowd. Maybe a group of 8 2000 ranked ladder players could have close spawn positions, complex opening strategies, wide open spaces and not see a com death in the first 20 minutes... I have never seen such a thing so I'm skeptical that its the case, but I wouldn't be blown away surprised if its the case either.

But there just isn't a large enough pool of that kind of player for anyone to bother building maps around that.

So you target guys 1000-1800 on the upper range for team maps, because that's your player-base. That's the range where the maps will actually fill if the map is decent. If you target above that good luck getting a game.

Don't ignore the impact of the lobby simulator on map design.

So tell me why a 1000 rated astro or dual gap player will want to play a 1k+ 2v2.

Uh... they do?

Hell, why would he want to play a wonder game.

Wonder is a popular choke-point map, I consider it to be in the class of maps that I'm talking about here. Its a team map, pretty easy to tell which mexes are yours, opening isn't overly complicated. There are central contested mexes, but they're so contested that taking them means little and isn't game determining. Players generally make it to the middle game.

If they didn't, no one would play it.

@BlackYps I am not certain the the devs intended Experimentals to work as they did at launch, all I know for certain is that they did work that way at launch, and I think that because they didn't step on the role of T3 land, this was a design choice that was intentional. There were significant advantages to having it work the way it did, and so I think it wasn't just random chance that caused them to function as they did at launch.

I played the game in the era of the sucker punch, and I remember frantically begging team members to scout because I was terrified that someone had a ML or GC building out there. Experimentals were scary units then. I also remember times when I had 4-5 bricks out and a teammate would start panicing because he scouted one and I could just tell them to calm down, we have enough production to deal with it.

Experimentals at that time built faster, but they were also far weaker relative to T3 units, so they were scary, but you could deal with them. If we just took experimetnals in their current state and cut their build times we'd break the game, but if they got nerfed in accordance with the buffs to their build times, they'd be impressive terrifying units but they'd be manageable as well.

As for gameplay, I recall an old tactic that people used during the "Sucker punch" era that doesn't really exist anymore. You'd pretend to build an experimental. You'd see a scout coming and would start building a ML so that the scout sees it. you'd then reclaim it. This was a useful tactic - I wasn't the only one using it at that time - because of the panic the sight of a building experimental would cause in the other team. They'd start major investments in PD, gunships etc to have a chance at dealing with the ML they they thought would arrive in a few short minutes. You could then choose to continue to eco, or build a nuke. Just threatening to build one would have an impact on the other team.

I suppose that good players would perform a double check, to see if they were really serious about building it or if they were just trying to get you to make poor investment choices.

But that's really not the major issue with gameplay, and I explained this issue already in my original post.

Having T4 step on the role of T3 land disincentives T3 land and T3 production. You replace large armies filled with an array of different units each serving different roles - and their production facilities - with a ML or GC escorted by AA. THAT is bad for gameplay.

And maybe you aren't comfortable with scouting or knowing the build timings of T4 and you don't like the idea of getting sucker punched with them from time to time, but you have to see that replacing a lot of the late game unit diversity with Chickens and Galactic Colossus is bad from a gameplay perspective.

@Morax said in M&M Map Vault Plans for Fall 2020 and On:

Popular does not equal good, Moses. Sorry.

Yeah... in retrospect it looks like this was not the smartest thing to post. I didn't know that such a firestrm might be created by my posting here, but the elitism is comical.

And the people leading the charge against the popular maps are the ones that don't know why popular maps are actually popular until its explained to them in detail.

@Mod_Councillor Yeah, if you could spawn a current strength ML with very short build times, sure... it would be crazy.

But T4 is stronger now relative to T3. If we nerfed T4 concordant with the buffs we give to their build times, they'd be manageable, have a niche, and not step on T3 land.

Also yes, appeals to authority suck. The buildtime nerf has not made the game into something unrecognizable to a player from base FA.

The change doesn't need to make the game unrecognizeable to be a bad change.

I noticed that this is happening. I play Cybran, Bricks used to be terrifying units, and against most things they still are, but against T4 they're pitiful.

Defending against T4 with T3 land is a poor strategy nowadays, and that is different than how things used to work, but more importantly it makes late game engagements less interesting than they used to be.

Nowadays, you micro your Monkeylord rather than your 20 unit T3 formation made out of Bricks, Loyalists, Decievers, Trebuchets, Bouncers, T2 flak and firebeetles. You do this because there is little to gain in investing in T3 production to the point where you can spend a significant portion of your mass on bricks. Why go the diverse formation route when a ML is better faster and requires less investment?

So is the game "unrecognizeable"? No, it isn't, but it has lost something that I think had value from a gameplay perspective. It lost something and I'm not even sure we ever really had a conversation about the fact that we were losing it, was it even a conscious choice to replace T3 formations with Experimentals or were we just fixing the "bug" of Experimental build times and OP T3 land?

I acknowledge that there is an argument for keeping things as they are, saying that the current system is superior. I get that, but I disagree with it, and I think this conversation really needs to happen.

If we're going to replace T3 land with experimentals, lets at least do so with eyes wide open.

I think you'll enjoy this post: https://forum.faforever.com/topic/171/i-think-the-relationship-between-t3-land-and-experimentals-should-be-re-examined/5

Which says a lot of the same things you're saying.

T3 land is suffering against assault experimental unless you're UEF. Experimental units are no longer surprise assault units. For many factions there's little reason to invest heavily in late game T3 production when you can just go T4.

Anyway, they're working on SCU changes that might go a long way towards fixing the problem...

@Jip said in Hives Need a Nerf vol. III:

You can't balance for all maps.

Sure you can.

Map choices certainly changes the game, but its very much still FAF.

It mostly just emphasizes a different stage of the game. Turtle maps tend to emphasize the late game. Large open maps, the early game.

A problem on a turtle map is likely to also be a problem in the very late game stages of 1v1 games.

If you stealth nerfed T3 static arty for all factions but one, the first people to notice would be your Gap/Isis players.

If you nerfed the mantis, gap players might never notice... but the ladder crowd would lose their minds.

The notion that they're somehow completely different games is just wrong.

Striving to balance all maps is effectively the same as trying to balance all stages of the game. You're going to wind up with a more balanced game if you take all maps into account.

Think of turtle maps as "canaries in the coal mine". If those players spot an imbalance, that imbalance may eventually show itself on ladder. You just have to have games played so evenly that they routinely make it to the extreme late game before you notice.

Will the hives imbalance that the OP is claiming exists appear in ladder 2-5 years from now? Maybe... As players get better and games become more even... maybe exploitation of hives will eventually show itself as a significant balance issue in the ladder scene.

Of course its going to appear on turtle maps first, every game on those maps is focused on the late game.

NOTE: Again I'm not trying to take sides on whether Hives are actually OP, just trying to provide perspective.

@herzer99 said in Points of Imbalance.:

Not some weird chokepoint teammaps

Interesting way to describe the maps people actually play.

@Turinturambar said in The SCU Rebalance:

Cybran

The Cybran have 1 rambo preset and 2 support presets. The rambo preset is the same as in current balance (AoE stun). The first support preset provides a large stealthfield (it can easily stealth an entire army). The second support preset provides a speed buff field, which increases the movement speed of its surrounding, non experimental, units. Both support presets also have the aa upgrade. Multiple speed buff fields do not stack.

I imagine no one is ever going to use a stealth SCU.

By the time you start building SCUs, stealth is no longer particularly useful. The enemy is going to have T3 spy planes and omni.

Beyond that, if you want stealth, you have other good options in deceivers and MLs. The Stealth SCU seems like it would only make sense when you're forwarding a unit composition through enough splash damage that deceivers are unlikely to survive, but can't yet afford a ML... in the unlikely case that keeping your reasonably late game army stealth really matters to you.

I get why stealth was a natural choice for a Cybran AoE ability. Its thematic... but I don't think its going to be a particularly useful unit. Good on paper, but I doubt it will be useful in practice. Maybe someone will use one to hide a massive drop from time to time on maps large enough for that to be possible...

If the unit is destined to be largely ineffective, it might be better to narrow your focus and just ensure that the speed field works really well. Much better to have a single SCU type with high utility than to have two where one is essentially useless - even if it is thematic.

Beyond that, if a unit just isn't useful, it shouldn't be in the game. No unit should be a "noob trap", a unit that players don't realize doesn't really serve a purpose that they have to learn isn't worth building. Am I correct in my presumption that a stealth SCU would be so useless as to have no purpose whatsoever? I can't sure of that, but I also believe that there won't be many instances where an expensive single SCU is better at providing stealth than several cheap deceivers or a spider.

If we really want to give Cybran two support SCUs, where both of them have a real purpose, trade the stealth for a shield. I know... Cybran's supposed to have crap for shields, and has never had a mobile shield, but maybe it doesn't need to have crap shields that deep into the tech tree.

If you disagree with this assessment, maybe you're right. I haven't used the SCUs. I could be wrong, but please consider this when evaluating the stealth SCU yourself.

TLDR: Either cut the stealth SCU concept because its not going to be useful in late game engagements and just focus on the speed field support SCU, or change the second support unit's AOE ability to something that would actually be useful for Cybran players - like a shield.

@biass said in Points of Imbalance.:

Straight up, if you think the t3 land nerf was bad for the health of the game you're probably not informed enough to be able to contribute to a reasonable discussion. I cannot reasonably beliieve that anyone would want to go back to pre nerf, that's insane. Wanting to go back to every single game being a t3 rush and watching a single harbinger or two smoke hundreds of t1 and t2 tanks with hardly any effort because of "i like t3 stage" is not a heatlhy state of mind.

@Arran said in Points of Imbalance.:

These changes add a lot of meaningful diversity to how you use LABs.
Now onto my main point --> Aeon tanks are the slowest and now they have the slowest LAB too. Perhaps swap the speeds of UEF and Aeon LABS and adjust their costs accordingly? This proposed change is to prevent Aeon lacking map control in the early game owing to insufficient unit speed across the board.

I don't see how 0.2 speed is going to "cripple" your raid ability outside of the largest of 10km maps. It's a bit of a waste of breath trying to instantly ask for changes on a patch without replays or etc.

Also, as most people have said: You need to come to the reality that factions (especially yours, because I know you main aeon) are not supposed to be good at every aspect of the game. Aeon have a defensive early t1 and then threaten to crush the entire game out unless enemy can make a reasonable counter, why should aeon crush both early AND late?

Don't know why I bother with you, you're happy to make bad faith arguments that misrepresent a position... however...

No one is arguing that we roll back 3696, the argument is whether assault experimentals should be nerfed to "finish the job" that the 3696 patch began.

We nerfed T3 assault bots primarily to fix them in relation to T2 bots - which you can see if you go back and look at discussions of that patch - but we never bothered to apply the correction to the next higher tier.

And that's why it was a sloppy patch. I don't necessarily disagree with the changes, its just that if you're gong to make such changes you have to finish the job. Nerfing T3 bots as hard as that patch did while leaving T4 unchanged aside from build timers and the world's most negligible mass increase just doesn't fix the problem.

Its not like people weren't building T4 assault bots like mad before the 2018 patch, to just nerf to T3 and ignore its relationship to T4 was negligent.

@Evan_ said in Points of Imbalance.:

You can hear Petric talk in the video about the changes in relation to T3. GC lost the ability to trigger its claws as fast. Monkey got a cost increase in addition to bt nerf. Other units like T3 mobile arty and sniper bots were nerfed after they were found to be a bit too strong vs T3 bots.

You neglect to mention that it was a 1/19th cost increase that the ML received. It now takes 10 seconds more mass production to build if you're rushing it. Its a completely negligible nerf.

Yeah, T3 mobile arty did get a nerf... not in that patch, the actually buffed mobile arty in that patch... but much later it got a nerf. Doesn't matter though as the issue is that T4 is just better in certain circumstances. So much better that people don't bother to build significant T3 land assault bots.

Ythotha got a cost increase and had its dps shifted so if can't one shot Percies/Bricks. That's not to say these changes put them back in line with old balance, that wasn't the point. T4 IS stronger now once it gets up, that's intentional. T4s are no longer a cheese strategy but a proper unit. The video explains it better, but to say it was a sloppy patch and that there wasn't any concern or that people didn't notice changes for T3/T4 balance is just wrong.

Looks to me like the Ythota just got nerfed in general, but the nerfs were fairly slight.

Look, here's the issue with your "cheese strategy" claim. Gyle and other casters actually recorded lots of games from pre-2018 and made them publicly available. When I got through those games, I see top tier players spamming lots of T4 units. They weren't a cheese strategy, but they weren't T3 land either.

Making them a "proper unit" means that they're stepping on the role of T3 land - which I think is a terrible way to deal with them - but if everyone's dead set on making them a normal unit rather than a suckerpunch (which isn't necessarily cheese) then for god's sake, drastically increase their build times. Right now they're no less cheesey when rushed than they ever were, but in addition to being a cheesy rush unit they're also mass competitive with T3 land assault bots which causes them to replace T3 assault bots in some situations.

Even with the changes, T3 land still beats T4 mass for mass with a good formation, especially with shields.

If its rushed, you aren't going to have the T3 land to beat it. If its not rushed they'll have T3 land too. Beyond that T4 assault bots don't need to beat T3 land to be relevant. They have other major advantages. In a lot of cases you can just run a T4 assault bot by a mass of bricks or percies, and if you build enough splash units you'll be just fine when the assault bots eventually make it to your base. Your opponent will have substantially less time to react however.

And that's not considering that while your opponent makes T4 you can win much of the map with T3, if not kill his whole base. Even when the T4 gets out it's an uphill battle to retake the map since you only have one experimental that can only be in one place.

You say this, and this is great in theory, but if you read through more of this thread you'll find that I've linked several replays where pro players are indeed building token amounts of T3 land and then spamming T4 bots.

And its happening on maps that aren't popular in team game circles so the absurd idea that we shouldn't balance around maps that people actually play doesn't apply.

T3 is also helped by ACU (which is the main thing early T3 has to contend with) being nerfed in some upgrades and having Overcharge made more expensive in power and storage. Also The build time increase is not trivial and has a direct cost in how long it takes to get up an experimental. You can't spread out an exp, you can't drop it (T3 drops have actually been made stronger with the ASF/scout nerf). And so on.

Jesus man, nerfing OC means that rushed T4 bots are even more powerful... not less.

And people don't generally talk about post nerf units vs pre-nerf units because the two will never meet.

Yeah, but that's an issue, because its one of the simplest objective ways to determine how extreme a set of proposed changes are. If you test them and find that you've double the effectiveness of a unit with your changes that might be enough to convince you that what your doing might have impacts beyond what you intend.

@FtXCommando

Its at least 2, you have factor in the opportunity cost of mass lost from choosing to build a TML rather than upgrading a mex.

Note that I'm not claiming they're not OP, just that its not as simple as you're portraying it.

@FunkOff

For what its worth, I build MMLs so rarely that I sometimes forget they exist.

"OMG, a com is T2 PD pushing me, and my T3 land factory hasn't been started so I can't get T3 mobile arty... I should make gunships!".

I don't think it could hurt to have a stronger T2 siege unit.

I build far, far more firebeetles than I build T2 MML.

@arma473

Petric made a video to make people aware of the changes: "Now let's discuss the biggest focus of this patch: the long-in-the-works T3 rebalance that was started by Zock in 2015 or so."

https://www.youtube.com/watch?v=emz9WwAOoxM&t=6m17s

I admittedly haven't watched the entire thing, but it seems very focused on T2/t3 balance, and seems to ignore T3/T4 balance. I'll watch the rest of it later to ensure that I didn't miss that discussion.

There was a forum thread about the balance changes with 283 posts. The very first response to that thread brought up bricks and percivals. https://forums.faforever.com/viewtopic.php?f=67&t=15809

That's the thread I was referring to. Yeah, they considered the balance between bricks and percies (which I haven't been saying is a problem) but going through that thread I didn't see significant concern for whether it would throw T3/T4 balance off - which is the claim that I'm making.

I also didn't see much discussion about the magnitude of those changes, about how they're multiplicative, about how effective a post nerf unit would be against a pre nerf unit.

@Psions said in Points of Imbalance.:

@Fletching People don't have the forethought to realise that by increasing BP on T4 it also increases the earliest time a T4 might be up, meaning there is a greater window to eco before building relevant defensive structures.

They just think ooh Ras spam hive spam insta monkey.

The reason why T3 is stale, is because T3 mobile arty do not break the main base, they a gimped units and a successful viper spam would simply work better. T2 artillery at t3 stage is just too devestating for t3 maa.

So what you have nwo at t3 is 3 useful units

Long range, Mid range and raid units.

Aeon v Aeon t3 is much more interesting because of shield disruptors.

I have to keep repeating this, because I don't think people have really accepted this yet, but the 3696 nerf was a massive nerf, which reduced the combat effectiveness of T3 units by around 50%.

Have two current bricks fight one pre-3696 brick, and they're going to come out a bit even, with the pre-3696 brick perhaps killing both with a small amount of health left. Assuming you allow the pre-3696 brick to kite until the current brick catches it. I could be slightly off here, but I imagine it would be close.

Certain nerfs are multiplicative meaning that they stack in such a way that they drastically change unit balance.

Seemingly small changes to health and damage, when combined with other changes to say range can together significantly cripple a unit.

So if T3 seems weak, perhaps we should acknowledge that its weak because of patch 3696.

When you nerf a unit at a low tier, you have to then nerf all units at higher tiers, or the unit will be underpowered in comparison to units at higher tiers. The balance team did not do that for one reason or another.

What's most hilarious about all this, is I hunted down the discussion thread for that patch, and there was very little relevant discussion about these nerfs. People didn't realize how massive those changes were, and seemed not to consider the impact such changes would have on T3/T4 balance.

I think its clear it was a sloppy patch. They may have "fixed" T2/T3 balance, but they drastically reduced the effectiveness of T3 land versus both T4 and static structures in order to do it.