Cost and buildpower analysis of all constructors in BAR

[late pagoda]

This is a breakdown of the attributes of builders and how it relates to their buildpower, i.e. which you should build.

This compares metal cost, energy cost, buildtime, movespeed of every constructor, and ranks them according to their total buildpower.

Most of this will not be surprising to experienced players, but you may notice some tidbits you never realized before.

The #1 spot, to my great surprise, was the T2 Naval Engineer. This guy fucks. If you happen to have access to T2 sea, there is no constructor that even comes close. It beats out every single option in metal, energy, and buildtime, and it's really fast!

The more realistic winner is, of course, the nano turret. Its obvious downside is that it's immobile, but its buildtime, metal cost, and energy cost are all better than all mobile builders besides the naval engineer. I was somewhat surprised by how much better it was than T1 cons. If your buildpower doesn't need to move, always make these.

Some further notes - the minelayer is not particularly efficient as a builder, but do note its incredibly high movespeed for a t1 builder.

T2 builders are very very inefficient sources of buildpower, and their terrible movespeed cannot be overlooked. Don't be fooled by the high buildpower of the T2 vehicle con, its cost makes them highly inefficient. Similarly, decoy commanders despite having the same BP as a commander have terrible efficiency as builders.

The mobile winners are the assist builders from T2. The biggest winner here is the Butler (Arm T2 assist bot), it has great costs across the board and high movespeed to boot. The only downside is the terrible build options. The twitcher (core T2 assist bot) is not quite as efficient in cost, but its incredible movespeed (fastest non-air builder!) and powerful buildset make it an all around fantastic choice.

So make nanos if you can, if you need it to be mobile make T2 assist builders, only make regularly constructors when you need the buildoptions!

[opaque imp]

Quote of the year

[woeful loom]

guess that partially makes up for the increased cost of underwater eco

slightly

[wary willow]

naval eco produces more E and M than land eco

[frosty panther]

Can somebody help me make sense of this table? I guess i'm having trouble figuring out the end goal of this, is it to find the best builder unit in the game? If so what does BuildPower per Metal mean?

[reef kettle]

every constructor has a build power stat (how fast they build things), and every constructor has a metal cost

[hardy lotus]

naval engineer not surprising, its cheap as hell

[frosty panther]

Ah ok so it's simply Build Power over the Upfront Cost. Thanks a lot I was already stressing over the possibility that you'd have to consider 2 BPs, one for Metal and one for Energy and that i hadn't really understood anything of the game

[forest parcel]

Also not mentioned and generally underestimated is the T1 naval constructor

[split moth]

good idea to make such table, but its wrong calculation: because it does not account for the fact that cons produce energy; eg con veh arm produce +10e !! this is huge, its literally walking half solar collector! so to calculate BP per m it makes more sense to first substract the +e; so e.g. good model to calculate would be to convert +e into cost of equivalent solar; so in con veh +10e is half solar so ~77m cost (given solar cost 155m); con veh cost 135m - 77m = 58m; so it could be argued con veh real cost for the purpose of comparing with nano tower is just 58m! and then it will turn out con veh (also con kbot) have much better BP/m ratios than nano tower itself! so arm con veh is 90/58=1.55 BP/m so way higher than anything else on the list

so @late pagoda conclusion is wrong, I think better conclusion is: build t1 con vehs or con kbots for cheapest buildpower; only build towers when you need scale and wanna save time on microing cons (eg so they do not stuck in movement) or already so many cons assist your factory that more cons would not have build range; also build towers for long range for the purpose of auto repair and reclaim; also build t1 cons for mobilility (eg assisting t2 mex build assisting far away) and for armour (t1 cons, esp veh have much more armour per m than towers); overall in most typical scenarios it might make sense to build ~10 t1 cons (for the purpose of factory assist and eco porc before making first towers) + additional cons for map expansion

[gusty island]

this idea is correct, but -77m cost for 10e income is not correct, because basic solar collectors are the least efficient energy production, replaced soon after better energy production is available and also almost never built on windy/tidal maps. so if you wouldn't build a solar then 10e income cannot be worth 77m.

my personal method to get a reasonable working value of the 10e was to find the 'cost of energy income' with a fusion reactor which is pretty efficient, average it out with the 'cost of energy income' with a solar collector which is not efficient to find something in between, converting the energy cost to metal cost with an arbitrary exchange rate of 15e:1m. this method of estimate is most accurate for 1v1s which usually end somewhere between solars and fusions, but it's least accurate for longer games where you end up with afus since those are so much more efficient and the 10e income from a cons is worth relatively less.

with my costing method above, i found the t1 veh cons to be ~10% more expensive than con turrets. the main difference is that the con turret has nearly half the build cost per build power gained, so it allows you to ramp up build power the fastest. there are many other pros and cons of either choice outside of this.

[split moth]

I dont agree with this fus method either 😄 because the whole idea of t1 cons vs nano towers is early game; late game it doesnt matter cause u need scale of turrets anyway! so if we talk only about v.early game you can only make decision like: build turrets +solar/wind OR build cons (which in itself work like part solar); thats why i compare to solars; also not to wind since cons give constant non-variable income just like solars

[gusty island]

a reasonable take. there's also midgame though, where t1 cons are still a good option for scaling build power, and where 10e income doesn't cost 77m

[split moth]

mid game (mass t2 for me), depending how u define it, but one of main things is you lack APM/focus time and thats why u want automation like towers; early game (i mean in 8v8s) you have more time to micro manage many cons

[torpid sluice]

@split moth @gusty island Necroing old thread.. :)

I have some suggestions for this.
We already know convertion formula for e/s <-> m/s, so if you have the con cost in m/s you can subtract how much it generates. Problem is m/s cost varies with build power used, and I don't have a good suggestion for a standard BP number to use.

Other problem is how much the con returns depends on lifetime.
It might be fair to simply check how much metal a con makes (or would make) in a minute and subtract that from its cost. This number corresponds with a buildtime of one minute with the m/s route.

Anyway, it doesn't seem to change the order much so it seems to be negligible in this regard.

[sharp plume]

A structure that produces energy's lifetime is just as relevant as a unit's lifetime in many cases. For t1 cons a reasonably reliable formula would be subtracting from their cost their asolar-equivalent in e production. For example, it takes 7.5 t1 veh cons to produces as much e as an asolar, so divide the price of an asolar by 7.5 and that's how much to subtract from the price of the con. With t2 units it would generally be more accurate to use fusion or afus % values, but that general rule is probably the closest you can get to accurately quantifying this.

this method was used for deciding corprinter's price and energy production. Its +20 e/s causes it to be more cost-effective than a nano if on an asolar economy, but less cost-effective than a nano once on fusion economy.

These exact values all go out the window on windy maps, but accounting for variable energy sources would make things too messy

[torpid sluice]

Hm I guess it is just an issue of conversion rates. You are kind of pigeonholed into asolar for T1 cons, but 50M rebate for a T1 veh con sounds kinda high imo.

[sharp plume]

it's why you shouldn't rush nanos on solar maps. 2 t1 cons are substantially more cost-effective than needing to build an extra basic solar with the nano

[coral harbor]

howso? underwater fusions are less efficient than afus?

[sharp plume]

People see maps having higher tidal number than average wind speed, and converters with marginally more efficient conversion rates, then proceed to incorrectly assume it's universally better.

The advantages of naval eco at t1/in general:

Tidals are consistent e income while often being more cost-effective than solar and more space-efficient than wind farms.

Underwater structures are more expensive to raid than land structures.

At t2:

Fusions are underwater (read: safer).

You can make mass naval engis instead nanos, saving some resources both up-front and with the ability to keep all of them active wherever you build things.

Disadvantages:

Tidal is completely blown (pun intended) out of the water (pun intended) by wind eco on all 8v8 maps i know of that have water, and most maps in general, particularly for arm but even cor to somewhat lesser degree.

No afus.

The e converters' increased conversion rates (both t1 and t2) are often barely enough to offset their own increased cost, let alone offset the reduced income you're getting from naval e sources compared to land ones.

No cheap way to defend floating eco structures from hovers.

[coral harbor]

naval engies are nice for fusions but fail badly when you try to boost converters with them; they just path derp for hours and trip over themselves

underwater fusions are a little safer yes, but the converters are not, especially considering they're vulnerable to almost anything

tidals are nice in their consistency, but are metal intensive / poor compared to wind on most maps

[sharp plume]

That's why i specified underwater stuff is safe. Floating structures absolutely suck to defend.

On top of all the listed stuff, it doesn't help that usually t2 shipyard can't happen until you've finished winning your lane, and it's not like you can buy a teammate's sub con to upgrade your underwater mexes on most maps either, so t1 eco just ends up lasting an eternity.

But who knows, maybe new experimental navy updates make t2 navy actually viable for winning sea, so eco actually gets to grow

[sharp plume]

I usually find that there's a certain critical mass of naval engis where they start getting stuck taking turns blocking the blueprint where it goes from a minor nuisance to "you need to micro these away from their con basically forever". A little trick that seems to help is to put them on repeat and give them an area repair command over wherever your eco will be, so they don't constantly try to go back to floating on top of the t2 con

[placid loom]

It would be nice if we also had a list that included the units that don't build things, like the rezzbot, webber, and uw rezzunit, so I could see what the best unit for repairing and reclaiming is. The chart is also missing the commander and commando.

[wanton parrot]

so spamming t1 air cons is more efficient than t2 air cons? interesting

[hot sage]

The workers generate e that makes it harder to quantify

Like on solar maps the e gen from vehicle con is pretty strong

[signal turret]

Where's the spreadsheet where you got all of these stats? It seems extremely useful!

[unkempt latch]

And nobody talks about how is the Hover con ?