MCV Balance Playtesting

this is coming from the guy who abuses them the most

Discussion about the game and its default mods.
Post Reply
OMnom
Posts: 255
Joined: Fri Oct 28, 2016 4:04 am

MCV Balance Playtesting

Post by OMnom »

Please go to page 10 to find the list of final changes.

Brief summary of playtesting

This thread has gotten very convoluted, and the information is all over the place. I've decided to start updating the OP with the updates and bumping the thread, rather than posting the new information in the most recent post.

These are the conclusions we have proof of thus far

-Increasing pillbox price from $400 to $600 decreased the average number of pillboxes in the first 10 minutes by approximately 5. Scaling of pillbox was reduced to about 1 static defense every 30s, on average. Minimal effect on strength of basepushing. Overall spamming decreased.

-Increasing pillbox build time past $600/15s makes it incrementally harder to get T2 tech. Infantry spam is buffed. Early game rushes become very deadly.

-Increasing MCV build time from 32s to 40s makes it harder to get T2 tech, and makes it very difficult to get T3 tech. Eco scaling is delayed, army sizes are larger, and production queues are typically reserved for more refineries/barracks. Increasing MCV build time to 48s slowed the development of economy significantly and made the resulting 1-base army even larger upon moving out.

-Decreasing MCV move speed slows the development of the economy. Similar effects to increased build time of MCV. Slowing MCV to 71 (heavy tank speed) makes it easier to catch the MCV and makes it more difficult to use in an aggressive base push. Slowing MCV to mammoth (60) base speed had similar effects to the 48s MCV, where the eco was signifcantly affected.

-$1600 WF buffs all vehicles in general, and speeds up the economy. This is a viable co-balance to negate the slower economy from 40s/48s MCVs

-Limiting barracks to a maximum of 7 made it easier to defend base pushes. However, by buffing the defender's advantage, games are more likely to develop into static/passive/stalemate situations.

-Moving refineries or barracks to defense tab buffed MCV basepushing by allowing barracks + refineries at the same time. This change also sped the pace of the game up drastically. Using the data from the pillbox price test, moving everything to 1 tab would signicant buff infantry spamming and early rushes due the indirect increase of the pillbox build time (cancel production + build pillbox).

-In general, all changes that protect or create a stronger economy helps support T2 tech, with a marginal effect on T1/pb spam. Making the economy more vulnerable or slower supports T1 spam and makes T2 tech harder to get to. This is inherently affected by map balance as well.

-TabWF edit allows Dome to be built 3m earlier, on average, relative to the current bleed. Tech center is built approximately 2m earlier. The scaling of static defense is, on average, 0.5 lower at all points between the 6-10m range.

-Reduced build time on Dome and Tech center has a minor effect (30s-1m) on the time when the Dome and Tech centers are built, relative to the current bleed. This slightly increases the impact of aircraft and v2/artillery, due to both players having less overall assets when T2 comes into play.

-T2 and T3 tech appears only when players are not fighting each other and when they have already optimized their economy. There is enough evidence to support that the risk of teching is too high compared to the rewards you get. The investment needed for tech to pay off is too high for it to offset the smaller army size, even with the TabEdit.

-Basepushing seems to be kept in check by V2/Artillery and by the speed of the game. Base pushing is the quickest way to dump all of your resources into assets, but it is focused only on 1 location. Trading cost efficiently is one way to stop base pushing. The other way is being able to spend your money

-It is difficult to spend all of your money all of your money in RA. The starting spending rate is $42/sec, with the exception of certain units, which translates to around $2.5k a minute. However, this rate increases with additional scaling, up to $5k/min. In most games, the money distribution will be maxed infantry ($5k/min), 1 war factory ($2.5k/min), and 3 MCVs ($3.6k/min * 2 queues), which adds up to around $14.7k of "maximum" resource spending. For reference, a 6 harvester economy will give you around 10k resources / min.

-$1000 yak, $1100 hind, $1800 LB, and $1700 MiG, +$100 on V2 and artillery. --- minimal effect in most games. Increased unit stats make T3 units more worthwhile to get, relative to their cost and how difficult it is to get there with a solid economy.


Things that still need testing and/or currently undergoing testing:

-TechEdit vs TabEdit with the 40s MCV and $1600 WF -- done.
-Buffed vs unbuffed tech -- done.
-TechEdit and TabEdit on a low eco map -- in progress
-MCV speed reduction with building placement delay -- done (definitely not good)
-Reduced build time of radar dome and tech center -- done
-Tank buffs -- done
-Con yard/MCV in defense tab -- done (very drastic change)
-Pillbox damage adjustments -- done
-Remove all build time reduction modifiers from MCV -- in progress

____

Original post:

As many of you may know, I've been spamming mass MCV games, partly as a new playstyle, but mainly because I wanted to gather enough games to assess whether or not MCVs are in need of a change. I think I've played enough of these games to figure out why MCVs are so good and how to change them properly.

The MCV is very strong because the opportunity cost of an MCV, as a function of strategical value, is lower than all other combat vehicles

Take the ore truck for instance: It takes 27 seconds to build, approximately 5-10 seconds of travel time to an ore patch, and about 2 trips to earn it's money back, which we will generously call 20 seconds. In total, it takes, at most, about 60 seconds for our investment to pay off.

Let's look at the MCV now. It takes 32 seconds to build, and generously speaking, about 30 seconds to travel to a new ore patch to deploy. Assuming you queued up a refinery so that you can place it immediately as you deploy your new MCV, and after accounting for mining time on 1 harvester, it'll take around 120 seconds for our investment in the MCV to pay off for itself.

As a third example, lets look at the Minelayer. It takes $800 / 20 seconds to produce; however, it will most likely take a moderately long time to travel to the enemy's ore patch...lets call this 25 seconds. It may take even longer for your opponent to even set up a refinery at that ore patch, so you'd have to wait anywhere from seconds to 10's of minutes for your opponent to go there. Alternatively, you can mine a strategic pathway, but unless you force your opponent into that path, the time it takes for your investment to pay off is indeterminate. This logic applies to all combat units; there is a significant amount of time that it takes for a combat unit to earn it's value.

Of course, in-game factors will change the timing and how quickly our investment starts to pay off, but for now, lets keep this idea in mind.

Problem #1: The MCV lowers the opportunity cost of all combat units

MCVs travel quicker than infantry. You can have a barracks ready to be placed the moment you deploy your MCV. Therefore, the quickest way to move infantry from point A to point B is to use an MCV to build a forward barracks to start production there. Late game, this applies to vehicles and aircraft as well. The quicker you start using your investments, the quicker you'll be able to get your returns (Note: this does not assume you have to trade cost-effectively to be able to get value from using a unit).

This is all without considering that you can build a pillbox/tesla/flame turret to secure map control, that each additional MCV reduces the build time of all structures by 7%, or that you can build barracks as meatshields for half-the price of a medium tank that also poop out infantry, at a fraction of the time it takes to just make a tank .

Now, I will confess, I love this mechanic, and it should absolutely stay in the game. It's one of the things that makes C&C unique. I've played many RTS's in the past, and this is a mechanic that is more or less unique to RA. But in it's current state, having one unit capable of reducing your opportunity cost, increasing your map control, and increasing your economic potential... its an easily abusable mechanic if you build multiple MCVs, and it pisses off a lot of people.

Problem #2: The opportunity cost of tech units and buildings is too high

Every serious game that is played involves at least getting a 2nd MCV before getting tech. Let's put aside the obvious question of, "How do we make both tech and eco viable" for one moment and take a look at the opportunity cost of getting tech.

Radar dome = $1800 / 44 s. Helipads are $500 /12 s each, and you probably want at least 2 of them. Tech centers are $1500 /35 s. The list goes on for a while. Tech units are expensive, and suffer from the same problems that normal infantry and vehicles do; it takes a significant amount of time for us to get our value back from our units. Not only that, during the time that a player is teching, his economy is going to be at a standstill since you can't build more refineries. The opportunity cost of higher tech is so high that some players opt not to tech at all, and rather invest everything into eco and more units. It's also the same reason why people don't like building mammoth tanks and why people love playing allies; mammoth tanks take forever to get to the front lines, and allies have cheap, quick, and cost-efficient spammable units.

Now don't get me wrong. I'm not advocating for the devs to nerf the MCV into oblivion or to buff tech units to the point where you have to get hinds before MCVs. What I hope to have accomplished is to convince people that 1) the MCV needs to be rebalanced and 2) why teching is not as good as getting another MCV. I will now provide 2 ways to balance this in a way that does not involve touching the MCV stats itself. This is important because any stat changes to the MCV will also affect the player who decides to tech before getting additional MCVs; in other words, nerfing any aspect of the MCV will either exacerbate the problem, ruin the effectiveness of the MCV, or will have no effect on the problem.

Solution #1: Increase Build time reduction for tech buildings (radar dome, tech center, IC, and chronosphere) to 15% per additional MCV, to a max of 50%.

This would leave the MCV's strengths alone, but I hope it will buff higher tech enough to the point where it would adequately match up against mass eco/mass map control styles of play. This is seemingly the the most innocent buff, but it'd also might be a dangerous one, since allowing for faster radar dome tech will open the door to mass artillery much more easily. All in all, a possible solution, but not my favorite.

Solution #2: Put the Radar Dome and the Tech Center in the Defensive building tab

Currently, tech and eco are both linked together in one tab...it's impossible to expand your eco to new ore patches and to tech up at the same time. While preventing people to go for both eco and tech at the same time may seem like a good idea, in reality, it's the main reason why defensive building spam is so prevalent.

Remember what I was saying about opportunity cost? When you place a defensive structure down in the middle of a battle, the only opportunity cost you have to worry about is the build time. Base pushing is essentially the quickest way to get value for your money. When I tech, I have to build a radar dome, and I have to wait 20s for 1 artillery to come out, not considering that I also have to move my artillery to the front lines. Not only that, I have to stop developing my eco and stop expanding my map control (technically, you could try to contest the map while teching, but it's very risky vs a base push). In that amount of time, I have enough time to set up a refinery with 1 MCV, and base push with 3 barracks + static defense on a different side of the map. In this case, I'm happy with giving you the tech advantage, because I now have the eco and the map control advantage -- a 2-for-1 deal. Even if I did hold a 2-1 advantage before I tech, it's just giving your opponent some breathing room while they get to solidify their position...these scenarios tend to be the ones with artillery stalemates.

Now let's move tech buildings to the defensive structure tab. Now, my eco advantage is negated because my opponent can build a ref while getting tech at the same time. I Know if he's teching, he can't build pillboxes, so in order for me to abuse this, I have to cut my eco and start base pushing/mass barracks immediately in order to punish him.
Or, lets say he decides to go for map control and tech instead. Since he can't build pillboxes here and is reliant on a mobile army for map control, I should be able to contest his map control advantage with multi pronged attacks without fear of pillboxes popping up like daisies. He'll have to use his mobile army and scale up while he's teching, so he'll have to fight for map control in order to deflect my attacks. During this time, I'd probably just eco with my MCV. In the end, what will most likely occur is that one of us has the eco advantage, while the other has the tech advantage, with both sides having a relatively equal share of map control.

Alternatively, creating a separate "Tech" production tab after building the Service Depot may also work, but could potentially be very tricky to balance.
Last edited by OMnom on Sun Mar 19, 2017 11:09 pm, edited 16 times in total.

User avatar
Smitty
Posts: 191
Joined: Mon Sep 26, 2016 1:33 am
Location: Oklahoma

Post by Smitty »

Good guy Omnom. Breaks the game just so he can show the world how to fix it. :lol:

User avatar
kyrylo
Posts: 193
Joined: Wed Mar 11, 2015 8:46 am

Post by kyrylo »

So why not just to make MCVs slower? If you base push, you can easily undeploy and retreat. With slower MCV base pushes would be much harder. You wouldn't be able to retreat as easily and you wouldn't be able to send another MCV quick enough to back it up (in place of your destroyed one). It's really easy to "dance" with your MCV as of now.

Speaking of the speed value, why not to make them as slow as the speed of mammoths from previous releases?

User avatar
Doomsday
Posts: 199
Joined: Thu Nov 10, 2016 11:45 am
Location: Helsinki

Post by Doomsday »

MCV has highest cost / build time value of all units ingame. Almost everything in game is consistently balanced around 40 $/sec. For some reason MCV is 62,5 $/s. Mammoth Tank has 50 $/s. (Numbers listed here are based on calculation.)

I looked at vehicles.yaml and apparently Mammoth Tank and MCV are the only vehicles with BuildDuration and BuildDurationModifier.

I don't really understand why MCV has a build time modifier. Perhaps the answer is something along the lines of "bad players who lose their first expand get less punished". In which case Omnom has proven it works both ways - good player can push this mechanic to his advantage.

EDIT: Slowing down MCV speed would have dramatic impact as players wouldn't be able to pack their MCV and run it to safety. I would not want to remove comeback mechanic of those games when opponent successfully denies your first expand.

klaas
Posts: 208
Joined: Wed Feb 06, 2013 9:38 am

Post by klaas »

Great analysis OmNOmNOm.
Doomsday wrote: I looked at vehicles.yaml and apparently Mammoth Tank and MCV are the only vehicles with BuildDuration and BuildDurationModifier.
Removing the modifier for the MCV sounds like a simple way to rebalance a bit. However, it does not diminish the fact that building tech comes at a considerable cost. Maybe a decrease in cost for the radar dome and tech centre would help.

Maybe we could also increase the power consumption of barracks significantly, to prevent barracks spamming to boost production of blobs early game, making the pure Eco/Spam build just a little bit slower.

User avatar
Smitty
Posts: 191
Joined: Mon Sep 26, 2016 1:33 am
Location: Oklahoma

Post by Smitty »

Maybe this should be its own forum post, but how the heck are balance decisions made? Who are the ‘deciders’?

A great concern of mine is that none of the players I talk to, even those at the top-level, agree on what needs to be fixed and how. Some even have entirely different visions of what direction the development of OpenRA should take.

Murto the Ray
Posts: 486
Joined: Mon Nov 10, 2014 4:34 pm

Post by Murto the Ray »

Smitty wrote: A great concern of mine is that none of the players I talk to, even those at the top-level, agree on what needs to be fixed and how. Some even have entirely different visions of what direction the development of OpenRA should take.
Because balance is a completely subjective case. One could say the game is balanced at the moment and that this is fine, another would say this is grossly imbalanced and a lot needs changing. There isn't really a central definition for what balance is for OpenRA so people will always argue.

Murto the Ray
Posts: 486
Joined: Mon Nov 10, 2014 4:34 pm

Post by Murto the Ray »

Personally, i think spamming MCVs works not because tech isn't good enough but because the power gap between structures/defences and units is too large. Units have a really difficult time bringing down defences and structures and as a result it is more worthwhile to spend your income on inexpensive turrets planted right beside your opponents base rather than an expensive army that gets shredded to pieces in the face of defensive turrets.

For half the price of a heavy tank i can get a turret that can take down a heavy tank and that i can repair over time with a little cash. With $400 (even $600 next release) i can buy a pillbox which tears down so many more infantry than i could have bought with the same cash - all whilst being able to repair it. For a few hundred $s i can buy walls to block you from entering my base. For $800 i get an aa gun with a huge view range and easily able to take down multiple $1000s worth of air units. For $600 i get a flame spouting cannon capable of doing considerable damage to infantry. for $1200 i get a long range anti-land electricity tower that can take down tanks with ease. For $750 i get a SAM..... er... yeah.

Defences are extremely powerful. Plant one next to each of your MCVs and their view range makes it extremely difficult to attack without significant losses. It may be taking the defenders advantage a bit too far. Defences are on a completely different level to units in terms of health/damage per $. Expansions are rarely risky and base pushing is usually worth it if you can afford a couple of barracks and a lot of defences. Perhaps making defences weaker and the repair faster would make sense as then once you have a critical mass of units they would burst down the tower before it repairs and you are able to breach with ease once you have enough units focusing a tower but have difficulty otherwise. Though I feel this is an issue to be discussed elsewhere perhaps as its the kind of thing that is very subjective and can screw around with the gameplay of OpenRA.

As for MCVs; I agree with what has been said about production time. The decision to make them build faster was likely made a long time ago when the competitive scene for OpenRA was completely different, it doesn't make sense to have it anymore.

Making MCVs slower has been mentioned but what about making them deploy/undeploy a lot slower instead/as well? Giving them a 10s deployment/undeployment timer would certainly make it much easier to punish base pushes and also catch out expansions unguarded. Especially if they couldnt be built on in that time.

I dont agree with tech buildings being made differently, they feel fine as they are. SoScared and i have discussed reducing the build time for late game units such as the mammoth, longbow, chrono tank, tesla tank, cruiser, missile sub without altering their cost in order to allow players to get high tech armies out in a similar time to a lower tech army but for a higher price. This will probably be mentioned in another post.

User avatar
AoAGeneral1
Posts: 597
Joined: Wed Apr 29, 2015 6:11 pm

Post by AoAGeneral1 »

*cough cough*

lucassss
Posts: 144
Joined: Mon Jan 04, 2016 1:55 pm

Post by lucassss »

Maybe the problem is not only with the MCV, but that most maps have too little ore per patch?

Medium Tank
Posts: 12
Joined: Sat Apr 30, 2016 4:35 am
Location: Arizona

Post by Medium Tank »

I have a thought about this. Thinking back to TD's balance with MCV's. You have a wait timer after you place a building and cannot place another structuure within the radius of that MCV for a few seconds. I think this is what could be a possible solution to the Basepushing problem that Om adressed. If you made the MCv have a longer deploy timer and buildings add an extra sceond or two it could negate the MCV plop and spam of structures. In a 2v2, once you deploy that MCV, theoretically, you can have a barracks, WF and 2 pillbozes within a second. another 15 seconds, and you could have 2 more pillboxes and 2 more barracks.

This is just my thought process to the thread. Maybe creating a system like TD's mechanics that help negate the impactfulness of the MCV. I don't think this would be a solution that is implemented. I posted this to stur the pot with ideas of other probably solutions.

OMnom
Posts: 255
Joined: Fri Oct 28, 2016 4:04 am

Post by OMnom »

Before you guys start shouting out "lower the cost of X" ,"change the build time of Y", or "lower the speed of Z", please consider the following:
1. People are needed to test these changes, and a large sample size of these games is needed to determine whether or not if those changes positively impact the game.

For this reason alone, I think moving the radar dome and tech center over to the defense tab is a better idea because its easier to test out and the changes will be clearly evident in the first 10 games. There are no numbers or fine tuning needed -- simply put the radar dome and tech center in the defense tab. If my change does not work, then we can start altering the value of the MCV.

2. There are a large number of combinations of different changes that could fix a particular problem. As such, there will be a large number of people who believe in fixing a problem in one way, with an equally large number of people who are partial to another.

There are a lot of ways to fix one problems, with some being worse than others. I may have presented my argument against directly nerfing the MCV poorly, so I will reiterate it again. If we were to nerf the MCV as a solution to nerf mass eco/base pushing, all it will do is slow down the game for the base pusher and make it even harder for the teching player to expand.
-Slowing down the MCV will just slow the game down...if you want to play a slower game, just set the speed on Slower.
-Increasing the MCV build time significantly may have unforeseen consequences to the stability of the game. Right off the bat, I will say that a significantly expensive/longer build time on the MCV will force people to move their original MCV out, which will inevitably lead to base pushing anyways.
-Having deploy delay or build placement delay is a good idea, but this is going to take a very long time to get to balanced numbers. My proposal is quicker to implement, and if it doesn't work, we'll know just as quick.

Short summary and reasoning for what I am proposing:

MCVs control tech, production, and economy in one production tab. If we put the radar dome and tech center into the defensive tab, now there would only be production and economy in one tab, with the other having defense and tech. This would balance the macro infrastructure among 2 tabs, which in turn would create a new "food chain."

The current food chain is Refs/pillboxes and Rax/pillboxes > Tech/pillboxes. If we move tech to defensive production tab, now we have 4 combinations: Refs/Tech, Refs/pillboxes, barracks /pillboxes, barracks/tech. Allowing Ref/Tech gives people another option besides spamming pillboxes. Also, allowing barracks + tech to happen at the same time means that you won't have to sacrifice your army size for more tech.
On paper, Ref/Tech and Barracks/Tech should have distinct advantages over base pushing defensive structures. If one is too slow, Ref/Tech will outlmacro the base push. If one is unable to do sufficient damage, barracks/tech will outscale the base pushing player.

I've made some quick edits to a map that puts the radar dome and tech centers into the defense tab. I've also required radar dome to require the SD instead of the refinery. I will be trying this out tomorrow.

TL;DR

Put all the tech in the defensive tab, and maybe people will actually use tech to against pillboxes instead of building counter pillboxes?
Smitty wrote: Maybe this should be its own forum post, but how the heck are balance decisions made? Who are the ‘deciders’?

A great concern of mine is that none of the players I talk to, even those at the top-level, agree on what needs to be fixed and how. Some even have entirely different visions of what direction the development of OpenRA should take.
I will make a different thread to address this later.
Last edited by OMnom on Tue Dec 13, 2016 10:16 am, edited 1 time in total.

Murto the Ray
Posts: 486
Joined: Mon Nov 10, 2014 4:34 pm

Post by Murto the Ray »

Until people can decide on what the issue actually is, testing solutions should be put on hold; its obvious that there isn't only one opinion here.

Blackened
Posts: 310
Joined: Sat May 21, 2016 6:27 pm

Post by Blackened »

Murto the Ray wrote: Until people can decide on what the issue actually is, testing solutions should be put on hold; its obvious that there isn't only one opinion here.
Bingo. There a so many balance issues and directions this game can go.

Lucassss brought up a good point that shouldn't go unnoticed. Name a map where 1 base plays are actually viable? The closest in the current map pool is sidestep as you have 1 large ore patch in the front a smaller one in the back. Even then, you will exhaust yourself of ore before tier 3 tech becomes viable.

If you do not expand you will lose.

Or you could argue that defensive structures aren't used for their intended purpose. I.E. defense.

You could twist Omnoms argument a bit and say that unit cost effectiveness has diminishing returns in tech. Where you would think tier 3 tech should be better than tier 2 and that tier 1 that really isn't the case in RA. Infantry are the most cost efficient units in the game. Infantry remain crucial at every stage of the game.

Lastly, my personal favorite to argue about is the lack of mobility in the game. There are far and few units that are mobile enough to poke at bases, raid eco lines, or back line harass. Those that do can easily be shut down. Mobility is a big reason why I think Fiveaces loves v2's and migs. Both can do the aforementioned maneuvers with little or no support and when they are used in conjunction with heavy support perform very well.

Mobility is a big reason why defensive structures are so powerful. Rarely can you choose to sidestep them.

You can't increase map size without hindering main army mobility more and causing base pushing/expansions to be even more prevalent. You can't decrease map size to make infantry more mobile either. It becomes to chokey especially with walls in their current state.

zinc
Posts: 657
Joined: Sun Feb 09, 2014 3:46 pm

Post by zinc »

Smitty wrote: Maybe this should be its own forum post, but how the heck are balance decisions made? Who are the ‘deciders’?

A great concern of mine is that none of the players I talk to, even those at the top-level, agree on what needs to be fixed and how. Some even have entirely different visions of what direction the development of OpenRA should take.
I think we should be able to download custom factions/rules in the same way as maps. That way (perhaps) different experiments could be easily tried out and we would see what style of games became popular.

Post Reply