This is an ambitious and comprehensive balance proposal that is meant to be read in full; that said, this is a serious read and there is a TL;DR for those of you lacking the willpower. The thread that developed into this article has garnered nearly 150 likes and a considerable amount of positive feedback in the MechWarrior: Online forums, and I encourage you to spam the developers with this proposal if you support it. I strongly believe that this is the best (and perhaps only) solution to most of our long-standing and forthcoming balance issues, and that’s why I’ve put such an inordinate amount of time and effort into this. I would like to personally thank Tombstoner for helping solve the biggest player communication issue (weapon group reticle colors) and Phaesphoros for the terrific HUD mock-ups.
MechWarrior: Online is my kind of ‘mech combat – it’s gritty, it’s brutal, and face-offs between skilled pilots can take minutes. Unfortunately, the introduction of host state rewind for ballistic weapons had an unintended side effect: extreme, pinpoint damage now reigns as the undisputed king. Even before the latest PPC and AC/40 craze, the Splatcat was all about putting a ridiculous alpha in a relatively small location by face-hugging. It’s a problem that has persisted all the way through Open Beta and will become exponentially worse with the introduction of the Clans.
I will be the first to acknowledge that many aspects of the game’s balance need fixing: SRM damage, LRM coring/damage, SSRM coring, pulse lasers, hit detection, etc. I also won’t deny that hardpoint restrictions, penalties for overheating, tonnage limitations, and encouraging players to run lighter ‘mechs would cut back a lot of the cheese, but none of them are sufficient to solve the largest and most systemic balance problem that MechWarrior:Online (and all of its realtime predecessors) suffers from.
The crux of the MechWarrior: Online’s major balance problems is being able to deal more than 20 or 30 points of damage to a single location in a single click. Separately, neither massive alpha strikes nor convergence is a bad thing. Together, however, they create a nasty scenario where a couple of clicks is enough to vaporize an opponent. It’s bad, both in terms of gameplay and from a Battletech lore standpoint. There’s absolutely no incentive to fire two shots at 20 damage when you could fire one for 40 damage.
The Coming Storm
Most of the solutions I see being thrown around are solutions to the current symptoms of our problem: PPCs. Heat and PPCs are being debated ad nauseum simply because they are the flavor of the month. PPC boats are a serious problem, but to ignore large ballistics is a dangerous mistake. I fear that the tunnel vision about the current metagame will result in PGI ignoring the impending problems that will be introduced by new ‘mechs and the arrival of the Clans.
Not many people think ballistics are a huge problem right now, but I’d argue that’s only because we don’t have a particularly scary ballistics platform in the game. The AC/40 Catapult has to sacrifice speed, and the AC/40 Jagermech is really squishy because of its profile and XL. They’re still cheesy and too good for their weight, but they have exploitable weaknesses. If, however, PGI decided to drop a Mauler, Devastator, or Thunderhawk on us, there would be untold amounts of whine.
That exact thing could be said for PPCs: if we didn’t have any assaults that could boat 4+ PPCs, they wouldn’t be nearly as reviled. Heavies have to make serious sacrifices to boat them and don’t have the armor to take serious punishment, just like the current scenario with ballistics. When an assault starts boating something unbalanced, it becomes immediately apparent.
The minute PGI releases a ‘mech that can mount 4xUAC/5s plus change and has the ability tank a good amount of damage, the community will be shitting its collective pants. Even if they add ridiculous heat penalties to the autocannons, the Gauss rifle will continue to dominate. 2xPPC + 2xGauss and 3xGauss builds are impossible to solve with heat.
Even more dangerous than new ballistic ‘mechs are the Clan ballistic weapons. The Clan UAC/20 is a mere 12 tons and 8 critical slots. The double-tap will be able to put 40 damage on a single spot. Now imagine a ‘mech with jumpjets that mounts two of them. You’re imagining a stock Hunchback IIC, and it’s absolutely terrifying. Something drastic must be done to prevent Clan assault ‘mechs from being able to kill an Atlas with a single alpha.
My suspicion is that PGI’s plan is to avoid any assault ‘mech that can boat large ballistics simply because they have no good way to balance them – which is a really shitty solution. There’s no reason that balance issues should prevent awesome ‘mechs like the Mauler from showing up.
And if ballistics weren’t bad enough, Clan missiles will bring a host of problems: half-tonnage, no-minimum-range LRMs will make previous LRMageddons look tame, and lightweight SSRM6 packs will make the Splatcats of yesterday a laughing matter. Though the arrival of the Clans is not an immediately pressing issue, it’s better to have a system in place now than to ignore the horrible balance problems for later.
Disclaimers: PGI has my complete permission to use this system as-is or with any modifications they see fit. I would be happy to sign anything needed to prevent intellectual property from being an issue. Also, none of this has anything to do with the “Targeting Computer” piece of equipment.
My solution is to implement a scale that represents the load on the targeting computer (TCL). Each weapon would, similar to heat, have an associated targeting computer stress value (TCS). When a weapon (or group) is fired, the stress value of all about-to-fire weapons are added to the load on the targeting computer. The targeting computer load automatically dissipates at a constant rate of 100 per second.
When the load is between 0 and 100 (inclusive), there are no ill effects. When it goes over 100, all missile locks and Artemis functionality are lost, convergence stops working, and you begin to take an accuracy penalty (cone of fire) to any shots fired. Locking capability, Artemis, and convergence are not restored until the load on the targeting computer reaches 100 or below.
From 101 to 200, the accuracy penalty gets progressively worse (the cone of fire expands). Each weapon fires at its own accuracy offset so that weapons mounted in the same component fire in different directions. The pilot can continue to drive the targeting computer load up to a maximum of 500 by continuing to fire, but the effects of a targeting computer overload reach their worst at 200.
To clarify, you can’t get away with one free alpha strike; TCL values are added and penalties are applied before the shots are fired. My proposed TCS values for all weapons can be found in The Numbers section.
Clarification: Loss of Convergence
Convergence is the system that makes all of your weapons aim at the same spot; without it, all of your weapons fire directly forward (parallel, never crossing, out to infinity) from the place they are mounted on your ‘mech. Because all weapons fire directly forward without convergence, it has the same impact at 20m as it does at 800m. It is for this reason that I see the immediate loss of convergence as a necessary penalty – without it, snipers are the only builds affected. A PPC Stalker or AC/40 Jagermech at 100m will barely be affected by all but the most ridiculous cone of fire; the loss of convergence affects all roles and ranges equally.
Though I suggest that going over 100 immediately deactivates convergence, I would not be opposed to the gradual loss of convergence (linear interpolation between non-converged and converged) when the TCL is between 101 and 150 if the original implementation is deemed too abrupt and severe. Either way, the loss of convergence is absolutely key to this system working properly; a cone of fire simply doesn’t affect short-range combat.
Clarification: Cone of Fire
A cone of fire is a small angular offset applied to each weapon. Many first-person shooters have recoil that causes your bullets to spread more – that’s a cone of fire. The spread has a greater effect the farther away you are from your target. The cone of fire I’m proposing would be tuned for around 500m. Snipers would suffer the largest penalty, while combat closer than 200m wouldn’t see a dramatic impact.
The cone of fire is necessary for a few reasons. First, it affects chassis that can boat high damage in a single component (HGN-732 with 3xPPC in the RT). The loss of convergence alone would not affect such chassis as severely. Second, it adequately punishes those who massively overload their targeting computer. Barely going over the threshold is one thing – alpha striking six PPCs deserves a far more severe punishment. Lastly, it forces snipers to exercise fire discipline. I’m okay with brawlers continuing to fight (albeit far less effectively) with their targeting computer maxed out, but I see no reason to allow sniping with an overloaded targeting computer.
The Heads Up Display
The HUD will require a few minor changes to clue the player in about what’s going on. There needs to be a meter showing the current TCL and indications for convergence loss and cone of fire. The most important thing is that the player should always know whether firing a certain weapon group will ever push him over 100% load on the targeting computer. The weapon group icons around the reticle have had their colors changed to represent the TCL value after that group is fired (blue for 100 or below, yellow for over 100, and red for over 200).
Keep in mind that these numbers are meant to demonstrate the spirit of what I’m going for. Though I think these numbers are a pretty good indication of what relative balance would look like, it’s important not to get caught up in the minutia; this is just a starting point.
My TCS values are based primarily on two things: damage and pinpoint capability. For instance, the TCS of four medium lasers (20 damage) is equal to the TCS of a single PPC (10 damage) because lasers take more skill and effort to keep aimed at a single component; unless you and the target are still, laser damage almost always spreads. The same logic applies for SRMs and LRMs, and it’s why their TCS to damage ratio is so low. My numbers may not be perfect, but they’re a pretty damn good starting point.
A Few Scenarios
The following are examples of how TCL would interact with various ‘mechs. They’re meant to paint a picture of how shots need to be staggered and how alpha strikes will affect different builds.
The Work Required
At this point, a lot of you are probably thinking, “This sounds like an awful lot of work…” As someone who programs video games for a living, I feel relatively qualified to speculate about how much work this would actually take to implement.
- HUD/Mechlab interface design (5 days – art)
- Program TCL scale (3 days – programming)
- Add TCS to each weapon
- In code, adding to the TCL (2 days – programming)
- In game data file / associated code (3 days – data/programming)
- In the ‘mechlab UI (5 days – art)
- Add and hook up all HUD elements (5 days – art/programming)
- Implement convergence loss and cone of fire based on TCL (5 days – programming)
- Random, terrible, and unforeseen things (3 days – art/programming)
All the time estimates for programming are very generous. The only possible problem I forsee is the way their firing / convergence system works. It might require some refactoring to make all weapons add their TCS to the TCL and have an accuracy penalty applied before they fire, but it’s difficult for me to imagine that would be a huge change.
The art side of things is a bit vaguer since I don’t have much first-hand knowledge. Based on my experience working with artists, I figure a week is a reasonable amount of time to get a few HUD and Mechlab icons planned out and drawn. The HUD could be an absolute disaster, but I would hope it takes less than a week to hook up a couple simple elements.
Testing would be, by far, the biggest time-suck. This clearly needs to be tested thoroughly before being released into the wild. Personally, I think this is the perfect candidate for the upcoming test server.
Worst-case scenario, we’re talking about three weeks for two programmers and two artists. If everything goes smoothly, it should only take one programmer and one artist a single week. Knowing what I know about game development, the worst-case scenario is probably more realistic. Regardless of how long it takes, you simply can’t put a man-hour price on game balance.
- The Problem: Most of our gameplay imbalances result from the combination of high damage and weapon convergence. Convergence is not a bad thing on its own and neither is high damage, but together, they’re killing game balance. Battletech was balanced with random hit locations, while a first person shooter needs reliable and meaningful aiming.
- The Solution: Implement a second scale (targeting computer load or TCL) to limit extreme, pinpoint damage. Each weapon fired raises the TCL, and it dissipates rapidly (100/second). If it goes over 100, convergence is lost, you take an increasing cone of fire penalty, locks are lost, and Artemis stops working. TCL penalties are applied before the shot is fired. You can do extreme, inaccurate damage all at once or you can do stagger your fire to remain accurate.
- The Numbers: The goal is to limit pinpoint damage to about 20 damage per second. You’ll see the penalties are less severe for weapons that tend to spread damage. Examples of what you’ll be able to fire simultaneously with no accuracy penalty: 2xPPCs, 1xAC/20, 1xGauss+Change, 8xMedium Lasers, SRM24, LRM40. See The Numbers section for more.
- The Good: Balances extreme alphas of all kinds (particularly pinpoint), it solves ballistics (unlike a heat penalty), retains (and increases) the value of aiming skill, improves combat pacing, keeps heat untouched, adds an interesting tactical choice, and goes a long way towards balancing Clan technology.
- The Bad: Added complexity and some work for PGI.
- Why It’s Worth It: A new system is needed to bridge the gap between tabletop balancing and the precision aim of a shooter. I believe my solution adds a believable layer of tactical depth to the game while solving a host of current and future balance issues better than any other alternative.
- Hang on Just a Minute: If you have reservations, disagreements (particularly if you believe this solution is too complex), or an alternative that you think works better, I highly encourage you to check out the Rebuttal sections below. I have exhaustively addressed every issue and alternate solution I’ve seen, and it will help to get a much better understanding of why I’ve come to this solution.
Rebuttals I: Disagreements
In this section, I have done my best to answer every possible concern about the TCL scale. If you have lingering doubts, they are probably addressed below. Feel free to disagree, but I challenge you to present a better alternative.
Rebuttals II: Alternatives
In this section, I offer rebuttals to every alternative system I’ve seen presented. If you think you have something that would work better to solve our current problems, chances are extremely high that I found it inferior for one reason or another. Even if you disagree, you’ll know why I stand where I do.