1
If you don't care who I am, scroll down to the next bold line.
First off, I can't make a legitimate claim to be smarter or better than any other gamer, and I don't know all the trade secrets you learn when you become a "real" developer, however, I want you to know that these silly ideas have at least some basis in experience. 22 years of playing just about every video game under the sun, starting with MW2 on my dad's IBM, piloting from atop a chair augmented with the kind of massive phonebook you only saw in the early 90s, to a 9 year old Steam account valued at nearly $6,000, combined with a mind that derives joy from deconstructing new systems have helped me form a pretty vivid picture of the very highest and lowest points video games have reached and the design choices that put them there.
That said, I have two confessions. 1: I was pushed away from MWO for several months by the packet overflow bug. So even though I started writing this when the scaling feature was first announced, the new maths thread is what pushed me to finish it. Better late than never, right? 2: I'm not going super deep into numbers and theorycrafting, it's not my thing. The purpose of this post is to propose something that's more intuitive and easier to understand at the beginning-to-intermediate gamer level, while still addressing the dominance of high alpha builds and even adding some spice to the game. That's why I'm just going to give some hypothetical examples and let the devs figure out the right values. Let's not talk about why my numbers would be unbalanced, I already know.
The Big Problem: Trapped in a Prison of Your Own Design
It's my opinion that the best games balance their numbers around a small set of rules that are easy to understand and scale evenly across everything they affect without the need for exceptions. This is opposed to creating game mechanics to specifically to address stubborn outliers that won't lay flat on the balance table (the WoW PvP approach), or flooding your game with so many options that a few viable builds emerge by law of averages (those ridiculously complex space 4x games). Obviously, because you need a variety of weapons and abilities to make the game interesting, there will be some that, while mathematically balanced, express certain strengths that are inherent and irremovable from the game's genre. Because of the nature of an FPS, sniping weapons will be more dominant over damage-over-time weapons due to the difference in the length of time the player must maintain their aim to use the weapon to its full potential, even if these weapons would be perfectly balanced in an RTS, auto-targeting type of system.
These strengths can be countered by nerfing their stats, but that makes them feel underwhelming and bland. If one weapon is made to be mathematically inferior to account for an assumed skill deficit in the player, than those weapons will go unused by players that have no such deficit. The alternative is to exaggerate their weaknesses, which introduces an element of danger that benefits more knowledgeable players on both the firing and receiving players' ends. This raises the skill ceiling for that weapon, giving lesser-skilled players more room to grow and competitive players more opportunity to make and capitalize on mistakes. On the downside (for the kind of players that prefer to play games at the surface level) it also adds a learning curve to the weapon, but that curve can be made approachable if the formula that defines it is simple and clearly communicated. When a mechanic applies differently to every weapon, it creates many different curves, instead of one master curve that works across the board. It's basically the difference between learning how to convert centimeters, meters, and kilometers, versus how to convert dollars, pounds, and kroners. Like that comparison, I can see max alphas and multipliers fluctuating like exchange rates on a regular basis to cope with the changing meta.
How It Affects MWO: Reward Outpaces Risk
We already have numbers to tweak with base damage and heat values, but TT fanatics aside, changing those won't fix the problem because weapon strengths are compound while their weaknesses are linear. Firing 4 PPCs as an alpha increases their power exponentially. On top of the fact that you deal 4 times more damage than a single PPC, you also reduce your odds of missing and inflict all the damage at once to a single component, destroying that component or mech sooner than you would have using chain fire and preventing potential damage that player may have dealt during the time your weapons were cycling toward the killing blow. If the target is firing on a mech in critical condition, a quarter of a second can mean the difference between life and death, and that mech may not die immediately in its next engagement. It may go unnoticed or ignored and go on to deal hundreds of more points of damage while in a critical state, or it may assist your team by capturing unguarded points, all because you dealt your full damage to its previous attacker in an instant, rather than over time.
Yet despite these potentially massive advantages, the heat cost you pay is a static 4 times normal, with no additional penalties emergent in the gameplay. Since there is absolutely no difference in the performance of your mech between 0 and 100% heat capacity, the only way the additional heat cost expresses its intended effect is when firing a full alpha would surpass the 100% threshold. All that is needed to overcome this weakness is to learn the exact amount of heat your alpha generates and fire at your personal safe heat number. If your alpha generates 60% heat, you know that firing as soon as you are under 40% has no mechanical disadvantage to firing at 0%, so why would you wait? This creates a no-brainer situation where firing at 40% is always advantageous to not doing so. There is no decision here, no strategy, no risk to balance the reward, and therefore no skill involved in managing heat beyond memorizing the heat % of each of your weapon systems and keeping an eye on the bar. There is SOME skill involved in that, but it's memorization and multitasking, not anything that would engage the player in critical thinking.
With that in mind, I'd like to discuss my thoughts on solving alpha strike dominance. First of all, we need a solution that doesn't contradict the intuitive themes of the weapon categories. Energy weapons are low-recoil and high heat, ballistics are the opposite, and missiles are a little of both plus their lock-on mechanic. I think the most intuitive method of adding risk to alpha strikes would involve amplifying the existing weaknesses, and begins with the first shot fired.
Energy: Run Hot or Die
For energy weapons, we introduce heat retention. In simple terms: the hotter a mech gets, the more dangerous additional heat becomes. When any weapon is fired the heat gauge increases as normal, but as it decreases you can see that there is now a second gauge laid over the first: your heat retention. Every time a weapon is fired, the heat retention gauge instantly matches the new current heat value of the mech, then decreases at some static rate (subject to testing) determined by your active heatsink count, similar to how your mech dissipates heat under current balancing. Since all of heat retention's functions scale the same no matter what the mech's actual heat capacity is (because the whole point is to create a mechanic that is simple and works the same way for every build), it is always calculated as a fraction of your capacity. 1.0 retention always coincides with the shutdown limit. 20% might mean 8 heat in a Jenner and 22 in an Atlas, but it always means 0.2 retention. While heat retention is above 0%, it interacts with heat in two ways.
First, the heat generation of any weapon fired while heat retention is above zero is increased by x% of the current heat retention value. To give an example: let's say x is 50. If the heat retention gauge is at 60% of your maximum heat tolerance when you fire a weapon, that weapon's heat is multiplied by 1.3. 50% of 60 is 30, so the weapon generates 30% more heat than normal. If the heat retention gauge is at 20%, then the heat generated by your next weapon discharge is 10% greater than normal. The exact amount of extra heat generated when a weapon is fired, x, can be tweaked by PGI to whatever feels right, but it would be uniform across all weapons. At first I thought that 100 would be too high, since that would mean that any weapon fired at 50% heat retention would generate 50% more heat than normal, but I ran some basic calculations and it's not as big of an increase as it first seemed. You'll see in the next example.
Now, for weapons fired as groups, there are two solutions: you could apply heat retention to each weapon in turn, starting with the hottest weapon. Say you fire 4 PPCs and 2 MLs, and we'll say that under current balance, each PPC is 10% of your total heat tolerance and each ML is 5%. Now, that would generate 50% heat. With heat retention set to 100%, the final heat increase would look like this: (10 heat x 1.0) + (10 heat x 1.1) + (10 heat x 1.21) + (10 heat x 1.331) + (5 x 1.4641) + (5 x 1.537305) = 61.42
The other solution is to multiply the total base heat by the total heat retention after all weapons are fired. In this case, the weapons generate 50% of your total heat capacity, so we multiply their base heat by 1.5 and your final heat generation is actually 75% of your capacity. This is obviously a bigger penalty, but I'm simply illustrating two ways of calculating how heat could scale. The difference can easily be addressed by changing how much retention affects heat generation (x from earlier). Personally, I prefer this method since it's simpler to calculate when you're setting up weapon groups, however it's much less forgiving about groups firing weapons with different heat values, so the first method will probably feel better on the field. That's is what really matters, since anyone interested in calculating this stuff can handle the math. Finding a system that feels more intuitive in the heat (pun not intended) of battle is why I'm arguing against alpha limits in the first place.
In order to avoid players gaming their way out of the compounding effect of group firing by setting up macros, we could use the same 0.5 second rule we already know the engine can handle. This is the only part where things get complicated, so I'll block it off for people who want to skip the technicals.
TECHNICAL STUFF:
Quote
If weapons A, B, and C put you at 30 heat after factoring in compound retention, but adding weapon D to the alpha would have put you at 42, then firing D 0.3 seconds later retroactively pulls the extra heat from A, B, and C to put you at 42, instead of whatever amount it would have generated under a fresh calculation. If E is fired 0.3 seconds after that, it performs the same calculation to get extra heat from D, but it acts as if the heat retention level when D was fired was 42 minus D's base heat, and that D was fired alone. Only each weapons' base heat and the retention value when fired need to be stored, and can be dumped after 0.5 seconds. You may still see players try using macros to fire every 0.26 seconds, but it will still be much more heat efficient to wait the full 0.5 seconds, and firing every 0.25 seconds or faster will still cause any third, fourth, or even tenth shots to pull extra heat from all the weapons proceeding them until 0.5 seconds have passed. Still, you may want to put a 0.1 second global cooldown or something to prevent macro users from overloading your servers.
END TECHNICAL STUFF
The second effect of heat retention would be to slow the rate of actual heat dissipation. As a reminder, heat retention will always dissipate at a static rate the way heat does now. The exact rate is subject to testing, but it would only change during a match in the event of damaged heatsinks. However, the rate at which your actual heat level dissipates will decrease as heat retention increases, returning to normal as retention depletes. To throw some hypothetical numbers out, let's say you subtract half your current heat retention from 1.0, then multiply your actual heat dissipation rate by that. If you're at 50% heat retention, then subtract 0.25 from 1.0 to get 0.75, meaning you would dissipate your actual heat at 75% normal efficiency. If you just barely toe your max heat tolerance and reach 100% exactly without shutting down, then you would lose your heat at half the normal rate immediately after firing, gradually cooling faster so long as you don't fire again. Let's say retention drops twice as fast as your heat drops at 0 retention (I tried having them drop at the same rate, but it would slow fights to a crawl). If you lose 10% heat per second at full efficiency, then you lose 20% retention per second at all times, meaning it would take 5 seconds before your actual heat dissipation was back to normal. Since heat retention always drops faster than your actual heat, and never exceeds your actual heat, the retention bar will always be equal size or smaller than the actual heat bar. That means that you only need to make them two different colors (red and amber?) and won't have to worry about the amber bar in the back (actual) being hidden by the red bar in front of it (retention).
So does this all sound very confusing? If a PGI dev is reading this, they're surely thinking "no normal human could wrap their head around this in the heat of battle, much less a new player!" I won't blame them, I'm kind of terrible at explaining things and everything sounds super complex when I can't give you guys some kind of video to demonstrate the idea, but in practice it would be a very simple concept to understand: the amber gauge is how hot your mech actually is, and the red gauge is how efficiently your heatsinks are working. When red is high, amber will rise faster and fall slower, because your mech is under stress. It's easy to understand when you think of it as a stress bar and a temperature bar. Players will learn that the key to maximizing sustained DPS is to allow the red bar to disappear before firing, slowing the pace of the game in general and leading to more low-heat brawling and DPS builds. This is the same goal that spawned the max alpha system. As a nice bonus, there will be much more tension during critical moments of desperation, when MechWarriors forsake heat efficiency for burst damage. "Run hot or die" will take on a whole new meaning once heat actually has an effect below the 100% threshold. Right now, I just learn my alpha strike's heat value, subtract it from 99, and keep firing at that number to stay just below overheating without penalty. I would appreciate the added depth of actually choosing whether to maintain low heat or accept retention as a calculated risk when trying to finish off an enemy or remove that pesky right arm on the Yen-Lo-Wang chasing me down. Obviously, base heat efficiency can be increased to compensate if retention slows the game down too much at low heat levels.
Ballistics: Hold My Beer
Now that I've explained heat retention, the answer to reducing the supremacy of group fired ballistics will be much easier to explain. It's a similar system of compounding penalties based on one word: recoil. We already have an effect that causes weapon convergence to go off-target: jump jets. Don't panic! We've already learned the problem of jump jets. The shaking screen made many people sick, and the amplitude of the reticle's vibrations was so high that it didn't make jumpsniping harder, it made it impossible. Recoil will not be as strong as jump jet shake, but it needs to exist. I'll explain why and how.
Certain ballistic weapons are meant for sniping, as spelled out very clearly by the functionality of the gauss rifle. However, sniping is exactly the sort of tactic that I was outlining earlier when I explained why alpha strikes are exponentially better than the sum of each weapon. Mechs that can mount multiple gauss rifles can multiply their power exponentially with no penalty to accuracy at all. This is also a problem when dual AC/20s are both guaranteed to hit out to their max range if either one does, effectively combining them into a single AC/40. The answer is NOT probabilistic cones of fire, it is infuriating to miss a shot that you are certain you lined up perfectly because RNG rolled against you. That's the kind of thing that makes people quit games entirely. Instead, look at Counter-Strike. The bullet spread on nearly all of Counter-Strike's weapons expands to such a degree that your third round of automatic fire has a 50% chance of hitting a barn at 20 yards, but the first shot is always on target. That spreading mechanic offsets the advantages of range by forcing players to control their rate of fire, whereas the superior DPS of spraying and praying can be a legitimate solution when you're face to face in a small room. We need a system, like heat, that presents a scaling penalty to group and rapid fire without affecting sustained brawling or preventing players who want to take risks from trying their luck and/or skill.
Here's the idea. When a group of ballistic weapons are fired, each one pushes the mech's arms and torso off center a random number of degrees in a random direction within a 10 degree cone related to where the weapon is placed. For example, a weapon in the right torso near the mech's center of mass will send the torso a very small number of degrees in a direction between 10 and -10 degrees on a unit circle. A weapon in the right arm near the mech's center of mass will send it a greater number of degrees off center in the same direction. Weapons above the center of mass (such as on a Jagermech or Blackjack) have their advantage of shooting over terrain offset by also pitching their torso up a very small amount, so the random angles may be centered around 30 degrees when firing from the right arm. Weapons in the center torso produce the least amount of recoil, and do so in a completely random direction. So long as no recoil is in effect, any single weapon will perfectly strike its target. When weapons with recoil are group fired, the weapon placed closest to the center torso gets the first shot and fires on target. If multiple weapons with recoil are being fired from components equally far from the torso, then the weapon with the highest base damage (or the one that weighs the most) fires on target. If those are equal, like on AC/40 Cats or Jagers, then a weapon is chosen randomly each time. Macros, chain fire, or weapon groups may be used to fire one weapon before the others, there is no retroactive recoil. From there, recoil is applied in the same order: distance from torso, base damage (weight), then random selection, and each weapon fires at the point determined by the recoil of the weapon preceeding it. Since it would absolutely make the pilot sick should their view teleport around while all these weapons fire simultaneously, it will instead lurch toward the point left over at the end of the entire sequence.
Keep in mind, the number of degrees you would be off center is tiny, and only enough to matter at long range, something like 20 damage = 5 degrees. To visualize the amount of recoil I'm talking about, if you were to fire two gauss rifles simultaneously at a target the maximum effective distance away, you would usually hit an assault mech, sometimes hit a medium, and almost never hit a light, assuming you are skilled enough to hit dead center with the shot on target. At the absolute maximum distance (3x effective for ballistics), you might hit an assault mech with the second slug if you're firing from the torso and get the minimum amount of recoil. Grouped AC/2s at effective range would usually hit lights and nearly always anything else with the second shot due to how little recoil they produce, but their recycle is so fast that you would be dealing with an constantly and randomly shifting reticle if you tried chain firing them for long periods. Dual AC/20s fired at their effective distance will always hit assaults and heavies, usually mediums, and sometimes lights. Remember, we're comparing 20 damage at 270m to 15 at 660m. The gauss rifle effective range is much greater, so even though it has less recoil, each degree matters more. Devs, I know you want to say that scaling recoil to damage is a bad idea, and I get that every weapon has different properties and it would hurt some weapons more than others, but I feel it's more important that these effects are based on publicly visible numbers that correlate to the power of each weapon. We also want to avoid adding yet another statistic to the weapon panel. If damage doesn't work, what about the weapon's weight? Whatever you choose, it should be extremely straightforward and intuitive. The bigger the cannon, the bigger the boom, the harder the push.
Obviously, recovering from recoil is part of this mechanic, so how long does it take to re-orient? Well, that would also scale with the force of the recoil. Obviously, we can't make the mech's lurch too fast as that would make players sick, but we can move the reticle very quickly so the player would know right away where the next shot is going. Your view would only rotate a small portion of the distance that the reticle and cockpit model would, since you were expecting the shock and are an experienced mech pilot, but it would still move slightly because hey, it's awesome to feel like you just fired a 400lb cannon round from your body. So your mech's torso would rotate from the force for a split second to absorb some of the shock instead of transferring it all to the pilot before the gyroscope compensates and steadies it. Once the gyroscope has control of the torso again it will return to your center of view at its normal turn rate. I imagine an AC/20's recoil would take about 1 second to completely stop before your reticle was back under control, but your whole cockpit would jerk hard by about 5 degrees and your moving decorations would be swinging all over the place. Since AC/20s have a 4 second recycle time, 3 seconds would be plenty to center yourself for the next shot. It would only prevent your ability to hit with other weapons. Other weapons would have more modest shocks and faster recoveries. The idea here isn't to make it take forever to get back on target, we're just trying to forcing snipers to make multiple shots instead of blasting through some poor light's CT in one click. In <100m brawls you'll be able to unload to your heart's content, through your cockpit with jerk and rock around while you're laying into them. I don't know about you but that sounds pretty metal to me.
The best part?
Quote
The Particle Projector Cannon (or PPC) is a unique energy weapon. PPCs fire a concentrated stream of protons or ions at a target, causing damage through both thermal and kinetic energy.[3] As such, despite being an energy weapon, it produces recoil.
http://www.sarna.net/wiki/PPC
PPCs (and missiles) have recoil. Lasers and flamers do not. It humors me to think of calculating recoil for MGs, but that's obviously a waste of resources, so of course MGs are exempt from recoil. What about mech weight differences? Well, smaller mechs have less mass, which means they require less force to move and therefore get pushed farther by that force than a mech with more mass would. However, light mechs have the same gyroscopes as any other mech, and because of their reduced mass, they are also easier to stop. So an Atlas may be as hard to get rolling as a boulder but it's also as hard to stop, while an AC/20 Raven may by as easy to throw as a baseball, but it's also as easy to catch. End result: recoil is the same for everyone.
Missiles: These Things Aren't Powered By The Cloud
Ok, so, can anyone tell me why I'm piloting a giant walking weapon of mass destruction that costs millions of C-Bills, and yet it still takes my computer several seconds to identify what weapons the mech standing in front of me is carrying? Or even what kind of mech it is? This is 3050, so I should have a supercomputer in the palm of my hand, right? Why do these cockpits still have so many mechanical switches and dials? Because 3050 isn't a perfect future. For those new to the BattleTech universe, the Succession Wars that raged almost nonstop from 2786 to 3025 destroyed so much of Inner Sphere civilization that the knowledge of how to build the advanced technology that mankind had developed during the Star League era was either destroyed, or stolen and then destroyed to prevent it being used against the faction it was stolen from. Many planets are cut off from their own governments because all of their space flight facilities and personnel were bombed to dust from orbit, and the coordinates of their worlds are either lost or the various states are so busy trying not to be annihilated that they don't have the time or resources to reconnect with worlds that no longer offer a strategic advantage. The concept of technology that no one left alive knows how to replicate is so common that they've coined a term for it: Lostech. So it's no surprise that our Battlemech's computers are kind of primitive for what we'd expect from the year 3050.
Now then, if that's the case, and it takes me seconds to analyze a target for such obvious information as its chassis (which I can confirm visually in an instant), why do our targetting systems have an infinite capacity for missile guidance? I don't think missiles really need to be weaker, they just have too high of a firing rate to the point that you're being rained on constantly until you find cover. You don't need to increase their recycle time. Instead, set a limit on the number of missiles that can be actively guided, equal to the sum of the numbers on all missile weapons you have mounted (LRMs and SSRMs counted separately). Have 2 LRM20s? You can guide 40 missiles at once. Have 4 LRM15s? You can guide 60. Only a single LRM 5? Only 5 missiles at a time for you. If you lose one of your launchers, don't worry. The targeting systems are in your head, so you can still fire whatever you have left from whichever tubes remain and guide the same number of missiles, subject to recycle times on the remaining launchers of course. To prevent wasting ammo, your computer will prevent you from firing missiles while you have a lock if you're at your guidance limit. As soon as you lose that lock, those missiles are no longer tracking their target, and you're free to either dumbfire (after a short grace period to release the key/button so you don't immediately throw ammo away) or obtain a new lock, even if the ones you were guiding previously are still in flight. You simply can't fire on the same mech while your lock is active until your first volley lands, and it would be a terrible idea to break your own lock to fire again since that would disable tracking on the volley in flight. Obviously, this isn't a problem with SSRMs since you're so close to your target that they hit pretty much instantly, but they still follow the same rule anyway, just in case. This change will do nothing to nerf the damage per volley or per ton of ammo of LRMs, but it will stop the persistent rain of death and bring burst LRM damage down in a consistent way, just like the above changes. If LRMs become too weak with this change, you could buff LRM damage slightly to compensate. This would give LRMs more power to punish mechs that stray from cover but run back immediately, while making the option of rushing into battle a little less suicidal for heavy brawler builds that suffer on maps like Alpine Peaks. On top of that, it also introduces risk/reward to LRM play. The closer the LRM mech gets to its target the less time it takes its missiles to travel to their target, and the higher its sustained DPS becomes. Yes, it's counter-intuitive for long-range mechs to close distance, but that's the very definition of risk/reward. There would be a choice there that doesn't exist currently.
And obviously, since launching missiles generates both force and heat, they are subject to retention and recoil, just like ballistics.
Conclusion
Well, my conclusion for now is that it's past midnight and I'm damn tired, but I hope you see how much it would improve the game to create gameplay systems that only need to be learned once, apply to all content consistently, and scale in a way that's very clear and intuitive to the player. I welcome all constructive feedback. Feel free to tell me if I'm wrong, just tell me why I'm wrong. Offer alternatives, but please stick to suggesting changes to what I've proposed. If you have a completely different mechanic in mind, that belongs in a separate thread. Thanks for reading, I'mma crash now.
TheBaron a.k.a. DreyfussFrost
Veteran MechWarrior of Clan Snow Raven, Zeta Galaxy
Edited by TheBaron, 19 August 2013 - 01:35 PM.
I think there would be room to buff LRMs if a single mech couldn't fire hundreds on cooldown and take down an assault mech alone from 1000m. Obviously, we want LRMs to hurt to discourage simply running around in the open, but charging a lone LRM boat up the field on Alpine in an assault mech should be risky and painful, not completely suicidal. On the other hand, you will very rarely hit a light with more than one volley. Once they know there are missiles incoming, they will run to cover. It would be nice if that one volley hurt a little more so they would be more cautious about stumbling into an LRM boat's view. This would also change LRM builds into more effective sentries and rear guards, emphasizing the currently underrepresented role of lookouts and support mechs, rather than the "no-skill" DPS tsunamis you see players complain about currently. Assaults are the class meant to deal massive DPS from a stationary position imo, and it would be fairer if they did this best with direct-fire weapons which put them in the open.
