Interview with Chris from PGI
One of the industry's oldest independent developers, Piranha Games is fast becoming a much bigger fish in the Free-to-Play game marketplace. We are alive and thriving in the pond because we're strong on talent, management, and the alliances we have made with industry leaders. Our core competencies lay in Excellence through Passion, Creativity, and Humility.
Founded in 2000, we started out as an independent developer with extensive multi-platform and multiplayer experience, making FPS action games for big-name franchises. With AAA licensed titles such as Need for Speed: Undercover and Duke Nukem Forever, we took a shift in direction towards the growing Free-to-Play market with MechWarrior Online. We have a strong vision and dedication to the product we make. And our talented team contributed as the key component to our success. Looking ahead, we are committed as ever to be one of the best Free-to-Play Development Studios in the world.
Can you tell us a little bit about your position at PGI? I.e., when you started, what you're doing, etc.
I am a Game Designer on the MWO team focused exclusively on game balance and have been with the team since Oct. of 2016. I’ve been working on a wide variety of things from weapon balance changes, ‘Mech quirk changes, skill tree node balancing, and gameplay feature changes such as the Engine Desync. Along with a handful of things I can’t talk about at this moment.
If it's not too prudent, where did you work before PGI? And why did you want to be a game dev?
I’ve been working professionally as a game developer for about 10 years now on a wide variety of titles across console and various Free to Play spaces. I wanted to be a game dev because I simply grew up around the industry. Growing up in the San Francisco bay area in the 80’s and working at a local import game store that served many of the game developers and media outlets like Tech TV back in its day had me grow up around not only a wide amount of games from across the world, but also with many of the people in the industry at the time, and it gave me the drive to enter it as a career.
How do you explain MW:O to your family and friends? What's their usual reaction?
BattleTech has been with me since the 80’s. I still have my Battle droids box and all of my leatherback Table Top rules compendiums, and my friends and I have burned through many days playing campaigns. So, joining the MWO team to them just seemed like I wound up where I was always supposed to be.
Are you a fan of any particular genres or game developers? What game has you hooked at the moment?
I’m a fan of most genres out there. Although I love strategy games and RPG’s, I find I don’t have the spare time to go through them as much as I would like. Witcher 3 is still staring at me in my steam library and I haven’t found a good day to just sit down and get into it. Like many others in the community, I am highly looking forward to Harebrained Scheme’s upcoming BattleTech game. As someone who used to run Table Top Leagues and participated in BT tournaments, that is very much up my alley and my old BT friends and I can’t wait for the Beta to go up.
What is your favourite lore 'mech vs. game 'mech?
Timberwolf definitely my favorite lore mech. Between how they initially introduced the clans on The Rock, Aidan Pryde’s last stand on Tukayyid, to some of Vlad Wards later fights, it’s just an iconic ‘Mech for the franchise. I don’t really have a favorite ‘Mech in MWO as I need to spend a lot of time playing with all the ‘Mechs. I’m still holding out hope that one of my favorites from Tabletop eventually makes it into the game. The Hellion.
There is a huge focus on MWO being a competitive PvP shooter, which obviously leads into competitive leagues, eSports etc. Are there any plans for in-client competitive play, similar to Ranked mode for Dirty Bomb, CSGO or Overwatch?
Have to respond to this with a “no comment” for right now. But I can tell anyone interested in an answer to this question should make sure to tune into Russ’ upcoming competitive roundtable this Friday
What is "balance" to PGI? I believe no matter what, you will always have bad mechs and even bad weapon systems. What are your views on this? What does PGI see as the best way to combat this for the long term?
I’m not prepared to really set in stone what the balance big picture should be like. Lots of different answers depending on what aspects of MWO you talk about. I will say that my role on the team has me looking at balance through three broad and equally important categories:
- Equipment Balance
- ‘Mech Balance
- Game Balance
Each of these categories has its own challenges, goals, and nuances that need to be considered with the greater sum of the whole of the MWO experience. As to bad ‘Mechs, there will always be ‘Mechs that are sub-optimal to whatever meta is out in the game at any instance, but from where I stand, it’s very important to question if the factors that lead to a ‘Mech being suboptimal are factors that should matter as much as they do in the current game. And what kind of core level changes can we target to improve the state of other ‘Mechs that suffer from the same shortcomings. With nearly 400 different ‘Mechs in MWO at this point, it’s important that balance doesn’t just come down to a single batch of common traits on a ‘Mech as being the only viable ways to effectively play the game, as this not only leads to greater balance discrepancy between the ‘Mechs that fit those traits and those that don’t, but it also leads to the entire “arms race” mentality that we are trying to avoid in MWO compared to other MechWarrior games.
We do not want ‘Mechs to simply come down to nothing more than skin swaps of the same load outs with only the most optimally placed ones being used. So rather than looking to homogenize the entire ‘Mech line up, we would rather go the route where each ‘Mech does have perks and drawbacks to the core design that are made up in other ways through the core game mechanics.
This is an aspect of MWO that can definitely use some improvement, and it is why future features are being produced in the way that they are. As we want much more give and take with the decisions you make to allow for more nuances in the game to be viable alternatives to approach a match. Because if you only design around a handful of options, then it’s very hard to continue to not only expand the ‘Mech choices, but also the weapon and equipment choices when it boils down to only a couple of desirable outcomes that will always result in a small handful of choices being better than others.
Ultimately, we need to make sure our game design justifies not only the large amount of ‘Mechs we have in the line-up but also the raw amount of options we allow players to customize around.
Does PGI want a balanced system where the meta of the game changes every few months, or are you looking for a more stable balance?
Meta will always be the result of what is effective within the established game mechanics. At the moment, I am more concerned with ensuring that these mechanics provide compelling amounts of give and take in their choices in how players customize their ‘Mech as opposed to actively attempting to preserve any specific type of meta.
The focus on give and take within the core systems is because this can lead to not only meta shifts, but more diversity in what people consider meta. In the long run, I would love to see this diversify in a way that meta revolves more around in-game strategy and reactionary counter play then chasing those handful of loadouts that essentially does everything you would ever want it to do. But for this to happen, more things need to change at the base level. Which is why these core systems are getting focused on in the way that they are.
How does PGI balance the game around 4 different games modes and the major different game mode FP? Does data show some mechs performing better in FP than QP, and how does that affect your balance?
Yes, the meta for Faction Play is much different from Quick play, which is much different from Competitive play. And all of these factors need to be considered equally when it comes to balancing the game as a whole. We do prioritise what we view at the top levels of play more heavily than the lower levels of play, as what is effective at high levels of Competitive play also tends to be effective within quick play and faction play. But while most changes are targeted at diversifying the scene at the top, this does not mean that every change is made exclusively because of the players at the top.
We simply cannot neglect the enjoyment of the vast majority of our player base that occupy Tier’s 3-5 or players that primarily wish to focus on the Faction Play experience.
How are you approaching 'mech balancing long-term? Obviously, this becomes different as more 'mechs get added, but how is the team approaching it? Again, is it short-term adjustments, or set changes that won’t move much?
Like many aspects of the game, ‘Mech balancing is heavily dependent on the core mechanics supported by various stat enhancements like quirks. New features can come and call for additional changes, while also, providing new avenues to tune ‘Mechs in different ways and try to diversify the ‘Mech line as a whole.
As we continue to expand the ‘Mech roster, this becomes even more important as we continue to not only expand our game mode selection, but also, we wish to diversify the overall skill sets that ‘Mechs can bring to the battlefield to again avoid chassis homogenization. This will usually come from a push of a major feature set followed by numerous month to month adjustments to ‘Mechs as we get metrics in and see which ‘Mechs need the most help taking into account the new variables that we introduced with a feature set. An example of this will be skill tree and engine desync. Skill tree in of itself will introduce many changes that we currently do not have, and will see a general scale down of things that we do not want to compound on top of the new skill tree system. As we examine metrics for how both skill tree and engine desync affect all the mechs, we will make adjustments accordingly as that information becomes available to us. Which will continue until the next major feature, such as the Civil War update, comes in and potentially changes everything again.
New weapons systems, and in old 'mechs, what do you think the hardest adjustment will be (player and PGI related?) Any time you introduce a large number of new equipment into a competitive environment there is the potential for disruption in the game balance. The hardest adjustment from my perspective is making sure, that the introduction of the new tech does not see the invalidation of what is already there.
While optimization with the new tech is always going to be a thing, we need to ensure that the viability spread between optimisation with the old tech does not get so outshines by optimisation with the new tech that someone that took a break from MWO comes back after July to find their entire ‘Mech collection can no longer compete with what is on the field and they don’t know why.
Are there any plans to help out old 'mechs with more hard-points, bigger engine caps, something else? The answer to this is very much “something else” for the foreseeable future. As I said earlier, with nearly 400 ‘Mechs in game now, we need to diversify what makes an effective mech. Hardpoints and engine caps should be something to design around, not further homogenize ‘Mechs into being nothing more than a skin swap of other ‘Mechs. To speak directly to the point of engine caps, this is why Engine Desync will be coming with the skill tree patch. Since engine selection is an aspect of the customisation we allow within the game, we need to ensure that the right answer isn’t always “go big or go home.” There needs to be a legitimate give and take to the system that sees different perks and drawbacks for each option available. Otherwise there is little point to us providing engine customisation as an option to the player when a majority of the options provided are trap options that puts players at a disadvantage. This is not to say that we would never consider it.
Nothing in design is ever off the table. But given the greater goals of ‘Mech loadout diversity, this is a point where we will be looking into baseline changes to better balance the underlying mechanics first before considering further hardpoint inflation or engine caps homogenizing mech loadout selection.
Can you explain the benefits of engine decoupling? I.e., how is decoupling torso agility from engine size, and decoupling acceleration/deceleration from engine size will be good for MWO?
As explained above, Engine Desync is primarily being pushed to further balance engine selection to be less weighted towards always wanting to directly upgrade your engine as the only “correct” way to customize a ‘Mech while remaining competitive. We want both ‘Mechs characterized by their lower speeds / lower engine caps, and ‘Mechs that choose heavier payloads over speed to be on much more of an even playing field when it comes to mobility against those ‘Mechs that have typically always had high engine rating / high speed options.
Competitive players are often surprised when we see a sub-par 'mech(s) nerfed. Can you give us any clarification into the initiating process of nerfing so we aren't so baffled?
We have a large amount of match / ‘Mech metrics that we track across the entire game through Quickplay, Faction Play, and Private Matches. We utilize the data in these metrics to track trends in ‘Mech performance that includes, but is not limited to, average match score, Kill / Death ratio, Win / Loss ratio, Average damage, ‘Mech loadouts, average match earnings, etc. We then focus on data around Tier 1 and Tier 2 metrics to compile an expected performance curve that we track along a per weight class level which creates an overall “range” that we like to keep the ‘Mechs operating within to ensure that ‘Mechs of similar tonnage / weight classes remain relatively even choices against other ‘Mechs within their respective weight brackets.
We also utilize loadout metrics to also monitor the use of individual weapon systems if we monitor a performance spike in the metrics directly related to common loadouts. We use this as an outline for identifying outliers, both under performers, and over performers. Once Identified, the ‘Mechs are evaluated. Typically for under performers, it’s easy enough to see what needs a bump and we can usually fast track some kind of boost to provide them. For over performers, we tend to put them under further observation, both to monitor to ensure that any metrics spike is the result of a constant and not just a month-long swing.
We tend not to nerf ‘Mechs unless the ‘Mech itself operates outside of what we consider acceptable variance away from the curve for multiple concurrent months. And even then, this is often after evaluating other ‘Mechs with similar loadouts to the ‘Mech that is targeted for a reduction. Often, if we find that multiple ‘Mechs with similar loadouts are outperforming their contemporaries in their weight class, this typically will call for a weapon performance evaluation and put the weapons in particular loadouts under scrutiny rather than the individual ‘Mech. For ‘Mechs that continue for multiple months to outperform the curve and where we find that similar loadouts of ‘Mechs do not see similar performance spikes, then we target the ‘Mech for a nerf. The severity of which is proportional to the amount it has gone over its metrics for the length that it was under observation. Quite often, ‘Mechs that are nerfed typically are operating outside of the curve across multiple “high value” metrics that we monitor compared to equivalent ‘Mechs within their respective weight bracket and weight class.
Can you give an example of how PGI tests new features or balancing changes before they go live? I.e., how do you know certainly that a change is tested in non-live environment(s)?
This all depends on what we are trying to change. A single weapon or ‘Mech value change is often done by examining weapon / ‘Mech metrics, establishing how far off the mark it is from where you want it to be, design a solution, test it here in the office, and then push it live and monitor the metrics to see if the incoming metrics pushes it to where its designed to be at. More complex features add to this with design team play throughs. With PTS tests being a final way that we can both test an extensive number of changes to more complex features, while also pushing test values to a bit more extreme measures to monitor PTS match metrics data regarding all things to refine before release. We do not make balance changes in a vacuum. Even the tiniest weapon change has a projected result that we are looking for when it gets pushed live. And no matter what we push, we are always monitoring how the net result affected actual match metrics. So, should a change pushed need any further tuning or a potential roll back, we are aware of it.
A lot of player units/corps would love to see the quick play (group) going back to 8v8. Balance wise, what are the pros/cons vs the current numbers that concerns PGI?
The thing about 8v8 is that while it works well for matches that are focused around direct combat of equally skilled sides, it is less than ideal for quick play like environment because it is much more susceptible to winning / losing a match because of snowballing off of a single kill or an unlucky disconnect then you see in 12v12. In this regard, 12v12 offers a bit more variety not only in what you see taken to the battlefield, but what you are able to do to adapt to a situation if you get an unlucky quick loss. It also allows for much more diverse objective based gameplay with the larger player counts. While objective based gameplay is something that we need to improve a bit on in MWO, the point is that having a larger number of players in match allows for a much larger amount of mechanics to be in play that would be too risky in an 8v8 match. The upcoming Incursion mode for instance works much better within the context of a 12v12 setup then it does an 8v8 setup. As having one guy grabbing power cells seeing your front-line fight 11v12 is not as big of a detriment as the same situation in an 8v8 where having one guy objective grab could see you fighting 7v8.
The thing to remember as well is that many of the ‘Mechs that we are bringing over from BattleTech itself do not always translate to direct combat roles. And as mentioned before, we cannot expect to keep expanding the ‘Mech roster and create redundant roles with nothing more than skin swaps of the exact same loadouts for ‘Mechs, so we need to expand the way that dedicated support ‘Mechs by design like the Mist Lynx or Spider 5V can actively contribute to a match in a positive way. This is a much more feasible task to do when the ‘Mech is 1/12th of a total force instead of 1/8th of a total force. Keep in mind that the way that many competitive leagues are structured plays to 8v8s strengths and this has not gone unnoticed by us. It is why we have structured the previous World Championship around 8v8.
Beyond that all I can say is that while it is not the most ideal way we wish to approach quickplay, that doesn’t mean that something else can’t come along and be more applicable to that structure.
Knockdowns; a balance nightmare or do you see it coming back? (Or is the troll factor too high!)
Knockdowns are not coming back in the foreseeable future. But this has much more to do with performance and HSR concerns then it has to do with balance. I’m not one to dismiss any kind of mechanic that can help diversify non-direct damage combat effects. As these effects can always be tuned to ensure that they have a sufficient give and take to balance them with the rest of the game. But as nice as these additional effects can be to the sim aspect of the game, they cannot come at the expense of core game performance.
Is there one particular mechanic that you're so attached to that, no matter how much feedback you receive, you’ll still keep it in the game?
I would have to say no. From where I stand, all mechanics are fair game for tuning, re-works, or removal if they are not accomplishing the greater design goal we are seeking to achieve through them implementation. Although I need to heavily stress that while all mechanics in my eyes can be fair game for examination, this does not mean that we can consider any change lightly. As feature changes can often take time and additional developers working on changes rather than new features. And as mentioned earlier, all changes first and foremost cannot come at the expense of core game performance. So, it’s a constant balancing act of what is efficient to produce to not impede other improvements we want to make to the game and what will work without causing issues to the core game play. I know this aspect can sometimes frustrate players who have formed all kinds of consensus on what would be the ideal way to handle certain aspects of the game. The best I can say is that we are listening, and should there be a change that allows us to explore some of those suggestions I would love to look into them as an option. But ultimately any change we make can’t come at the expense of performance. So just because it’s a popular opinion doesn’t mean it’s always technically feasible for us given the amount of complexity already in the current game.
Thanks for giving us an insight into your work, is there anything else you wanted to say/want us to know?
Thanks for having me on, and good luck to all the teams participating in this season of MRBC. Very much looking forward to the matches. I want to announce that I will be continuing to do patch day stream drops with Phil of NGNG for the monthly patch drops. So, if anyone has any questions feel free to swing by the stream and ask. Also, be sure to tune into Russ’ Competitive town hall. This Friday 3/31 to hopefully get some answers on some of the things I can only gloss over here.