Regarding the new map vs. fixes/perf thing, that is actually a fair point. Yes, we chose to spend a lot of the donated money on a new map. Maybe we should have spent it on contracting one programmer (probably fsfod) to optimize the game for 2-3 months? That's how long it probably would have taken to get a significant bump in perf - it would basically involve changing 50% of Lua code, everywhere. But, that still would've been risky - would the perf actually get much better after that? How much better? How many MORE bugs would it introduce? It would certainly break ALL mods. So, at the end of the day, you can see why making a new map is a much safer bet than a complete overhaul of the code.
Just to add to this, while we fully acknowledge that many people have performance problems with the game, there are also at least an equal amount of players, or more, who either do not have problems or they at least don't notice or care, and it doesn't keep them from having an enjoyable experience. Those people would likely have felt cheated if we spent that much time and money on ONLY performance fixes (that might barely have been noticeable) instead of new content that really makes a big gameplay difference.
And, when it comes down to it, we do have to consider the cost vs impact of the updates to NS2. Sure, performance improvements are nice, but does that bring in new players? Does that bring back players who have gotten bored of playing the same maps over and over? Can you make a high impact trailer that gets the attention of people with just performance improvements? When we spend that kind of money, we want to make sure it makes a difference in keeping NS2 alive.
I'm glad you guys took the time to respond to this topic and appreciate that you read our posts related to this issue.
I love your game and have played it since age 12 (NS1) and will continue to play it until it dies. I am not just another competitive player complaining in this post. I have played public games in NS1 and NS2 for a very long time.
I have quite an opinion on this issue (performance and NS2's problems) so if you do decide to read.. prepare for a wall of text.
I couldn't agree more with your last statement there Cory. Spending a ton of money on something that keeps NS2 alive should definitely be the focus but I have a different opinion on where the focus should be. I completely respect that you have a ton more game development experience than me but I still want to give my opinion here which is largely based on my own observations in games that are currently accomplishing something that NS2 is not. Playerbase retention. I'm not out to prove you wrong here but I do want to bring your attention to something I feel has a bigger impact on playerbase retention than more maps.
To answer your questions there Cory.. I say YES. It not only brings back players that had issues previously but it also allows a larger amount of people to play the game at all to begin with. Example: a player that would get 15-30 FPS in common scenarios in NS2 upon trying the first time now gets 30-60 FPS after Build XXX patch. Unplayable -> playable. Planetside 2, for example, had similar issues to NS2 and they made big improvements to performance and quite a lot of people came back to see how it fared on their PCs. In addition to that, they have quite a neat system in place regarding skins and content that is along the lines of the Kodiak update in terms of community members benefiting from customization alongside the developer.
A great map like Kodiak or Eclipse might be safer to invest money into from your point of view but it doesn't really halt or relieve the massively declining player base scenario that has happened on multiple occasions to Natural Selection 2. In the long term.. these types of additions to the game fall short in helping resolve the biggest issue. CS:GO which I will refer to several times here.. has a good amount of maps but I think everyone knows how much people play Dust/Dust2. I don't buy that people quit games because of limitation in the maps available. If the gameplay and other basic functionality aspects are there, like in CS:GO for example.. people will keep playing your game for a VERY long time. Maps are nice and all and I really personally admire your concepts and art design/direction but are not something that fix a big issue like a declining playerbase.
You can see it clearly on the SteamGraph in regards to the playerbase issue. I don't think I need to link it or provide #'s but it's pretty clear NS2 is not on the up and up.
Before I continue on I want to say this: I don't say any of this with the intention of being sour about the topic or NS2 development. I love NS2 and NS1 gameplay/conceptually and it's the only reason I'm still around. I don't say any of this with the intention of bashing you guys as developers either but I had a lot of hope and expectation that performance was going to be a big focus. I only give my thoughts and criticism here in hopes that NS2 will retain players in the future and not end up a game that barely anyone plays anymore.
Content brings people back temporarily to try it out thanks to flashy and professional trailers but it doesn't keep people playing. Gameplay, performance (and thus accessibility) and certain features like matchmaking I think have a far larger impact on a consistent playerbase. Going F2P at this point in a games life also increases accessibility by a huge amount but ONLY if performance allows it to.
Of course performance increases don't guarantee "everyones playing ns2 now" but think of it this way:
Shortly after release there were a ton of things that people expected from you guys. (First person spectate, and many other improvements all being asked for on stream chats)
Obviously you couldn't fulfill all the demands instantly but one thing stood out.. performance.
When you put up a poll on what people wanted the most after release - it was an astounding landslide in the favor of performance.
You put out a patch featuring LuaJIT which was quite a good increase in FPS across the board (also reduced frame jitter) and then I had a hard time understanding why improvements stopped there. It's been almost 2 years since release / end of beta days and the performance improvements have for the most part halted entirely after LuaJIT. If F2P is on the table then you really need to think about the type of machines that run F2P games and try to cater to those people when you do decide to go F2P. (via increases in performance)
My feeling on it is that by increasing performance you have very little to lose (bugs [which are fixable, and I know requires time, money and effort to fix but in theory it should be worth it - will explain below] like you mentioned).
You allow a larger spread of people (more people.. more money obviously) with varying machines to play your game without experiencing performance issues on top of the increasing the amount of people that would be able to play at all on the super low end machines. People can play CS:GO [a comparable FPS to some degree] (and many other games) on a potato and get their refresh rate's worth of FPS. (60fps on 60hz for most people) NS2 - I play on an i5 with 4.8ghz and a GTX580 and still get astoundingly low framerates in common scenarios (40's-50's on my kind of machine with the CPU as the bottleneck) in public games - even on 18 slot servers with the lowest settings. You guys made great improvements after release but I can't help but think - why stop there..
If you think long term here - if more people can play your game potentially.. when or if you decide to go F2P (the route TF2 successfully went) or have more sales that is more people that can run your game acceptably - which hopefully stick around and purchase skins/DLC.. which can sustain development (plus tell their friends about it and even more sales).
It's an investment that puts your game on the level of other games in terms of accessibility. It might be risky but when you see your playerbase is rollercoasting with each sale that should be a sign that something isn't quite right and a plan to change that should be thought up. Almost every update that you guys have put out that has included a map and other features you see the playerbase skyrocket on the steamgraph temporarily and then everyone quits again. And I have never once heard anyone in the 2000+ hours I've played in this game tell me that they dislike NS2 or quit because of too much Summit.
Continuing on the note of performance though - I have read, agreed with and understood the statements from Hugh that Source has been around for ages and that NS2 is using a new engine. I would never expect Spark/NS2 to compete with Source/CS:GO within the 10-20% margin in terms of performance but I know Max, SteveRock and others from your team can kick some ass and if you guys put money and time into it - I think you could make it pretty damn close to the point that most people wouldn't notice that it's better/worse.
I don't expect you guys to work miracles here but a friend of mine played your game for about 2,000 hours (I've played about the same/more) without really any performance complaints after LuaJIT.. then Reinforced released.. The game went from 30-60FPS average on my friend's machine to severe hitching issues and sometimes as low as 15 FPS in public games. That friend of mine has quit and has taken up CS:GO for his FPS game that he pubs regularly and buys skins in. He really enjoyed NS2 and preferred the gameplay but the performance issues wore him out. Forcing texture streaming on I believe was the root of his issue and he hasn't played the game since nor purchased a Reinforced package / the most recent DLC (obviously, because he quit and his issues were never fixed). Many people that I know that played NS2 have a similar story.
That's an example of how you lose players and lose potential $$$ because content was prioritized over performance. People don't leave your game in droves because you aren't putting out tons of new maps and weapons [ I think people have much more tolerance for that stuff ]. I think people leave your game when they either A) can't run it as well as they would like to enjoy it [or enjoyed it previously in my friend's case] stop having fun with it due to difficulty or lack of interest or C) another game grabs their attention for a while or permanently. A combination of all three is probable as well. One of those things is definitely within your control as a developer. B is arguably in control as well but I think Sewlek for the most part did a good job with that bit for public games.
If you look at a lot of successful games.. not just FPS games, I think you'll find that the top games had the basic functionality NAILED from the get go - and then all the cool content and features came later with new versions of the game / expansions / sequels AFTER people were hooked on the gameplay and able to even play it acceptably performance-wise in the first place (accessibility tied to performance in NS2's case).
I don't think all of NS2's issues related to people leaving are due to performance but I would wager that it is a big part of it. I certainly don't think that I know how to solve all of NS2's problems and that performance increases in huge amounts is going to save the game.. but basic functionality is something quite a lot of people expect to be a given. After that I think most of the people that left were casuals that didn't stick with the game because it was too difficult or something else caught their attention. But there are far more not-so-intuitive games out there than NS2 that don't have this rollercoaster "try and quit" playerbase issue.
In conclusion here, I think that performance is definitely something that has needed attention for a long time coming. I also think it's a good idea to look at your competition and see where you fall short. I will repeat that I am not expecting you guys to be equal in all regards to a Valve title with X amount less money and manpower but it's still something worth noting.
CS:GO vs NS2
Gameplay - about as good or slightly better/worse depending on who you ask. I personally think NS gameplay has just as much potential or more to be what CS:GO has become in terms of esports and public gameplay.
Performance - far worse and much more inconsistent based on the PC and servers in question.
Matchmaking - extremely good with skill placement and progression versus non-existent.
TL;DR - I had hoped for more focus on performance after release and I feel like the focus wasn't in the right place after LuaJIT. This new strategy for development involving Kodiak and DLC has me hopeful but I would like to see a stronger focus on performance over content still. (and finishing matchmaking - I feel like that was a great idea that was just forgotten about and never put into place when the playerbase was at it's height [no one uses it now because there's not enough players to really make it viable, plus it doesn't work]) I might be naive to think this but I think NS2 can still progress out of this cycle of 'patch - people come back - people quit again'.
New content is always welcome, but I believe the main issue with NS2's success has been its performance or lack thereof. I love the game and hope to see the community grow.
In short, Better performing game = Larger possible player base who can actually enjoy the game.
I think you are underestimating how many people would've come back with improved preformance.
Of the 10 people I got to play NS2, none are still here (but me!). Of those 10 people, they all* say they would come back if performance was better.
*one says that he would only come back if the rifle-butt knock-back was re-introduced, so thats a lost cause
I understand the desire/need to move on to other projects, but perhaps if you started another Reinforced program, explicitly for performance focus......?
Great thread, albeit with some pretty ignorant posts. But that's also NS2's playerbase i guess. Good to see UWE respond.
Here's why i stop playing NS2 everytime, shortly after an update.
I never had (fps) performance issues. My main gripe is always related to hitboxes and hitdetection:
-Being a Skulk or Lerk in a vent somewhere, or behind some cover, thinking no marine can see me, when actually he sees the top of my character and so kills me.
-Running or flying away when fired upon, and then dying behind a corner after still taking damage. Although i was not in the line of sight of my attacker anymore.
I have a ping that's less than 60 on those servers. Also i monitor the server tickrate in game.
While there may be perfectly logical explanations for those issues, they don't happen in many other fps games i play (*). The first, regarding the character's point of view, i don't even understand why UWE has it that way. It gives the player a wrong suggestion of his character's height.
You probably don't realize it Zeep but dying around corners is directly related to a performance limitation of NS2's netcode.
The current implementation of 100ms interpolation and lag compensation is the source of that issue.
The 30tickrate you're looking at is related to AI according to Max, the client update rate is 20. That is what is limiting the server performance as well.
The hit detection in NS2 is actually quite good. There's a few issues with it but I won't go into detail.
+1 fix/improve the game before adding content, content is good, but even KODIAK is hated by some in the public servers because of low FPS, early versions of biodome in falls had huge performance drops even on my machine to the point where i would want to avoid that room all together or rather the whole map until it was fixed
Regarding the new map vs. fixes/perf thing, that is actually a fair point. Yes, we chose to spend a lot of the donated money on a new map. Maybe we should have spent it on contracting one programmer (probably fsfod) to optimize the game for 2-3 months? That's how long it probably would have taken to get a significant bump in perf - it would basically involve changing 50% of Lua code, everywhere. But, that still would've been risky - would the perf actually get much better after that? How much better? How many MORE bugs would it introduce? It would certainly break ALL mods. So, at the end of the day, you can see why making a new map is a much safer bet than a complete overhaul of the code.
We have definitely learned from the performance woes of NS2 (yes, we absolutely acknowledge it as an issue), and for future projects that use Lua, we will architect things from the ground up with LuaJIT in mind. As for NS2, unfortunately, it's hard to make those kind of changes this late with a gigantic code base.
with $100k, you will
A. create new content like ns2_eclipse "for free" and then milk money from existing miniscule (and declining) playerbase with kodiak dlc ($5.99 x 1000 players) OR
B. fix the underlying problem of the game - performance - and therefore
1. retain old players who quit due to performance issue (+~6000 players x $5.99)
2. earn new players from sales/adverts (+? players x $29.99)
3. encourage current players who now aren't afraid to introduce a good game (previously unplayable due to high pc requirements but now every average pc plays awesome! <3) to their friends (+1000 players x ? friends x $29.99)
4. make modders happy in helping to create a stable game which hopefully stops breaking their mod in almost every new patch. old modders will thus also be motivated to fix their mods to be compatible with a much better performing game (+? players who like modded games)
with so much focus on client-side performance, let's not forget better multicore/multithread support for servers so it's economically viable to host servers (consumer grade hardware and 4.5GHz just to host one instance of 24 slots WTF?!) around the globe especially in the Asia Pacific region where infrastructure is so expensive
games like team fortress 2 (f2p) or left4dead2 (not f2p) don't have that many maps either, but they do have a very large player base. <sarcasm> i wonder why...? </sarcasm>
let's hope the company still have another $100k for a wise turnaround strategy
(consumer grade hardware and 4.5GHz just to host one instance of 24 slots WTF?!)
I had to comment on this. This is my server load with three 24-player instances running: cpu load
Since this screenshot was taken I have replaced the cpu with another one of the same model that overclocks higher at a lower voltage. And I know that Combat servers doesn't require as much as normal NS2 but I don't have a screenshot of the server running more instances.
games like team fortress 2 (f2p) or left4dead2 (not f2p) don't have that many maps either
lol, what? tf2 has over 30+ official maps and probably thousands of user created maps. same with l4d2, there are over 10 campaigns with each having 4+ maps each and again, and a ton of community created maps.
(consumer grade hardware and 4.5GHz just to host one instance of 24 slots WTF?!)
I had to comment on this. This is my server load with three 24-player instances running: cpu load
Since this screenshot was taken I have replaced the cpu with another one of the same model that overclocks higher at a lower voltage. And I know that Combat servers doesn't require as much as normal NS2 but I don't have a screenshot of the server running more instances.
what are the servers doing though? i'd love to see the tickrate of all 3 servers while they're all packed and being played on at the same time.
games like team fortress 2 (f2p) or left4dead2 (not f2p) don't have that many maps either
lol, what? tf2 has over 30+ official maps and probably thousands of user created maps. same with l4d2, there are over 10 campaigns with each having 4+ maps each and again, and a ton of community created maps.
how many tf2 payload maps are there? if you add up all other maps for other modes (how many official modes does ns2 have?), then yes it's a fairly large number. nonetheless, relative to their player base, it's not THAT many maps! if you're going to do benchmarking, never look solely at absolute numbers.
also, where do community created maps come from? among others like game concept, a good confidence in the game stability and performance! we don't see that many for ns2, do we?
(consumer grade hardware and 4.5GHz just to host one instance of 24 slots WTF?!)
I had to comment on this. This is my server load with three 24-player instances running: cpu load
Since this screenshot was taken I have replaced the cpu with another one of the same model that overclocks higher at a lower voltage. And I know that Combat servers doesn't require as much as normal NS2 but I don't have a screenshot of the server running more instances.
that is a haswell at 4.4GHz. i may have exaggerated on my previous post and that was my experience of trying to host one during the b229 days, so while things may have improved significantly, that is still a consumer grade high end cpu. does a 3 x 24 player tf2 require just as high end cpu?
anyway, it seems like they aren't keen in giving ns2 an overhaul anymore and just wants to milk every last bit of it. so i assume their objective is just new payable content and new payable content with little regards to fixes (unless critical) from now.
for future projects that use Lua, we will architect things from the ground up with LuaJIT in mind. As for NS2, unfortunately, it's hard to make those kind of changes this late with a gigantic code base.
good luck to a dying community. i really loved the game concept back from ns1 and now but i'm just not satisfied with where they place their focus.
games like team fortress 2 (f2p) or left4dead2 (not f2p) don't have that many maps either
lol, what? tf2 has over 30+ official maps and probably thousands of user created maps. same with l4d2, there are over 10 campaigns with each having 4+ maps each and again, and a ton of community created maps.
how many tf2 payload maps are there? if you add up all other maps for other modes (how many official modes does ns2 have?), then yes it's a fairly large number. nonetheless, relative to their player base, it's not THAT many maps! if you're going to do benchmarking, never look solely at absolute numbers.
also, where do community created maps come from? among others like game concept, a good confidence in the game stability and performance! we don't see that many for ns2, do we?
who ever said anything about game modes? at the end of the day it's content no matter how you want to twist it to try and win a forum argument.
nonetheless, relative to their player base, it's not THAT many maps! if you're going to do benchmarking, never look solely at absolute numbers.
What?! What twisted kind of logic is that? "The more players we have, the more maps we need. If we have a 100 million players we will need a thousand maps or else they'll get bored!"
What matters is the individual player. If the individual player is bored on 10 maps, then 20 more maps could change that. Same goes for the next individual player. Just because you increase the playercount from one to two doesn't mean those two guys suddenly need double the amount of maps!
anyway, it seems like they aren't keen in giving ns2 an overhaul anymore and just wants to milk every last bit of it. so i assume their objective is just new payable content and new payable content with little regards to fixes (unless critical) from now.
Game is over a year and half out since release... Why on earth would you expect an "overhaul?" Why shouldn't they support, and try to get the most out of their product in what ways they can?
Regarding the new map vs. fixes/perf thing, that is actually a fair point. Yes, we chose to spend a lot of the donated money on a new map. Maybe we should have spent it on contracting one programmer (probably fsfod) to optimize the game for 2-3 months? That's how long it probably would have taken to get a significant bump in perf - it would basically involve changing 50% of Lua code, everywhere. But, that still would've been risky - would the perf actually get much better after that? How much better? How many MORE bugs would it introduce? It would certainly break ALL mods. So, at the end of the day, you can see why making a new map is a much safer bet than a complete overhaul of the code.
We have definitely learned from the performance woes of NS2 (yes, we absolutely acknowledge it as an issue), and for future projects that use Lua, we will architect things from the ground up with LuaJIT in mind. As for NS2, unfortunately, it's hard to make those kind of changes this late with a gigantic code base.
How mechanical is the process of refactoring to take advantage of LuaJIT? Is it the sort of thing you can do with only making local changes to the codebase? Do you guys have a profiler that you trust? Is there enough test coverage that you'd have a way of trustIng community contributions?
I'd really enjoy working on this periodically if its a tractable thing to do. I'm not familiar with either Lua or LuaJIT, but if its just a tedious process rather than a difficult one, it wouldn't be hard to learn.
nonetheless, relative to their player base, it's not THAT many maps! if you're going to do benchmarking, never look solely at absolute numbers.
What?! What twisted kind of logic is that? "The more players we have, the more maps we need. If we have a 100 million players we will need a thousand maps or else they'll get bored!"
What matters is the individual player. If the individual player is bored on 10 maps, then 20 more maps could change that. Same goes for the next individual player. Just because you increase the playercount from one to two doesn't mean those two guys suddenly need double the amount of maps!
exactly my point. you don't need many maps to keep a big player base. don't make my argument sound like a twisted logic - what i've tried to point out is those two games have such a huge player base relative to the number of official maps. what @ns2isgood pointed out is that big games are successful as a result of content i.e. precisely what you quoted.
games like team fortress 2 (f2p) or left4dead2 (not f2p) don't have that many maps either
lol, what? tf2 has over 30+ official maps and probably thousands of user created maps. same with l4d2, there are over 10 campaigns with each having 4+ maps each and again, and a ton of community created maps.
tf2, l4d2 have small number of maps. so does ns2, ergo you do not need new content to sustain a big player base.
anyway, it seems like they aren't keen in giving ns2 an overhaul anymore and just wants to milk every last bit of it. so i assume their objective is just new payable content and new payable content with little regards to fixes (unless critical) from now.
Game is over a year and half out since release... Why on earth would you expect an "overhaul?" Why shouldn't they support, and try to get the most out of their product in what ways they can?
correct. my point being there's only going to be new and more new content with little technical support.
games like team fortress 2 (f2p) or left4dead2 (not f2p) don't have that many maps either
lol, what? tf2 has over 30+ official maps and probably thousands of user created maps. same with l4d2, there are over 10 campaigns with each having 4+ maps each and again, and a ton of community created maps.
how many tf2 payload maps are there? if you add up all other maps for other modes (how many official modes does ns2 have?), then yes it's a fairly large number. nonetheless, relative to their player base, it's not THAT many maps! if you're going to do benchmarking, never look solely at absolute numbers.
also, where do community created maps come from? among others like game concept, a good confidence in the game stability and performance! we don't see that many for ns2, do we?
who ever said anything about game modes? at the end of the day it's content no matter how you want to twist it to try and win a forum argument.
then why are you comparing apples with oranges? there's no relevance to your argument so why do you even want to twist my logic? i've never wanted to "win a forum argument". i just analyzed that this was a failed direction and indeed uwe has learned from it.
games like team fortress 2 (f2p) or left4dead2 (not f2p) don't have that many maps either
lol, what? tf2 has over 30+ official maps and probably thousands of user created maps. same with l4d2, there are over 10 campaigns with each having 4+ maps each and again, and a ton of community created maps.
You then claimed that this amount of maps is low, compared to the userbase - just like with ns2.
how many tf2 payload maps are there? if you add up all other maps for other modes (how many official modes does ns2 have?), then yes it's a fairly large number. nonetheless, relative to their player base, it's not THAT many maps! if you're going to do benchmarking, never look solely at absolute numbers.
Ergo, you say the amount of map a FPS needs is relative to their playerbase, which brings me to my original point...
nonetheless, relative to their player base, it's not THAT many maps! if you're going to do benchmarking, never look solely at absolute numbers.
What?! What twisted kind of logic is that? "The more players we have, the more maps we need. If we have a 100 million players we will need a thousand maps or else they'll get bored!"
What matters is the individual player. If the individual player is bored on 10 maps, then 20 more maps could change that. Same goes for the next individual player. Just because you increase the playercount from one to two doesn't mean those two guys suddenly need double the amount of maps!
games like team fortress 2 (f2p) or left4dead2 (not f2p) don't have that many maps either
lol, what? tf2 has over 30+ official maps and probably thousands of user created maps. same with l4d2, there are over 10 campaigns with each having 4+ maps each and again, and a ton of community created maps.
You then claimed that this amount of maps is low, compared to the userbase - just like with ns2.
how many tf2 payload maps are there? if you add up all other maps for other modes (how many official modes does ns2 have?), then yes it's a fairly large number. nonetheless, relative to their player base, it's not THAT many maps! if you're going to do benchmarking, never look solely at absolute numbers.
Ergo, you say the amount of map a FPS needs is relative to their playerbase, which brings me to my original point...
nonetheless, relative to their player base, it's not THAT many maps! if you're going to do benchmarking, never look solely at absolute numbers.
What?! What twisted kind of logic is that? "The more players we have, the more maps we need. If we have a 100 million players we will need a thousand maps or else they'll get bored!"
What matters is the individual player. If the individual player is bored on 10 maps, then 20 more maps could change that. Same goes for the next individual player. Just because you increase the playercount from one to two doesn't mean those two guys suddenly need double the amount of maps!
how many tf2 payload maps are there? if you add up all other maps for other modes (how many official modes does ns2 have?), then yes it's a fairly large number. nonetheless, relative to their player base, it's not THAT many maps! if you're going to do benchmarking, never look solely at absolute numbers.
payload maps - 10 maybe? player base for just payload maps? BIG
ns2 maps - 10 including the new kodiak. player base? small
lol, what? tf2 has over 30+ official maps and probably thousands of user created maps. same with l4d2, there are over 10 campaigns with each having 4+ maps each and again, and a ton of community created maps.
People continue to say we should have spent the money on performance improvements instead of a new map so I Just wanted to reiterate that 100K does not necessarily automatically equal a big performance improvement. Most of the big wins in performance were made already, and at this point it would require much more time and effort for very diminishing returns. As Steve said it would involve most of the game being rewritten to take full advantage of things like Luajit, something which would likely have huge repercussions throughout the game, breaking stuff left and right, with an unknown end result. We could spend months and months on it and end up with a gain of 5 fps. Sure, some people would notice that, but most wouldn't.
We decided to spend the money on things that were more guaranteed to be noticeable, like new content. But don't forget that while Eclipse was being worked on, on the art and mapping side, programmer time was still being spent fixing bugs and issues with the game.
We could endlessly argue over the impact of performance in the game right now. I don't expect to change anyone's minds on the matter, there will always be people who view it as the silver bullet that would bring in tons more players, keep people from dropping the game, and magically bring everyone back to the game who stopped playing. Maybe this was more true at the initial release of the game, but there have been huge performance gains since then.
I have visited the Steam forums of other successful games and just as for NS2 there's a large percentage of people complaining about performance in those games. Just something that is always going to be the case for PC games.
I believe that what keeps people from playing NS2 is not the performance, but the overly punishing game mechanics that prevent many new players from sliding easily into the game. Face it, NS2 is a brutal game for new players, with a steep learning curve and limited rewards to encourage these new players to keep playing. We've learned a lot making NS2, that our future projects will certainly benefit from, but at this point, to make a big difference to player counts NS2 would probably need to be reworked from the ground up.
All that said, we are very thankful that NS2 was successful enough to allow us to keep making large updates to it a year and a half after it's release, which is pretty rare in the game industry.
People continue to say we should have spent the money on performance improvements instead of a new map so I Just wanted to reiterate that 100K does not necessarily automatically equal a big performance improvement. Most of the big wins in performance were made already, and at this point it would require much more time and effort for very diminishing returns.
that's fine. an investment can never be guaranteed to make returns. and returns are always uncertain. this is more so in the game industry.
We could endlessly argue over the impact of performance in the game right now. I don't expect to change anyone's minds on the matter, there will always be people who view it as the silver bullet that would bring in tons more players, keep people from dropping the game, and magically bring everyone back to the game who stopped playing. Maybe this was more true at the initial release of the game, but there have been huge performance gains since then.
thank you, i have personally witnessed less crashes and better gameplay albeit with ugly infestations and squarish textures. @SupaFred had also proved there had since been huge performance gains.
until performance issues have been fixed to an acceptable level (the average low~mid-end modern pc can play), people are just going to buy the game and then complain about the lag before dropping it.
uwe had weighed on this and decided that content is more important than fixes. so be it.
but if it had another $100k, would it have been spent prudently to make the game more playable for the average pc?
We decided to spend the money on things that were more guaranteed to be noticeable, like new content. But don't forget that while Eclipse was being worked on, on the art and mapping side, programmer time was still being spent fixing bugs and issues with the game.
We could endlessly argue over the impact of performance in the game right now. I don't expect to change anyone's minds on the matter, there will always be people who view it as the silver bullet that would bring in tons more players, keep people from dropping the game, and magically bring everyone back to the game who stopped playing. Maybe this was more true at the initial release of the game, but there have been huge performance gains since then.
I have visited the Steam forums of other successful games and just as for NS2 there's a large percentage of people complaining about performance in those games. Just something that is always going to be the case for PC games.
i'm not sure if dota2 as a successful game or even minecraft as a niche success requires a high end pc to be enjoyable - but that's just a few exceptional cases, right?
I believe that what keeps people from playing NS2 is not the performance, but the overly punishing game mechanics that prevent many new players from sliding easily into the game. Face it, NS2 is a brutal game for new players, with a steep learning curve and limited rewards to encourage these new players to keep playing. We've learned a lot making NS2, that our future projects will certainly benefit from, but at this point, to make a big difference to player counts NS2 would probably need to be reworked from the ground up.
you are correct. just like any other RTS and MOBA games, punishing game mechanics are what make games so addictive and fun. that's why i have waited over 10 years for ns2.
All that said, we are very thankful that NS2 was successful enough to allow us to keep making large updates to it a year and a half after it's release, which is pretty rare in the game industry.
thank you too. i was able to rekindle with some players i met from ns1 because of ns2.
People continue to say we should have spent the money on performance improvements instead of a new map so I Just wanted to reiterate that 100K does not necessarily automatically equal a big performance improvement. Most of the big wins in performance were made already, and at this point it would require much more time and effort for very diminishing returns. As Steve said it would involve most of the game being rewritten to take full advantage of things like Luajit, something which would likely have huge repercussions throughout the game, breaking stuff left and right, with an unknown end result. We could spend months and months on it and end up with a gain of 5 fps. Sure, some people would notice that, but most wouldn't.
Why should it not result in a huge gain? Why would NS2 only gain 5 FPS?
Earlier you explained how investing into a mapping team works and how it yields quality maps. You have several mappers and prop designer/graphics designers involved and the entire team works together to yield a big awesome result.
Why is that not possible for performance? If you spend 100k on a team to do work for any aspect of a game I would expect a great end result. How is performance the huge dice roll that you are making it out to be?
I agree that 5 FPS would not be worth it but I have a hard time believing that would be the "common result" of a developer putting 100k or something similar into increasing performance for their game. That just doesn't make sense to me. Is performance just one of those weird things where X time * X effort does not = Y result? Or is this just specific to NS2 because of some of the features that have been implemented that are resource hogs?
We decided to spend the money on things that were more guaranteed to be noticeable, like new content. But don't forget that while Eclipse was being worked on, on the art and mapping side, programmer time was still being spent fixing bugs and issues with the game.
We could endlessly argue over the impact of performance in the game right now. I don't expect to change anyone's minds on the matter, there will always be people who view it as the silver bullet that would bring in tons more players, keep people from dropping the game, and magically bring everyone back to the game who stopped playing. Maybe this was more true at the initial release of the game, but there have been huge performance gains since then.
I don't think spending money on maps was a bad idea for sales.
However, it certainly wasn't what the community was asking for the most. Bugs and fixes are always welcome and I haven't forgotten how relieved I was after some of them were implemented. I was never saying that you completely dropped everything but maps.
I don't seek to endlessly argue here. I'm mostly trying to make the point that performance in NS2 has always been a larger issue and more important than the developers of NS2 seem to realize.. or openly admit in most cases.
I have visited the Steam forums of other successful games and just as for NS2 there's a large percentage of people complaining about performance in those games. Just something that is always going to be the case for PC games.
I believe that what keeps people from playing NS2 is not the performance, but the overly punishing game mechanics that prevent many new players from sliding easily into the game. Face it, NS2 is a brutal game for new players, with a steep learning curve and limited rewards to encourage these new players to keep playing. We've learned a lot making NS2, that our future projects will certainly benefit from, but at this point, to make a big difference to player counts NS2 would probably need to be reworked from the ground up.
All that said, we are very thankful that NS2 was successful enough to allow us to keep making large updates to it a year and a half after it's release, which is pretty rare in the game industry.
Visiting the Steam forums isn't something I would consider valuable evidence in this kind of discussion on performance. I'm in no way insulting you or anything along those lines but if you want to get down to the hard facts it's as simple as loading up NS2 on my friend's PC and getting 15-30 FPS average in an average round of NS2 public and then loading up CS:GO and getting stable 60+ FPS the entire game; every game.
Complaints on Steam forums are a given and are going to happen whether or not the game actually runs well in general or not.
The fact is that most people can run Source games on a potato. It's basic functionality and it's guaranteed. That's what NS2 doesn't have.
Going the route of "all PC games have this issue" sounds like a cop out to me. Go load up any Source game or current gen first person shooter and benchmark the game and then go benchmark NS2 on the same hardware. If you don't believe me that there is a huge difference then I will gladly do it and display it here.
I think that new players definitely have a hard time in this game but if you believe the reasoning that people leave due to that then why was the 100k not put into something that resolves that issue instead of more maps? Whether it be peformance or a lack of matchmaking/new player friendly features.. it seems like the direction of the improvements and patches was not in the right place.
All of that said, I'm really grateful for your reply and for making the game I enjoy so much.
I'm sure that if UWE says that they will be using the Natural Selection franchise as their launch game for their engine people will be happy.
Rumor has it that Spark 2.0 is in the works, so maybe we'll see a new Natural Selection title on their new-and-improved engine.
A Natural Selection that uses the knowledge they've gained from NS2, and with an engine that has been improved and fleshed out (also with a MUCH shorter dev time ).
I think most people just want to be reassured that while NS2 development has slowed down, there may be plans for a continuation of the game that includes all the knowledge, experience, and improvements made from the making of NS2.
I know that I would buy that Natural Selection a few years down the road, and keep playing this one until then.
Why should it not result in a huge gain? Why would NS2 only gain 5 FPS?
Earlier you explained how investing into a mapping team works and how it yields quality maps. You have several mappers and prop designer/graphics designers involved and the entire team works together to yield a big awesome result.
Why is that not possible for performance? If you spend 100k on a team to do work for any aspect of a game I would expect a great end result. How is performance the huge dice roll that you are making it out to be?
I agree that 5 FPS would not be worth it but I have a hard time believing that would be the "common result" of a developer putting 100k or something similar into increasing performance for their game. That just doesn't make sense to me. Is performance just one of those weird things where X time * X effort does not = Y result? Or is this just specific to NS2 because of some of the features that have been implemented that are resource hogs?
I see where you are coming from, but there are two main reasons that the outcome is not as clear-cut as spending the same amount of time/money on adding a new level.
The first is that a level is a relatively self-contained object. You can get all sorts of bugs and issues inside a level but it's extremely unlikely that the level will have an adverse affect on another part of the game. So assuming that the level itself is of appropriate quality, this makes adding it to the game a relatively low-risk proposition. This is in contrast to making changes to code in NS2 (or any large, complex software project). There's a huge amount of dependency. Let's say you identify a potential performance gain in class 'ClassA()'. Making changes to ClassA has potential and sometimes startling effects on every other class that references or relies upon it. In a project of this size and complexity, that's usually going to be rather a high number. What this means is that there's an inherent risk in making code changes.
Secondly, as Cory already brought up, we've already made all the obvious changes. All the "low-hanging fruit" that might lead to significant performance gains have already been dealt with. What's left is either rewriting very large chunks of the game or making a lot of small changes all over the codebase for very small performance gains per change. The risk of taking these options are naturally multiplied by reason one.
These factors increase in risk the more mature a project gets. Hopefully this has made it clearer for you why there isn't a 1:1 equivalency for value and risk in adding a new map against making sizable code changes to a project as mature as NS2. If there was we'd definitely already have done it!
Comments
@SteveRock @Squeal_like_A_Pig
I'm glad you guys took the time to respond to this topic and appreciate that you read our posts related to this issue.
I love your game and have played it since age 12 (NS1) and will continue to play it until it dies. I am not just another competitive player complaining in this post. I have played public games in NS1 and NS2 for a very long time.
I have quite an opinion on this issue (performance and NS2's problems) so if you do decide to read.. prepare for a wall of text.
I couldn't agree more with your last statement there Cory. Spending a ton of money on something that keeps NS2 alive should definitely be the focus but I have a different opinion on where the focus should be. I completely respect that you have a ton more game development experience than me but I still want to give my opinion here which is largely based on my own observations in games that are currently accomplishing something that NS2 is not. Playerbase retention. I'm not out to prove you wrong here but I do want to bring your attention to something I feel has a bigger impact on playerbase retention than more maps.
To answer your questions there Cory.. I say YES. It not only brings back players that had issues previously but it also allows a larger amount of people to play the game at all to begin with. Example: a player that would get 15-30 FPS in common scenarios in NS2 upon trying the first time now gets 30-60 FPS after Build XXX patch. Unplayable -> playable. Planetside 2, for example, had similar issues to NS2 and they made big improvements to performance and quite a lot of people came back to see how it fared on their PCs. In addition to that, they have quite a neat system in place regarding skins and content that is along the lines of the Kodiak update in terms of community members benefiting from customization alongside the developer.
A great map like Kodiak or Eclipse might be safer to invest money into from your point of view but it doesn't really halt or relieve the massively declining player base scenario that has happened on multiple occasions to Natural Selection 2. In the long term.. these types of additions to the game fall short in helping resolve the biggest issue. CS:GO which I will refer to several times here.. has a good amount of maps but I think everyone knows how much people play Dust/Dust2. I don't buy that people quit games because of limitation in the maps available. If the gameplay and other basic functionality aspects are there, like in CS:GO for example.. people will keep playing your game for a VERY long time. Maps are nice and all and I really personally admire your concepts and art design/direction but are not something that fix a big issue like a declining playerbase.
You can see it clearly on the SteamGraph in regards to the playerbase issue. I don't think I need to link it or provide #'s but it's pretty clear NS2 is not on the up and up.
Before I continue on I want to say this: I don't say any of this with the intention of being sour about the topic or NS2 development. I love NS2 and NS1 gameplay/conceptually and it's the only reason I'm still around. I don't say any of this with the intention of bashing you guys as developers either but I had a lot of hope and expectation that performance was going to be a big focus. I only give my thoughts and criticism here in hopes that NS2 will retain players in the future and not end up a game that barely anyone plays anymore.
Content brings people back temporarily to try it out thanks to flashy and professional trailers but it doesn't keep people playing. Gameplay, performance (and thus accessibility) and certain features like matchmaking I think have a far larger impact on a consistent playerbase. Going F2P at this point in a games life also increases accessibility by a huge amount but ONLY if performance allows it to.
Of course performance increases don't guarantee "everyones playing ns2 now" but think of it this way:
Shortly after release there were a ton of things that people expected from you guys. (First person spectate, and many other improvements all being asked for on stream chats)
Obviously you couldn't fulfill all the demands instantly but one thing stood out.. performance.
When you put up a poll on what people wanted the most after release - it was an astounding landslide in the favor of performance.
You put out a patch featuring LuaJIT which was quite a good increase in FPS across the board (also reduced frame jitter) and then I had a hard time understanding why improvements stopped there. It's been almost 2 years since release / end of beta days and the performance improvements have for the most part halted entirely after LuaJIT. If F2P is on the table then you really need to think about the type of machines that run F2P games and try to cater to those people when you do decide to go F2P. (via increases in performance)
My feeling on it is that by increasing performance you have very little to lose (bugs [which are fixable, and I know requires time, money and effort to fix but in theory it should be worth it - will explain below] like you mentioned).
You allow a larger spread of people (more people.. more money obviously) with varying machines to play your game without experiencing performance issues on top of the increasing the amount of people that would be able to play at all on the super low end machines. People can play CS:GO [a comparable FPS to some degree] (and many other games) on a potato and get their refresh rate's worth of FPS. (60fps on 60hz for most people) NS2 - I play on an i5 with 4.8ghz and a GTX580 and still get astoundingly low framerates in common scenarios (40's-50's on my kind of machine with the CPU as the bottleneck) in public games - even on 18 slot servers with the lowest settings. You guys made great improvements after release but I can't help but think - why stop there..
If you think long term here - if more people can play your game potentially.. when or if you decide to go F2P (the route TF2 successfully went) or have more sales that is more people that can run your game acceptably - which hopefully stick around and purchase skins/DLC.. which can sustain development (plus tell their friends about it and even more sales).
It's an investment that puts your game on the level of other games in terms of accessibility. It might be risky but when you see your playerbase is rollercoasting with each sale that should be a sign that something isn't quite right and a plan to change that should be thought up. Almost every update that you guys have put out that has included a map and other features you see the playerbase skyrocket on the steamgraph temporarily and then everyone quits again. And I have never once heard anyone in the 2000+ hours I've played in this game tell me that they dislike NS2 or quit because of too much Summit.
Continuing on the note of performance though - I have read, agreed with and understood the statements from Hugh that Source has been around for ages and that NS2 is using a new engine. I would never expect Spark/NS2 to compete with Source/CS:GO within the 10-20% margin in terms of performance but I know Max, SteveRock and others from your team can kick some ass and if you guys put money and time into it - I think you could make it pretty damn close to the point that most people wouldn't notice that it's better/worse.
I don't expect you guys to work miracles here but a friend of mine played your game for about 2,000 hours (I've played about the same/more) without really any performance complaints after LuaJIT.. then Reinforced released.. The game went from 30-60FPS average on my friend's machine to severe hitching issues and sometimes as low as 15 FPS in public games. That friend of mine has quit and has taken up CS:GO for his FPS game that he pubs regularly and buys skins in. He really enjoyed NS2 and preferred the gameplay but the performance issues wore him out. Forcing texture streaming on I believe was the root of his issue and he hasn't played the game since nor purchased a Reinforced package / the most recent DLC (obviously, because he quit and his issues were never fixed). Many people that I know that played NS2 have a similar story.
That's an example of how you lose players and lose potential $$$ because content was prioritized over performance. People don't leave your game in droves because you aren't putting out tons of new maps and weapons [ I think people have much more tolerance for that stuff ]. I think people leave your game when they either A) can't run it as well as they would like to enjoy it [or enjoyed it previously in my friend's case]
If you look at a lot of successful games.. not just FPS games, I think you'll find that the top games had the basic functionality NAILED from the get go - and then all the cool content and features came later with new versions of the game / expansions / sequels AFTER people were hooked on the gameplay and able to even play it acceptably performance-wise in the first place (accessibility tied to performance in NS2's case).
I don't think all of NS2's issues related to people leaving are due to performance but I would wager that it is a big part of it. I certainly don't think that I know how to solve all of NS2's problems and that performance increases in huge amounts is going to save the game.. but basic functionality is something quite a lot of people expect to be a given. After that I think most of the people that left were casuals that didn't stick with the game because it was too difficult or something else caught their attention. But there are far more not-so-intuitive games out there than NS2 that don't have this rollercoaster "try and quit" playerbase issue.
In conclusion here, I think that performance is definitely something that has needed attention for a long time coming. I also think it's a good idea to look at your competition and see where you fall short. I will repeat that I am not expecting you guys to be equal in all regards to a Valve title with X amount less money and manpower but it's still something worth noting.
CS:GO vs NS2
Gameplay - about as good or slightly better/worse depending on who you ask. I personally think NS gameplay has just as much potential or more to be what CS:GO has become in terms of esports and public gameplay.
Performance - far worse and much more inconsistent based on the PC and servers in question.
Matchmaking - extremely good with skill placement and progression versus non-existent.
TL;DR - I had hoped for more focus on performance after release and I feel like the focus wasn't in the right place after LuaJIT. This new strategy for development involving Kodiak and DLC has me hopeful but I would like to see a stronger focus on performance over content still. (and finishing matchmaking - I feel like that was a great idea that was just forgotten about and never put into place when the playerbase was at it's height [no one uses it now because there's not enough players to really make it viable, plus it doesn't work]) I might be naive to think this but I think NS2 can still progress out of this cycle of 'patch - people come back - people quit again'.
New content is always welcome, but I believe the main issue with NS2's success has been its performance or lack thereof. I love the game and hope to see the community grow.
In short, Better performing game = Larger possible player base who can actually enjoy the game.
I think you are underestimating how many people would've come back with improved preformance.
Of the 10 people I got to play NS2, none are still here (but me!). Of those 10 people, they all* say they would come back if performance was better.
*one says that he would only come back if the rifle-butt knock-back was re-introduced, so thats a lost cause
I understand the desire/need to move on to other projects, but perhaps if you started another Reinforced program, explicitly for performance focus......?
Here's why i stop playing NS2 everytime, shortly after an update.
I never had (fps) performance issues. My main gripe is always related to hitboxes and hitdetection:
-Being a Skulk or Lerk in a vent somewhere, or behind some cover, thinking no marine can see me, when actually he sees the top of my character and so kills me.
-Running or flying away when fired upon, and then dying behind a corner after still taking damage. Although i was not in the line of sight of my attacker anymore.
I have a ping that's less than 60 on those servers. Also i monitor the server tickrate in game.
While there may be perfectly logical explanations for those issues, they don't happen in many other fps games i play (*). The first, regarding the character's point of view, i don't even understand why UWE has it that way. It gives the player a wrong suggestion of his character's height.
(*) Tribes Ascend, Arma3, RO2/Rising Storm, TF2, BF2:PR, PS2, Insurgency, DayZ
You probably don't realize it Zeep but dying around corners is directly related to a performance limitation of NS2's netcode.
The current implementation of 100ms interpolation and lag compensation is the source of that issue.
The 30tickrate you're looking at is related to AI according to Max, the client update rate is 20. That is what is limiting the server performance as well.
The hit detection in NS2 is actually quite good. There's a few issues with it but I won't go into detail.
i5 3570k @4.0 gtx680
with $100k, you will
A. create new content like ns2_eclipse "for free" and then milk money from existing miniscule (and declining) playerbase with kodiak dlc ($5.99 x 1000 players) OR
B. fix the underlying problem of the game - performance - and therefore
1. retain old players who quit due to performance issue (+~6000 players x $5.99)
2. earn new players from sales/adverts (+? players x $29.99)
3. encourage current players who now aren't afraid to introduce a good game (previously unplayable due to high pc requirements but now every average pc plays awesome!
4. make modders happy in helping to create a stable game which hopefully stops breaking their mod in almost every new patch. old modders will thus also be motivated to fix their mods to be compatible with a much better performing game (+? players who like modded games)
with so much focus on client-side performance, let's not forget better multicore/multithread support for servers so it's economically viable to host servers (consumer grade hardware and 4.5GHz just to host one instance of 24 slots WTF?!) around the globe especially in the Asia Pacific region where infrastructure is so expensive
games like team fortress 2 (f2p) or left4dead2 (not f2p) don't have that many maps either, but they do have a very large player base. <sarcasm> i wonder why...? </sarcasm>
let's hope the company still have another $100k for a wise turnaround strategy
I had to comment on this. This is my server load with three 24-player instances running: cpu load
Since this screenshot was taken I have replaced the cpu with another one of the same model that overclocks higher at a lower voltage. And I know that Combat servers doesn't require as much as normal NS2 but I don't have a screenshot of the server running more instances.
lol, what? tf2 has over 30+ official maps and probably thousands of user created maps. same with l4d2, there are over 10 campaigns with each having 4+ maps each and again, and a ton of community created maps.
what are the servers doing though? i'd love to see the tickrate of all 3 servers while they're all packed and being played on at the same time.
how many tf2 payload maps are there? if you add up all other maps for other modes (how many official modes does ns2 have?), then yes it's a fairly large number. nonetheless, relative to their player base, it's not THAT many maps! if you're going to do benchmarking, never look solely at absolute numbers.
also, where do community created maps come from? among others like game concept, a good confidence in the game stability and performance! we don't see that many for ns2, do we?
that is a haswell at 4.4GHz. i may have exaggerated on my previous post and that was my experience of trying to host one during the b229 days, so while things may have improved significantly, that is still a consumer grade high end cpu. does a 3 x 24 player tf2 require just as high end cpu?
good luck to a dying community. i really loved the game concept back from ns1 and now but i'm just not satisfied with where they place their focus.
who ever said anything about game modes? at the end of the day it's content no matter how you want to twist it to try and win a forum argument.
What?! What twisted kind of logic is that? "The more players we have, the more maps we need. If we have a 100 million players we will need a thousand maps or else they'll get bored!"
What matters is the individual player. If the individual player is bored on 10 maps, then 20 more maps could change that. Same goes for the next individual player. Just because you increase the playercount from one to two doesn't mean those two guys suddenly need double the amount of maps!
Game is over a year and half out since release... Why on earth would you expect an "overhaul?" Why shouldn't they support, and try to get the most out of their product in what ways they can?
How mechanical is the process of refactoring to take advantage of LuaJIT? Is it the sort of thing you can do with only making local changes to the codebase? Do you guys have a profiler that you trust? Is there enough test coverage that you'd have a way of trustIng community contributions?
I'd really enjoy working on this periodically if its a tractable thing to do. I'm not familiar with either Lua or LuaJIT, but if its just a tedious process rather than a difficult one, it wouldn't be hard to learn.
exactly my point. you don't need many maps to keep a big player base. don't make my argument sound like a twisted logic - what i've tried to point out is those two games have such a huge player base relative to the number of official maps. what @ns2isgood pointed out is that big games are successful as a result of content i.e. precisely what you quoted.
tf2, l4d2 have small number of maps. so does ns2, ergo you do not need new content to sustain a big player base.
correct. my point being there's only going to be new and more new content with little technical support.
oh look, i just repeated myself
then why are you comparing apples with oranges? there's no relevance to your argument so why do you even want to twist my logic? i've never wanted to "win a forum argument". i just analyzed that this was a failed direction and indeed uwe has learned from it.
First you said
ns2isgood then responded that those games do, indeed, have a lot of maps.
You then claimed that this amount of maps is low, compared to the userbase - just like with ns2.
Ergo, you say the amount of map a FPS needs is relative to their playerbase, which brings me to my original point...
......
let's do this is in plain, broken english.
tf2 and l4d2 - wow, successful games - huge player base! number of maps? not many. ok.
ns2? successful too! big player base! but only initially
but why are they successful?! number of maps? no! it's good game concept (which ns2 has), AND good performance (which ns2 has failed terribly).
how to repair this state of dropping player base crisis?!
A. fix the performance
B. add new dlc like skins and maps
uwe's answer: B
payload maps - 10 maybe? player base for just payload maps? BIG
ns2 maps - 10 including the new kodiak. player base? small
ns2isgood had indirectly suggested that in order to gather a larger player base, more content i.e. maps are needed. i have never said so.
We decided to spend the money on things that were more guaranteed to be noticeable, like new content. But don't forget that while Eclipse was being worked on, on the art and mapping side, programmer time was still being spent fixing bugs and issues with the game.
We could endlessly argue over the impact of performance in the game right now. I don't expect to change anyone's minds on the matter, there will always be people who view it as the silver bullet that would bring in tons more players, keep people from dropping the game, and magically bring everyone back to the game who stopped playing. Maybe this was more true at the initial release of the game, but there have been huge performance gains since then.
I have visited the Steam forums of other successful games and just as for NS2 there's a large percentage of people complaining about performance in those games. Just something that is always going to be the case for PC games.
I believe that what keeps people from playing NS2 is not the performance, but the overly punishing game mechanics that prevent many new players from sliding easily into the game. Face it, NS2 is a brutal game for new players, with a steep learning curve and limited rewards to encourage these new players to keep playing. We've learned a lot making NS2, that our future projects will certainly benefit from, but at this point, to make a big difference to player counts NS2 would probably need to be reworked from the ground up.
All that said, we are very thankful that NS2 was successful enough to allow us to keep making large updates to it a year and a half after it's release, which is pretty rare in the game industry.
that's fine. an investment can never be guaranteed to make returns. and returns are always uncertain. this is more so in the game industry.
thank you, i have personally witnessed less crashes and better gameplay albeit with ugly infestations and squarish textures. @SupaFred had also proved there had since been huge performance gains.
until performance issues have been fixed to an acceptable level (the average low~mid-end modern pc can play), people are just going to buy the game and then complain about the lag before dropping it.
uwe had weighed on this and decided that content is more important than fixes. so be it.
but if it had another $100k, would it have been spent prudently to make the game more playable for the average pc?
from this, i deduce it's no.
i'm not sure if dota2 as a successful game or even minecraft as a niche success requires a high end pc to be enjoyable - but that's just a few exceptional cases, right?
you are correct. just like any other RTS and MOBA games, punishing game mechanics are what make games so addictive and fun. that's why i have waited over 10 years for ns2.
thank you too. i was able to rekindle with some players i met from ns1 because of ns2.
Why should it not result in a huge gain? Why would NS2 only gain 5 FPS?
Earlier you explained how investing into a mapping team works and how it yields quality maps. You have several mappers and prop designer/graphics designers involved and the entire team works together to yield a big awesome result.
Why is that not possible for performance? If you spend 100k on a team to do work for any aspect of a game I would expect a great end result. How is performance the huge dice roll that you are making it out to be?
I agree that 5 FPS would not be worth it but I have a hard time believing that would be the "common result" of a developer putting 100k or something similar into increasing performance for their game. That just doesn't make sense to me. Is performance just one of those weird things where X time * X effort does not = Y result? Or is this just specific to NS2 because of some of the features that have been implemented that are resource hogs?
I don't think spending money on maps was a bad idea for sales.
However, it certainly wasn't what the community was asking for the most. Bugs and fixes are always welcome and I haven't forgotten how relieved I was after some of them were implemented. I was never saying that you completely dropped everything but maps.
I don't seek to endlessly argue here. I'm mostly trying to make the point that performance in NS2 has always been a larger issue and more important than the developers of NS2 seem to realize.. or openly admit in most cases.
Visiting the Steam forums isn't something I would consider valuable evidence in this kind of discussion on performance. I'm in no way insulting you or anything along those lines but if you want to get down to the hard facts it's as simple as loading up NS2 on my friend's PC and getting 15-30 FPS average in an average round of NS2 public and then loading up CS:GO and getting stable 60+ FPS the entire game; every game.
Complaints on Steam forums are a given and are going to happen whether or not the game actually runs well in general or not.
The fact is that most people can run Source games on a potato. It's basic functionality and it's guaranteed. That's what NS2 doesn't have.
Going the route of "all PC games have this issue" sounds like a cop out to me. Go load up any Source game or current gen first person shooter and benchmark the game and then go benchmark NS2 on the same hardware. If you don't believe me that there is a huge difference then I will gladly do it and display it here.
I think that new players definitely have a hard time in this game but if you believe the reasoning that people leave due to that then why was the 100k not put into something that resolves that issue instead of more maps? Whether it be peformance or a lack of matchmaking/new player friendly features.. it seems like the direction of the improvements and patches was not in the right place.
All of that said, I'm really grateful for your reply and for making the game I enjoy so much.
Rumor has it that Spark 2.0 is in the works, so maybe we'll see a new Natural Selection title on their new-and-improved engine.
A Natural Selection that uses the knowledge they've gained from NS2, and with an engine that has been improved and fleshed out (also with a MUCH shorter dev time
I think most people just want to be reassured that while NS2 development has slowed down, there may be plans for a continuation of the game that includes all the knowledge, experience, and improvements made from the making of NS2.
I know that I would buy that Natural Selection a few years down the road, and keep playing this one until then.
I see where you are coming from, but there are two main reasons that the outcome is not as clear-cut as spending the same amount of time/money on adding a new level.
The first is that a level is a relatively self-contained object. You can get all sorts of bugs and issues inside a level but it's extremely unlikely that the level will have an adverse affect on another part of the game. So assuming that the level itself is of appropriate quality, this makes adding it to the game a relatively low-risk proposition. This is in contrast to making changes to code in NS2 (or any large, complex software project). There's a huge amount of dependency. Let's say you identify a potential performance gain in class 'ClassA()'. Making changes to ClassA has potential and sometimes startling effects on every other class that references or relies upon it. In a project of this size and complexity, that's usually going to be rather a high number. What this means is that there's an inherent risk in making code changes.
Secondly, as Cory already brought up, we've already made all the obvious changes. All the "low-hanging fruit" that might lead to significant performance gains have already been dealt with. What's left is either rewriting very large chunks of the game or making a lot of small changes all over the codebase for very small performance gains per change. The risk of taking these options are naturally multiplied by reason one.
These factors increase in risk the more mature a project gets. Hopefully this has made it clearer for you why there isn't a 1:1 equivalency for value and risk in adding a new map against making sizable code changes to a project as mature as NS2. If there was we'd definitely already have done it!