An Art of Balance

posted on November 23rd, 2011, 12:32 pm

Dont forget about the warp in which required only engineering (not science also) and 25 perc faster build times for mason if i remember correctly. Or the wonderfull full firing arc e2 with extra acuracy against medium i think, and was faster also.

yeah mayson got reduced build time on all stations. so the strat was rush warpin while spamming norways for death coil, then if the enemy doesnt gg, go for e2s that had practically no drawbacks. mayson NEVER had to build engineering or do chassis 1.

mayson could get warpin while enemy had 1 or 2 ships, and maybe a single turret. and this was back when you could get 2 of the stronger warpins (including steamy). and e1s werent really as easy to kill as now. basically if they got 2 steamy they could just pick one of your buildings (other than starbase) and delete it. usually this was yard.

posted on November 23rd, 2011, 1:40 pm

Thanks for the insight Dominus. I will try to remember these points the next time I start to get impatient waiting for the next patch.  I appreciate all the work that goes into making the game work well - both in regards to balance and stability.

@ Myles - I remember some of those. My little brother gets sad when I tell him how powerful some ships used to be. Then he looks perplexed when I try and explain that it was needed to keep things fair. Then he says he just wants to blow thing up.

posted on November 23rd, 2011, 1:57 pm

Thanks for the insight Dominus. I will try to remember these points the next time I start to get impatient waiting for the next patch.  I appreciate all the work that goes into making the game work well - both in regards to balance and stability.

I know these points, but that doesn't stop me from being impatient :p

posted on November 23rd, 2011, 2:26 pm

I know these points, but that doesn't stop me from being impatient :p

True, I can only try. Sometimes I still get a little impatient, but the wait for the finished product is worth it. I recently bought a game that was released before it was done - I was very disappointed.

Just know developers - Thanks for the hard work and we will still be here when the next release is ready.

posted on November 23rd, 2011, 2:54 pm

Plasma coil was the worst by far lol. You could just spam those and warpins and never lose.

posted on November 23rd, 2011, 10:23 pm

Meanwhile, thanks for the insight into the proccess It's one thing to whine "why are three dodeca's wiping out my Descent?!" but it's another thing entirely to actually be shown "this is the process we use, this is how long it takes to test out a certain feature, please be patient" 

thanks!

posted on November 23rd, 2011, 11:45 pm

Very good description, Dom.  I'm hoping to become a game designer myself when I graduate, and I learn a lot from various gaming forums including Fleet Ops.

I would love to get involved with the behind-the-scenes work, but so would a lot of people around here.  Still, thanks for the insight into your methods.

posted on November 24th, 2011, 3:17 pm

Maybe the fleet ops devs should create a beat group comprised of volunteers to test out beta builds and work out imbalances, i think this would get the best results for future releases.

Altho i think the response to this is going to be they cant keep uploading new beta builds because its too time consuming but how long has it been since a release big or small, quite some time (5-6 months i think), don't want to sound as if im complaining but the online community is getting less and less (tunngle) and the last thing i want is people moving on to other games and not returning, the death of this game is the very last thing i want because its so awesome.

Maybe instead of working on huge builds they should do as they announced and do smaller ones, what ever happened to that?

posted on November 24th, 2011, 4:10 pm

We released small builds for patches 3.1.0-3.2.5, but at a certain point we needed to have an extended patch to release the redoes: it's just not possible to release them in small bits, as it's quite impossible to balance like that (the redoes are quite significant changes to balancing across the board, and the groundwork to both implement them and test them takes awhile). 

We do balance Fleet Ops as a group of course - it's not as if just one of us does the balancing 

Before every major patch people leave, and after every major patch is released, droves return or new people come - this time seems to be no different, although the tunngle community is still pretty healthy. Especially when, say, compared to the gap between 3.0.7 and 3.1.0 (where we had less than 10 people as a total community).

[Tryptic]

posted on November 24th, 2011, 4:38 pm

Last edited by Tryptic on November 24th, 2011, 4:40 pm, edited 1 time in total.

I know this would be a large and unorthodox step, but it seems to me that given the small, closeness of this community, it would be totally fine to share your beta builds openly and let us play them.  Like Wesnoth or Warzone 2100, you could have a core build and a test build available for download, and update the test build as often as you feel like it.  People will be willing to re-download it if they know that's what they signed up for.

For me and many others in this community, we would LOVE to test your broken, half-transition builds.  We don't care if it's broken, we play small-community games specifically because we like to give feedback and feel like we're part of the creation process.  If the build is known to be incomplete and unfair, that doesn't make it less fun for me.  After all, I chose to play the test build and I can go back to the core build at any time.

The simple act of switching back and forth between playing the stable and test versions when you get tired of one tends to be enough to keep people entertained almost indefinitely, and then you would have more people and more productive feedback during the cool periods before a patch.

If you're willing to trust us, and occasionally release your work in progress, I think you won't be disappointed by the helpful attitude and useful response you'll see from this community.

posted on November 24th, 2011, 5:08 pm

The team is small and we already give out beta-releases. If a release i broken, you will get hotfixes right a day or two after. The devs have a plan on it and as for now, we are rebuilding larger elements like whole fractions that cant be released point-to-point but only at once.

After this large release, there will be more frequently ones with smaller updates.

Wesnoth is already in a state, where the dev-release is more popular than the stable release. If you have a larger team, you might do it this way - but the devs here are few in numbers so I rather take more content per time instead of having bleeding edge every day without real progress.

posted on November 24th, 2011, 5:11 pm

that method even works for large development stuff like firefox, they publicly release all their half baked stuff, if you install a nightly/alpha/beta you cant really complain its unfinished.

it might end up being more work for the devs though, and i'd rather they spend more time working on the game than on readying half finished things for the community.

[Dominus_Noctis]

posted on November 24th, 2011, 5:24 pm

Last edited by Dominus_Noctis on November 24th, 2011, 5:48 pm, edited 1 time in total.

From our experience, the community actually hates to play buggy, unbalanced, and problematic things  ^-^ . In 3.0.0 there was also a series of more open betas - but they were quite some work to maintain community-involvement wise. However, there are of course a number of different issues:

(grossly oversimplified) The vast majority of people don't test bugs or report things regularly or completely - meaning you end up with a whole bunch of "I got a crash, fix it!" or "Feds are too weak" . There's far too much community management involved with that style of beta, as who do you "trust" to report bugs and not? For instance, you could look at the OOS testing as such an example: there are a few (2) relatively dedicated people who have tested that, but that's out of the total size of the Fleet Ops players - that's quite small and thus of course the OOS stuff is going nowhere fast  :blush: . Similarly, who to "trust" for balancing information and not? And how do you discuss such things in a community of over 150 just online gameplayers (not to mention the few hundred offline or LAN party gamers)?

Likewise, there is the issue of simple logistics - I often make a bunch of changes even between "official" alphas/betas: how would one disseminate those as well? When the version number is changing sometimes several times per day soon you'll have a lot of people playing different versions without knowing which one is correct and it'd be difficult to even ask people to make sure when they report bugs that they are on the correct version.

Of course, if releases were open like that, well, it wouldn't be really a beta - it would be an open version . Furthermore, it'd mean there would be less people playing the last full version, which means you get a whole bunch of gamers very sad  :'( .Releases are meant to be complete things that the community can enjoy "right out of the box" with all the surprises and accoutrements to boot

EDIT: firefox can definitely get away with that for a few reasons. Probably better version control, more developers, and of course - very importantly - it's not a multiplayer game  . That last point of course means that you don't have to synchronize for other than 1 'player' with firefox, but you do for an RTS like FO. Also, on the 'goodies-factor': how many people do you know would want to go back to 3.0.0 Borg when they have access to 3.1.0 style Borg?

posted on November 24th, 2011, 5:50 pm

Okay, that makes sense.  I guess I was under the impression that you did most of your balancing in matlab without actively testing, apparently I was wrong.  Still, normally in a community this small you would expect to see a lot more communication with the players; I've been around for almost 2 years now and I've yet to meet a beta tester (or if I have, I didn't know it)

Can you at least tell us who they are so we can kill them in their sleep  worship at their feet  thank them for their work, and establish a closer relationship with more friendly conversation. 

I mean, what's the point of serving a community if they don't get to see who you are or whether you're actually working?

rabblerabblerabble

posted on November 24th, 2011, 6:04 pm

Matlab is good for technical calculation you do with assumptions staying still. But in a game you have a player, that tries to get around borders, frontiers and who will always strive to make max. damage while recieving less. Matlab is just a way to calculate raw numbers (also in a complex way) but it will never take a strategy into account.

To do this, you have to play the game, human vs. human, where you can see the real advantage or disadvantage of a strategy and go on balancing it depending on this.

The brain of the player is a huge self-learning-matrix and as we all know, after you have given away the input signal and convolute it with the "brain" you can never re-create the input signal, because the process is not linear. (To speak in technical terms).

So you have to balance it the hard way "by hand". Sue, Matlab can help, but the tweaking in done by thinking with this non-linear system. Braaaainz