Tremulous Forum
General => Announcements => Topic started by: Norfenstein on January 09, 2010, 11:50:13 pm
-
Well, the dust is beginning to settle for 1.2. I think we've all had a fair amount of time to adjust to the major changes, and, having let GPP phase two marinate for several weeks longer than phase one, I think we can trust that the statistics collected over this period more-or-less accurately reflect the genuine state of the game. And there's been an improvement.
Reading the graphs: Each cross is one game, the distance from 0 on the vertical axis indicates how close a game was to the mean duration for all games on the graph (the weighting for how much this affects the lines is half as much as we used in phase 1). The numbers in the bottom right are the min/median/max/standard deviation for stage up times and game duration.
Both servers combined
(http://imgur.com/DowVD.png)
Euro server
(http://imgur.com/fz837.png)
US server
(http://imgur.com/MIV8r.png)
The lines for stage 1v1 and 2v2 are the most disappointing, but I'm relieved to see that the endgame (3v3 plus SD, which account for 35% of all games) is as close to perfect as I could hope for. This suggests that humans make better use of stage advantages (and neither team seems to stage up significantly faster than the other). So even though the win ratio isn't perfect (53 aliens to 47 humans), I'm quite satisfied overall. Which means that the gameplay changes I've decided on for phase three are based more on my own judgement and popular sentiment than the objective statistics.
Gameplay Changes for Phase 3
First of all, the flamer is still manifestly useless. In GPP phase 2 it was used for less structure kills than every other human weapon, and only managed to beat out the blaster, grenade, and painsaw in player kills. For a weapon best suited to killing low-level aliens, this is quite lousy. I do like where the damage is at now, but it's clear that the real issue is how much it hurts the wielder. Since kills from splash damage account for less than a quarter of all flamer kills, I think reducing its splash damage will actually be an improvement.
- Flamer splash damage reduced 25 -> 12
Second, a frequent complaint I've heard is that humans have too little stamina for sprinting and dodging. This hadn't really seemed like an issue to me, but I don't disagree with the negative feedback. See this (http://tremulous.net/forum/index.php?topic=12511.msg185466#msg185466) post if you want to know how I decided what to change.
- Reduced sprinting's drain on stamina 8 -> 6
- Changed the stamina threshold for jumping and dodging to the slowing threshold (-500) plus the cost of jumping or dodging (250 for both)
- Increased the stamina restoration rate for standing 25 -> 30
Lastly, the dragoon has probably been the biggest point of contention since the start of the 1.2 gameplay preview. I think everyone has had plenty of time to get used to the new bite range and it seems like a significant number of people still feel it's too short. Moreover, I've come to agree that in bigger games it does feel a little inadequate (my guess is because it's more important to be able to pivot between targets). I think a small change is warranted.
- Dragoon base chomp range increased 72 -> 80
The pounce, in contrast, seems to be considered by a lot of people to be overpowered. I don't necessarily agree that it's too overpowered, but I have felt for a while that it's slightly too hard to dodge -- mostly when the goon doesn't actually make contact with you, but flies past and hits you on the way. A range decrease (mechanically, the pounce works just like every other melee weapon) should make these attacks a little harder while having no impact on direct hits.
- Dragoon pounce range reduced 54 -> 40
There will be a GPP phase four. If none of the above changes turn out to be regressions my plan is to just turn on friendly fire on both servers and see what happens.
-
Both US1 and EU1 are updated with these changes.
-
I like all these changes, this will end up making it more fun for the majority of players.
-
Yep, the changes are great, but the flamer is used a lot right now. I have a feeling this won't happen so much when ff is turned on.
-
Most of the changes seem good, but how exactly is the pounce range used? From middle of bbox to corner would be 45.9 and 56.3gu for goon and +goon respectively (I think the origin point is higher, making it easier to hit targets above, and harder to hit below?). So is the new value enough to hit diagonally?
-
Most of the changes seem good, but how exactly is the pounce range used? From middle of bbox to corner would be 45.9 and 56.3gu for goon and +goon respectively (I think the origin point is higher, making it easier to hit targets above, and harder to hit below?). So is the new value enough to hit diagonally?
your numbers are off by a bit, but yeah it's a lightle tight to hit targets near corners with a pounce.
-
Oh the +goon has smaller bbox :-x From middle of bbox to corner not edge: sqrt(26^2x2+27.5^2)= ~45.9 and sqrt(29^2x2+33^2)= ~52.6 (not 56.3).
-
you also got the viewheight wrong, and you're not taking into account the width of the attack.
-
Dunno those, but viewheight is above center which makes hitting below even more difficult, and the width would have to be huge if +goon also has 40 pounce range and wants to hit diagonally.
-
Any specific reason to reduce pounce range instead of pounce damage?
-
Any specific reason to reduce pounce range instead of pounce damage?
I suppose people were complaining more about being hit awkwardly than being hit hard.
-
Most of the changes seem good, but how exactly is the pounce range used? From middle of bbox to corner would be 45.9 and 56.3gu for goon and +goon respectively (I think the origin point is higher, making it easier to hit targets above, and harder to hit below?). So is the new value enough to hit diagonally?
I think it's more like sqrt(31^2 + 26^2 + 26^2) = 48.0936586256 for the dragoon and sqrt(39^2 + 29^2 + 29^2) = 56.5950527873 for the advanced goon (which currently gets +3 to its ranges over the base), in the absolute worst case scenario (below and beside your target). I didn't account for the vertical axis when choosing a range, but I did test this thoroughly and couldn't make myself not hit because of the bounding box, probably because you don't get just one chance to hit -- if you're pouncing into a target there's almost guaranteed to be a point at which you're in a good enough position to connect.
But yes, it's a little lame that it's theoretically impossible to hit a target under the right circumstances when you deserve to, but in practice you won't be able to distinguish between when this happens and when you really didn't deserve to hit. And these sorts of fly-by pounces are exactly what the change was supposed to make harder. It can still be changed if it turns out to be a problem.
Any specific reason to reduce pounce range instead of pounce damage?
Yes: partly because I don't want it to be impossible to one-hit kill unarmoured humans, but mostly because I don't believe there's any problem with the damage past stage one -- and I did agree that dodging pounces would often feel futile. Changing it this way makes skillful dodging more possible and skillful pouncing less trivial, which should be more interesting than just lowering the time it takes to die from pouncing.
-
So far these are interesting proposals; it will be interesting to see how they play out in practice.
I'm glad that human stamina is taking a step in the right direction, as are the various tweaks for dragoon.
-
ehh, I misinterpreted the meaning of goon pounce range. I thought it meant the distance a could pounce, not the range of the pounce attack to hit. Sorry.
-
Good grief.
Just played a couple games, and holy crap Flamer is Overpowered.
Humans were winning rounds in like 5 minutes just by getting s2, then having like 2 flammers and a lasgun rushing, game over man, game over.
-
flamer isn't overpowered. it just wasn't used as much. its like s1 humans vs s1 aliens, 3 rifles vs 3 dretches and rifles with good aim will win. flamer has a pretty high price so its pretty good where its at now. people didn't really utilize it and now that the flamer's reach has been elongated, it can turn into a useful weapon.
-
What irritates me most about the flamer is that there's really no way to dodge it.. it hits just about everything whether you aim for it or not =/
-
"The Noob Luci"
-
What irritates me most about the flamer is that there's really no way to dodge it.. it hits just about everything whether you aim for it or not =/
Truly. I intend to fire up fraps and post a comic video depicting fleeing tyrants accompanied by a comic, well fitting song. And I'll use only the flamer XD
-
I'm very satisfied with these - on paper. I wish I could playtest them. :/
Anyway, thank you!
-
Can we download somewhere newest qvm?
-
http://downloads.mercenariesguild.net/gpp/
-
I like the new changes a lot, it seems like to hit with goon pounce you have to aim in the center of the human which is good. Also, I think that the flamer is just a tad bit overpowered but with friendly fire on it might balance it out.
-
Yes, FF on will take care of the flamethrowers [smiles to self]
-
Yes, FF on will take care of the flamethrowers [smiles to self]
Not exactly sure how it would work, but maybe if there was a 300 hp damage limit against team kills that resets when you refuel at the arm. i say 300 hp damage so that the player cant go around and put the whole team on 20 hp, whilst rushing for example. Obviously the flamer would be a useless weapon to use in a team game if the quota didnt reset in these circumstances. Hopefully it would prevent flame grieving should FF be switched on.
Just an idea open to criticism.
EDIT: Hrmm i just thought it through a bit, and players could sit in base next to arm and kill teammates. Perhaps the quota resetting every 40-50 seconds is a more viable option.
-
:hovel: SAVE THE HOVEL :hovel:
-
:hovel: SAVE THE HOVEL :hovel:
-
The hovel is only useful on the server i play on because it can be spammed in sd around important structures, and to deny access to ramps.
-
I've found goon pounce to require near perfect aim now. Also humans can dodge all day, and sprint across the whole map now without ever slowing down. The overall effect of the new changes is that you need perfect aim, or near perfect aim to be effective. It's just not fun for me, so I've stopped playing, as basically the players with better aim but less overall skills are just dominating. This contrasts completely with Trem 1.1 where strategy was superior to aiming abilities, and was one of the main reason I liked trem. Of course the players who can aim well probably like the changes alot and maybe the point was to make a few players dominate more then they did in 1.1 without allowing successful resistance by smart players on aliens. I don't know, that's just a gameplay design decision by the devs, I can't really declare one mechanic better (rewarding smart play vs rewarding skilled play), but the new version is geared toward rewarding the skill of aim over all other considerations.
-
I agree a lot with phungus, the range nerfs especially in pounce are a bit much. You do have to be really close and have perfect aim to use anything effective. I think we should find a happy medium between what we are now and what we were in 1.1
-
Once I realized that starting a pounce mid-air cancels the previous one, I was able to pounce pretty well; but the cancel is really annoying. Its previous range made pouncing over-powered.
I'm really happy for the range nerfs, especially tyrant's. If anybody's open to convincing, I can try to show you how good aliens can be even with the new range in a 1vs1; you can find me on irc. Of course, many players are better than me. imo, Aliens' range is lenient enough.
Also humans can dodge all day, and sprint across the whole map now without ever slowing down
If you look at tremulous.h, you'll realize that sprinting takes as much stamina as it did with larmour in 1.1. It used to take even more stamina. When I play, I often use up about half of my stamina before I run into aliens, which leaves me with only two dodges left, if I still have at least half left, since "dodge" takes as much stamina as jumping.
-
The pounce being canceled may be what is causing me to think that (and I agree the old pounce was OP, but the new one just never seems to land, at all against a decent dodger). Basically I just always assumed my aim must be off. Regardless, while I can't put my finger on any specific change that is causing it, the current 1.2 gpp rewards the specific skill of aiming way more then 1.1 (or even previous gpps), where smart play seemed to have a more drastic effect, and perfect aim just wasn't as important. Like I said that's a design decision by the devs that's not necessarily a negative or a positive one, but a noteworthy change I personally dislike and want to point out.
-
I noticed the pounce cancel thing and that improved my pounce A ton. I still, however, think the chomp nerf is a ton. I would pounce people into corners and if I was even slightly off it was a complete miss. I do agree it was way to easy to hop through hall nabbing hs's in 1.1. This is just too much though, it seems to me the people who suck, suck even worse at aliens now because they miss twice as much. I think the aiming needs to be a little bit easier. Let me clarify on pounce range too, since I got a little sloppy description on that. I don't mean how far you can go, I still find I can get up to high places and go quite a distance, but its also a little easier to chase down running goons. So I like the distance better. I mean, when I'm pouncing to a human, a lot of times, I even collide with them and am mostly aimed at them(a little to the side) and the pounce misses and I just slide against them and off. That's bad, the whole idea of pounce is your giant goon body slamming them, it shouldn't have to do anything with biting, so it shouldn't require complete forward aim, as long as your body hits them thats all that should matter imo. Since thats what causes the damage/knockback. Besides if you hit them with you're side, it's way harder for you to land a following hit, so that alone is more than enough encouragement to be aimed as straight as possible when pouncing someone.
I just don't like pouncing into a flamer, to find that even though I slammed into them, nothing happened, and now I've got this small delay of movement as I'm landing where I'm getting roasted to death unable to defend myself, fall back, or damage them.
-
Besides the flamer, the pounce range is the only thing I'm still thinking about tweaking for the next phase. It does feel a little too short to me, though yesterday I saw someone still dominating with it like before, so I'm still not sure. But I'm leaning towards bumping it 48 (is 40 now, was 56 in GPP phase 2).
-
Can I ask something? I haven't tested it out yet fully, but do you have to full charge a pounce to do same damage, or is it like in 1.1 when a quick pounce still did the same damage you just didn't go as far.
-
Can I ask something? I haven't tested it out yet fully, but do you have to full charge a pounce to do same damage, or is it like in 1.1 when a quick pounce still did the same damage you just didn't go as far.
in 1.1 a quick pounce did less damage, and it still does.
-
Can I ask something? I haven't tested it out yet fully, but do you have to full charge a pounce to do same damage, or is it like in 1.1 when a quick pounce still did the same damage you just didn't go as far.
in 1.1 a quick pounce did less damage, and it still does.
Really, is it in the same ratio? I can't believe I never noticed that, but I almost always did full pounces in 1.1
It seemed 1.1 quick pounces were a lot more effective for me in 1.1 than 1.2 though.
-
in 1.1 adv goons charged pounce a lot faster than they do now (this only changed a couple months ago)
-
Besides the flamer, the pounce range is the only thing I'm still thinking about tweaking for the next phase. It does feel a little too short to me, though yesterday I saw someone still dominating with it like before, so I'm still not sure. But I'm leaning towards bumping it 48 (is 40 now, was 56 in GPP phase 2).
Could you keep the range the same, but get rid of the cancellation so that goons can still reflexively charge pounce when attacking? I think that would have the desired effect. Like I said old pounce was OP, and now that I know about the cancellation it's probably not the range that's the problem, it's the fact that pro alien players are trained to constantly recharge pounce, so they are getting the damage cancelled mid flight.
Edit:
Also I'd like to add I think dodge is crazy strong, and will end up making humans overpowered once people start realizing it's there. I still see very few players use it. This conclusion is based on my own experiences. In 1.1 I've played abut 2/3rds of my time on aliens, so I have much more experience and am just better with them. In 1.2 after experimenting with dodge I've found playing humans to be equally good, and for my case this shouldn't happen, I should still be much better with aliens. Granted this is only subjective, but I've found dodge to be too effective and too easy, and very few players use it yet, which I think are skewing the game balance results (though maybe people will never learn to use it, so it's not). Perhaps in the next gpp humans could have a notification (like how aliens see "press q to evolve") for dodge, so they see a "press v to dodge" notification somewhere, similar to aliens notifications. This would get more players to use it in this balancing stage so it's effects would be more visible.
-
If you don't want to allow mid-flight charge, make pressing it not cancel but just queue the next charge, so it starts as soon as it can, preventing wasted time.
-
If you don't want to allow mid-flight charge, make pressing it not cancel but just queue the next charge, so it starts as soon as it can, preventing wasted time.
Why not just allow charging mid-flight without cancelling the previous charge?
-
Because that would lower the effective repeat rate, altering balance.
-
I agree with phungus.
In 1.1 I could get 5-1 kill ratios on aliens on AA all the time. I was lucky, however, to manage 2:1 with humans. Thats playing against all noobs too.
In 1.2 I barely manage 2:1 on aliens at my best, and I get 2:1 sometimes 3:1 on humans and I don't play it all right yet(don't utilize dodge or forward bases well on H). Humans are so much OP now it's not even funny. Base wise, it seems if you don't have an excellent builder, one adv mara wins, so that could explain why the humans seem to lose equal to more actual matches. But kill-wise, it's not contest. I care about the balance for when scrimming comes around. So once standard bases pop up and everyone starts using them, I have a feeling scrims will no long be fun as humans will clearly over-dominate.
-
I think a big issue with these gameplay tests is that the majority of the testers (us, the community, the oldfags) are way too used to 1.1 to effectively play 1.2. A newcomer, yes, would suck, but after getting over the learning curve would begin his path to clan level gaming. Us, on the other hand, basically have to unlearn 1.1 just to start learning 1.2. It'll be a lot easier to build well and play well if we didn't have these old habits hardcoded into our systems. Not sure how games are going to go once 1.2 gets it's large influx of the dreaded Release: Day/Week/Month One Newbies, but I think that'll be the true test of balance (post final version, unfortunately [barring, obviously, the common 1.2.x bug fixes & minor patches]).
-
Once I realized that starting a pounce mid-air cancels the previous one, I was able to pounce pretty well; but the cancel is really annoying.
I don't know what you guys are talking about. The payload for a pounce is saved at the start of a pounce and is thereafter separate from the charge. So whatever your charge is when you leave the ground is what determines how much damage you do, regardless of whether you begin charging again in the air or not.
-
Once I realized that starting a pounce mid-air cancels the previous one, I was able to pounce pretty well; but the cancel is really annoying.
I don't know what you guys are talking about. The payload for a pounce is saved at the start of a pounce and is thereafter separate from the charge. So whatever your charge is when you leave the ground is what determines how much damage you do, regardless of whether you begin charging again in the air or not.
It seems to me if you start another charge in mid-air the damage from the previous charge is cancelled.
-
Once I realized that starting a pounce mid-air cancels the previous one, I was able to pounce pretty well; but the cancel is really annoying.
I don't know what you guys are talking about. The payload for a pounce is saved at the start of a pounce and is thereafter separate from the charge. So whatever your charge is when you leave the ground is what determines how much damage you do, regardless of whether you begin charging again in the air or not.
It seems to me if you start another charge in mid-air the damage from the previous charge is cancelled.
Strangely, I was unable to reproduce that when I last tried it. The code straightforwardly copies STAT_MISC.
For some reason, I was unable to do any damage when I pounced during one game, even to buildables. When I didn't recharge my pounce immediately after I pounced, I was able to deal damage. Perhaps the server was running a QVM with a bug in it.