Gameplay Preview Phase 7 Results

[NotYarou]

For the first time ever, I actually enjoyed playing GPP.
I really feel that these balance changes are excellent; GPP no longer feels as slow as it used to.
You've managed to capture the deathmatch style I loved in 1.1, while still keeping the greater focus on teamwork introduced in GPP.

Have you not played in a while, or were these relatively small changes enough to make the difference? Either way, <3

It's the latter, the locdamage/armor region tweak really made quite a bit of difference IMO.

[scrape]

I've been thinking about the flamer for the last month or so, and I think this addition could
improve its use:

Humans lack a DoT(damage over time) like aliens, If the flamer would add
"Burning" from contact with the flamer (same tick times and refresh) it could be
useful against fleeing aliens.

[jm82792]

I've been thinking about the flamer for the last month or so, and I think this addition could
improve its use:

Humans lack a DoT(damage over time) like aliens, If the flamer would add
"Burning" from contact with the flamer (same tick times and refresh) it could be
useful against fleeing aliens.

THis could be very nice if it's integrated properly.
The weapon would need a compromise, but it would make the flamer more unique thus theoretically more fun.

[swamp-cecil]

The flamer Needs, like mentionned before DoT (damage over time), so like 1damage per 100 msec. We can turn the current damage into DoT by:
Current Damage Divided By Current Fire rate = X
and X would be in msec the fire rate of doing one damage.
...And because im too lazy to open GPP-SRC and do it myself.

[Nux]

The flamer Needs, like mentionned before DoT (damage over time), so like 1damage per 100 msec. We can turn the current damage into DoT by:
Current Damage Divided By Current Fire rate = X
and X would be in msec the fire rate of doing one damage.
...And because im too lazy to open GPP-SRC and do it myself.

I pressume you mean dividing the current repeat time by the current damage per shot to get the repeat time for 1 damage DoT hits doing the same average damage over time.

Why do you want this burning effect to do 1 damage a hit at the same overall rate as with direct shots? Only the painsaw does more damage than it over time. That's much too effective.

If TF2 is anything to go by, it's better to have the a fairly weak but long acting DoT effect and a much stronger direct hit damage.

[Pete]

the flamer would add
"Burning" from contact with the flamer

Great idea! Then what if the granger spit could put out the fire?

[GeneralScott]

Then there would be way too many battle grangers. How about goon spit has a 50% chance of doing no damage and putting out the fire and a 50% of killing the TK target xD xD

[OhaiReapd]

Imo the shotgun was fine

I agree, if you ignore its cost (which is easy to do for such a cheap weapon). My opinion was that it was no better than the rifle (and that there was room to make it better without needing to change anything else).

The shotty was fine how it was, but if you want to buff my favorite weapon and give me more kills, that's fine with me. <3

[Rubuco]

And when will be tremulous 1.2 finally? Because with this phases test you are loosing people in the game all the time. (Anyway they are uselles because there will alwyas be a people which won't enjoy the game)

[UniqPhoeniX]

1.2 will be released When it's done (TM). Currently I think only some sounds (you can help with these if you know how to make sounds) and maybe some models are needed. And these phases are not useless because the gameplay is improving.

[Rubuco]

1.2 will be released When it's done (TM). Currently I think only some sounds (you can help with these if you know how to make sounds) and maybe some models are needed. And these phases are not useless because the gameplay is improving.

So diplomatic. And how can phases improving gameplay if you still are on the same things? And what will happen when there will be nobody to play your game? 

[Dracone]

Don't be stupid. This game would be buried and forgotten by the devs if they didn't have a plan (despite that it may be a minor/simple one) to bring new players in once 1.2 is officially done, given that at the moment it's in a state that's pretty much dead.

Doesn't matter anyways, you're wrong as fuck imo. The game's on a slow path to reviving (through GPP) as it is, despite that 1.2 is not finished yet.

[Pazuzu]

A lot of it depends on whether Stannum is still around to work his magic. Without a graphical revamp of some kind, the sad truth is that few people will notice the other changes.

[Prince_Andrei]

I think it's possible that the human/alien win-loss statistics are not being used correctly and that changes to create more balance are not creating a more balanced game. From scanning through individual player stats, it looks like the more experienced, better players are playing alien more often than human. Similarly, newer and overall worse players tend to play humans more often. There are clearly exceptions to this. However, as a general trend, this makes sense. If your background is with other FPSs, humans are simple to pick up, but learning to use a basilisk or the narrow goon chomp is a trick.

How does this impact the stats? If the trend I'm describing is true, aliens win more often because better players are playing aliens, not necessarily because aliens are more powerful. Rather than simply charting win-loss stats, you need to statistically control for the quality of the players in the games. You could use each player's average score per game as a proxy for their skill. It wouldn't be perfect, but it may help you better understand the game's balance. The technical way to see whether this trend is true is to use a statistical method called logistic regression to test how well the score-per-game of a team's players predicts wins and losses. Norf, if you can help me get the data in an easy to use format, I'd be happy to do this.

[Meisseli]

I think it's possible that the human/alien win-loss statistics are not being used correctly and that changes to create more balance are not creating a more balanced game. From scanning through individual player stats, it looks like the more experienced, better players are playing alien more often than human. Similarly, newer and overall worse players tend to play humans more often. There are clearly exceptions to this. However, as a general trend, this makes sense. If your background is with other FPSs, humans are simple to pick up, but learning to use a basilisk or the narrow goon chomp is a trick.

How does this impact the stats? If the trend I'm describing is true, aliens win more often because better players are playing aliens, not necessarily because aliens are more powerful. Rather than simply charting win-loss stats, you need to statistically control for the quality of the players in the games. You could use each player's average score per game as a proxy for their skill. It wouldn't be perfect, but it may help you better understand the game's balance. The technical way to see whether this trend is true is to use a statistical method called logistic regression to test how well the score-per-game of a team's players predicts wins and losses. Norf, if you can help me get the data in an easy to use format, I'd be happy to do this.

I sincerely doubt that.

[Dracone]

The reason for the alien/human time imbalances are because it's much easier to do a large amount of damage to the enemies by yourself if you're on the human team than on the alien team. You pretty much can't win games on aliens without your team helping you now, unless the humans have a pathetic base and can't defend it as well.

Humans is a different story. I can confidently assert that if you don't have at least a couple skilled players on your alien team, I will annihilate your team and your base once I get S3, or even S2, by myself.

The point is that its each team's capabilities to attack per individual player that are causing the times to appear that way, but I guess that so long as the aliens are actually still winning more games this doesn't matter.

[Prince_Andrei]

Humans is a different story. I can confidently assert that if you don't have at least a couple skilled players on your alien team, I will annihilate your team and your base once I get S3, or even S2, by myself.

Great point. This is probably part of the reason for the time imbalance, but it reinforces the point I was making! The game results appear more even because more skilled players are playing aliens. This is all conjecture, though, since the stats page doesn't show Dracone's a-game and h-game counts--just the total times. I know I play aliens much more often than humans, if only because of all of the human-only players. The only way to know for sure if this causes the game balance to appear different than it really is is to look at the data.

[sirshiz]

I think it's possible that the human/alien win-loss statistics are not being used correctly and that changes to create more balance are not creating a more balanced game. From scanning through individual player stats, it looks like the more experienced, better players are playing alien more often than human. Similarly, newer and overall worse players tend to play humans more often. There are clearly exceptions to this. However, as a general trend, this makes sense. If your background is with other FPSs, humans are simple to pick up, but learning to use a basilisk or the narrow goon chomp is a trick.

How does this impact the stats? If the trend I'm describing is true, aliens win more often because better players are playing aliens, not necessarily because aliens are more powerful. Rather than simply charting win-loss stats, you need to statistically control for the quality of the players in the games. You could use each player's average score per game as a proxy for their skill. It wouldn't be perfect, but it may help you better understand the game's balance. The technical way to see whether this trend is true is to use a statistical method called logistic regression to test how well the score-per-game of a team's players predicts wins and losses. Norf, if you can help me get the data in an easy to use format, I'd be happy to do this.

I sincerely doubt that.

Meisseli, usually you are more helpful in your critiques. Exactly what you doubt and why would be valuable information. Thanks!

[khalsa]

I doubt player score is a valid metric to weight against, and I don't think you have enough data to even claim a valid result on a logit model, but I'm no mathologist.

[Plague Bringer]

The "newbs", as it were, flocking to humans over aliens is a trend that most people I know recognized a few years ago in 1.1. It's given that newer players will (most of the time) chose humans over aliens, creating what can become (but perhaps, is not inherently,) a relatively large skill imbalance. While public statistics are a fantastic way to address balance, statistics are biased by the interviewees, who are, in this case, the general public. I think that some stats based on competitive clan-play, which I've heard is starting to move over to 1.2, would be both interesting and helpful, as I'm sure Norf & Co. want to cater not only to the masses, but also to the more exclusive and hard core clan-scene. After all, if not for some of the oldest and clan-oriented players, I think Tremulous would have been abandoned already.

[c4]

However, there are people like ArxInsantium who have only played for a couple of weeks, but almost exclusively play aliens

[Meisseli]

Meisseli, usually you are more helpful in your critiques. Exactly what you doubt and why would be valuable information. Thanks!

I sincerely doubt all he wrote. There have been countless people claiming that newbies join humans and vice versa. Then they have a grand theory on how this affects balance. There hasn't been any significant evidence to prove this, mostly the differences between the gametimes of "experienced players" in the two teams are a couple hours, which is well within the margin so one really can't say experienced players play team X more. Then people like him like to make a grand theory to support their view that either team is underpowered.

I also doubt he can make a valid measurement of player's skills with any approach. For example, some people can be considered skilled but they don't necessarily tip the balance significantly in anyone's favour since they don't bother to attack bases.

What comes to clangames, humans have always been the more powerful team. It's a lot easier to work as a team with them, and small maps like UTCS, Tremor and ATCS are biased towards humans when both teams have experienced players. For example, our individual round statistics:

Code: [Select]

15:44 <river-tam> Map            aW  aL  aD      hW  hL  hD
15:44 <river-tam> arachnid2       6   0   0       6   0   0
15:44 <river-tam> atcs           39  17   1      48   6   2
15:44 <river-tam> atcszalpha      2   0   0       2   0   0
15:44 <river-tam> fort5           0   1   0       1   0   0
15:44 <river-tam> gloom2          1   0   0       1   0   0
15:44 <river-tam> karith          5   0   0       5   0   0
15:44 <river-tam> niveus         20   1   0      22   1   0
15:44 <river-tam> orion           2   0   0       2   0   0
15:44 <river-tam> sectorb17       1   0   0       0   0   0
15:44 <river-tam> thanatos        1   0   0       1   0   0
15:44 <river-tam> tremor         21  13   0      32   2   0
15:44 <river-tam> uncreation      1   0   0       2   0   0
15:44 <river-tam> utcs           20   6   0      23   2   0
15:44 <river-tam> veddak          2   0   0       2   0   0
15:44 <river-tam> Total         121  38   1     147  11   2

[Nux]

I don't see anything wrong with Prince_Andrei trying to test his hypothesis. The very fact he wants to test it makes him stand apart from all the speculation shouters. If he's anything like the stats geek he says he is, he will state his assumptions and give his method of analysis and then we can decide whether it's fair or not.

The current balance stats are pretty limited in thier explaining power; far too much information is pooled together in my opinion. I believe they can be greatly improved upon (timbo himself said how they 'aren't very scientific') and if ever I get the data from timbo I'd be happy to try my hand at some statistical analysis on it too.

If anyones worried about how to define 'a skilled player' or other vague concepts like it, this has always been a necessary part of analysis. The best you can hope to do is pick a standard that enough people agree upon and see what the numbers tell you under that assumption.

[Prince_Andrei]

Assuming I can get the data, I will make all the assumptions and limitations clear, like a good mathologist. Regarding "skill"... For this purpose, we don't really care about true skill. We care about game influence, which is related but different. And we don't even need to quantify it. You can statistically control for multiple things that overlap with game influence, like average score per game, kill/death ratio, etc. If people who score a lot and kill much more often than die tend to play one side versus the other, that would be a nifty thing to know. I'm not trying to convince anyone of a grand theory, and it's very possible that I'd find the exact opposite of what I expect. At a bare minimum, if I get the data you'll see some very cool looking charts.

[RAKninja-Decepticon]

Assuming I can get the data, I will make all the assumptions and limitations clear, like a good mathologist. Regarding "skill"... For this purpose, we don't really care about true skill. We care about game influence, which is related but different. And we don't even need to quantify it. You can statistically control for multiple things that overlap with game influence, like average score per game, kill/death ratio, etc. If people who score a lot and kill much more often than die tend to play one side versus the other, that would be a nifty thing to know. I'm not trying to convince anyone of a grand theory, and it's very possible that I'd find the exact opposite of what I expect. At a bare minimum, if I get the data you'll see some very cool looking charts.

some numbers may be contrary to what you expect.  i usually score in the mids and lows, but sometimes, i'm in the top three - all points wise.

if you'd seen my k/d chart, you'd likely not believe me. i'd like to see a shiney chart of each player's k/d plotted against score, per race.

[gimhael]

If anyones worried about how to define 'a skilled player' or other vague concepts like it, this has always been a necessary part of analysis. The best you can hope to do is pick a standard that enough people agree upon and see what the numbers tell you under that assumption.

If you really want to make a player ranking, you should take a look at Microsoft's Trueskill(tm) algorithm. I think it would fit Tremulous quite well, the required inputs are only the members of the two (or more) teams and the performance of each team (i.e. win/draw/loss). As it ignores 'number of kills', 'score', etc. it should really only measure how much each player contributes to winning the game (resp. how much more likely it is for a team to win if this player is part of the team), even if he for example only builds.

[F50]

the required inputs are only the members of the two (or more) teams and the performance of each team (i.e. win/draw/loss).

No. Bad. Bad idea. The last thing Tremulous needs is a rating system that goes by the amount of games you won. Instant stackage.

[UniqPhoeniX]

The rank doesn't have to be published with names, it can still be used to see how much players that significantly prefer 1 team affect win/loss ratio.

[gimhael]

the required inputs are only the members of the two (or more) teams and the performance of each team (i.e. win/draw/loss).

No. Bad. Bad idea. The last thing Tremulous needs is a rating system that goes by the amount of games you won. Instant stackage.

Trueskill is based on the ELO system used in chess ranking etc, the system awards more points if the expected result (based on the estimated player skills) and the observed result differ.

So by heavy stacking you'll be in a game where you can win nothing, but maybe lose a lot.

[F50]

the required inputs are only the members of the two (or more) teams and the performance of each team (i.e. win/draw/loss).

No. Bad. Bad idea. The last thing Tremulous needs is a rating system that goes by the amount of games you won. Instant stackage.

Trueskill is based on the ELO system used in chess ranking etc, the system awards more points if the expected result (based on the estimated player skills) and the observed result differ.

So by heavy stacking you'll be in a game where you can win nothing, but maybe lose a lot.

Fair enough. Unfortunately, this sort of compensation only works when playing against players who's skill you can reliably calculate. There are several good players who like to remain anonymous. However, that actually sounds half-decent, if in fact some means of rating players is deemed necessary.