The amount of ammo for lower ammunition weapons is capped at 25. In Q3 it was 200 and it looked ridiculous.

consistent spread when firing the shotgun

Most obvious answer is you can strafe jump up stairs

Things Q3 did better than QL:

- 15 minute duels
- 15 seconds Weapon respawn
- 900 ups Rockets
- better Spawn logic
- gtv
- cg_nomip 0
- better lg (in osp at least)
- better netcode

Things QL does better than Q3:

- more balanced weapon damage
- 250 fps (although, it exists in cpma now)

Q3 speed rockets would make the game all shaft. It's not like the old days where 30% LG was considered good. And I actually like 10min duels. Might decrease chances of a comeback but it's better spectator wise.

Then you simply change the lg rate of fire/damage.

And 10 min duels is no good, because comebacks are much harder. I remember from myself, losing games, where I was in the lead for like 8-9 Minutes. Thats simply not possible anymore.

I do think the spectator would enjoy more epic comebacks, would he not? :)

How about TDM and QLs weapon damage? Is it good there too?

TDM in QL is fine.

Comebacks have almost nothing to do with the timelimit. What is responsible for them in Q3 is the spawn system. 50% furthest from enemy works in CPM because the speed there still allows you to force engagements off spawn. That system will not produce comebacks even in 100 tl in slow games like QL.

You only post that so that you can shit on non CPMA game

That and to remind idiots it's called CPM.

Its Q3:CPMA. Never forget your roots.

I had to check the date before replying, GTV really?

Actually, if the tournament is organised just like this Qcon. Where you have the option to watch all the Players you like and all their matches throughout the groupstage. (Just like how it was in GTV times) Then we dont need GTV anymore.

But it wasnt always like this, wasnt it?

You forgot:

-Announcer that sounds badass and not gay as fuck.
-Being able to play any map and mod you want
-Gibs in all servers
-Cool colourful nicknames
-Being able to spam the chat for fun with the epic chat noise
-Overall a lot more customization
-r_picmip "5" with r_overBrightBits "1" and r_mapoverBrightBits "5" made all maps look nice and clean and not completely retardedly bright as in half of QL maps so I have to have custom r_mapOverBrightCap for EACH FUCKING MAP.
-and finally and most importantly: the game was FASTER in OSP. When they "fixed" the fps movement dependant thing the game became slow as fuck.

You're alright
Don't go to ztn tomorrow

250fps also existed in ra3

Lag compensation

Edit: Well, okay, it depends whether you're considering mods or not.

X E R P

CPMA only

It's probably better to compare a specific mod to QL.

I think QL uses xerp as well, it's just not a available for users as a variable.

I believe QL is more or less the same as Q3 but with backward reconciliation up to 80ms (you know the thing where it keeps a history of player positions on the server and rewinds time when you fire hitscan), similar to the Q3 mod "unlagged".

I'd have to look through some old forum posts but I want to say that this is incorrect (I could totally be wrong here) as if I remember correctly arQon really didn't like unlagged, and QL and CPMA had close ties.

I'm happy to be proved wrong, but I am quite sure, and it is something I have paid attention to over the years.

I also remember that arQon didn't like unlagged, but he was working for id, right? He'd have to make what they wanted, and I'm not sure I ever read any formal statement about the extent of his involvement with it.

QL tries to interpolate between two snapshots (rather than extrapolate beyond the latest one) unless you set timenudge negative. This is the same as Q3 behaviour, and that interpolation gives you the most accurate player positions at the expensive of highest lag. The interpolation delay is considered an acceptable default since BR is being used to compensate for it. This is a common method in a lot of games, which have the "favour the shooter" philosophy.

If you do set timenudge negative, QL will do what looks like pretty much the same job of extrapolation that Q3 does, and players will look as if they are being drawn at lower frame rate, which is exactly what xerp in CPMA is supposed to fix---CPMA does a better job of extrapolation.

The unlagging effect has a clear cut off at 80ms in QL, which would not be necessary if it were only client side extrapolation.

Besides, QL has the old "shot behind corner by high pinger" effect that is the hallmark of unlagged netcode.

Thanks for the explanation. Maybe Syncerror could provide some clarity on this issue, but I believe he no longer visits these forums :/

Arqon was hired to do ql's netcode, but never finished it due to illness. Ql doesn't have xerp, it's built-in laghax and timenudge. Taichou is right that the lag compensation doesn't work beyond 80ms.

 

Thinking back to pre-9/11 QL off the top of my head:

- The content delivery / client update / whatever was all centralized on quakelive.com and easily accessible with one click, pretty much "click and play", to the point where QL was graced by players who would have otherwise never even got past the second step of setting up a functional Q3 client with RA3. Even with the piss-poor handling of the project's marketing and management, this feature alone gave the game a vast playerbase. Sure, most of them remained in CA but a decent % eventually ventured out into other modes.

- Pub CTF servers back then, with a healthy dose of Team Arena maps in the pool, was just pure bliss. Those TA runes struck the absolute perfect balance between "quake" and "classes", which is the approach QC should have used if they still had a brain between them. Oh yeah, Kamikaze item was jizz to my eyes and ears.

- The "server difficulty" ratings while flawed, were a good idea, especially for newcomers.

- Pub server coverage and stability were pretty good in the beginning from what I remember, before they started to take locations offline one by one.

- 80 dmg rail, 10% increase in rocket speed, mg dmg to 4 and maybe even lg dmg to 6 were all good changes imo. There was a certain small balance shift needed towards the beginner end of the spectrum and I think these "minor" changes were all that were needed as far as modifying game mechanics went. 10 min duels on the other hand maybe should have only stayed at pub server level, and not in the competitive scene.

- Other modern amenities such as friendlist, chat, training area, clans (all with room for improvement) were all great additions with potential and increased the social aspect of the game, especially for the aforementioned newcomers unversed in the ways of IRC.

I know there's more elements to QL which made 2009-2010 a golden era of pure hope and prosperity with a bright future. But those golden days are OVER bitches, driven into the ground by procrastination and incompetence. Better stop reminiscing now.

jizz to my eyes and ears

:D

Oh how the times have been fucked.

Edited by ShadyAK at 14:19 GMT, 1 February 2013 - 133514 Hits
Hall of Fame: Hilarious (added to hof by ShadyVoltaire)
64% [+]

^ clicked on your link, saw this, got sad :c

2009-2010 QL was also a shitfest of people complaining it's worse than Q3 with shit netcode, then the OSP CTF plappers were complaining that how better netcode makes the game too easy, poormode pros "lol@no CPM movement"(before they added the horrible PQL) and "lol@shit homosex imitation of CPM movement" (after they did), then when they added the pro/premium account to "start your own servers" and with content that could be reached for free on Q3, i've had clanmembers literally rage at the audacity and just quit the game for good (never to be seen again, RiP cLimber et al). Every change (shaft being taken down to 6/6/6, addition of "new features" which were copy+pasted from mods 10+ years old at the time, "new maps" from 3wave and osp RA map packs, the ineptitude of saying "NEW UPDATE GUYS, WE HAVE SPENT 3 MONTHS AND 800 BILLABLE MAN-HOURS, AND NOW YOU CAN CUSTOMIZE THE COLOUR OF YOUR SKELETON'S NUTSACK WITH CG_MODELNUTSACKCOLOR X/BRIGHT". Every change seemed laughable, every idea of "porgress"(deliberate spelling c:) was with hope for the better and how we'd see it as a laughable blip in time and they will finally get things right, and it will be a better game. A bit like most of the shit in September 2017 (i played QC for about a week, but driven by experience thought "is already shit dead game, I shall uninstall now)... turned out to be a right call.

Perhaps the real difference was the friends we made along the way c:

"Things Q3/QL Do Better than Quake Champions"


Please continue on a separate sheet of paper :o)

All the maps in one client, you don't need to install 27 mods and map packs before being able to play, the weapon balancing is more adequate to modern hardware, (except the infamous 6 damage lg with 7 damage knockback, should be the opposite because current version punishes good dodgers), no satanic symbolism wich makes it better as an international esport platform, many ql maps are a tiny bit more balanced like ql's version of DM13, the settings menu is a little more... existent.



What we REALLY need is a Q3 mod based on the cpm client, but with all the gameplay settings and maps of quake live. And the spiterbots bots ai.

That would mean user spawned servers all around the world, xerp netcode (but lock it to sane settings that can't be tampered with too much), that slick slick engine instead of quake live's "awesomium" frame parasite, etc. etc. etc.

I would PAY for such a mod. And so would some others i guess.

If i won 1 million at the lottery, i'd spend some thousands dollars to pay one/two coders to make such a mod for me and all of you. That would be the dream.

yeehaaa cowboy

What about the ioQuake3 client? Any experience with that?

I played open arena maybe 10 years ago, from what i remember it even lacked most features q3 had

no dreamcast version of quake live :<

Is the PS2 port of Q3A any different than the Dreamcast one?

Quake 3 Revolution was so bad. Still played the hell out of it due to no choice for a time. A dark time. =(

Yeah.. It was a total PAIN in the ass playing or trying to do with the Dualshock2. I just get to Klesk.

Anyway, Quake may suck on the PS2. But Black was an other history. Holy fuck that game is like an xbox 360 shooter... On a PS2!!

no dreamcast version of quake live :<

lags more
sucks more
Sibs more

Having players.

HARD MODE: Things that QL did better than Q3 that wasn't dropped or fucked to bits in QC.

dude, where are youuuuu

miss ya too, buddy.

Steam integration

Well q3 is on steam too, and since years

Ah ok

Well, being a game on steam and having "steam integration" are not really the same thing.

Steam integration would mean the game accesses steam features, like being able to see your steam friend list within the game.

The steam workshop stuff would fall under steam integration.

I don't actually know whether the steam version of Q3 has been modified to have any steam integration.

Ah ok #2

In that case, no.
Q3A doesn't have any steam integration, no workshop as well.
In any case, you can still play the game free downloading the demo and play on demo servers, i think massacre 1.16 demo servers are still up and ppl are playing too.

Added results to the initial post

everything apart from the LG nerf

Friend list (through steam integration)

Things i like in QL more than in Q3.
- Movement, it feels smother than VQ3 (just a bit)
- Training maps, they are easy and have youtube tutorial videos for newcomers to learn basic things like bhop or stjumps
- Easy installing maps and simple mods via Workshop, still no complex modifications like QRally.
- Server Browser, have a tag system but i think is way slower than the Q3 SB


Things that Q3 don't have and QL yes
- Friends list, like Xfire/IRC in Q3 2005 days anyway
- Installing a plugin and just click and play (2009)
- Player Profiles and Clan Pages (2009)
- Achievements, the old ones just looked way nicer than the steam ones.

Do you know any hard facts why the movement feels smoother?

is just an opinion, for me, QL is little more smother than VQ3, i guess maybe better mouse integration or different values in mouse commands. So no, i don't know any hard facts...

:(

Because of airsteps.

(this is what makes some jumps easier, or even possible, and allows normal strafing on stairs, esp. when combined with autohop and that quarter-doublejump I forgot correct name for)

CPMA dealt with stopping on stairs differently (pulling you down to step below instead of stepping up), QL is waaay more natural and fun to jump around.

I liked a lot of stuff in Q3 more than in QL:

In Q3 the netcode favoured the people with good hardware/connections and still you had to aim ahead of the enemy according to your ping.
In QL lpb's rails go directly through the enemy and hpb's rails go around corners, players with good connections have to aim behind the enemy, the netcode favours people that have a crappy upload which makes them unhitable due to warp, while they can aim normally and therefore the netcode is designed for 3rd world warpers.

The LG felt useful in Q3, in QL it is just shit, especially due to the low damage and high knockback in combination with the netcode problem described above.
The RL is fully retarded in QL.

Q3's mod support and the community driven development were heaven, compare that with QL's 9/11-scrubdate.

I hate the fact that QL tracked my statistics and made them public, while elo just made matchmaking worse and destroyed the fun.

I am really not trying to bash on QL, it is just I have a hard time finding something that QL did better than Q3.
Even from your list I disagree on all, except:
- constant shotgun spread (unless the bullets get stopped by air)
- stairstrafjumping

I hate the fact that QL tracked my statistics and made them public, while elo just made matchmaking worse and destroyed the fun.


elo is fun and made things super competitive

Not at all. Your opinion heavily differs from mine.

First part of my statement is pure subjective opinion, having the data for myself is useful to monitor progress, but it ain't nobody else's business what weapons I use or which servers I play on.
Where the external data tracking objectively went full retard, was to display on a player's page what server's IP he is currently playing on, this information allowed the camelface "vincent_vega" to DDOS the sunday cups.

Competitive Quake died years before elo-tracking even began, you are always top 20 when only 17 pro players exist.
Elo was designed as a virtual win probability scale for pro chess, not for casual online gaming under different connection/hardware-setups nor team modes.
But the retards misinterpreted it as some kind of skill measurement and started to get butthurt about it.
How many of those basement beasts wanking to a value they didn't even understand went to LANs in order to win big price money?

I clearly remember all the fun I didn't have with those elo-obsessed morons.
Tierslumming in duel to farm elo, server owners aborting games to prevent elo-loss, winjoining, 5000 shuffles before every map of ctf, deliberate vpn-warping and the eternal whine.
Yeah, I totally need elo for matchmaking, because I can't remember the average performance of the last 17 players, LOL.

I also recall my favourite match of ctf.
Several shuffles bored me to death, then the game finally starts and within minutes it is clear we will 8-0 them easily, so they start to quit.
They were 2 players less and some captures behind, I felt sorry for them and switched teams.
In Q3 without elo nobody would have given the slightest fuck no matter how the game would end, but this is what happened instead:
We equalised the scores and finally won, just as I thought "Wow, nice comeback, I am surpised we did that, what a good game!", the team that I left while in the lead and then lost in the end starts to call me out for quitting and winjoining because they lost some elo.
Winjoining a loosing team? How?
I had several such games, which made me realise that those idiots got obsessed with a virtual value, they don't even understand and which they cared more about, than about having fun while playing the game.
That was the death of casual fun Quake and the birth of crazy tryhard Quake of wannabe competitive players who ironically prefer 8-0 blowouts.

What disproves you best, is how we all witnessed the player "komap".
He systematically winjoined himself up to the #1 spot of the elo-rankings in 3 different game modes.
Have fun explainging how this systematic winjoining helped
a) the fun for the involved players,
b) the competitive aspect of the game,
c) the matchmaking purpose of elo
or
d) the conclusiveness of elo, considering the #1 ranked player in 3 different game modes was an unskilled shit noob.

I really laughed my underwear brown, as qlranks shut down and those fanatics started to whine, because it got painfully obvious how worthless those values were they worked so hard for and how they didn't even play the game for fun anymore, but instead just to farm elo in order to boost their nolifes egos.
But those lunatics quickly found a new purpose in life, by implementing qlstats as their new graven image, a system even more flawed than the previous.

Anyway, there is not really a point in further discussing this dead horse (arguments aswell the game itself).

"But the retards misinterpreted it as some kind of skill measurement and started to get butthurt about it."

it is a skill measurement, a 2100 duel player will beat a 1400 duel player the majority of the time etc

"What disproves you best, is how we all witnessed the player "komap"."

wat, i'm not vouching for someone who used a glitch to get lots of elo

That wasn't a glitch, it is in there by design.
You don't need to vouch for "komap", he simply disproved that it is a skill measurement, which you are unable to understand.

I've never played ffa or ctf so wasn't really aware of win joining, at least in CA it's not something you can do as you have to play for more than half of the game for it to count.

This virtual example someone once brought up (I forgot who deserves the credit for this):

There are two players that were born and live on an oil platform, they have internet to update the stats tracking, but their ping to every server is so bad, that they only play each other on LAN.
They are on an equal level, so each of them wins/losses 50% of the games in an alternating manner.
They practise really hard, day in and day out, after decades of playing 24/7, they both reach a level far beyond the rest of humankind:
They bridge2rail forward, backward and sideways as warmup. Their flick rails hit pixels, that don't even exist on 8k-displays. Their timing breaks the spacetime-continuum and armors don't spawn in 25[s] intervalls anymore, armors spawn instantly whenever one of them is nearby.
...
If elo measures skill, then why it doesn't factor in their movement, aim, timing, ..., and why they will be stuck forever at the same initial elo value they started with, although they both are better than everyone else? Please explain.

I hope you will one day comprehend this:
https://en.wikipedia.org/wiki/Elo_rating_system#Theory

the system doesn't work for people on oil rigs isolated from the playerbase, I agree

i'm not saying it's perfect but you can't deny that A MAJORITY OF THE TIME a 2200 duel player would beat an 1800 and an 1800 player would beat a 1300

again, i'm not saying it's perfect but I think it's shown itself to be a good system in terms of measuring skill in quake.

The oil platform example is just an abstract scenario, that explains how elo measures a relative virtual win probability value instead of actual skill; but it applies very well to isolated communities, which explains why you can't compare a specific elo value xyz across Europe, North America, Australia, Africa and so on.

Not really, the 2200 elo player from Africa will lose to the 1800 elo player from Europe in a majority of the time, see above.

But since you keep repeating this "elo = skill measurement" crap again and again like you are unable to understand it, let me show you some other examples where the algorithm for 1v1 tournament chess fails at 1v1 casual online Quake:

Unique person vs. accounts?
The 2200 elo player randomly meets the 1800 elo smurf account of Bulat, he loses.

Chess board vs. Quake maps?
The 2200 elo player tries a brand new map for the very first time, he loses.

Chess Board + clock vs. PC hardware + internet?
The 2200 elo player forgot to do his homework, so his mother pulls the internet plug, he loses.

Competitive vs. casual?
The 2200 elo player's telephone rings during the match, it is his girlfriend that talks 30 minutes without breathing, he loses.

But let us be real, you will never get it, you will forever be stuck at applying a competitive metric to matches in which your opponent went afk in order to prepare tea in the kitchen, only because you misinterpret the metric and like what this misinterpretation tells you.
This is a waste of time like applying comon sense on conspiracy theorists.

In what way am I misinterpreting the metric? my interpretation of it is that it's an imperfect ranking system when applied to Quake but it's still an effective enough one by definition as the majority of the time a higher ranked player will beat a lower ranked player, a statement most Quake players would agree upon as I'm sure pure match statistics would show.

Some of your examples of criticism could be applied to any realistically implementable system ranking online games.

But let us be real, you will never get that I'm agreeing elo in Quake is imperfect and you'll continue to point out imperfections that haven't in reality proven themselves drastic enough to have meaningfully impacted the effectiveness of the system, nor stopped incentivizing people from wanting to win games to gain more elo points nor stopped elo being an effective predictor of who will win a duel..

Ehm, "komap"?
But I like how your vocabulary switched from "skill" to "rank", that is a step in the right direction.

Yes indeed, no matter if Elo, Glicko, TrueSkill, URS, Chessmetrics or whatever, they basically have the same problems from
- being designed for tournament chess and not casual online games (except TrueSkill, afaik it was designed for online games)
- the behaviour of the underlying mathematical formulas (especially at the very first matches where the calculation intervall/chronological order is a huge factor)

imperfections that haven't in reality proven themselves drastic enough to have meaningfully impacted the effectiveness of the system

Ehm, "komap" again? At the end they had to manually remove him from the #1 spots of the 3 gametypes. So in reality the system's effectiveness got meaningfully impacted by drastic assfuck.

nor stopped incentivizing people from wanting to win games to gain more elo points

Ironically, it prevented me from even playing games and made me uninstall. Suddenly I couldn't join games on duel servers, the bot kept putting me into spectator and demanded me to first register at qlstats. At that timepoint I was already so fed up by id's incompetence, the elo-farming tierslummers playing their hardcore +back rl-only wifi-warping and broken external stats tracking, that instead I rather just uninstalled:
https://steamcommunity.com/id/smackalodon/

Don't mean to be pedantic but quake rating is: https://en.wikipedia.org/wiki/Glicko_rating_system

GB is right in the general sense, it's a good metric for generic player "skill" comparisons. In most cases, with a high enough resolution of "fair" games (e.g. 1000), rating 2000 > 1900 > 1800 > 1700 > (...)

I know, ... don't mean to be pedantic but: Which Quake, which Glicko and with what parameters?

I am really amused how you try to sell elo as some sort of skill measurement and in order to do so, you must list a couple of unrealistic restrictions.

Anyway, this elo discussion is as boring and dead as QC and Esr, see you in a year for next Quakecon, if there is any. >:)

You do realise a lot of games have the concept of a "skill" parameter which they use for matchmaking / team balancing?

Just because they hide the "data" or "number" away from the player doesn't mean it isn't there. Rating systems are used in lots of games, not just Quake.. and I highly doubt they're all civilised enough to consider "every" aspect of a player, i.e. movement, etc etc. It's purely a generalisation and estimate

https://en.wikipedia.org/wiki/Whataboutism
I know, ... basically every competitive game I am aware of, uses such algorithms for matchmaking, but I can't recall a single one of them that falsely calls the metric "skill".
So I am really wondering what the aim of your Whataboutism is; did you find a random game the falsely calls the metric "skill" or you just try to evade the circumstance that you can't answer my questions?
Or, ... do you really want to talk about the absurd ranking results after the 10 placement matches of CS:GO's Wingman? Really?

Just because they hide the "data" or "number" away from the player doesn't mean it isn't there.

The reason developers completely hide the values or overlay them with ranks/tiers is:
Prevent the kids from getting obsessed about a metric they don't even understand, as seen in Quake.

I highly doubt they're all civilised enough to consider "every" aspect of a player, i.e. movement, etc etc. It's purely a generalisation and estimate

Yes indeed, because those algorithms don't measure skill.

I'll try to articulate this concept as simply as I can, as there appears to be some difficulty comprehending the underlying concept.

Consider a fixed-sized regional data-set: EU CA players
Consider a predefined minimum number of games: 500
Consider ignoring outliers (i.e. KOMAP)

On average, someone with 2000 "points" is better than someone with 1700 "points"

From your earlier point, yes, of course a 2000 player can lose to a 1700 player in duel, or on a map they don't know or whatever, but talking in the general case this logic doesn't apply. The same shit happens when tier1 teams lose to tier2 teams and other upsets in tournaments.

You can more or less "see" that the glicko calculation is psuedoaccurate as the top20 players are actually top20 players.

This ain't simple, lolca (aswell as lolctf) is the worst example you can pick.
I don't play it, but I remember players complaining that qlstats underrated frags and final win, while overrating damage delivered; combined with a wonky damage delivered calculation (I have 1 hp, you rail me, your damage delivered counts 80 hp).
I also remember Yakumo mentioning T1hol as example for stats focused gameplay in lolca: +back and let your team die, just to clean off afterwards with high damage delivered.
Since I don't play lolca, I am the worst person to ask, but what I can tell you for sure, is how damn stupid the algorithm in lolctf underrated someone who caps 8 times and makes his team win single handedly, compared to overrating someone who deals a lot of damage due to useless mid ffa.
(Okay okay, I know that I just Whataboutismed. :)
As you can see, what you call "better than" depends on the algorithm's parameters, which rate specific game statistics, you can even write an algorithm that calculates glicko based on the number of different colours in a player's nickname.
The problem with team modes is even more fundamental, because the algorithms were designed for 1v1 chess.
So the approach was to take the team's overall average rating, for team balancing aswell as calculating the expected outcome that the algorithm uses.
So
team A (5000 + 1000 + 1000 + 1000 = 8000)
and
team B (2000 + 2000 + 2000 + 2000 = 8000)
have both a sum of 8000, result in the same average of 2000 and therefore will be considered as equally likely to win, lol.

I'll try to articulate this concept as simply as I can, as there appears to be some difficulty comprehending the underlying concept.

Consider a fixed-sized regional data-set: EU CA players
Consider a predefined minimum number of games: 500
Consider ignoring outliers (i.e. KOMAP)

On average, someone with 2000 "points" is better than someone with 1700 "points"

I think I now understand your underlying concept:
Ignoring all undisputable arguments that disprove a thesis, makes the thesis correct.
Nice try, just I am not down with that: Komap, was he the best at 3 gametypes? No? Then it ain't a skill measurement. The end.

So ignore casual team modes like lolca and lolctf and let us focus on your last statement.
(Shit like hardware, ping, casual, fake accounts, unexpected upsets, isolated communities etc. aside to establish your unrealistic concept.)
Yes, for 1v1 the algorithm delivers very accurate estimates of outcomes, it still can't measure skill (see Agent's aim, player movement speed or whatever), but the relative virtual win probabilities it generates are an excellent indicator of who you should bet your money on.
Even when you tryhard more than you just did in order to extrapolate skill from probabilities, it still ain't a skill measurment, it doesn't have an absolute scale or an unit, it is a relative virtual comparison of likelihoods:
Your wife might be a 17 in your circle of friends, but compared to my wife, your darling is just a 5. (See, no absolute scale like 5/10 and no unit like [errections/day].)
Or, when everybody would be as good as Rapha while you start the algorithm, everybody is actually top skill but gets rated as medium skill.
It doesn't measure skill, deal with it.










You are still here? Just stop playing and spectate your glicko decrease. You wonder how that decay in "skill" got determined? It got pulled out of a juicy ass.










Omfg, you are still here? Then start warping and collect your free +200 in "skill".

Woohoo!
XD
HAHA, BROHEME

You studied logic, you parroted everything therein, and you learned nothing.

Glicko, in fact, by measuring probabilities, measures average performance level at attaining the win. Almost like, a skill, if you will.

You are restricting your definition of skill to only admirable skills (I truly wish it were so) like aim, movement, timing, and strategy. Lag hacking is a super easy/high return skill because of selfish, masturbatory-gleeful opportunism by the only ones willing to use it-- worthless drains on human life force within human society-- but nevertheless, these kinds of skills can in fact get wins, the probability for which is what is being measured, tracked, and predicted. Successfully, BASICALLY. That's the reality and the actual claim.

You accuse of whataboutism, then try to "INVALIDATE OMG" by using literally only outliers.

IN THE FACE OF WHAT IS BEING CLAIMED, WHICH IS A MEASURING OF AVERAGES/PROBABILITIES, AS YOU YOURSELF SAID, OUTLIERS... is... WHATABOUTISM, you fucking faggot.

When you and others accuse your opponents of exactly that thing of which you yourselves are actually guilty, I feel safe in inferring your politics.

Suck a dick, you ugly, retarded Marxist.

you fucking faggot

Very strong words, from someone who writes in such a gay manner:

Lag hacking is a super easy/high return skill because of selfish, masturbatory-gleeful opportunism by the only ones willing to use it-- worthless drains on human life force within human society-- but nevertheless, these kinds of skills can in fact get wins,

But I am concerned that you wannabe intellectual don't even understand the words that you are writing:

You accuse of whataboutism, then try to "INVALIDATE OMG" by using literally only outliers.

"Whataboutism" is when Supremee runs out of arguments and tries to switch the topic from Quake to other games.
"Well argued" is when I make an argument about the rating system and instantly back it up in the same post with undisputable evidence, like komap as the living proof.
And I am really amused about how I am allegedly "using literally only outliers", because you idiot didn't comprehend any of the other discussed topics besides komap.

IN THE FACE OF WHAT IS BEING CLAIMED, WHICH IS A MEASURING OF AVERAGES/PROBABILITIES, AS YOU YOURSELF SAID, OUTLIERS... is... WHATABOUTISM

Actually it was being claimed that Elo/Glicko measures skill, which is wrong and a single example is enough to disprove that claim. Retard.
But it was really cute, how you moron mixed up player-focused average of performance with the population-focused average of players, thanks for emphasising your lack of mathematical understanding.

You studied logic, you parroted everything therein, and you learned nothing.

And you obviously studied art or philosophy and sucessfully became a proper unemployed faggot.

Suck a dick, you ugly, retarded Marxist.

Hm??? Why so mad? When did I hurt you?

I feel safe in inferring your politics.

Well, thanks for trying, I wish you had more luck. Maybe next time, ... when again your everlasting butthurt sends you on another kamikaze mission. Until then, keep living in my shadow.

I think I love you.

Getting more into networking and writing my own code for my projects, I've realised that haven't an amazing connection is actually such a hinderance in FPS games.

Like you said, with the way the netcode currently works in QL even though we'll never see anymore updates, the fact that shit hits when it should and often I have people rocket jump at me, in a straight line and my rail or rockets go through them at chest height, still blows my mind. Yet I can sometimes hit invisible pixel rails and just wonder.

So yeah, some interesting shit indeed. Maybe things will get better in the future games coming out.

If Quake survives after the abortion that is Quake Champions.

Yeah, everybody experienced those kind of shots going right through the model and shots behind the model magically hitting.

If a "bad" connection is playable depends on certain characteristics: When your download is smooth but your upload is crippled, you will see everything normal (downloaded snapshots) while everybody else will see you warping around (uploaded packets), which is a huge advantage and the reason they had to block modifying cl_maxpackets (as if there weren't external methods for achieving this effect, lmfao, with "deliberate vpn-warping" i meant a ctf-player from Spain who instead of using his normal connection to DE-servers with a stable 39ms ping, preferred to pay extra money for a VPN that gave him consistent 50-55ms warping and masterclass behind-corner-rails).

I once played my brother and after that I watched the demos of both perspectives multiple times because I couldn't believe it and I still don't understand it:
I get the better start with more stack, the first damage exchange is pure LG vs. LG, I start firing first and don't stop firing, I hit the higher accuracy and still, somehow I die.
Higher stack, first strike, firing without pause, hitting more and still losing, how?

Another favourite match I watched several times on demo (only my perspective available) was vs. a player that started heavily lagging halfway through Bloodrun. Everything started normal, but then he did the Houdini:
He goes into the lower teleporter at rail and exits it, from GL I spam rockets into that hallway, he goes to the PG room and disappears, so I decide to rocketjump to RA, I pick it up and instantly get telefragged on the RA-spot.
My theory is that the server didn't recieve his current position updates due to his upload having packetloss and the first packet the server saw from him after that, was him standing on RA.
The rest of the match was funny aswell, he stands still with 999 and the strafe animation for a split second while he releases multiple plasma projectiles at once and then teleports to a new location again doing the moonwalk with firing several projectiles at once while my LG-tracking goes full retard.

In ctf I also captured double rail demos on 3 different occasions, one even including a reincarnation:
I frag someone with RG and in the same moment he frags me with RG, the console shows both messages following each other, I thought that is because the packet I sent of me firing must have had the same timestamp as the packet he sent of him firing, this happened 2 times and I assumed it is normal until on the 3rd occasion I saw the opponent reappearing with knockback, it looked like this:
I rail him and get the frag message for a split second, then I die from his rail and the console shows both frags, but then he reappears on my deathscreen in the spot I railed him with getting knocked back from my rail but without taking damage (because he should be dead according to the console).

My brother sent me a demo of himself firing a grenade at an enemy on Japanese Castles and the grenade gets trapped between the head's hitsphere and the torso's hitcylinder, it orbits around Keel's head before it explodes due to reaching the end of its countdown, it looked like the moon (grenade) is orbitting around the earth (Keel's head).
I look at the timestamp within the demo and the date shows that it happened 2 days after the update where the changelog claimed that they have fixed the 0-damage-direct-projectile-hits.

And this is just the magic of QL's serversided unlagged netcode/hitdetection. QC's clientsided hitdetection is even better, I saw this on Toxjq's stream:
He fights Xron on Bloodrun, Toxjq runs low on health so he decides to run away into the oneway teleporter at RL, you can see him exiting the teleporter at the Tribolt stairs, but then out of a sudden he gets teleported back infront of the teleporter entrance at RL and dies.
In QC this ain't due to bad hitboxes, bugs, warp or lag; it is in the netcode by deliberate design.

I've had weird shit like you've mentioned happen over the years as well.

I guess the devs and creators at large, don't play the hundreds of thousands of games and scenarios that play out in a game like quake, to see all of the shit that happens. Even if the code and design is technically perfect. There will always be a situation where some madness can unfold.

My guess is something like this: original Q3 netcode was aimed at making the game playable with lots of packet loss, fluctuating ping etc. The 50 ms smoothing was introduced for that purpose.

The game was played this way for some years, internet was advanced by the providers, the most used connection changed from modem to isdn/dsl.

Then I think the cpma guy (arqon?) came along and build his unlagged netcode. the 50 ms smoothing was removed and hit detection was 'improved' by making it not only server side but a mixture of server and client side.
During testing this worked well when two friends are playing with good connection. But heavy packet loss and fluctuating ping was not considered, the main reason the Q3 netcode was so smoothing and required aim to compensate for lag was exactly the point that was now mindlessly thrown out the window by unlagged netcode.

Everyone noticed there were some 'weird' occurrences, 'something' was off every other game. But it was never thoroughly investigated. Maybe because the netcode had become too complex and only a few gurus new how to tweak it.

Years later some reminaing players systematically tinkered with QL connection settings, routing to server and limiting uploads to give themselfes an advantage on purpose, because that's what players do, they optimise their setup to play as well as possible. As long as the game allows it, the fault is on side of the game, not the players tinkering with their connections.

The end.

What do you think?

I've always found that playing people with 50/60ms when I'm on 20/30ms - my shots feel as if they have to trail behind to hit them.

I don't know enough about the Quakes netcode to how this might feel but if I play people within a 5/10ms range of my own ping, all shots feel 100% right.

As far as I know, they indeed use a variant of netcode by the CPMA guy, but it ain't 1:1 the same as the CPMA version and relating console variables are blocked. Also from my understanding QL's hitdetection is purely serversided, but it is true that as arQon vanished there was nobody left at id with enough competence to work on it.
Abusing the negative side effects and the resulting rage began before QL was even released:
http://www.esreality.com/post/1528988/n-a/
https://esreality.com/post/1520770/cg-xerpcli...-gameplay/
If I remember correctly, cg_xerpclients is hidden in QL under another name; cg_smoothclients, g_laghax or something, I forgot so you will have to ask Yakumo or Lorfa about the name and value.

That is true, I don't expect the developers to have as many shit-tons of playtime as the players, but at least they should listen to suggestions and bug reports from those players, which didn't happen.

For example the scrubdate reduced the MG starting ammo in all gametypes to 50 rendering the gun useless, a couple of days later I have met Stinkerror on a CTF server and it basically went down like this:

Smack10: Your stupid update reduced the MG starting ammo in duel to 50 which makes the gun useless.
ShitPyle: Really? Well, that is not a problem, because you will quickly restack MG ammo on every of the new "grey unified ammo boxes" we have introduced in that update.
Smack10: There are no "grey unified ammo boxes" in duel.
ShitPyle ...

Another example is the removal of custom tickrates. As they initially gave the community back the ability to spawn their own client servers you could set the console variable sv_fps to a maximum of 125, which my brother and I did on my decommissioned simulation PC that was used for us as private duel server or public FFA/duel server when we didn't play. It was really nice, projectiles (GL, RL & PG) suddenly look bigger when you fire them, because the server sends you the information about their trajectory earlier. Effects that need serversided confirmation react much quicker, which makes cg_crosshairhitstyle feel damn responsive. And of course the hits you make/receive are 1/3th less delayed aswell. Because my brother, the server and I were located in Berlin, we pinged around 10...15ms, it felt like LAN and everything was fine. Then out of a sudden the next update forces sv_fps to 40 on all servers, effectively making the console variable useless. I visit them on IRC and it went down like this:

Smackulaas: Your stupid update disabled custom sv_fps settings which makes the variable useless.
SpongeBob: We tested it and setting it to 125 increases CPU load and triples bandwidth.
Smackulaas: My server runs on an octa-core and is hooked to a 50/4* Mbit/s connection, I don't give a shit. If your cheap 3rd World servers can't handle it, just run them at 40 like you always did, but give the community back the ability to use whatever their setups can handle. What is the problem?
SpongeBob: ...
*I forgot what contract it was back then, 50/4 or 100/10 Mbit/s, I didn't have to care because even as 8 players FFA server at 125 ticks it didn't come close to the upload limit and the protocols didn't show any other bottleneck (CPU/RAM) aswell.

Yes, even with technically perfect code you can get into trouble with an unexpected combination of unpredictable boundary conditions, but the things is, their code is shit and they are just too incompetent to fix absolutely banal stuff:
Over the period of about 4 years I occasionally reported them the broken crosshair images, 11 and 13 aren't in the center of the screen and can't be fixed with cg_crosshairx/y due to differing reference matrices, 15 and 18 have broken asymmetrical images, and so on. I was so annoyed that I fixed the images of 11 and 13 myself in 10 minutes by simply mirroring 1/2 of the images and offered them to send the corrected images with handing over any rights for commercial use for free. They declined because that would open Pandora's Box and then anybody would want to contribute (as if community contributions were so bad for Quake in the past). After roughly 4 years of reminding them, they finally worked on it: Out of the nearly 10 broken crosshairs that I reported, they worked on 2, namely the same 11 and 13 that took me 10 minutes to fix. They successfully fixed crosshair 13, they also tried to fix crosshair 11, but unfortunately they fucked it up in a new manner during the process, it now has a corona beyond the outer dark ring.
The other crosshairs I reported are still as broken as on day one (off center/asymmetric/...), try it yourself: Stand infront of a bright wall, use cg_drawcrosshair 18 with a non-black color and set an extremely high cg_crosshairsize so it fills the whole screen, you will notice that the crosshair has asymmetrical angles of the arrows, the arrows are inconsistent and the dark border is broken on some arrows.
They are far away from technically perfect code, when they worked 10 years on an AFPS that still doesn't have its crosshairs in the middle of the screen.

If they let the community get involved, the game would have been improved and grown imho, but hey, we know what happened now. We're stuck with QC which is at a point of it being too late to be aborted and there's no real way to make the game better/more popular.

Yes, in my opinion the key feature of Quake titles is the mod support. The people who work for free in their free time in order to perfect a game that they love are highly motivated fanatics.

Yes, QC is going down the same sink as QL, but QL had the better community influence due to the OSP-clone-start. Nonetheless the problem is the same: No mods, means you are forced to endure them do stuff nobody asked for but everybody hates.

I will forever remember QC as the game, where they hosted an 1 million dollar tournament on Parkinson LG and instead of listening to the players regarding the removal of round based duels, abilities and chimpions; id rather tinkered around on chimpion physics with a precision in the realm of parts per million.

I said that they should have taken that massive prize pool and used it to create smaller micro-tournaments, which could also be used to pool fan donations in that could have been used as a starting point for a Quakecon tournament, as well as hosting Quakecon in Europe instead of Texas, cause let's be honest. There has always been a greater/more active base of Quake players than the whole of the America/s

Those ideas sound very good, if just id would ever listen. :D

id *Americans

Fixed*

:)

-registering the rails i hit behind my opponent

What about Double Jump?

What about jump height?

I got into OSP super late, like four years ago or so.

I must say that I really do favour that floaty feeling-- especially with the e-dawn.org executable, which took out another source of engine input latency! Oh, the mid-air rocketage. And the actual OSP chain-jumpage (on stairs, you know?).

it deinstalls faster, because the client is smaller then in q3, i think.

Haha ok nice

is wrong,

you can install q3 just by copying a folder and start the game from there.

so no need to uninstall, just delete. what is faster than that?^^

By the way, I really wish Camping Gaz would come back, and give us The One Quake. Or its physics, I should say.

cityy, you keep making sweet maps and studying game level pattern language, mang.

Guys, we need a procedurally generated Quake zombie-mod platformer, to really break through to the big time!

I recently reinstalled QL. What is about the UI/X that makes it feel like totally disjointed, cheap crap? I couldn't find anything easily. I get that the UI in Q3 is utter shit as well but it was also designed in 1999.

Maybe cuz it is. I don't play other games though, besides nethack, so it seems very acceptable next to Q3, except for nostalgia regarding the menu sounds.

- Steam integration
no.