CS2KZ - What would you change? - GOKZ

CS2KZ - What would you change?

444

skz with wad + prekeep plz PLEASE

tatska

I don't think we should completely wipe KZT out, I think the RNG element creates excitement for players, although making grinding short maps more unenjoyable.

zer0kz

Could always play VNL if you want the sweet RNG because it's basically the most RNG KZ mode ever existed without having to write any code

Nooral

I definitely agree with splitting it into VNL & MOD, but it would be magical if when writing the code it could be written in a way that is friendly to adding other game modes (similar to how it is now where you write your own gokz-mode-simplekz.sp but here you also have to tweak files in other places).

I personally love SKZ's greens and consistent crouchhops, but its hard to dismiss how much more popular KZT is and how everyone else loves to gamble for 380s. That being said the changes I would like to see, which I think would make for some creative gameplay mapping, would be:

The fall stamina change mentioned by zer0
10aa (surf is not kz if kz cannot be surf)
low-skillceiling perfs similar to SKZ

BallisticBacon

yo what up

I have high expectations for what a next generation KZ plugin will contain, and I have a few ideal things I'd like to see.

Map Globalling

I will state this shortly, in CS:GO the Map Approval Team has never been the reason that there are too many maps despite yearly revisions to increase the strictness of approval AND map denial rate, it has only been due to an increase of mappers (e.g. an average of 27.3 maps per month submitted for three months compared to 15.2 per month before, as said in my post here).
Now, I have doubts on the viability of a distinct competitive map pool because of subjective reasoning around who determines what's competitive, but I don't outright oppose it if there is effort put, there's already the competitive pool talk here.

Porn Rooms

For some reason in the global pool there are 2 furry porn rooms, 9 hentai rooms, and 6 IRL porn rooms. This is a movement gamemode. You should be moving ingame, not AFK with movement IRL.

Tiers

9 tier system, plus an unranked 10th for uncleared, possible, and difficult maps which goes to tier 9 when cleared. Tier 6-7 is too small of a range in CS:GO.
Tiers are subjective. If you debate tiers for an extended period of time I think you would want to put your own vision directly into the gamemode. Map Approval Team tier + inbuilt averaged community tiering vote in the plugin if someone wants to switch between, with a max delta between voted tier and Map Approval Team tier? e.g. MAT Tier 6, community tier 5.8, with max vote possible 5 or 7.

Points

Maps should give a flat amount of global points based on tier. Right now we have a large portion of people who want to complete harder maps, and for first completions they usually get balls times. Flat bonuses help for this.
Global rank shouldn't be based off total points. The difference between top 10% and top 1% in the current system is massive and the amount of points given is too much, so there's a lot of people who have high ranks despite being nowhere near a top player. I'm reminded of score ranks in osu! compared to performance point ranks where only the performance points of a player's top 100 counts. You can switch between both leaderboards as well and there are "score farmers." Having ranks be based off average percentile of your top x maps is something I think can be looked into.
If there are player-owned servers, local points should not be phased out. Competition between players on a global scale is in the thousands, so new (and committed) people aren't able to see much beyond how many points they obtained. Local points gives them a reason to stay on a server for more than ping reasons and work towards a goal. I don't really care about making the ranks system the same as what I proposed above, maybe giving server owners a great ability to tune their systems can be good.

Maptesting-Adjacent Plugins [Third Party Plugins or Commands in CS:GO Shown]

Being able to show triggers and playerclips would be very useful, because mappers suck at making things skip-free.
[showtriggers_toggle, r_drawplayerclips 2, sm_showtriggers]
An inbuilt teleport menu to timers would be extremely useful, CS:GO isn't as good here because there is no !b for an end timer, and there can be multiple start and end timers (I believe doing !end on kz_floatingislands puts you inside of a wall far away).
[sm_tptotimer]
If something like r_novis 1 exists then it should be able to be toggled on without cheats somehow. Maybe an extra version of noclip that would enable it?
[r_novis 1]
Since I'm talking ideals, I've been pestering people for years for a plugin that shows where trigger_teleports or climb_teleports end up with arrows and to show what trigger_multiples are told to do. You do not know how many hours of my life I've spent noclipping into bhop triggers or random teleports to see if there's a wrong warp.
Cheat commands? These massively help with testing what you can do.
[sv_autobunnyhopping, +jumpbug, sm_autostrafe]
Segmented modes help with routing, they're also somewhat used in Bhop for competition, and saveloc doesn't have timers. You can possibly integrate a mode or modifier to the modes like that.
[sm_simpletas, sm_ztopwatch]
Disabling specific triggers is a complete godsend, too bad it barely worked for me without removing the triggers.
[ent_pause, ent_resume]
Listing/Rendering only a specific type of brush? There are uses to finding out what is a trigger_softbarrier, trigger_push, trigger_multiple, ladder brush, etc.
[sm_raytrace]

Plugins

CS2 will have a much more robust HUD system (at least I hope), so a fine-tunable speed and timer system that doesn't need a third-party plugin would be good.
[sm_mhud, sm_gcspeed, sm_speed]
A route command that doesn't suck would be great. The KZTimerGlobal !route command with the blue balls is one of the least helpful commands that should be helpful. I'd say optimally the WR route would be used, but like KZTimerGlobal you can also use the server route, maybe switch between both.
[sm_route]
Robust measurement systems are needed for maptesting and general routing. I have to look at GameChaos' spreadsheet on distance vs height for even knowing if something can be possible, but I have no idea if something can be inbuilt to alleviate that.
[sm_measure, sm_gcmeasure]
Mapping API ASAP. Gravity antibhop triggers give me nightmares.
LJ TPs in Mapping API.
[sm_lj]
Nightvision.
[sm_nv]
Splits. Influx timer was chopped up for their zone split feature on GOKZ, but one version of the chopped up timer completely wrecked VNL airacceleration and caused the kz_theaquila bonus to have a glitched time for years. Having splits be non-third-party would be useful.
[sm_zones, sm_ztopwatch]
Complete removal of timer tech and only allowing timer zones is amazing. Do that.

KZT

I think this plugin should be mostly the same, but I can see removing Our Father tech as something people want. KZT is the only KZ mode with a prestrafe that doesn't follow other modes, where you gain pre by having any horizontal mouse movement instead of a specific amount of turning. Prekeep is incredible though for comboing, it's purely skillful with its pre cycling and timings, plus prekeeping after crouching gets technical and difficult.

zer0kz

BallisticBacon If there are player-owned servers, local points should not be phased out. Competition between players on a global scale is in the thousands, so new (and committed) people aren't able to see much beyond how many points they obtained. Local points gives them a reason to stay on a server for more than ping reasons and work towards a goal. I don't really care about making the ranks system the same as what I proposed above, maybe giving server owners a great ability to tune their systems can be good.

It's probably a good idea to make local rank a "lite" version of global rank and give them local-exclusive name tags as a reward.

BallisticBacon Robust measurement systems are needed for maptesting and general routing. I have to look at GameChaos' spreadsheet on distance vs height for even knowing if something can be possible, but I have no idea if something can be inbuilt to alleviate that.

For straight jumps small offsets it is feasible, but it starts to fall apart when a drop starts getting very long (eg. some people can spam 280s but can be really bad at drop because their strafes slowly fall apart after a while). Once it gets to more sophisticated jumps (eg. boxtechs, things that aren't even jumps like surfs) it becomes really hard (if not unfeasible) to robustly define a measurement system. The best way to measure difficulties that I can think of is using jumps in existing maps as a reference point for each tier (eg. "this jump is similar on that other jump in map X, and since it's a tier 3 map and this jump is way harder it must be at least tier 4").

BallisticBacon I can see removing Our Father tech as something people want.

Interestingly, some people think Our Father tech is THE best feature of KZT (though I don't get it tbh). It is possible to remove the whole "wiggling to gain speed" without removing prekeep along with that though, so that wouldn't be too bad.

DanZay Pro Run Culture
Maybe only PRO runs should be ranked. At most TP runs could award flat points but to be rewarded for your rank, or given a rank at all, it should be PRO only.

I think the PRO run culture is a weird one, because speedrunning culture was always around the Any% category (which involves save/load whenever it's necessary). I am aware that it is always like that for surf, though it could be argued that save/loading on surf will make everything ridiculously way too easy if it is allowed, which isn't necessarily the case for KZ.

DanZay Permanent bans might be replaced with a temporary ban but a permanent mark on their profile that may be disputed if they really, really care. Less ban appeal effort.

I hate permament bans, so this is a +1 for me. Bans should always be temporary with increasing duration for repeat offences, and undisputed bans should resolve in invalidation of previous runs.

DanZay Ranked map pool might be determined by community survey or statistics instead of very time consuming opinion forming. Less map approval team effort.

I think maps can be ranked without direct team approval if it is not rejected and has a significant enough amount of completion (eg. 2 completions for death tier, 100 for very easy,...)

Also I have some visions for the new API:

You can no longer create infinite global private servers for free (though the first one is free). Public servers are always free though.
- 2nd private server costs $5/year.
- For $25/year you can create 10 private servers.
- Private servers must be password protected to allow players to filter out public/private servers in the server browser.
Locally hosted servers are allowed but they must have 99% uptime over 90 days.
If a global database exists then might as well incorporate surf and bhop. It's time to take over.
Subscription for supporter for minor perks. The money goes straight to keep the database running. 2 tiers of supporters, let's just call it tier 1 and 2 right now.
T1: $3/month
- Custom name tag
- Access to previous PB replays
- Friends and country leaderboard (not a big fan of this one tbh)
- 1 more private server (to a total of 2)
T2 $10/month
- Everything from T1
- Even more custom name tags
- Ability to save all replays and not just PBs
- 2 more private servers (to a total of 4)
- Dev server access

DanZay

Pro Run Culture
Maybe only PRO runs should be ranked. At most TP runs could award flat points but to be rewarded for your rank, or given a rank at all, it should be PRO only.

Reduce Dependency on Volunteers
Long term it's important for the system to basically look after itself. This may mean that server approvals are automatic based on some criteria applied to the server owner. Could be something simple like whether they have lots of times or are a mapper, and if the server is public and active. Less server approval effort.

Permanent bans might be replaced with a temporary ban but a permanent mark on their profile that may be disputed if they really, really care. Less ban appeal effort.

Ranked map pool might be determined by community survey or statistics instead of very time consuming opinion forming. Less map approval team effort.

BallisticBacon

DanZay Pro Run Culture
Maybe only PRO runs should be ranked. At most TP runs could award flat points but to be rewarded for your rank, or given a rank at all, it should be PRO only.

Given a majority of the CS:GO playerbase will transition to CS2 and there's a large set of people who go for hard map completions, I would look into something like a dual point system for PRO and TP globally, but maybe only have PRO rank show (or something like PRO PRO (lol) and TP PRO as a setting someone can set for themselves). Although, in the points system I proposed it wouldn't make sense to have TP runs be included at all unless there are no PRO runs.
Full freedom mode would be to not care at all and force maps to adapt to a PRO-biased system, where a map would be globalled after someone pro runs it. On a similar note there's a system like Super Mario Maker 2, where mapper verification runs are required to essentially PRO run the whole map and from all placed checkpoints. Certainly would make the "this map is too hard / unfair for PRO runs" people happy.

DanZay Ranked map pool might be determined by community survey or statistics instead of very time consuming opinion forming. Less map approval team effort.

A return of the maptesting team would be cool. Tying it with a community-defined "competitive" map pool is interesting as well (maptested gets put into unranked, community puts it into ranked).

zer0kz Once it gets to more sophisticated jumps (eg. boxtechs, things that aren't even jumps like surfs) it becomes really hard (if not unfeasible) to robustly define a measurement system.

Yeah, although I am an addict to thinking about batman's TAS tool where you have frame advance, precise fine-tunable angles, and position visualization, it's just not practical that you can make something that predicts any of it. I was mostly just thinking about straight jumps because of that, but then it gets to finding distances humanly possible instead of autostrafe distances (which is what I used the spreadsheet mostly for).

zer0kz I think the PRO run culture is a weird one, because speedrunning culture was always around the Any% category (which involves save/load whenever it's necessary).

There's a bit of a spectrum with it. I see TP runs akin to taking savestates in a pause screen, which is an action not allowed in a large amount of speedruns. Popular speedrun games like Mario 64, SMW, Odyssey, Celeste, Hollow Knight, etc wouldn't allow that and loading save is similar to falling on a PRO run and being sent back to a specific previous area, but FPS games like HL and Deus Ex do have on-demand saves which I do understand. I think it's actually less common that you can save on demand to retry from any position instead of being sent backwards.

GameChaos

DanZay Permanent bans might be replaced with a temporary ban but a permanent mark on their profile that may be disputed if they really, really care. Less ban appeal effort.

Permanent bans actually reduce ban appeal effort for me. I'm a big fan of perma bans.

apprehensivebob The main change to mechanics I'd like to see is asynchronous state update, i.e. infinite tick rate. CS2 might be a good opportunity for this with the new sub-ticks. If the plugin update can receive a timestamped buffer of all user input that occurred since the last tick (multiple mouse and keyboard events) then you can effectively do infinite tick rate. Maybe this helps for anti-cheat and anti-bind as well as improving motion accuracy.

"Infinite" tickrate doesn't have any benefit to movement. it's also a complete nightmare to implement seeing as how even valve is struggling with it.

apprehensivebob On the KZT vs SKZ discussion, I don't think there has been a fair analysis on the claim that SKZ has better mechanics. The big argument for SKZ is that KZT bhops have randomness and randomness = bad. That's not true in general, it just means there is a probabilistic element to the game. Yes for some specific maps the best route has a small number of random jumps, and the map is long, so you can't build up enough samples for your time on the map to not be random... That is a problem, WITH THAT MAP. Overall KZT does not have a problem with it being random which player performs better. I think the probabilistic element is important to the game, it's the difference between being able to follow a static learned trajectory vs having to react in the moment to the hop you got. This is very similar to the distinction in car racing game TrackMania between fullspeed maps and non-fullspeed or dirt maps, same deal with speed either being known or having to react. Both are fine, but fun for different players. If there neeeds to be a single kz mode I would want to give mapper discretion on mechanic, e.g. "dirt" blocks that have KZT behavior, auto-bhop blocks, etc.

This is a very interesting discussion. People like to play kz for different reasons on different kinds of maps. I for example am a big fan of perf grindy maps, but other people really don't like that at all and prefer to focus on other types of kz like micro maps (which I like a lot as well). I think SKZ is less "versatile" than KZT because it's designed for fewer playstyles (though KZT probably wasn't designed with much thought). SKZ is great for climb/micro type maps and awful on bhoppy maps (imo). This is also probaly the reason why no one can agree on what good KZ mechanics would be, everyone enjoys KZ for wildly different reasons. Sorry for the scatterbrain, I hope I got some kind of point across.

Things that I would change for CS2:

Consistent jump height of 57 units
plekz tech/perf height fix
sv_ladder_dampen 0 which makes ladders easier to use imo (removes 30u/s cap sideways when looking into the ladder). Which is ironic cos it was implemented to make ladders easier: "Made ladders easier to climb. Added cvars sv_ladder_dampen and sv_ladder_angle."
Begone timer tech, hello zones.
Possibly 10 airaccel with the airaccel fix (like in cs 1.6) cos that's fun.

Also mapping is going to be much more difficult in CS2 because lighting compilation in hammer is gated raytracing gpus and it's a completely new engine and workflow. I won't be expecting a lot of high quality new KZ maps any time soon after launch.
KZ mapping discord is still alive though if anyone's got a raytracing capable gpu and wants to start mapping after CS2 is publicly released!

zer0kz

BallisticBacon I think it's actually less common that you can save on demand to retry from any position instead of being sent backwards.

Yeah that might actually be the case, I was mostly thinking about speedrunning on fps games using source/goldsrc engines (half life, portal, etc) and not necessarily other genre of games. KZ is right now somewhere in the middle where you can retry but not from any position.

Geuser

Eh, I'll speak up for fellow bad players and say TP runs should award points and count towards rank. Though i could see having them be totalled separately and/or weighted significantly less. e.g. pro runs give a max of 1000 points for wr, tp runs give a max of 250 or something.

Mostly thinking about higher tier maps, it just doesn't sit right with me to think about someone completing oloramasa or spacemario but then go "ah but i see you used some checkpoints, therefore this completion tells us nothing about your skill level at all, and you get no reward/recognition for completing something that the vast vast majority of other players will never be able to do at all lol"

Also as a WAD enjoyer. Would it be possible to keep it, but maybe force the wiggling aspect to less free? E.g. you have to "strafe" much wider and/or faster

BallisticBacon

Geuser Also as a WAD enjoyer. Would it be possible to keep it, but maybe force the wiggling aspect to less free? E.g. you have to "strafe" much wider and/or faster

Surprised more WAD users aren't here talking about it. I don't dislike Our Father tech, but I don't see making the esoteric tech difficult to do as a viable compromise. I think it would have to be the same as current KZT, like SKZ prestrafe with prekeep, or like SKZ prestrafe without prekeep.

DragonCat_

@DanZay In my opinion, it's not suitable that just cancel previous modes like KZT or SKZ.
A new mode is welcomed,but it doesn't means that the previous modes should be replaced. We want something new in order to have extra fun.It's change appealing rather than revolution .Giving up the previous ones is not a good idea.Maybe it'
s could be added as a new mode,and these 3 modes are included in Modded.Though it might be a burden for database server,but it doesn't make sense to just simply abandon what we enjoyed before.As for ''unifying the community'',excuse me? It's not a competitive game like competition in CS.Actually it's a causal part of CS in community.We play KZ for fun.

zer0kz

DragonCat_ As for ''unifying the community'',excuse me? It's not a competitive game like competition in CS.Actually it's a causal part of CS in community.We play KZ for fun.

I think the problem is that skz and kzt are too similar to each other that one mode just invalidates another mode on 95% of the maps. I think kzt and skz (and maybe legacy vnl csgo) can stay but only the new mode and new cs2 vanilla should be ranked (see nooral's custom mode idea). Meanwhile each style on surf guarantees a totally different style of gameplay.

Geuser

BallisticBacon

Surprised more people aren't using the forum? lol : )

Gotta keep spamming the discord, so people more relevant than me can start contributing their input. Specifically GC, as it seems like he's going to be implementing a lot of this crap anyway.

But just to clarify my wad suggestion: the idea isn't to make it disgustingly hard. But just to make it require slightly more turning. I like esoteric tech in videogames. And i like mappers having the option to force it occasionally. But i think there's a perception that its currently too easy and too optimal compared to regular pre because you can just run perfectly straight. But if you actually had to "strafe" instead of just wiggle, it could be more lj-like, where theres a tradeoff between getting good gain vs optimal ~~air~~groundpath (perfectly straight).

Again dont know if thats feasible, just throwing it out there.

Zhivotniy_Instinkt

KZT is perfect, I wouldn't change a thing. The other game modes are irrelevant. Also I agree with Geuser on ourfather and prekeep being important.

DanZay

Would a point system like this be good?
https://docs.momentum-mod.org/guide/ranking_cosmetic_xp/
What about the same system without the formula component (looks like it would result in a lot of database churn)?

I also have doubts about map pool size kinda inflating points for people who go for completion. Would think limiting the number of maps that contribute to your rank like Bacon said sounds goodish.

zer0kz

DanZay

Cosmetic points could be interesting but I don't see KZ adding any custom cosmetic anytime soon apart from name tags, mostly because Valve doesn't seem to like custom player cosmetics in their game.

RandomEye

I really like zer0kz's idea of separate map pools. CSGO maps (that have been ported to CS2 or otherwise) should be in their own Legacy pool, and CS2 maps should go in two pools, Unapproved and Approved. All new maps start in the Unapproved pool. Each map goes through a vetting process before being added to the Approved pool, but this vetting process should be community driven instead of being controlled by a small team. One solution would be to let every player score a map on some categories after they complete it. A map should only be added to the Approved pool once it has been scored enough times and met a minimum average score in each category. I propose these categories for consideration:

- Accessibility:

Accessibility is important. We don't all have the same monitor with the same settings, same IRL lighting conditions, same eyesight, colorblindness, ect. Some players may have trouble with a low accessibility score map where other's don't. A high accessibility score means few or no players have trouble playing your map due to external factors. An example of a low accessibility score would be a map with poor lighting, areas that are too bright or too dark to see, poor use of textures such that it is hard to judge distances or see where the next jump is (I have a lot of trouble with this on flat color textures maps).

- Beauty:

Self explanatory. KZ can be beautiful! CS2 KZ maps should have a stronger focus on visual appeal. A map with a high beauty score means the mapper put in a clear effort to make the map visually appealing. No more low-effort ugly maps!

- Flow:

Flow is hard to pin down, but we all know that being lost on a KZ map is no fun. A map with a low flow score would make players guess where they have to go next. The next jump should be obvious at all times. A high flow score would mean a player playing the map for the first time spends little or no time trying to figure out where to jump next.

DanZay

zer0kz
I was more referring to the ranked points system on that page.
https://docs.momentum-mod.org/guide/ranking_cosmetic_xp/
It consists of WR/top10 points + formula points + group points. I am personally a little uncertain about the formula points component presented but seems good otherwise.

apprehensivebob

The main change to mechanics I'd like to see is asynchronous state update, i.e. infinite tick rate. CS2 might be a good opportunity for this with the new sub-ticks. If the plugin update can receive a timestamped buffer of all user input that occurred since the last tick (multiple mouse and keyboard events) then you can effectively do infinite tick rate. Maybe this helps for anti-cheat and anti-bind as well as improving motion accuracy.
It would be nice to be able to pause your run, teleport to someone to demonstrate a jump, and then continue your run.
For ports of maps especially those that exist in both CS and CSGO I think it's important to get expert opinion on any changes. Maybe map approval could mandate that the porter of the map provides additional approvers that have history with the map. Maybe there won't be any significant changes though, depending on KZT vs SKZ discussion.

On the KZT vs SKZ discussion, I don't think there has been a fair analysis on the claim that SKZ has better mechanics. The big argument for SKZ is that KZT bhops have randomness and randomness = bad. That's not true in general, it just means there is a probabilistic element to the game. Yes for some specific maps the best route has a small number of random jumps, and the map is long, so you can't build up enough samples for your time on the map to not be random... That is a problem, WITH THAT MAP. Overall KZT does not have a problem with it being random which player performs better. I think the probabilistic element is important to the game, it's the difference between being able to follow a static learned trajectory vs having to react in the moment to the hop you got. This is very similar to the distinction in car racing game TrackMania between fullspeed maps and non-fullspeed or dirt maps, same deal with speed either being known or having to react. Both are fine, but fun for different players. If there neeeds to be a single kz mode I would want to give mapper discretion on mechanic, e.g. "dirt" blocks that have KZT behavior, auto-bhop blocks, etc.

zer0kz taking map difficulty into account for map points

If failed run stats were collected that could help - e.g. list all the failed runs after the slowest run, 1000pts to fastest player, and e.g. 500pts to slowest player if there are 500 players with failed runs. "Failed run" could be defined something like: made it x% through the map in y minutes and didn't finish, if both x and y are over some threshold that's a failed run.

BallisticBacon

apprehensivebob It would be nice to be able to pause your run, teleport to someone to demonstrate a jump, and then continue your run.

Adding onto this, an inbuilt feature to go in freecam would be very helpful especially for first times running maps. CS:GO already has this with going into spectate, but if there are other players in the server it goes to them in first person, then you press space until you go into freecam, then you fly to where you were, and then you have the always-looming possibility of joining the wrong team accidentally then losing your run.

apprehensivebob That's not true in general, it just means there is a probabilistic element to the game... I think the probabilistic element is important to the game, it's the difference between being able to follow a static learned trajectory vs having to react in the moment to the hop you got.

This is pretty much what my argument in favor of KZT has always been, as well as the fact that being able to scroll well is a skill that is overlooked. I see the crux as the fact that KZ has always been a precision-based gamemode, and inconsistencies impede the focus on precision. I think of other speedrun games that have (truly) RNG elements that players just deal with and those games have a place as well, and I think KZT and SKZ are just like different types of speedrun games in that way. Perfect bunnyhops are deterministic anyways as it's a 1/128s timing (which is why I've argued so much about players calling them RNG when speedrunners would call it a type of frame perfect trick), its issue is that it's relied on so heavily in many maps which at the highest level of competition can make it brutally annoying to play if there are large amounts of blockskips or forced skips in the route.

apprehensivebob If there neeeds to be a single kz mode I would want to give mapper discretion on mechanic, e.g. "dirt" blocks that have KZT behavior, auto-bhop blocks, etc.

The mapping API was an incredible idea put into place too late in CS:GO's life. Mode forcing/switching, checkpoint/teleport disabling for TP runs, turning off certain triggers with a timer running (e.g. turning off other course teleports that bring you near the end of another course, instead of a point_clientcommand that makes you do sm_stop), autobhop + uncapped speed, disabling prestrafe, etc. I believe more interesting things being added to the mapping API would be one of the most important additions for mappers.

zer0kz

BallisticBacon This is pretty much what my argument in favor of KZT has always been, as well as the fact that being able to scroll well is a skill that is overlooked. I see the crux as the fact that KZ has always been a precision-based gamemode, and inconsistencies impede the focus on precision. I think of other speedrun games that have (truly) RNG elements that players just deal with and those games have a place as well, and I think KZT and SKZ are just like different types of speedrun games in that way. Perfect bunnyhops are deterministic anyways as it's a 1/128s timing (which is why I've argued so much about players calling them RNG when speedrunners would call it a type of frame perfect trick)

While it's technically true that they are all frame perfect trick, I think you should take into account how long that frame is. The typical frame perfect execution is 17ms for 60Hz games, which is much much more than CS:GO's 8ms window. On L4D2, you can do frame perfect perfs with spacebar surprisingly consistently because the game tickrate is really low (20 tick > 50ms perf window). So not all frame perfect tricks are equal.

apprehensivebob On the KZT vs SKZ discussion, I don't think there has been a fair analysis on the claim that SKZ has better mechanics. The big argument for SKZ is that KZT bhops have randomness and randomness = bad. That's not true in general, it just means there is a probabilistic element to the game. Yes for some specific maps the best route has a small number of random jumps, and the map is long, so you can't build up enough samples for your time on the map to not be random... That is a problem, WITH THAT MAP. Overall KZT does not have a problem with it being random which player performs better. I think the probabilistic element is important to the game, it's the difference between being able to follow a static learned trajectory vs having to react in the moment to the hop you got

You can say that the problem is with the maps, but when a huge portion of the maps are affected by this issue, then I think it's a problem with the mode.

I agree being able to adapt and react to RNG is a skill, but the map must be either long enough to have enough samples to minimalize the luck aspect, or be designed by a way so that other skills are much more important than getting lucky in order to go fast.
A good example of the former is bkz_underground_crypt_v3 (where the entire map is bhop but there are so many bhop sections that you have to react and can't just get lucky on all of them for a good time), while a bad example is bkz_apricity_v3 (the difference between WR and #2 run is literally just hitting the lucky perf at the right spot on those maps and other perfs do not really matter), and most maps are closer to apricity than to underground_crypt.

That or you have to make death tier maps where jumps are so hard that the RNG aspect of KZT doesn't matter if you can't oneshot everything, and even then you can be lucky and bruteforce hard bhop sections by having the right perfs at the right place.

DanZay

apprehensivebob On the KZT vs SKZ discussion, I don't think there has been a fair analysis on the claim that SKZ has better mechanics. The big argument for SKZ is that KZT bhops have randomness and randomness = bad. That's not true in general, it just means there is a probabilistic element to the game.

I made a video >3 years ago regarding KZT vs SKZ especially regarding perfs. To summarise the video, although perfs have their skill-based aspects (as you describe), they have an impact on run time far greater than execution of other "tech" like block speed, crouch control, and even strafing. So the level of skill you pour into those other "tech" becomes less important, which IMO is bad design considering that hitting perfs is practically just flipping a coin, and the other "tech" is much more engaging.

zer0kz And I think the downside of perf significantly outweighs the benefit.