The Payload can Skip Checkpoints

[@milesupshur9614]

Turn payload maps into push maps with this one simple trick! Non-mobile heroes hate it!

[@EggyB]

LMAO

[@Nick.11]

xD

[@sakucelty]

casual push players: the Doomfist with top 500 level rollouts:

[@Lani_Voices]

@@sakucelty the chadfist

[@youreinlovee]

SHH DON’T GIVE THEM IDEAS!!!!

[@justaharrowmain6853]

"Perhaps given an infinite number of Overwatch games, every possible bug is bound to happen" -Ow2 hot fix motto

[@beybladetunada5697]

"And thats why we are cancelling PVE, we need to focus on bugfixes!!1!"

[@lnemf]

It's like the monkey with a typewriter thought experiment

[@lemonlemon7932]

@@lnemf I mean, yeah. "Within infinity, everything"

[@defectivepikachu4582]

@@lnemf no the idea is that the monkey actually does something

[@aetheriox463]

@@beybladetunada5697 they were stripped of their resources and were expected to make the same game, what did you expect?

[@aeloh6921]

Oh my god, I had a comp game a while back where the enemy team just randomly won before capping the 3rd point. Now I can finally retire knowing that this mystery, the source of so many sleepless nights has been solved. Thank you

[@quantumsno]

I can't believe you managed to replicate this one tbh. Actually insane

[@SethJV]

It's... fairly easy to replicate. In custom games, I mean.

[@lennywhere]

@@SethJV figuring out how it happens is the reason why it would be difficult

[@Marblr]

I had some help from other users -- namely Coathar!

[@Weeblon]

@@Marblr You obviously don't have to but it might be nice if you'd add people that helped you in the video description. I mean who doesn't like being appreaciated, right?

[@mr.rabbit5642]

I saw it happen once in my comp match, on last point of Numbani. It kinda remained uncaptured, but the sides swapped right after (?) I was starting to think its just a visual bug on match proggress bar

[@Coathar]

Imagine staying up until 4am brute forcing locations for this, couldn't be me

[@spookiboys]

Smile

[@GirusBetterThanVirus]

@@spookiboys :)

[@cedricksusername]

I want this to happen in OWL so bad

[@SeraYGO]

For Real

[@miguelcruz7826]

Would they reset the march if it actually happened?

[@phil1500]

​@@miguelcruz7826 yeah absolutely they'd replay the map

[@jplayzow]

@@phil1500 probably ban everyone and then declare it a draw for existing near a bug

[@bawbehh]

event organizers should orchestrate this so OWL is actually somewhat interesting

[@charleslim6514]

imagine being a blizzard dev and finding out your tasks through Marblr

[@CaptainBen10]

Marblr should be hired by Team 4 since he clearly knows this game better than the devs

[@alexandredasilvavilasboasrosas]

imagine being a blizzard dev PERIOD

[@GlavGadLive]

lmao

[@willmcclard206]

“hm ok let’s see uh nothing to do today. um. hrm. oh Marblr uploaded, i guess i’ll watch it, might not fix it if it’s too complicated.”

[@LightsJusticeZ]

Don't think it's uncommon for game devs to find bugs through social media.

[@octoprime8019]

It's truly amazing how you find these and the reasons. Keep up the overwatch science

[@DeertickDaniel]

He gets them sent to him but the way he finds out why is impressive

[@focusprx]

when you know the games mechanics it’s easy to replicate bugs like these

[@absolutewisp]

​@@focusprxand that knowledge is impressive. While we have some tools to peek into the game a bit, like Workshop, a lot of it is left to speculation unless you're really into it.

[@cyberfish6507]

Due to how difficult this is to replicate and there have been seemingly no changes to affect how payloads work, it's possible, and even likely, this has been in since OW2's release, maybe even OW1. The fact that it's been there for at least a year and is only just now discovered highlights how insanely rare it is.

[@fidel_soto]

Been there since day 1. It happened to me at least twice and I only played Ow1.

[@tucker2840]

@@fidel_soto oh yeah it happened to me 5 times and i played overwatch one time

[@GoldenKirby]

Sometimes things exist for a while but nobody knows until somebody popular covers it

[@kid10249595]

I think this happened a couple times in some matches that I was in. It's extremely rare, but it is blizzard so I'm not surprised they let a bug like that slip through the cracks.

[@GoldenKirby]

@@kid10249595 to be fair it's a bug one can't normally test for, and due to its rarity and lack of big content creator calling it out it probably went unnoticed

[@gavinsfriend5632]

Imagine having a shit losing streak and on your “final” game of the night where things are finally going greatly, this glitch just pulls up and shatters your hopes

[@lesarchivesdeval8477]

I do believe it happened once during a multi-streamer event in France back in 2018-19. On a custom competitive game on Hollywood, the gate after the 1st point didn't open. They had to redo the game. If i recall this event correctly, then this bug is present in the game since Overwatch 1, so your theory about the infinite number of games seems pretty solid x)

[@oskain]

link? :3

[@Marblr]

I'll be looking at this bug on stream if you want to ask questions or just hang out: www.twitch.tv/marblr :)

[@marzman4]

Hello. Do u plan on expanding your content to other games eventually?

[@Marblr]

@@marzman4 There are a couple other games I'd love to explore!

[@philogl]

Heya! Just wondering if there is there a way to visualize how close to the cart you need to be to push it? I've noticed it is easy to c9 the cart on Dorado on the back side, yet on the front you trigger it from pretty far away, it feels like the model and the zone to trigger a push are misaligned

[@ecargnz7762]

@@philogl i have the same issue too on dorado .

[@Marblr]

@@philogl the push volume is a sphere centered on the payload's origin point (the dot shown at 0:39). I believe the Dorado payload's origin is toward the front of it, so a sphere would extend much further in front of the cart than behind it. I don't know what the radius of this sphere is offhand.

[@ToonDoge]

I think i actually had this happen once before and was wondering why the payload clipped through eichen's castle doors before the animation had a chance to play. Very interesting lmao

[@bigdoggetom6549]

That's got to be the most common place for it to happen tbh, I've not seen a bigger clusterfuck of a checkpoint ever

[@ccricers]

That's odd. Someone else in the comments saw a similar thing happen with the gate doors in Hollywood. But instead of clipping through the doors, the car got stuck in the doors

[@Damonj17]

6:56 "Perhaps given an infinite number of Overwatch matches, every possible bug is bound to happen eventually." Man, I hope I end up in the Overwatch match where the "lucio gets buffed" bug happens

[@Guarrow]

Had to watch multiple times to understand the competitive desync, the imagery really really helped, especially the gray dot that's representing the goal Awesome work as always

[@PieMan061]

“You’ll have to walk a marathon to get to the payload”. Devs saw this and thought it was actually a good idea. Thus push maps were born.

[@thesalad943]

And flashpoint

[@SunAndMirror]

Hey we heard u like 2cp and holdingW from spawn so we made 5cp with walking element of push. Heres flashpoint, enjoy

[@drizztdourden4855]

My guess as to the fact that you have to push it right as it starts rolling backwards is likely logic to prevent nodes from being re-triggered in error by the payload moving backwards. IE if the junkertown payload moves back over the gate closing node, it will not attempt to close the gate again, since it detects the payload is moving backwards. And thus this applies to checkpoints, for a single tick, the checkpoint is detecting that the payload is moving backwards so it doesnt activate that checkpoint node

[@talymenombrenuit3659]

That would mean that nodes would get triggered again if you let the payload roll back and then over the node again though (in a normal situation without the glitch)

[@SoWhom]

@@talymenombrenuit3659 that is what happens as demonstrated in the video

[@talymenombrenuit3659]

@@SoWhom No, what I meant is, in a normal game, if you got CP1, then the payload went backwards to before CP1 then went over it again, you'd get the extra time again That doesnt happen

[@JackAllpikeMusic]

just wanna say that although I don't have any plans on playing OW2 again - these videos are still super fascinating and are so well put together. Keep it up, love the content. :)

[@Stev4iK_FeeD]

Apart from ability to find and reproduce bugs of any complexity, I've always been delighted with your video editing skills

[@gplthebeast]

I completely purged myself from anything OW related, except for this channel, these videos are great.

[@zc8673]

These are literally my favorite OW videos, i love this stuff.

[@Oreokillerz14]

Man that point skipped harder than the OW2 developing team skipped PvE

[@allencunningham9002]

My favorite overwatch content! It's interesting seeing how the game works behind the scenes.

[@D_Quinn]

3:00 hey look, it turned into the push gamemode!

[@chewie__]

I don't even play Overwatch anymore, but I continue to come back to your videos. Your prose and straightforward editing style makes each video a really enjoyable experience.

[@tajwar9547]

It's all good, the developers are really busy working on PVE so these bugs are bound to happen.

[@crangejo]

DOES HE KNOW?

[@georgerice6488]

Any day now pve will be released perks and all. I cant wait

[@cwj2733]

no one tell him

[@hallowedchromee310]

@@georgerice6488 I love the part where Mei cryofreezed all over the place and froze PVE progress forever. May it rest in peace.

[@AtomicAlchemist]

@@crangejo Probably, that comment is dripping with sarcasm

[@ofischial]

I was always about to ask “well what if….” And then you answered it. Great job!

[@Vason_Gaming]

Blizzard should look at all of Marblr’s KZbin vids like this, it shows such an amazing tutorial on the bug, a good intro, good editing and explanations. Like someone give this man a raise ffs.

[@kanashisa0]

“Escort, Hybrid, and Push are temporarily disabled. Only Control maps are available in Competitive.”

[@unfairdev8197]

lmao

[@codegeek98]

fr?

[@b2n101]

Hey Marblr, really missing the videos, always loved watching your style of videos. Hope you make a new one soon!

[@jdude4264]

This is the first bug of yours ive actually experienced in game, and its a rare one too

[@bianglegaming1127]

Oh god, this is a fun one. Aside from not being able to win, it's worth pointing out the the spawn room not changing **combined with** no extra time being added makes for a truly uphill battle.

[@grankino24]

The infinite monkey theorem states that a monkey hitting keys at random on a typewriter keyboard for an infinite amount of time will almost surely type any given text, including the complete works of William Shakespeare.

[@TamWam_]

imagine a game where this happens, and they reach the end without winning… will it just be overtime til the attackers c9?

[@EggyB]

Imagine in Total Mayhem, the attackers wouldbe so tilted

[@dikiyhuy]

​@@jessicastjames6202 🤓

[@A_Minor_Mistake]

​@@dikiyhuybruh

[@PunkZombie1300]

​@@jessicastjames6202 Now I'm wondering what happens if on a payload map the first team were to do this glitch and make it to the end with zero points, like once they make it to the very end they just forfeit the rest of the round. Like does the other team win as soon as they reach the first checkpoint and score a point? Or would they have to make it to the end while also using the glitch? Since the distance is x meters away from the last checkpoint, I think the second team wouldn't be able to win (purely based on distance rather than points) if they hit the checkpoints, since the first team has the distance of the whole map while the second team would keep resetting their distance and only have 1/3 at most of the first team's.

[@tatri292]

@@wargex Considering that the attackers would have far spawns, yeah that's never happening for a long enough time

[@kyspace1024]

"Perhaps given an infinite number of Overwatch matches, every possible bug is bound to happen eventually." (Marblr's ergodic hypothesis)

[@sjshdjdjesj5628]

Why and how do u know that word

[@kyspace1024]

@@sjshdjdjesj5628 I do physics

[@changuito2024]

Your videos always remind me of one match I’ll never forget where I was playing Zarya on Dorado. We were fighting on that 2nd checkpoint where the gate opens, and both teams were contesting point for extremely long. All our stats were so high, we really contested point for so damn long. Eventually we died to a junkrat ult. 😢One of the funnest matches ever.

[@KissedToge]

National Marblr Upload day!

[@codegeek98]

5:06 I'm floored by the efficacy of that animation in explaining what's going on

[@AutumnlLeaf]

Dude your voice is so soothing. I could listen to you explain game bugs all day

[@glitch3579]

0:04 I can't believe you pulled a Hazy Maze Cave. Triggered my fight or flight with that one.

[@djdhjdnf]

this is so well edited for a payload glitch 😂

[@sirdurtle9519]

I know im a year late to the party, but I have a pretty good guess as to why this is. The logic for this moment likely looks something like this: if payloadDistance equals checkpointDistance; checkpoint = captured; When the first line should be: if payloadDistance >= checkpointDistance; This code is likely only run when the payload is actively being pushed. In other words, in order for the checkpoint to be captured, it must be actively pushed, and have its distance equal the checkpoint distance, which works fine in most cases. But, depending on the exact order of operations, what ends up happening is this: Tick 1: payload is being pushed Between Tick 1 & 2: payload stops being pushed (game will not register this until Tick 2) Tick 2: payload reaches checkpoint, and is no longer being pushed, so checkpoint is not registered Tick 3: Payload is being pushed, but distance no longer equals checkpoint distance, so it is skipped (because the payload has started moving) The fix in this case being that the checkpoint should be recieved if the position is >= the checkpoint distance, or alternatively have it collect the checkpoint regardless of whether its being pushed or not

[@AllOfARCH]

The visuals came in clutch. Especially with the Comp Desync cos I didn't understand at all without it 😅

[@skiper7795]

Marblr is giving us better content than dev team itself lol

[@theredgamer211]

Great, high quality video!

[@xenophanessonic821]

Wake up babe, marblr posted

[@Sangoreborn]

Very well produced video, explanation and easy to hear voice. Well done!

[@ArmyBlinkStayOnceMidzyMoaMY]

Marblr putting more time in ow than devs themself

[@veretxnerd983]

Pve

[@Zen-db5pp]

@@veretxnerd983 What pve?

[@javerious24]

nah the devs put they life into this shit. u should be mad at blizzard executives

[@phil1500]

idk this is so rare and i've literally never seen it happen in 1k+ hours. It might have even existed in ow1 tbh. They might not even know about it or its so obscure its not a priority

[@woodybob01]

@@Zen-db5pp It stands for Player Vs Everything. He's referencing the gamemode that blizzard promised would be coming with Overwatch 2 but the executives scrapped the entire thing which was 4 years of work.

[@Snowshill]

holy fricking heck I have seen the junker town gate happen, and not just one bu a few tiems so THIS is why D:

[@case3177]

as someone who has done lots of editing with the sm64 castle theme as background music seeing the ending sync was a treat

[@seatyourself7082]

production hero is unreal! Very well done

[@jordanshafer]

Cool vid! Also I played in a game w you earlier on emonggs stream. Had so much fun! -screamyposum

[@theonlymateus]

honestly love your content, thank you

[@AdrianManChild]

WAKE UP BABY, MARBLR RELEASED A VIDEO.

[@clarity-1850]

Nothing better than Marblr uploading

[@frezzingaces]

"Perhaps given an infinite number of Overwatch games, every possible bug is bound to happen" - this is an actual testing strategy called 'fuzzing'.

[@joelalexander928]

Best OW2 new content is the bugs we can't comprehend how the devs created them

[@spybgon4597]

Wealth over here in TF2 we don't need to worry about people doing that as a bug

[@ttantrum7948]

this explains all the times that we're near the 3rd checkpoint and I'm spawning at the default and having to pin across the entire map, losing a whole minute as tank

[@CasperVonGhoul]

I lost a comp match to this. I even still have the clip saved on Xbox and shared it to the subreddit and they just understood it to be a bug

[@PlNGUlFY]

So glad Icould be in this video, what a cameo

[@SessesSniper]

Hello Marblr. Can you make a video explaining the practicality of the "recoil recovery aim compensation" option. I feel like everyone has a different take on this and I'm not sure whether it's beneficial or not.

[@aeloh6921]

Yeah that would be great.

[@claymeistereu]

Meanwhile in tf2 you gotta make the payload go 100+ mph for desynchs, and more

[@dojelnotmyrealname4018]

I'm guessing what happens is that there's a race condition where the payload advancing triggers before the node, and the capture node triggers on exact match instead of match-or-exceed, which results in the node not triggering cause the payload has moved on by the time it's check happens.

[@TheFlyingSailorYT]

Great game, Blizz. Game itself don't work but the shop works just fine!

[@dannycoleii1030]

im throwing all my comp games like this now thank you!

[@ericocypriani2308]

this looks easy to fix. The progress is still registered even if the checkpoint is missed. The game logic should define if the checkpoint was reached by checking the distance the payload traveled, if it is greater then the distance from the checkpoint to the beggining of the path, then the payload reached the checkpoint.

[@baum279]

Blizzard should be paying this man for doing a lot of their work for them

[@Yoshiezzz]

I love this KZbinr

[@TonyBroski1]

I could listen to you explain Overwatch bugs forever

[@MrInfinityGD]

Payload just phasing through the door on Eichenwalde as the crusaders watch in horror

[@LuKiZenith]

Not sure if it warrants enough for a whole video, but Ive noticed a lot of "unspecified" tags relating to the point or payload instead of A B and C, usually when it says things like "attack point A"

[@myseglis]

The timing sounds so precise on this one it has to be one of the rarest bug, but losing a ranked game because of it would drive me crazy

[@whatis112-s2i]

what if you put mei ice block in between payload and eichenwalde doors as the payload skips the checkpoint?

[@SpacemanX6]

High quality content yet again! Keep it up!

[@facesareblurredx]

love the content. i sure hope blizzard can fix this quickly since they are focusing more on PvP. best company fr!

[@AegisAuras]

They need that manpower for the shop 💪😎

[@Skazzy3YT]

Excellent music choice

[@chairyoudown2656]

My theory for this is that the payload point travels beyond the checkpoint in one frame, thus bypassing it. Pushing the payload from its reversal state might "kickstart" the payload forward, making it travel further than it normally would in one frame so thats why you need pixel perfect timings.

[@johnhawkins5314]

If I had to have a *guess* at the reasoning I would imagine the logic is something like this Is Payload at Node? Yes. -> Node type check. Node = Checkpoint -> Is payload velocity >0 (i.e. is it moving forward). No. -> Do not change checkpoint captured flag. So on the single server tick that pixel occurs, the information taken is that the payload still is moving backwards (especially if referenced from the previous server tick), and the game thinks, oh it's just rolling backwards, don't do anything

[@johnhawkins5314]

If it used this logic, then maybe a potential fix would be to wait for the next server tick and re-check the velocity

[@ahmadaldajani2059]

Can you please explain how the ranking system work : like why do the rank get lower with 5 wins and two losses

[@ThemGrayJackel]

This channel is amazing. It’s basically 101 reasons to not play OW2 and still have reasons to laugh at it.

[@Buglin_Burger7878]

The reason it works is likely due to order of events and how the payload doesn't immediately capture the checkpoint. The payload basically appears to stutter on a checkpoint. So when you stop it on the checkpoint before it captures it the payload has to check for movement direction to figure out if you should or shouldn't capture it. If you time it perfectly though it will falsely get "backwards" but them move ahead. By doing this the game likely either thinks the payload is rolling back and was stopped by the checkpoint or that it never was able to catch it since the area to catch it is too small. Heck it could even be as simple as the checkpoint is literally between units so this tick perfect timing allows you to move through it.

[@ItGirlCaliiopx]

Daddy Marblr is back with another banger

[@DetroisHD]

the timed hazey maze theme on the intro is a 10/10

[@ScrDesh]

ik the PvE news is devastating but you're like the best OW content creator pls dont leave 😭😭

[@Omaryllo]

As a developer I can take a guess as to why this happens: I think the key is that the payload moves slower backwards. I believe you have to touch point the exact same tick that it starts to move backwards. Depending on the code, it may mean that it will move both backward and forward in the same tick. Depending on the code again, I believe the most likely occurance is that it's moving forward beyond the checkpoint, then moving back slightly, but still beyond the checkpoint, then finally it checks for whether it has *crossed* the checkpoint. With this order, the last position update was backwards, but just barely not across the checkpoint, which means it didn't cross the checkpoint in the last positional update. Checking whether it has crossed a checkpoint involves checking whether the checkpoint is within the current and last position. This is a code quality bug. The correct way to check for a checkpoint capture would be to check whether the checkpoint is between current position and the position at the end of *the last tick*, but another way to mitigate this is by simply not allowing multiple moves in the same tick, but that's likely not intended behavior. Then some actions will have to take priority.

[@sselesUneeuQ]

7:00 at this last part I got a couple second skip on the youtube video as you said "perhaps" and had to go back to see if it was the editing or just youtube being weird.

[@Kiwi-lj9jc]

thanks for the heads up marblr

[@lastnamefirstname8655]

cool little rare bug. thanks maxor.

[@gold7152]

Babe wake up, Marblr uploaded

[@scu5555]

Didn't this happen in OWL a few years ago? Back then no one could explain it. I remember Super looking at a clip from the OWL match on Eichenwalde. Back then he was still on San Francisco Shock.

[@lygmabalz8540]

ah yes, OW2, a great game with no bugs and an amazing PVE that exceeded everyone's expectations

[@christopherg2347]

6:56 This is not a "perhaps". That is a certainty. "Everything that can happen, will happen - if you program is just run long enough by enough people." This is murphys law, applied to questions of concurrency.

[@jazzmane6171]

The gate trick in dorado happened to me once in competitive, we were all confused as heck

[@Netryxa]

I love your work, please keep it up !

[@vulpixie__]

just had smth similar happen on a capture objective, where the maximum-progress tick got stuck at 67% even though the enemy team got ~89%.

[@cactuscian]

Blizzard hire this man

[@jolness1]

Love the ssbm music you use.

[@shipweck6253]

if i were to guess as to why this happens, it would be because the cart "lurches" when it starts being pushed after rolling back. I have no clue how the inner working are coded, i assume that a velocity is applied to the cart to roll back. Once a player begins pushing the cart, a push velocity is applied. If this happens in the same server tick, maybe the velocities stack? or override each other? or the cart realizes it should be getting pushed and "lurches" forward, pushing itself over the checkpoint and this skipping it? Im curious what others think, please tell me your theories

[@ericocypriani2308]

00:01 me in driving school