Aiming, Fast or Slow? - A Scientific Analysis
Вставка
- Опубліковано 21 лис 2024
- The short answer is "slow", the long answer is... a chart.
Support me on Patreon: / resprit
Twitter: / ggnoresprit
Music List: docs.google.co...
Postmortem: / aiming-fast-or-97926020
---------------
Ah, yes, the bottom of the description. I still don't know how many of you read this, presumably not a lot. But some of you do - neat!
Most of my thoughts on this video are summarized in the postmortem, so you should read that for more behind-the-scenes stuff. This video was supposed to be short, but instead it ended up being a large project!
StarCraft is next.
![[chess]puzzles, special rules|#chess](http://i.ytimg.com/vi/4oiuQYxsYH8/mqdefault.jpg)
Man casually drops the most in depth and informative, math-based aiming video and expects the community not to notice
"First you get good, then you get fast" - gray beard maxim
The blender animations are so cool holy moly
Still waiting for that part 2 of the starcraft documentary. Nice video like always though.
It is next :)
@@RESPRiTYay!
@@RESPRiTWOOT
Awesome motion graphics and great analysis!
I think you are the most underrated UA-camr I have ever watched and expect you'll do very well if you keep rolling the dice.
Like always, looking forward to your next video, especially the StarCraft series!
These are really high quality math visualizations for conveying the intuition. I also really like your little sketchy illustrations and faces. Looking forward to whatever you have coming next!
I wonder how the graph changes if you were to consider things like fire rate, because an awp duel is won by the first to hit, but since it takes so long for the second shot to fire, it's almost guaranteed you'll lose, but when you have something like the ak from cs, you can miss the first shot, but if you manage to shoot another before your opponent you'll win while being less accurate than 100%, iunno if what i said makes much sense, or i didn't pay enough attention to the video :P
Seeing the production of this on Patreon was so informative and eye opening. Thanks for your entire process in describing the details you see. ❤
seeing a geometric series equation pop up had me going "omg omg omg i know that one" at the screen hahah, great video I love everything you make
This video was super fucking cool to watch for such a nerdy and niche topic, even by your standards
MY GOAT FINALLY RETURNS
The quality of these videos you just drop out of nowhere is damn near unmatched. Well done again, RESPRiT! 😁
Honestly feels like the beginning of a legit scientific paper
He used to be a phd student lol
Cool video, and great graphics and animations! I would have appreciated a different color scheme for the Win Rate By Accuracy Percentage graphs; Red-Green color blindness makes it pretty hard for me to get any insight from it. Maybe monochrome next time?
haha it's great to see this comment because I am actually also red-green colorblind, and you're totally right - it's something that I didn't think to change until it was too late to, but I definitely had the realization of, "whoops this is definitely not colorblind-friendly" while working on it!
Love the graphics good job
The tribe called quest reference at 1:21 is appreciated. Great video!
really interesting math, I love the way you quantified this problem. just really awesome overall
This video is incredible. The production. The humor. The perfect use of resources from Baba is You. Easily the best video I have seen all year. Please keep it up! 🎉
The quality of your work is always very impressive man. I believe in your growth.
Accuracy is likely not linear, at the absolute fastest, there is likely a "floor" of accuracy around 30-50%, and then an asymptote of accuracy where spending additional time quickly reaches 80-95% range, but adding that same time again only goes from 95% to 98%, because its not possible to have over 100% accuracy, and likely never possible to truly have 100% accuracy, particularly against a moving intelligent target.
We live in the 3D. When it comes to 3D shooting. Not a lot of things are linear, and our movements have depth. I believe that through superior positioning, we can overcome a shootout.
But if you spam enough you introduce reload time
The presentation here is incredible. You’re making me want to pick up Blender.
Criminally underrated video
RESPRIT VIDEO ON AIM THEORY!? LETS GOOOOOO
goddam another absolutely cracked RESPRiT vid
wouldn't have watched this if I didn't see who it was uploaded by
did not disappoint
I love the mixture of Baba is you music and the squiggle vision grease pencil animation, it matches really well.
Thought lost this gem of a channel. Good thing can just always search for your sc vid.
Bro this video is so good
I had this theory. It is what I used for creating my book when it comes to action scenes. I use position as the ultimate factor in deciding the fight. Despite someone skillsets. When it comes to my protagonist, he's one to take an advantageous position every time. He uses accuracy and superior positioning combined to best his foes. In shooting games, it's sometimes best to get a better angle when shooting, instead of just alerting and missing your shots and running after the guy who knows now where and who is shooting at him. Which will end up with him killing you just by you running into his area.
GOOD 🅱️IDEO GOOD 🅱️A🅱️A IS YOU
such a cool video… i know this took tons of work, great job!
You know a vid is good when it drops that Baba is You thinking music.
I love that the result of the simulation is intuitive to understand - I always struggle maintaining the discipline to hold off on a shot to get better accuracy. Looks like I'm in the beginner category!
I'm not much of a tactile shooter kinda guy, i rock shotguns and SMGs like nobody's business. I treasure movement, how many shots, and/or kills I can get in before dying!
Though it is best to know *how* aggressive you need to be whether it's with a teammate or going in solo. It's just one of those things you start thinking about after playing any game over time.
So sick. Gj. Informative and entertaining. Great visualization of formulas and their relationships
the animation on this video are amazing
Lovely video! If you ever find yourself considering a companion follow-up to this video someday, I'd be interested to see some asymmetrical analyses using your formula though-specifically, I wonder if you could work in, say, an Awp/Op's ability to kill from a body shot by increasing W, and seeing if the final conclusions drawn are different if you're sniping vs. an AK (or equivalent).
the fun thing about science is that the scenarios are so wild and unappliable to actuall stuff. when i play shooter then i dont see my opponent most of the time until they hit me first or vice versa. that is because im bad and dont play much
Man, your videos are so great. Keep it up!
I've never really thought about aiming in a game theory context, but it's a great idea. It would be cool to see if anything else interesting comes up if you give one person an advantage over another, like a head start in terms of distance and/or time to move their crosshair
always happy to see a new respirit upload :)
god ur such a good artist dude
Now the next question is. What happens to the table when you factor in shots that aren't going to kill instantly? You could probably make an even more complex formula for that if you make it modular enough to calculate with different weapon accuracy, firerate, and damage values in mind.
super cool video, the visual presentation was lovely as well, only issue is (maybe) being too fast for those of us who eat while watching youtube and don't want to miss anything while looking down to pick up a chicken nugget
This is one of the best videos Ive ever watched.
where did we go?
absolutley goated vid top 10 on the site imo
Now, let's just apply this math to fix Storm Troopers' aim and Bam! Instant empire.
this is such a fantastic video!
Very good analysis! Do you have the code for the simulation available somewhere? I would love to mess around with some of the variables 😄
I'll get it up on GitHub later this week! Fair warning that the code was not written to be public-facing though haha
@@RESPRiT Thank you very much! No need to worry though, we all have code like that 😅
@@Brenopms I know this is super late but I just randomly remembered this comment haha, sorry it took so long: github.com/RESPRiT/1shot
My favorite content on YT as always
Very cool video
me getting jumped ingame: "OK, remember 65% accu-" splat
finally, ANSWERS
Great video and animations.
It would be interesting to ask the inverse question. "How often should I be shooting first?"
Both first bullet accuracy, and first shot frequency are the easiest metrics to use to determine shot quality.
Your video does a great job at answering the first question, but for the second question, I'd assume you'd need some data to determine the first shot speed at various ratings.
Incredible video
young king
Wow what an amazing video
excellent video.
The absolutely most important thing you overlooked is that enemies will not stand still.
When I used to be one of the bets snipers in a niche game you never heard of that went and died in microtransaction hell ...(yeap) I sniped by basically drawing a line from my current crossair position, far past the enemy, where the line would just go across the enemy head. And I would 'aim' by just shooting at the right time.
If instead you tried to actually aim for your opponent, they could just dodge and now you have to reaim, or they could move unpredictably and you wont be able to even keep your crosshair on them. A good example I have from the same game, one time I was stuck with a shotgun that had to reload against one of the other best players at about 3m distance. I dodged his entire assault rifle magazin (~6 seconds). (He still killed me even after I could finally reload 🥲)
So what im saying is the best 'aim' is to eliminate as many variables as possible. To do that you have to be VERY fast.
amazing
So glad i subscribed
Wow this is so cool!!
lets hope this one gets a second wind in the algorithm
Damn bro good vid
fascinating
damn just making me feel worse for my impossibly slow reaction time
Can you make a video breaking down how to made this? :o
Good video, very well executed and acknowledge your disclaimers. But, as a layman, I think its not quite right. I think this would only apply to semi-automatic weapons (seemingly implied in video too). Mostly tactical shooters revolve around automatic rifle vs rifle. This means that there is no “delay” between shots, ie if I an faster, i can miss the first shot in exchange for recovering for 2nd or 3rd shot. Realistically, there is still delay due to weapon RPM, but there is not as many discrete outcomes as this video suggests (rather continuous). I think this makes the issue a little more complicated, even if generally the finding stand. This is further challenged when considering body shots/no one shot headshots such as the classic M4A4 vs AK47 duel. In that case the AK probably wants accuracy to hit a one tap whereas the M4 wants speed to land 3 body shots as soon as possible. Again this is a nitpick but for counter strike at least, AK vs M4 is by far the most common fight. Anyway, still a great video, and very well executed
really good video, but what happend to the lil res podcast?
uhhh that was good
sick
love the vid
This is awesome O:
are you gonna submit this to the 3 blue 1 brown competition?
Can this be used to make a better SBMM?
This is actually accurate tho
Isn't there a problem with the geometric model. It assumes that both players would shoot the same amount in a given set of trials. But the faster shooter would actually get more shots off in a given amount of time. This lowers the expected value of faster shooting
Nevermind, you just addressed it
I guess I have to start aiming faster 😅
donkey kong
Collab with battlenonsense
speed and accuracy are treated in this video are treated as being mutually exclusive which should never be the case