I Watched 200 Overwatch Team Fights and This is What I Learned. Tips and Tricks

[@Starmojii]

interesting! Your objective approach to analyzing Overwatch gameplay is refreshing.

[@ubadman1]

He has mathematically proven that bronze is playing a different game. 😂😂

[@kelvintraylor7551]

😭

[@arthurlloyd5165]

TLDR: wait for your f**king team lol

[@kmk66666]

It sounds so simple but so many people struggle with it lol

[@arthurlloyd5165]

@@kmk66666 people just have the psychology of "we're fighting for that point" or "need to move the payload to win" but you need your team to do that, and you don't help, only hinder when you rush to the point only to die in a 2v5 and make your team gave to wait EVEN LONGER to get a meaningful team fight in. Definition of insanity.

[@joelalexander928]

​@@arthurlloyd5165 it doesn't help that there are automatic voice lines which encourage stupid objective play

[@joelalexander928]

"we must stop the payload" - while whole team is dead

[@mylegguy8115]

Ow2 dev team must add automatic voice lines that say “we must regroup together, we are stronger in numbers!”

[@darkcircle899]

Super interesting video there aren't many people bringing data analysis to improving at OW. It definitely reminded me just how important ult timings are to wining fights and ultimately games.

[@leonzchannelz]

So group up, soft reset if the fight is lost, get the first pick, use more ults than the enemy, use your ult first in the fight

[@kmk66666]

Pretty much! And if you think you’re already doing these, check the replays to see if you can do them even better

[@gaboqv]

I would say using ult first is the only not gertain thing, unless your team really capitalizes

[@kmk66666]

@@gaboqv yup! That’s why I think the percentage goes up in higher ranks, because they’re less likely to waste the ult

[@chuansonn7617]

There might also be some bias. Fights in which a team used ults first probably correlate with the same team using more ults than the other one. So the causality becomes a bit more murky.

[@kmk66666]

@@chuansonn7617 I still think this means you should ult first more often in general. I believe you are correct as a team fight might go red>blue>red>blue>red in a fight, which according to my data would mean red is more likely to win the fight. It also gives you a good chance to get a first pick by using an ult before them, which gives you a good chance to win the fight. If you always ult after they do you aren’t being proactive enough to get the first pick

[@pedromorgado628]

When data science meets gameplay. Amazing video and Idea, and that's a great effort to help others understand what can be done.

[@lucasschuh4009]

I work in sports and love sports analytics and always wanted some type of it for overwatch, and this is it. Keep up the great work

[@kmk66666]

We need some baseball junkies to get more into gaming to come up with a bunch of new categories to track 😅

[@KKJazz]

I beseech the almighty algorithm to take this to all my teammates in the metal ranks 🙏

[@kmk66666]

lol yeah it can be difficult, especially with most people being out of voice these days

[@heavenheathern]

I've heard the "~80% winrate when getting the first pick" thing from Adder but now I have a statistical confirmation :D Gotta keep this video in mind when I'll get back into the game, thanks

[@kmk66666]

Yup! It kinda just checks things we already knew, but it’s good to have that confirmation

[@skyalexander8361]

I really like the charts and graphs, shows the gradual changes from bronze to GM very well. Hopefully more people get to see this great info!

[@SenatorSlime]

Biggest thing ive learned from high level scrims is that going first is the most important thing to do 90% of the time

[@Switch_X_Back3884]

Biggest thing. Survivability. Yeah, I noticed how when higher elos get a pick they have that "Smell blood in the water" and push up quick since they have an advantage. They always take advantage of an advantage.

[@Taaxmaan]

thats not true. ive had a few games where we’ve won OT just because i ran in there, killed maybe a soldier or speedy character then got killed but still won. same with teamfights, if i kill one healer and know i can just full send the other one, kill them then die then we’ll probably win that fight. ive been gold and ive been top 500, survivability only matters really in master+ games

[@mylegguy8115]

In lower ranks players lack in capitalizing on advantages but also your getting 1-2 picks and dying so your clutching 3v2 into a 1v1 fight.

[@aeloh6921]

Great vids man, love the data driven approach

[@axel_r_]

Good job. KMK. I enjoyed this analysis, representation and conclusion of the data you have collected. You have some interesting key findings here that I will explore in my own gameplay.

[@dazey856]

I was not expecting this deep of analysis but this is the type of stuff I have been looking for, because like you said they for the most part what you expected and I think I know things, but there isn’t much data out there to know how true stuff like how important is getting first pick, it’s advantageous but how much of the time do teams win from it. I like math so this video was really good

[@shariqarain]

Another thing to note is that many people that have ults that have potential to get lots of kills (reaper, hanzo, genji, and most other dps/some tanks/some supps) will end up waiting a long time to use their ult for that “perfect moment” where they’ll get a quad kill potg, when in reality that reaper or genji only need 1-2 kills with their ult in an even 5v5 fight to win that team fight and get free objective progression. That’s why farming ults especially as dps is super important, as more ults, and especially ults that matter (5v5s) can help win more fights, so rather use them more often if it means a free kill or two than wait for an opportunity to get more kills that probably won’t happen often, if, at all.

[@501Labsmusic]

Me being called a try hard for spamming "Group Up" and "fall back" by the lone DPS player in full attack mode on the frontlines by himself; 5 seconds before he dies

[@starblaiz1986]

This is a great video and really solidifies with data things that a lot of us do or feel instinctually. This is definitely some stuff to focus on 😊 Now the next question is - how can you encourage your team mates to do the same? Spamming "Group Up" helps with the grouping up thing, but people never seem to listen to "Fall Back" and stuff like that.

[@kmk66666]

Yeah it’s tough without people in voice comms. So the best you can do if solo is work on forming good habits yourself, and letting your team die if they refuse to group up. Otherwise make a stack and work on it together.

[@omegzero45]

Nothing like screaming into the voice comms calmly like Dumbledore is in the Goblet of Fire movie 😂

[@markhughes3483]

Get this man to 1000 subs , such a great video , the Ow community needs more of this data analysis, such a benefit to the metal rank players to more understand what’s actually happening during the craziness of an Ow match!

[@calebirwin3964]

I don't think anyone has been able to answer all these questions. Very well done

[@Chad.Commenter]

This is so much effort. Thank you for your hard work!

[@timmckim6549]

Great video!

[@profoundproductionss]

i watched 201 teamfights over 14 games

[@triantulas]

Yeah wtf trash video bullshit from the start Not a good start mate Straight out of the gate with a fuckin lie.

[@Waseemmmm]

14.35714285714286 Teamfights per game, so probably comp matches

[@bornsniper9531]

Got to remember those games are also divide. From bronze to gm

[@Tiamatria1]

I know I was hoping the sample size would be bigger. This isn’t very scientific

[@Waseemmmm]

@@Tiamatria1 Scientific is 9999999 teamfights, we'll do that just for you

[@fjeldmats]

These statistics are very interesting! I'm curios what the data might look like if you did the same analysis for OWCS games, i am assuming the trends would continue in the same direction for most of these categories, but it would be interesting to see how drastic the difference was, or even if there were any differences! Another aspect that would be interesting to explore is if some of the statistical differences across the ranks could be caused by the difference in team composition. For example could the increased time between fights and shortened team fight length be explained with higher ranks tending to play more dive while the lower ranks tending to play more rush or poke? Hope you do more content like this in the future :)

[@kmk66666]

Yeah I did notice a difference in team fights being slowed down a bit when there was a Mauga lol so I’m sure comp does have something to do with it

[@NightwishNick]

Good data and information to know thanks for compiling this, would also like to see even more detailed info like which role's ultimates would lend themselves to victory more often etc. :) Also love that you used Home as the background music!

[@bdtzmrv]

you seem really chill man, great vid and very interesting perspective

[@kmk66666]

Thanks! Working on the next stats vid now

[@JasonBoring]

Well this was an instant subscribe. Great content, more data please.

[@tykong55]

This should be a mandatory watch before queueing for any ranked match in this game

[@tykong55]

I swear my teammates in plat/diamond do not soft reset ever

[@rent0]

Oooooo yes please, I love this approach to overwatch. Always want more

[@synktrain]

Hey, just commenting to let you know that I really enjoyed this statistical breakdown format. It's unique and refreshing. Just subscribed to your KZbin channel, and I really hope to see more videos like this in the future!

[@F1R3EMBLEM]

This video is big brain. Literally melting overwatch down to maths.

[@akaiYaMa9]

Great video. Nothing better than data backed up with science rather than opinions. Would love to see more vids like this that go into team comps, hero matches ups and win rates. Too many inaccurate "hero tier lists" being uploaded by biased people and the community just accepts thats what the meta is cuz some streamer told them. Anybody remember how moira became meta when the devs said she had a high winrate this season and how she was considering shit tier before that?

[@Desmond-Dark]

Data analysis ftw

[@trentmcintosh4923]

Love data driven things about improving great video

[@TheJonnyCon]

The power of stats! Thanks for sharing your insights, man, I appreciate it!

[@jfm3178]

This is awesome, would love to see more videos like this. Well done 👍

[@ThePaulGomes]

My strategy has mostly been to “group up and go for the supports”. If enemy team has no supports to heal, you increase the chances of winning most team fights. So I mostly try and have my team keep pressure on supports by diving them.

[@dr.flatface6276]

This is awesome, I wanna see what other pros think of this video

[@Jensen23]

Very informative information! I will try to improve my skills through this video and hopefully make it back to gm!

[@yugshende3]

I think a more useful video will be how you to the best of your knowledge accurately tracked these stats. I’ve been doing some how did I die tracking of my own recently and it’s astonishing how op sombra is in the lower ranks especially when you’re on support. I came back after almost 15-18 months of hiatus and used to play OW1 a lot and OW2 is a completely different game now

[@ABadGamble]

high quality analysis. great job!

[@ethidian3444]

This is the best Overwatch content

[@michakoaczkowski4878]

bump for algorithm. Very well made video!

[@ALLinDRAMA]

Cool vid idea

[@Joshh.K]

I think the high "team fight win rate when ulting first" in GM and Masters is probably because the skill gap is very close at the higher levels, people making very little mistakes that it takes something that guarantees something to actually sway it one way or the other. This is really interesting too, 75% is a HUGE amount, and I guess if you was creeping into or trying to retain those ranks you'd want to focus on that if you're already close.

[@mcpremier3544]

bronze only being 52% first pick winrate is CRAZY lol

[@weeblyrobo2325]

I like the video, gets the overall data and point across very well with evidence, which I rarely see in general. A quick question, could some of the data be skewed by some outliers, and/or be further accurate with more game analysis?

[@kmk66666]

Yup! It could always change with more games looked at. They just took so long that I didn’t wanna do any more 😅 the team fight length can also change based on meta. Dive being faster, Mauga mirror taking longer, etc

[@kmk66666]

But overall I think the trend lines will stay going roughly the same direction

[@weeblyrobo2325]

@@kmk66666 I can see that being the case as well. I wouldn’t want to analyze another game knowing I already watched 14 other games

[@liamwylieMMA]

I’ll glaze for algorithm keep it up ❤

[@T1J]

very interesting!

[@illestvillain1971]

Hush, don't expose our nerd corner 😆

[@tylerbacoka4248]

great video dude, smart man, at least we have one....

[@sageraptor12025]

Amazing video homie

[@Hiro6543]

Tank here. I often tell people that i spam "group up" and "fall back" for their own benefit. This provides data for that.

[@fractalelf7760]

It’s remarkable that people love to trickle in… I’ll never understand the lack of self-control. You can scream it on voice and the same smooth-brains will run right in as you’re speaking.

[@ethanwhite82]

This is an amazing video, thank you :)

[@leethomas1400]

Please condense this to a short so i can play the audio in voice chat in my gold games

[@kmk66666]

Lmao just record yourself screaming “group up” and put it on a soundboard, ez

[@DiamondTurtleGamer]

This kinda just visualizes what people in gm already know, which is really funny how people figure out these types of things, without actual data (previously)

[@brian856]

Bump for the algorithm

[@FullMetalSheen]

I love watching my friends screenshare their metal rank games on discord so I can talk them through and give them in-game coaching. The thought of using a support ult first simply to engage hasn't crossed my mind in a long time. Nowadays I'm used to that sexual tension between egotistic Lucio players who don't want their beat forced first.

[@d1o2c3t4o5r]

The best way to climb in Overwatch is to beat your rank's meta with a more effective tactic, such as the real meta. Bronze, silver, gold and plat struggle with incorporating map cover into their gameplay, so their supports get taxed too much. A very effective tactic is to choose a hero that excels at standing in the open because metal rank teams usually play like that.

[@RelexGG]

Guys remember that some of these stats are skewed as it doesn’t take it account the effect on the next coming team fights eg the clicking more Qs than the enemies in one team fight

[@kmk66666]

? I looked at each fight individually. If one team used more ults than the other in that fight they won that fight 70-80% of the time. This has nothing to do with any fights future or past

[@starblaiz1986]

I mean, technically yes, but the data of "you win more team fights if you use more ults" is still valid. If you use more than the enemy team one fight, you are more likely to win. But then next team fight the other team is likely to have more ults and so they will be the ones to ult more and therefore be more likely to win according to the stats, which makes sense. Like, I get what you are saying, but I think you are overthinking it and tying yourself into a knot as a result.

[@kmk66666]

@@starblaiz1986 exactly. I talked about this a bit when I mentioned that you can turn a fight around 4v5 by ulting more than the other team, but you should do this for important fights because it’s only a slight advantage and puts your ult economy in a bad spot for fights later in the match.

[@RelexGG]

@@kmk66666 Yes but as i said your stats dont take into account the hidden variable which is ult economy, next coming fights if your team only has 1 ult and the enemies have all 5 u will lose a couple of team fights in a row until u regain ur ult economy. Im not saying that your statistic is wrong, but the whole picture is skewed as you might lose more fights if you invest too many ults in a single teamfight.

[@RelexGG]

@@starblaiz1986 i guess it depends on what level you play at, as staggered ultimates can give a lot of ult charge to your teammates because you just run over the enemies in that particular fight, which means u pretty much recycle ultimates. But yeah it might be too much for just casual ranked.

[@davidw789]

While I knew all these things in theory, having a video with stats tell me means I'm now more likely to think about these while playing and hopefully winning more

[@TheHmongol]

The one thing that urks me the most is that people don't pay attention to kill feeds. So many times I find (mostly my tank) still strafing and trying to LoS at entrance when our teams picked off 2 enemy players. Or when I'm tank and we pick 2 guys off so I rush in only to find my teams are still keeping their distance . I don't understand why it's so hard to understand that overwatch is heavily based on the numbers game. Take advantage when you have the number over the enemy. There should be no reason your team is still hanging around the choke point or push point when it's 5v3 or less. When enemy teammates are picked off and you rush in. Almost almost all the time the rest panicks in fear and disorientation. This results in overwhelming them and getting easy picks and it also separates their spawn time so they cannot regroup or rejoin the fight.

[@mitchellhall847]

Good analysis!

[@scoutee]

Great refreshing content

[@DanielFlores-fo1ee]

Say what you want about this game, but there is no other game on the market that will give me that rewarding pvp experience that ow gives me. My favorite pve game of all time and it’s not even close. Maybe league ig

[@musicxmp3533]

Great video, very interesting! Thanks for the work! I have one question that has very little to do with the content of the video, more out of curiosity as I see it in so many videos and from streamers too - why do you call it "Masters" and not "Master" but then the s is dropped for "Grandmaster". As a non-native speaker i'm just curious if there's some logic behind that or if it's just because so many people call it that.

[@kmk66666]

Honestly, I’ve never thought about it 😅 would have to ask somebody else. English was always my worst subject in school lol

[@musicxmp3533]

@@kmk66666 Alright, thanks for the reply though and again, love the video!

[@StephanieMiller-r3m]

​@@musicxmp3533Honestly, as a native speaker, it just sounds.. better? It isn't exactly correct to call it Masters, but many native speakers add the s, including myself. English is a language with many rules and exceptions to those rules, sometimes the exceptions that are used aren't the "correct" way of saying it, but when multiple people start to use it and it sounds better, it becomes normalized. That is actually how many modern English language rules have been created.

[@StephanieMiller-r3m]

​@@musicxmp3533Also, master is a word with multiple meanings, and is used in some.. different context at times. Calling someone a master can sound weird with the way it is often used now, but saying someone is in Masters can't really be misinterpreted as easily.

[@musicxmp3533]

@@StephanieMiller-r3m that makes a lot of sense, thank you!

[@UNWANTEDMINDS]

awesome content bro

[@TrentTube]

great content!

[@teddyispg]

this is pretty good data showcase of why metal rank players seem to struggle alot. even if the data is not 100% correct, the analyzed outcome is. chances are, that even with this data and conclusions, most people will be hardstuck gold anyway.

[@MrXxblackopsxX]

While this is a lovely idea for a video, the data is from a time period in overwatch much different to where we are right now. This patch has really changed the way that each role is played, and the game as a whole. Would love to see this same video in different patches? Maybe every year?

[@kmk66666]

A small Mauga, Sym, and Sombra nerf did that for you? The games were from right before and right after the mid season patch, and right before the Mauga hotfix.

[@vcp2077]

Great analysis. I just want to know if you used data from just NA region? Asian regions have a much different skill rating in bronze and silver. I was in mid-upper bronze for a long time before I got bumped up to silver and gold in recent seasons and its not much different here.

[@llllTatollll]

Crazy how all this boils down to is 5>4 and 99% of the player base doesn’t get it

[@Blunkfist]

You're taking a lot of correlation for causation. For example "first ult wins": They don't win because they ult first. They ult first because they win. Almost all ults are offensive in nature and even the defensive ones bar Zen are great for using them offensively. If someone uses the first ult of the teamfight it is most likely because they saw the potential to get a pick and committed

[@kmk66666]

Exactly, and higher ranks are better at doing this thus the higher rate. Either way the team that ults first wins more often no? Im not saying people should just go and waste ult, I’m telling them to be more proactive with them and look for chances to use them instead of holding onto it forever like low ranks

[@amoblahblah]

Great video!!!!! 🙏🙏🙏

[@overratedfool6900]

Nice video. As somebody new to OW, I think there's something about having some stats to back statements up that makes it sink in that bit more. I don't want to be the guy that blames his team, but bronze three is an insulting rank. Match wins seem way too correlated with the team (tank particularly) that ran in by themselves less. The amount of games I've played already in the few weeks I've had the game where we'll lose and I'll have the highest dam, elims, healing and least deaths on the team. As my multiple endorsements roll in I'm just sitting there wondering how my mate who I've hard carried to hundreds of WZ wins who started playing OW at the same time as me is silver. I think the ranking system is just not accurate enough at bronze. I think that's a large reason why the bronze stats seem so random. The games are very random. I've literally just played 10 games total of DPS (ranked and QP) and it just placed me in gold. There's literally zero chance that my Reaper is better than my Baptiste lol. The matchmaking is all over the place.

[@katyperryaf8876]

telling myself its a zoning ult is how I sleep at night.

[@wanderer316]

BIG LIKR👍 Thank you for you service, spending your time to research ow2 to be better for us.

[@willthethrill8661]

I love how his video is of him with pocket mercy. Like obviously you gonna be running.

[@kmk66666]

Not the point of the video but okay 😅 didn’t wanna put a game where I got crapped on

[@Frostedhuh]

Wish I could send this video into my teammates brains before every match lmao

[@bradyfalor5122]

Damn I did so good following this advice holy moly

[@TheShapingSickness]

If several of you die, don't leave spawn. It's of no use, you'll only get dove by a Dva and a sombra

[@lukasg4807]

Is that 4v5 statistic counting a loss as the enemy team soft reseting? Because the enemy choosing not to take the first fight might skew a 60-40 to a 80-20. The same question applies for ulting first etc.

[@mihumiz.ow222]

You’re the goat 🐐

[@KX_OW]

is the match you spectating on console ? or does that Sojourn have a soft aimbot ?

[@LEWIS1992]

I would love to hear the stats based on 100+ games. 14 games is such a small sample size.

[@kmk66666]

I agree, it just takes like an hour per game and I work full time 😅 once I hit 1000 subscribers and can make some money off of the videos I may be able to justify putting more time into them

[@davec8473]

I think if anything I'm too quick to give up on a fight if we are down one. In the past I've retreated and then realised that actually my team is turning it. But overall I guess my always regroup mentality has prob been beneficial. Something to work on I guess.

[@kmk66666]

Yeah I think it partially depends on your positioning. If somebody gets picked right outside of your spawn doors it’s best to wait, but if you are already fighting on point it might not be a bad idea to fight it out

[@gettem6341]

it be interesting to see how many flanking will contribute to a won team fight

[@SëriålHëålër23]

Great data thanks for sharing 🙏

[@Skatche]

Brother, you missed the most interesting part of your statistics. Almost all of them trend upward from bronze to about gold; dip down until about diamond; and then rise again. That suggests there are several distinct stages of development in climbing from bronze to GM. Like for example, maybe you first learn mechanical skill, which allows you to get fat ults on the enemy team, so the first ult percentage goes up. Then people learn better positioning, so the chance of getting fat ults goes down again. Then you learn to manage your ult economy, baiting out a few enemy ults and losing the fight so you can win the next fight easily with only one ult. Not saying that's the actual explanation, only that there's some multi-meta thing going on.

[@dolopo9976]

If you’re shooting an enemy, and you get their health low PING THEM and someone should focus them. It’s a domino effect after you kill one.

[@peepeeking483]

Make this video famous

[@griglog]

Cool, but... Did you have an advantage because you did the first pick or did you do the first pick because you had an advantage? (Correlation does not imply causation)

[@zc8673]

If only my teammates would watch this video

[@Daddelmeista]

Amazing, things like this should be taught in OW school or at least more visible in-Game. Who cares about dmg numbers anyways? (gold player: side eye)

[@mikeruano4858]

how do you group up when people don’t understand the message group up?

[@kmk66666]

Play in a stack or just make sure you aren’t contributing to the problem yourself if you solo. It can be tough in lower ranks

[@AntLeonardi01]

Playing DPS & having good mechanical skill is the only way to win consistently. Also I wouldn’t soft reset over 1 death, but 2 or more of course. Can easily overcome one support or dps dying early though

[@kmk66666]

I’m a console player and I decided to play PC for a week and one tricked Moira to masters 3 with a 75% win rate. And all I can say is my mechanical skill is pretty mid on pc lol so there are ways around it using specific heroes and having good game sense. Hell, I climbed to GM for the first time on Torbjorn season 13 OW1 because he wasn’t mechanically demanding and I was still pretty new to the game

[@patrickream8062]

The stats at the beginning tell me people in diamond are bad at the game

[@illestvillain1971]

Legitimately in ow2 they probably are

[@Stanky_Foot]

diamonds are funny, mostly big egos but no skills to back it up, respect if you dont have a lot of time in OW2 and are diamond, but if you've been playing since OW1 and have thousands of hours on the game and think you're hot shit compared to the rest of the lobby, the joke is on you

[@SehnsuchtYT]

It's very difficult to do this in lower ranks when your teammates just don't listen or care

[@brian856]

Skill issue

[@SehnsuchtYT]

​@@brian856 This would actually be the one instance in which it ISN'T a skill issue, but nice try

[@brian856]

@SehnsuchtYT cap. If you deserve to climb you will. Just simply get better at the game and you be in gm lobbies.

[@SehnsuchtYT]

@@brian856 I never said I wasn't climbing. I didn't even say what rank I was!

[@brian856]

@@SehnsuchtYT your initial comment would infer your rank is low sr in the metal ranks

[@MrJoaoGarrett]

What Souljourn skin is that?

[@kmk66666]

Uhhhh default I think? lol got the gold gun though

[@poolmanbrown6459]

Good stuff