erbgorre I think the AI is constantly calculating its expected chance to win, and how that would change depended on the moves it makes. I wouldn’t be surprised if the developers hard coded it to surrender when the win chance is lower than like 1%. (But don’t quote me, I’m not sure)
The AI's aim is to win. It might have the capability to surrender, but it will have learnt almost instantly that that is a guaranteed loss. It will never do that.
Alphastar: My first reaper always dies, so I'll make another... mwahahaha Thank you LaughNgamez , I really enjoy your expert level knowledge and dry wit. :-)
Yes or something like that. That cannon rush defense was perfect. But the follow up sucked. So it/her/him/they (the bot), probably did a follow up to cheese to cheese. I think the bot studied to many matches. Instead of just doing ladder high level games. This is a mix of the worst and the best of Gsl
If theres one thing i've noticed watching alphastar is that it is very good at controlling where its enemies army goes. dragging you back to your bace to defend and then it relies on killing the 3rd bace and then turtles up in hope for a gg or an over aggressive retaliation. in recon if you dont protect you mineral line too much and just keep it moving forward. like a zurg bace rush that would probably throw it off guard maybe.
This is a pretty good example of the advice you always give on cannon rushes when you complain people arent pulling enough workers. Alpa star pulls enough workers and is able to hold with the worste micro ever and scvs just standing for way to long.
So the GM went for fast +1 after cannon rush because he already had a Forge. That combined with Glaives to make Adepts trade well with Marines. Basically Adepts can 2 shot Marines with +1 so that +1 attack is like a 50% damage buf. Alphastar went +1 shields as an attempt to make Adepts have to 3 shot marines.
If you guys are wondering why alphastar doesn't say "GG" at the end, it's because all chat text had been scrubbed from the replays. In the actual game, alphastar does say "GG" and "gl hf"
The thing with calling the human player the "Terran" was pretty funny! You could have also called Alphastar a Purifier if you wanted to follow that theme :)
13:10: 10 siege tanks. 8 in the natural, 1 in the main. Alphastar with the truly groundbreaking strats here! (but seriously though, thanks for the cast)
@@RedBelgradeStar I would like to play against it but haven't had a chance to so. Based on what I've seen and heard it seems like AlphaStar is indeed quite weak against cannon rushes.
Thanks for making this video. I just subscribed. Please let us know if we're watching AlphaStar FINAL, the final and best version of AlphaStar. You can see this info when you look at the replays that have been released by DeepMind. And please put the word "Final" in your title so we can search for AlphaStar Final videos, and not merely videos showing the older and inferior versions of AlphaStar. Can't wait to watch more AlphaStar Final videos!
After that one engagement, Alphastar Lost. No economy, lack of units. I understand why Alphastar kept going for hidden bases. It's not just hidden, you're taking your enemy's future expansion mineral patches. That way, in a long stretched game, your enemy can't have good economy even if they expand.
@@filipeguimaraes3596 Dont look at the comments section before watching the video men, what did you expect, that there wont be at least 1 guy that would make a spoiler? You lack patience, It's your own fault for being too curious and look in the comment section before the end of the video :)
A good cannon rusher would have made more probes. He only had 16 when ask was said and done. Basically no economic advantage, even though he killed 6+ SCVs.
I was a little worried when they announced alpha star that i would come to find that this game that i love for being incredibly deep can be broken by an A.I. relatively quickly..... im glad to find out that I had nothing to worry about XD its going to take quite some time and effort to get it to master sc2 after all. Take that machines! still cant beat us all at sc2 :P
Well they did cap it's APM, which is where machines excel instead of humans. Also I don't know how long alphastar has had to practice terran and zerg because it was build up on protoss first so that's where most of its experience should be.
@My Account by "reduce its capability" you mean: They capped its apm, and they didnt allow it to have omnipotent awareness (they made it actually move its screen around like a human would have to). But I would argue that is just making it actually play by the intended rules of the game. Itd be like if they designed a chess-bot that got to make 2 moves for your every 1 would that really be the bot being "better" than the real player? Also note that I said I was worried it would be broken "relatively quickly". They gon break it, I know that. The A.I. will eventually have enough data to make almost "perfect" decisions every time. Im just glad it wasnt that easy and it took (/is going to take, im not sure how good alphastar is currently) a lot longer than other seemingly complex games
Hint: it's not meant to play like a human, in case you thought we were the end game. AS will develop its own class of tactics (optimal order of mechanics) when its mechanics are matched that probably won't look human. Until then it is an aimbot.
Just so people know, this version of alphastar has an apm limiter and also only can view through the camera like a player, whereas the original unveiled version that nobody could beat had uncapped apm and also could view the entire map at once, hence its godlike micro carrying it's likely less optimal builds. Now with its apm limited and it can't micro perfectly these less optimal builds are getting exploited by players with more thought out builds. It is more at a master, or low gm level now, whereas uncapped it was at a pro level. Not saying I could beat it even in this state, just saying that's why it seems to be worse since the original show matches.
"what alpha star is: it is basically a super deep-minded learning program like ai" - yeah let's just link some buzz-words and hope it means something:) but nice video!
@@Broockle But these assumptions seem fairly accurate. They were statements about strategy, which is exactly where AlphaStar should have severe problems. AlphaStar knows way more than any human on how to micro and given that macro is somewhat of a misnomer (it is playing micro but on economy) it's also extremely good at that , but the hard part is the deep strategic knowledge, knowing when and where to engage, what army composition will work etc. For humans this is all fairly intuitive, because we are absolutely genius when it comes to pattern recognition. When intuition fails we can learn from each other. Watching tutorials or reading discussions on strategy will make out skills skyrocket, because we are actually able to understand that stuff or at least understand the underlying principles. A neural network AI will not understand these things. If its strategy is wrong, it will still try to fix it with better micro, more harass, better macro etc because it has learned that this is what wins him games against itself. It will have a very hard time assuming that it is simply wrong in its entire approach to the problem. DeepMind has been able to show that a neural network in principle is able to get to the point where it does seem to understand strategy a superhuman level, but this is in games where the amount of variables is fairly restricted, so it's really just sleight of hand. It's just tactics (micro) expanded to a level where it is indistinguishable from strategy for humans. Chess and Go are the games I speak of. Starcraft II is orders of magnitude more complex than Go and Chess at the strategic level.
It seems Terran AlphaStar still struggling with most basic things. Like building Medivacs, protecting economy, properly turtleing up, or knowing even how to respond to situations. "Oh, he has High Templar and Psionic Storm? Build more Marines!"
The protoss could have probably had a quick win if he built his first cannon on the low ground. Not sure why he didn't do that. Could have also won faster if he just freaking got a cannon in each expo. Still he played very well, and got a somewhat lucky fight to clinch a solid win.
I've seen maybe 10 alphastar terran games now. This agent (if there are multiple behaviors) always goes for banshee, marine, tank and tries to take hidden expansions regardless of what the opponent is doing. Needless to say, it sucks.
AlphaStars Terran seems to be the weakest from all of them. Maybe because Terrans are the most complex to play. But Terran was the only A.I. that Serral could win against while he lost 3 - 0 vs. Alphas Protoss.
Lol just watch the composition and say "well, let's go void rays for fun" :) Seems alpha star is mainly GM because player fail to recognize basic counters below master, after 10 years playing same builds :)
Yes, they limited it APM also forced it to play with human like camera and controls. The initial version didn't have these limitations and destroyed top players about a year ago.
@@chammy2812 sad that they limited the engine...would love to see its improvement and then seeing it compete vs other computer engines. Dont know why we dont see more of these..loved to see dota selflerning engine I think it improves the way of thinking about the game so much more....shure one might say its a cumputer dude we cant compare it to pro scene but I think we still can imply sth.
16:40 "And yes I did just assume the robot's gender" - Ha! Not impressed by Alphastar's play here. Looks very similar to how it played Terran in the 50 games a month or so ago - which is basically to camp at home with tanks and send small commitments out to do damage (bio medivacs before, banshees in this game). It's essentially hoping its opponent will a-move into a siege tanks or that it takes a downright amazing engagement when it finally decides to push with all its siege tanks.
If you look at other AlphaStar games cast before this latest batch, it does say gg when it loses. I assume the chat from the replays was removed just like the player names. Prevents any off-color stuff in Google's professional material, but with the side effect of making AlphaStar look bm.
@Cgi Dude Who is talking about build orders? I know they are important, however, you still need to scout. The cannon rush it was trash execution. Maybe the guy is really Gm maybe not. This "sad kid" thing is not offending, is a pathetic line that everybody say in the internet. You have to be more creative.
Pylon in the mineral field by a Grand master protoss? Huh?? And this cannon rush was executed nothing but silly. GM player are those pro players like stats and classic, will they cannon rush like this???
SkyNet (X) >>> Legion (X) >>> AlphaStar (possible future) ... any one catch this .... give a like .... also.... ive warned u guys.... be prepare.... D-Day is coming.... sooner than u think....
![Gucci Mane & Sexyy Red - You Don't Love Me [Official Music Video]](http://i.ytimg.com/vi/j7IBqyFKT4c/mqdefault.jpg)
deepmind employee: "so alphastar why didnt you have medivac for your marine push?"
alphastar: "Human life is meaningless. they are born to die."
jesus christ he was targeting zealots before stalkers also...
...what have we done?!
"let's break 20 likes"
RUclips algorithm: hold my beer
Came to the comments for this comment
"Im about to start this mans whole career"
Came to comment this comment and saw this comment
FIRST!
I didn’t come to comment but the I had to
youtube algorithm makes alphastar viral, i think the machines are LEARNING
Cross platform AI teamwork. We're so fucked.
It's 16 minutes into the game and AlphaStar has decided, "Right, let's build a medeevac"
the moment an AI accurately judges when to GG out of a game is when we should really start paying attention
erbgorre I think the AI is constantly calculating its expected chance to win, and how that would change depended on the moves it makes. I wouldn’t be surprised if the developers hard coded it to surrender when the win chance is lower than like 1%. (But don’t quote me, I’m not sure)
@@pigenator5814 Just needs to GG like Idra's algorithm mate.
@@joewarwick430 I really wanna see alphastar rage quit after seeing a hallucinated army
The AI's aim is to win. It might have the capability to surrender, but it will have learnt almost instantly that that is a guaranteed loss. It will never do that.
@@Asdayasman Thats not how it works thought.
The catch: GrandMasterPlayer was an Alphastar AI, too.
...and AlphaStar was a platinum league kid trolling.
Me watching this game: "Am I AlphaStar?..."
Alphastar: My first reaper always dies, so I'll make another... mwahahaha
Thank you LaughNgamez , I really enjoy your expert level knowledge and dry wit. :-)
AlphaStar Be like:
"I will take the top right corner or perish in the attempt!"
Obviously Alphastar learn how to counter cannon rush from bronze league heroes
Bronze league strats requires bronze league counters
Yes or something like that. That cannon rush defense was perfect. But the follow up sucked. So it/her/him/they (the bot), probably did a follow up to cheese to cheese.
I think the bot studied to many matches. Instead of just doing ladder high level games. This is a mix of the worst and the best of Gsl
when alphastar plays terran all I can picture is Archer yelling "look at it, it thinks it's people" hahahahaha
If theres one thing i've noticed watching alphastar is that it is very good at controlling where its enemies army goes. dragging you back to your bace to defend and then it relies on killing the 3rd bace and then turtles up in hope for a gg or an over aggressive retaliation. in recon if you dont protect you mineral line too much and just keep it moving forward. like a zurg bace rush that would probably throw it off guard maybe.
petition to add 3 alphastar AIs to the global finals next blizzcon, one protss, one terran, one zerg, maybe a random too?
When alphastar doubles down on another second base I was laughing so hard.
*double hidden base
This is a pretty good example of the advice you always give on cannon rushes when you complain people arent pulling enough workers. Alpa star pulls enough workers and is able to hold with the worste micro ever and scvs just standing for way to long.
would be a solid argument if he'd won.
It got the +1 bio armour because the protoss got glaives and +1 weapons, which allows fast 2 shotting of marines. +1 Armour stops that.
"AlphaStar just knows how to STIM and make Marines, which is good if you are playing Terran" lololol
That's seriously a bronze level cannon rush :P
I wonder if this was Beastyqt. I know the Deepmind devs were teasing him at Blizzcon for cannon rushing Alphastar.
Hey man, nice cast. Haven't heard your casts before but its always nice to find another good caster!
That awkward moment when your cannon rush ends up with 3 pylons and 0 cannons.
So the GM went for fast +1 after cannon rush because he already had a Forge. That combined with Glaives to make Adepts trade well with Marines. Basically Adepts can 2 shot Marines with +1 so that +1 attack is like a 50% damage buf. Alphastar went +1 shields as an attempt to make Adepts have to 3 shot marines.
If you guys are wondering why alphastar doesn't say "GG" at the end, it's because all chat text had been scrubbed from the replays.
In the actual game, alphastar does say "GG" and "gl hf"
When the second thing finishes I'm like, OMG AlphaStar just lost, then I realize it's a pylon, I'm like OMG Protoss just lost
Beckerzinho thing? Lol
The thing with calling the human player the "Terran" was pretty funny! You could have also called Alphastar a Purifier if you wanted to follow that theme :)
13:10: 10 siege tanks. 8 in the natural, 1 in the main. Alphastar with the truly groundbreaking strats here!
(but seriously though, thanks for the cast)
“Avilo Star” 😂🤣😂🤣👏🏾👏🏾👏🏾
wish I could thumbs up 100 times! lol I died when he said that.
I want to see all the alpha star games
go to deepminds homepage and you will be able to download the replays
I think this Protoss player was cripseil, the French cannon rusher.
I don't think so, this canon wasn't GM level
, Cripseil's cannon rush are just amazing
when are you going to play against AS? I think your cannon rush would destroy it easy
@@RedBelgradeStar I would like to play against it but haven't had a chance to so. Based on what I've seen and heard it seems like AlphaStar is indeed quite weak against cannon rushes.
Thanks for making this video. I just subscribed. Please let us know if we're watching AlphaStar FINAL, the final and best version of AlphaStar. You can see this info when you look at the replays that have been released by DeepMind. And please put the word "Final" in your title so we can search for AlphaStar Final videos, and not merely videos showing the older and inferior versions of AlphaStar. Can't wait to watch more AlphaStar Final videos!
Alpha star needs some manners programming in
it has been training on the EU ladder thats why ( ͡° ͜ʖ ͡°)
alphastar learned it doesnt need manners, probably learned that trick from its opponents
all chat was removed from replay
After that one engagement, Alphastar Lost. No economy, lack of units.
I understand why Alphastar kept going for hidden bases. It's not just hidden, you're taking your enemy's future expansion mineral patches. That way, in a long stretched game, your enemy can't have good economy even if they expand.
Thanks for the spoilers man
@@filipeguimaraes3596 Dont look at the comments section before watching the video men, what did you expect, that there wont be at least 1 guy that would make a spoiler? You lack patience, It's your own fault for being too curious and look in the comment section before the end of the video :)
Nicolas Buisson sorry dad, but yeah, you are right
"GrandMasterPlayer" truly deserves his Nickname.
FYI AlphaStar does not plan. It evaluates every position and decides the next action like nothing previously happened.
I have a feeling that he could win at the begining if he did cannon rush a bit better
Maybe it was just to harass and delay the economy, It seemed like a good trade regardless.
A good cannon rusher would have made more probes. He only had 16 when ask was said and done. Basically no economic advantage, even though he killed 6+ SCVs.
"Yeah 10 siege tanks, 8 in the natural, 1 in the main" lol
Got you your 20 likes man. Keep up the good work
I was a little worried when they announced alpha star that i would come to find that this game that i love for being incredibly deep can be broken by an A.I. relatively quickly..... im glad to find out that I had nothing to worry about XD its going to take quite some time and effort to get it to master sc2 after all. Take that machines! still cant beat us all at sc2 :P
Well they did cap it's APM, which is where machines excel instead of humans. Also I don't know how long alphastar has had to practice terran and zerg because it was build up on protoss first so that's where most of its experience should be.
@My Account by "reduce its capability" you mean: They capped its apm, and they didnt allow it to have omnipotent awareness (they made it actually move its screen around like a human would have to). But I would argue that is just making it actually play by the intended rules of the game. Itd be like if they designed a chess-bot that got to make 2 moves for your every 1 would that really be the bot being "better" than the real player?
Also note that I said I was worried it would be broken "relatively quickly". They gon break it, I know that. The A.I. will eventually have enough data to make almost "perfect" decisions every time. Im just glad it wasnt that easy and it took (/is going to take, im not sure how good alphastar is currently) a lot longer than other seemingly complex games
@My Account Yeah hopefully we do some more useful things with those before we get down to kicking everyones ass in sc2 XD
No human would ever play like this. Alphastar might have good mechanics, but it still has no understanding of tactics
The Protoss play is the only good one.
Hint: it's not meant to play like a human, in case you thought we were the end game. AS will develop its own class of tactics (optimal order of mechanics) when its mechanics are matched that probably won't look human. Until then it is an aimbot.
Just so people know, this version of alphastar has an apm limiter and also only can view through the camera like a player, whereas the original unveiled version that nobody could beat had uncapped apm and also could view the entire map at once, hence its godlike micro carrying it's likely less optimal builds. Now with its apm limited and it can't micro perfectly these less optimal builds are getting exploited by players with more thought out builds. It is more at a master, or low gm level now, whereas uncapped it was at a pro level.
Not saying I could beat it even in this state, just saying that's why it seems to be worse since the original show matches.
at 8:15 it burrows a widow mine where it wants to make a building :)
hey just got recommended this video and you're a really good game streamer! keep it up
Protoss player took forever to build cannons to protect his mineral lines, even when he was sitting on 1100 minerals.
what idiot would send more banishes?
"what alpha star is: it is basically a super deep-minded learning program like ai" - yeah let's just link some buzz-words and hope it means something:) but nice video!
ye he did make a lot of assumptive statements about deep mind.
It doesn't know this and doesn't know that...
@@Broockle But these assumptions seem fairly accurate. They were statements about strategy, which is exactly where AlphaStar should have severe problems.
AlphaStar knows way more than any human on how to micro and given that macro is somewhat of a misnomer (it is playing micro but on economy) it's also extremely good at that , but the hard part is the deep strategic knowledge, knowing when and where to engage, what army composition will work etc. For humans this is all fairly intuitive, because we are absolutely genius when it comes to pattern recognition. When intuition fails we can learn from each other. Watching tutorials or reading discussions on strategy will make out skills skyrocket, because we are actually able to understand that stuff or at least understand the underlying principles. A neural network AI will not understand these things. If its strategy is wrong, it will still try to fix it with better micro, more harass, better macro etc because it has learned that this is what wins him games against itself. It will have a very hard time assuming that it is simply wrong in its entire approach to the problem.
DeepMind has been able to show that a neural network in principle is able to get to the point where it does seem to understand strategy a superhuman level, but this is in games where the amount of variables is fairly restricted, so it's really just sleight of hand. It's just tactics (micro) expanded to a level where it is indistinguishable from strategy for humans. Chess and Go are the games I speak of. Starcraft II is orders of magnitude more complex than Go and Chess at the strategic level.
avilostar lol that got me
It seems Terran AlphaStar still struggling with most basic things. Like building Medivacs, protecting economy, properly turtleing up, or knowing even how to respond to situations. "Oh, he has High Templar and Psionic Storm? Build more Marines!"
Iridium I guess the learning algorithm agrees with toss players. Just make marines = gg ez
feels good to see a human winning from time to time
The protoss could have probably had a quick win if he built his first cannon on the low ground. Not sure why he didn't do that.
Could have also won faster if he just freaking got a cannon in each expo.
Still he played very well, and got a somewhat lucky fight to clinch a solid win.
You have not noticed AlphaStar lost Infantry attack upgrade in 16.25 cause his ebay was burning xD
20 likes for more AlphaStar
So... do we get all the AlphaStar videos now?
how to play with alphastar help please . I wanna know
if this was a GM player, the execution of the rush was super bad
yeah no cannon on lower level, only on upper.
they can't be all as good as florencio
Ive seen better cannon rushes from randos in silver
Terrans is usless race even for AlphaStar. It is fun how terrans army disappeared in main fight.
my face when I saw like amount after he said "lets break 20 likes" ...
against the banshee it would have been worthwhile to have 2 cannons protecting each mineral line
8:42 why takes the banshee no damage? I call HACKS!!!
The protoss lost detection while the stalker projectiles were mid flight. Yeah weird stuff!
GoOdy Style :) New Panzergeneral
Worst cannon rush ever? You be the judge.
I cant wait untill someone figures out an adversarial attack vs. alphastar, should be interesting or at least funny.
Where are these replays located? I have searched on the internet but haven't found it.
Somewhere on the deepmind website
It's on the deepmind website under research/open source. You need to download them as a zipfile.
Author, whatis your keyboard key switch?
It's a Razer Black Widow Chroma
wtf such a clean execution from the Terran player, its like its a fucking machine or something..........
I've seen maybe 10 alphastar terran games now.
This agent (if there are multiple behaviors) always goes for banshee, marine, tank and tries to take hidden expansions regardless of what the opponent is doing.
Needless to say, it sucks.
AlphaStars Terran seems to be the weakest from all of them. Maybe because Terrans are the most complex to play. But Terran was the only A.I. that Serral could win against while he lost 3 - 0 vs. Alphas Protoss.
Lmao @sucks, it had a gm winrate.
Lol just watch the composition and say "well, let's go void rays for fun" :) Seems alpha star is mainly GM because player fail to recognize basic counters below master, after 10 years playing same builds :)
The Terran, I mean, Human player
Lol "20 likes" Good cast man!
Alpha star reminded me of Uniden
Alpha star should really say gg before it quits
ragequit as a real gm, great progress!
If cannon rush works to harm alphastar, i say its all good!
didnt they limited its apm? Also it doesnt develop in this experiment to my mind
Yes, they limited it APM also forced it to play with human like camera and controls. The initial version didn't have these limitations and destroyed top players about a year ago.
@@chammy2812 sad that they limited the engine...would love to see its improvement and then seeing it compete vs other computer engines. Dont know why we dont see more of these..loved to see dota selflerning engine I think it improves the way of thinking about the game so much more....shure one might say its a cumputer dude we cant compare it to pro scene but I think we still can imply sth.
I think Alphastar had losed a bet, so he had to, he had to!!!!!!
Why is it taking too much time for the AI to learn, cant wait till Alphastar become the #1 SC2 player on earth...
ha ... i like it when the cannon rush gets owned
GrandMasterPlayer should be Bly. He's Zerg tho :)
I always go armor first and it feels even better when I'm behind
it's funny, AlphaStar plays the way I try to play...
There will come a day when AlphaStar gets revenche on all the people that forced him to play all these games over and over and over.
30 workers in 6minutes....? Grand master player had obviously a very bad day....
Great work. I really enjoy the AlphaStar videos. However, saying ''Like and Subscribe'' 5 times in one video, is a bit too much i think.
16:40 "And yes I did just assume the robot's gender" - Ha!
Not impressed by Alphastar's play here. Looks very similar to how it played Terran in the 50 games a month or so ago - which is basically to camp at home with tanks and send small commitments out to do damage (bio medivacs before, banshees in this game). It's essentially hoping its opponent will a-move into a siege tanks or that it takes a downright amazing engagement when it finally decides to push with all its siege tanks.
Maybe the production tab open would be nice
Was this beastly qt? He mentioned being tilted and cannon rushing alpha star...
still didnt learn to get medivacs :D
it thinks?? pull the Plug!!!!
Oh heck you upload consistently with niche content I want to see? Take my button pressing.
hmm deepmind should prob just make AlphaStar say GG before it taps out? shouldn't be hard right?
If you look at other AlphaStar games cast before this latest batch, it does say gg when it loses. I assume the chat from the replays was removed just like the player names. Prevents any off-color stuff in Google's professional material, but with the side effect of making AlphaStar look bm.
No GG? Very rude of you Mr. Alphastar xD
Well. If that was a GM player Im not so sure. The cannon rush execution was pretty bad, he had lack of constantly scouting and attacking.
@Wischmopps since protoss won the match and tarren lost, we really need to nurf Zerg....LOL
Cgi Dude scouting and doing pushes is most important part of been GM. Even masters 1 do that.
@Cgi Dude No. what it makes sense is that you belong in Bronze, so I'm going to ignore whatever you have to say.
@Cgi Dude Who is talking about build orders? I know they are important, however, you still need to scout. The cannon rush it was trash execution. Maybe the guy is really Gm maybe not. This "sad kid" thing is not offending, is a pathetic line that everybody say in the internet. You have to be more creative.
Cgi Dude how you know you ahead in macro if you don’t scout?
"GrandMasterPlayer"
"AlphaStar"
I see "DeepMind" enjoys PascalCase. I dig it
We demand camel case aaaaaaa
@@SleepLessThan3 That's heresy
@@MegaOgrady INVERSEcAMELcASE
👑
Pylon in the mineral field by a Grand master protoss? Huh?? And this cannon rush was executed nothing but silly. GM player are those pro players like stats and classic, will they cannon rush like this???
*It's gonna...it's a robot, not a he :P
Imagine if alphastar ai is implemented in all rts games,,, goodluck hahaha
Alpha star? no no no. Lets just call it, what it really is. Skynet.
Skynet vs Bronze league player :)
I need FFA Alphastar action
SkyNet (X) >>> Legion (X) >>> AlphaStar (possible future) ... any one catch this .... give a like ....
also.... ive warned u guys.... be prepare.... D-Day is coming.... sooner than u think....
Alpha Star is learning from our games our love for violence. How the response is from there is unknown.
Alphastar loses late game TvP after countering a cannon-rush.
As terran, I feel better now.
AlphaStars Terran is terrible tbh. The client simply doesn't use units like ghosts or ravens with great effect and is always playing quite passive.
well it wasn't gaulzi for sure
MOAR
U deserve my like and sub 😀