… if EVERYONE on both sides of the field played their heroes to their full potential, I’m talking 100% accuracy, god-tier positioning and awareness, etc… what team of heroes would be the best?
well assuming they all had aim bots then hanzo / widowmaker or ashe with a zenytatta to discord targets + he should be drilling all his right clicks in the head. a kiriko cuz im assuming she will be drilling nothing but head shots as well. tank probably ramattra cuz assuming he has 100% accuracy he can also snipe long distance with his nanites… or i suppose orissa wouldn’t be bad either for 100% jav / left click.
Hanzo would be one of the best assuming he hit 100% of the shot and has less charge requirement than Widow.
However, that is not possible even with aim bot unless it’s like an honing missiles curving his arrow around, which is not in his game design. With projectile, the enemy always has a window of time to change their direction.
Therefore, I’ll go with high damage hitscan like Widow/ Ashe, maybe a flanker at the back, Zen and Ana, tank Hammond/DVA.
Remember, the Tank playing also knows how to Tank the best. Meaning mitigation. Take that into account.
yeah, DVA matrix can eat infinite amount of damage as well as eating ult, it just on a bar instead of breaking like rein and sigma shield. I think that has a very high potential.
I put hammond there because the team comp I listed look like dive meta during ow1. I would put winston but his dps is very low (since we’re talking about 100% accuracy here).
1 Like
symmetra could be a tank of sorts… if we are assuming these players can make decisions within a single frame of perfect then she could literally block being head shot with her turrets. she could draw enemy fire / defend people with turrets as a shield if they go sniper.
1 Like
Both teams would be able to foresee the inevitable draw and just wait in spawn until timer ends
2 Likes
I’ll answer this one for the graduate class.
We’ve done simulated playouts.
The strongest play occured along one of two motifs.
Extreme turtle brawl comps.
The brawl-heavy, payload-heavy tactics. Stuff like goats but not goats (if 222 or 122 is forced). In general, those comps win by cd trading or more direct/powerful space confirm. It’s less kill-confirm and more space-confirm. It’s slow and methodical but it wins.
During adversarial AI training, these teams would play for some kind of late-fight mismatch. They jostle around with bait movements to trigger a cooldown or other resource defecit. This in turn is minmaxed and punished. But, since playing as near-perfect AI vs. near-perfect AI, the punish isn’t revealed until several depths later, resulting in things like 5% more ult economy for the better team. This would translate into something forced principal variation that converts more often, leading to more wins.
Range wombo comps.
Basically the reverse of the above. Perfect micro combos (usually from range) with very little macro.
Basically a team of aimbots with the highest possible on-demand burst.
They assess an escape deficit or force a positional mismatch and 3-2-1 all the 1shots arrive at the same time. The dmg spike is through the roof, and the minmax looks something like if LoS=true headshot=true hp_low=true → dead or forced defensive cd=true into completion cd/dmg = kill = true = push to next choke (push for free), and lather rinse repeat.
But in terms of perfect aim/micro (aimbots on most aimbot-scaling heroes), the better aimbot team wins using art-gallery rollouts. Basically, they learn the best geometry to grant the most awareness, most coverage, for least resources. That’s why our aimbots will always beat your aimbots. A lot of math (you dont gitgud at math sitting in lowly GM ranks). And it’s really cool to interpret the results.
So the best probably depends on the map, but the skill of the player? That’s over with. AI is supreme - better aim, better brains. Just not in Blizzard games but I’m psure they have some decent bots going around.
1 Like
I would honestly love to watch it play out. OWL isn’t enough, to see two nearly perfect built teams doing everything 100% correctly, and having it become a game of timing… sounds beautiful to even think of. Since you didn’t state any heroes though, I assume every hero is viable of doing this in their respective sub-role (brawl, dive, poke)?