Thread has been deleted
Last comment
AI Predictions.
 | 
China SalttyNoodles 
Why are they only in post plants/saves? It's always some lopsided shit like 90-10. We should get it at the beginning of round or after the first few engagements.
2021-01-19 18:50
Topics are hidden when running Sport mode.
THEN IT WILL BE 50-50 LUL
2021-01-19 18:53
1 reply
Nah less util and guns and go diff
2021-01-19 19:00
i bet when zywoo had like 3 hp in that 1 v 4, the AI would have said 100% failure or 99% failure, then he clutched it, so yea never trust ai
2021-01-19 18:54
10 replies
cause those would be the actual chances, ai would be quite correct, wdym
2021-01-19 18:57
??? Those are actually the odds, what are you trying to say?
2021-01-19 18:59
Even if it's zywoo he will still fail that 99 times out of 100 lol.
2021-01-19 19:01
NA math
2021-01-19 19:14
1 reply
Some NA kids are dumb af.
2021-01-19 19:18
#21
 | 
Brazil vittj2
what?
2021-01-19 19:24
3 replies
He's the 60iq NA kid.
2021-01-19 19:25
2 replies
that's a lot for NA
2021-01-19 20:35
1 reply
For people in households that have more than 70k a year it's low.
2021-01-19 21:08
ok
2021-01-19 20:33
because the model has to be fricking trainned dummy
2021-01-19 18:55
5 replies
ehrmmm no
2021-01-19 18:56
4 replies
ofc yes, it's way easier to predict post-plant than 5v5 round starts
2021-01-19 19:00
3 replies
but AI in this "plugin" or whatever has already been trained and developed from thousands of matches. If its designed to predict only postplants then ok, but i recall it can predict outcomes in any scenario
2021-01-19 19:12
2 replies
lol, predict it will always predict bro but the issue is accuracy metrics that you are not taking in consideration. For example, once the main neural model is trained, we can use its frozen representations at certain layer and learn a linear classifier on it by probing it. This essentially checks if the representation in the given layer has enough information for the specific task. What I mean here it's that round starts provide not enough information, and the model will likely predict a 50-50 win chance. Whereas rounds that are deeply disbalanced in terms of fire power, you will see AI predicting the obvious likely winner (e.g. team with more guns/nades etc). Thus, it's way more interesting the modelling of post-plant situations in which small details can be way more impactful. In other words, post-plant actions will produce larger perturbations which will have, on average, a larger effect on the neural network prediction.
2021-01-19 19:30
1 reply
ok i see, i can agree with that
2021-01-19 19:25
#6
 | 
Russia Syonamaru
Make ai great again!
2021-01-19 18:58
its for shits and giggles. Some random poor soul in soulless corporate environment had to suggest something that would "improve viewer experience" which was on his this year's targets so he came up with this dumb shit. However corporate dumbos did not thought nor probably cared that the thing like this that would be at least a bit accurate is insanely hard to implement since there are so many factors that impacts the outcome in round and you would need vast amount of data analyzed and then find a way to gather all the relevant data during the match and analyze it on the spot.
2021-01-19 19:03
1 reply
You know Blast didn't even make the air they just mad each "challenge" which was just free work.
2021-01-19 19:08
it's a pin point in the round, is it after every bomb plant
2021-01-19 19:02
AI the most overrated piece of technology since bitcoin
2021-01-19 19:07
1 reply
Here comes TOR
2021-01-19 19:14
They might only have a model for that particular situation
2021-01-19 19:15
1 reply
That's stupid considering they hired professionals with PhDs to do a challenge and the winner can't even compute beginnings of rounds.
2021-01-19 19:34
+1 AI Predictions doesn't have any purposes rn.
2021-01-19 19:27
Lilmix
1.11
megoshort
6.00
Astralis Talent
1.07
Vilay
5.10
Anonymo
1.59
SKADE
2.21
Bet value
Amount of money to be placed
Winning
Odds total ratio
-
Login or register to add your comment to the discussion.