|
Post by ChloeB42 (Alexcalibur42) on Jan 22, 2015 1:02:12 GMT -5
I call bullshit, those are just made up pictures generated by "Fakemycode.SHG" Now THOSE look more legit, but as mousey pointed out, that's not klingon. Are you saying you're affluent in Klingon?
|
|
Will
True Bro
K/D below 1.0
Posts: 1,309
|
Post by Will on Jan 22, 2015 1:14:21 GMT -5
What's even happening anymore? Is Danger Close 10% or 40%? WHO CAN WE EVEN TRUST? ? (answer: marvelous marvel)
|
|
|
Post by ChloeB42 (Alexcalibur42) on Jan 22, 2015 1:19:18 GMT -5
No one and everyone. Trust, but verify, Comrade
|
|
|
Post by I Am Hollywood5 on Jan 22, 2015 14:48:53 GMT -5
There are people who still believe the code is in klingon?
|
|
Will
True Bro
K/D below 1.0
Posts: 1,309
|
Post by Will on Jan 22, 2015 15:01:08 GMT -5
The far back of my head had a crazy theory that the Klingon was smoke and mirrors while Marvel tried to work out a financial arrangement with Drift0r... but I didn't want to sound like a lunatic.
Then I posted a bunch of other stuff instead that made me sound like a lunatic anyways
|
|
|
Post by Megaqwerty on Jan 22, 2015 15:38:53 GMT -5
Yeah, like this nonsense about Ghosts being a zero out of ten.
|
|
Will
True Bro
K/D below 1.0
Posts: 1,309
|
Post by Will on Jan 22, 2015 15:47:02 GMT -5
Yeah, like this nonsense about Ghosts being a zero out of ten. Prestige levels brah
|
|
|
Post by Megaqwerty on Jan 22, 2015 15:48:25 GMT -5
Yeah, like this nonsense about Ghosts being a zero out of ten.
|
|
|
Post by bmac39 on Jan 22, 2015 18:47:34 GMT -5
I gave Ghosts a [Dead Space 3/10]
|
|
wwaa
True Bro
PC / PS4 / X1
Posts: 2,086
|
Post by wwaa on Jan 24, 2015 15:46:48 GMT -5
Re-rolled carepackages have a chance of getting the exact same crate type. Personally this is sloppy code even there's a VERY VERY VERY VERY VERY VERY low chance of getting the exact same streak, the new crate should have been stored and compared against in the while loop rather than going through the loop, if possible 100 times. this code is ok, a new crate is generated, stored and compared vs the original one (previous_crateType), I am not sure what do you mean? there must be an end to random generated results (up tp 100), otherwise this "while" loop could be infinite, if something with generator goes wrong. "Re-rolled carepackages have a chance of getting the exact same crate type." - yes, you cannot eliminate it if using a "pseudorandom generator". //I mean: theoretically it is possible to eliminate the same crate, but in practice different (strange and unbelivable) errors appear (memory mgmnt etc.) ... "Good/Better aka less pseudorandom" generator requires more CPU/memory, while players prefer better graphics.
|
|
|
Post by Megaqwerty on Jan 24, 2015 16:54:27 GMT -5
Right, but this is very simple, computationally, and it's sloppy. The first roll could have been stored and re-rolls done until a distinct re-roll is obtained. Alternatively, a modified drop list with the first roll could have been used.
|
|
|
Post by brontesaur on Jan 24, 2015 21:49:24 GMT -5
Right, but this is very simple, computationally, and it's sloppy. The first roll could have been stored and re-rolls done until a distinct re-roll is obtained. Alternatively, a modified drop list with the first roll could have been used. Not sure what you mean, the code does do what you said. Let's go through what happens when you double tap F to reroll: 1. Sends a message from this object that it's been double tapped (probably to the UI code.) 2. Makes a note of what the first crate you got was. 3. Makes a reroll attempts counter. 4. Goes into the while loop for the first time. 5. Sets the counter to attempt 1. 6. Rolls a new crate and stores it. 7. Compare the new crate to the first crate you got. If it's a different crate then finish. 8. If it's the same crate and the number of attempts is less than 100, keep rolling.
|
|
wwaa
True Bro
PC / PS4 / X1
Posts: 2,086
|
Post by wwaa on Jan 25, 2015 4:48:22 GMT -5
_maybe_ a good thing is that a copy constructor (prev_crateT = self_crateT) and "==" operator are used before a new crate is generated. so if there is a problem with the current crate (not 100% properly constructed somehow / hacked / damaged / unfinished) it won't waste time. maybe. we have really not enough info to get to any final conclusion/opinion here I suppose : )
what is "better"? (it depends on a lot of factors around):
a) 1. generate random crate 2. compare it with the current crate & solve problems.
b) 1. copy and verify current crate 2. generate random crate and compare it with verified one, if problems: return the original one.
|
|
|
Post by brontesaur on Jan 25, 2015 5:20:25 GMT -5
OK I think you're overthinking this.This code snippet doesn't error check the crate, and there's no reason to really since this code is not likely to fail. Even if it does need to be checked it should be the job of the getnewcrate, not this function. This is simply a snippet that rolls randomly until you get a new crate.
Should they have done it properly by rolling a new crate with the current crate excluded? Maybe, but this is simple and it works.
|
|
|
Post by zimpoo on Jan 25, 2015 8:50:39 GMT -5
So am I more likely to get a good scorestreak if use "reroll" or "better odds" ?
|
|
Will
True Bro
K/D below 1.0
Posts: 1,309
|
Post by Will on Jan 25, 2015 9:14:24 GMT -5
I converted these to percentages, for the dumb layman like myself:
Normal Odds 16.8% - Map Scorestreak 16.8% - Recon Drone 16.8% - Missile Strike 16.8% - Sentry Gun 10% - UAV 9.8% - Vulcan 4% - Warbird 3% - Assault Drone 2% - Paladin 2% - Bombing Run 1% - Goliath 1% - System Hack
Better Odds 13.5% - Map Scorestreak 13.5% - Recon Drone 13.5% - Missile Strike 13.5% - Sentry Gun 11.5% - Vulcan 9.9% - UAV 5.9% - Warbird 5% - Assault Drone 4% - Paladin 4% - Bombing Run 3% - Goliath 3% - System Hack
|
|
|
Post by zimpoo on Jan 25, 2015 9:42:42 GMT -5
I did some math, could you game scientists tell me if my reasoning is ok ? To me, scorestreaks A to K are shit, and L to W good. Using the numbers provided by the OP, we have
| Good
| Shit | Normal CP
| 0.1419214
| 0.8580786
| Better Odds CP
| 0.26539278
| 0.73460722
|
Now, if I use reroll, the only outcome I don't want is "Shit + Shit", so the odds of getting at least 1 good scorestreak are the following : With Reroll
| At least 1 good out of 2
| Normal CP
| 0,263701112
| Better Odds CP
| 0,460352234
|
I usually rolled "Support + Trap + Better Odds", but seeing that I can get a good scorestreak out of half my carepackages, I think I'll use "Support + Better Odds + Reroll" now. Also note how reroll and better odds almost have the same effect. (PS : I'm not having quite the same percentages as you Will, but I think you didn't take the Map streaks into account if I'm not mistaken)
|
|
|
Post by brontesaur on Jan 25, 2015 9:43:32 GMT -5
Hmm I think you forgot to include the mapstreak in there. I think these are good though for the maps without mapstreaks.
|
|
Will
True Bro
K/D below 1.0
Posts: 1,309
|
Post by Will on Jan 25, 2015 9:45:02 GMT -5
Hmm I think you forgot to include the mapstreak in there. I think these are good though for the maps without mapstreaks. Oh shit yeah you are right. I'll edit it. Done. Ooo the percentages round off much nicer now
|
|
|
Post by brontesaur on Jan 25, 2015 10:00:33 GMT -5
I did some math, could you game scientists tell me if my reasoning is ok ? Your maths looks good, two things I want to point out though: 1. Killstreak A is good sometimes and something you want to pick up, (eg Walker Tanks) and sometimes you don't want to pick it up. (eg Defender Anti Air) This changes it slightly. 2. When you reroll, you can't get the same thing again technically. So your chances somewhat increase if you have shit to start with compared what you have there.
|
|
|
Post by zimpoo on Jan 25, 2015 10:39:24 GMT -5
Yes indeed, I'll try to fix it (Walker Tanks OP btw)
|
|
|
Post by I Am Hollywood5 on Jan 25, 2015 14:56:09 GMT -5
I'm almost positive this is not real game code. I was just playing today and re-rolled my care package into the exact same scorestreak. The original reward was a recon drone, I re-rolled, and it was once again another recon drone. Now let's be honest here: this would mean the game re-rolled my care package 100 times, and it got the exact same scorestreak every single time. The chances of that happening are so infinitesimally small, I just don't believe that could actually happen. Either this is faked code, or I have the most unbelievably terrible luck in the history of mankind.
|
|
Will
True Bro
K/D below 1.0
Posts: 1,309
|
Post by Will on Jan 25, 2015 15:10:33 GMT -5
I'm almost positive this is not real game code. I was just playing today and re-rolled my care package into the exact same scorestreak. The original reward was a recon drone, I re-rolled, and it was once again another recon drone. Now let's be honest here: this would mean the game re-rolled my care package 100 times, and it got the exact same scorestreak every single time. The chances of that happening are so infinitesimally small, I just don't believe that could actually happen. Either this is faked code, or I have the most unbelievably terrible luck in the history of mankind. There are two different recon drones possible though. Both are "blue" (two modules). One has paint + assist, the other has cloak + assist. The code to avoid rerolling the same crate just checks the "crateType", and the two versions of the recon drone have different "crateType", one is "type f", the other is "type g". You probably got one version of the recon drone, then rerolled it into the other version. There is no way of knowing for sure because they are both "blue", and you can't see the modules until it is picked up.
|
|
|
Post by zimpoo on Jan 25, 2015 18:12:27 GMT -5
I'm like 90% sure (only 90% cause it was in the heat of the moment) that I had 2 green recon drones in a row this afternoon.
|
|
Will
True Bro
K/D below 1.0
Posts: 1,309
|
Post by Will on Jan 25, 2015 18:27:56 GMT -5
I'm like 90% sure (only 90% cause it was in the heat of the moment) that I had 2 green recon drones in a row this afternoon. The one that comes in a common reinforcement is green.
|
|
Will
True Bro
K/D below 1.0
Posts: 1,309
|
Post by Will on Jan 27, 2015 2:53:45 GMT -5
I converted these to percentages, for the dumb layman like myself: Normal Odds16.8% - Map Scorestreak 16.8% - Recon Drone 16.8% - Missile Strike 16.8% - Sentry Gun 10% - UAV 9.8% - Vulcan 4% - Warbird 3% - Assault Drone 2% - Paladin 2% - Bombing Run 1% - Goliath 1% - System Hack Better Odds13.5% - Map Scorestreak 13.5% - Recon Drone 13.5% - Missile Strike 13.5% - Sentry Gun 11.5% - Vulcan 9.9% - UAV 5.9% - Warbird 5% - Assault Drone 4% - Paladin 4% - Bombing Run 3% - Goliath 3% - System Hack I did these wrong - I gave the map streaks the same weight as the "bad" scorestreaks failing to notice that there is only one chance of getting the map streak, but two chances of getting each "bad" scorestreak. Pooh on me. The map streak will be half the chance of the "bad" streaks like recon drone.
|
|
asasa
True Bro
fuck
Posts: 4,255
|
Post by asasa on Jan 27, 2015 8:54:35 GMT -5
So that "extra" 8.4% is just split proportionally amongst everything, then (including the map streak) I take it?
Btw, for smart programmer people, why do they use weights rather than, say, a rng + varied instances of each streak?
I.e. rng 0-999 1 uav a 2 uav b 3 uav c ... 10 mapstreak 11 Vulcan a ... 999 paladin a
I'm guessing it's easier to make adjustments for, in that you could delete a streak and it would adjust on it's own, while noob method relies on thinking it through before writing the code?
|
|
|
Post by brontesaur on Jan 28, 2015 3:09:49 GMT -5
So that "extra" 8.4% is just split proportionally amongst everything, then (including the map streak) I take it? Btw, for smart programmer people, why do they use weights rather than, say, a rng + varied instances of each streak? I.e. rng 0-999 1 uav a 2 uav b 3 uav c ... 10 mapstreak 11 Vulcan a ... 999 paladin a I'm guessing it's easier to make adjustments for, in that you could delete a streak and it would adjust on it's own, while noob method relies on thinking it through before writing the code? Yeah if they need to readjust the weighting of certain streaks or delete/add streaks then it's much easier. (we don't have the code to know what they're doing exactly) Hard coding streaks to numbers is pretty inefficient to say the least.
|
|