Get the Desktop App for Battle.net Now
- All your games in 1 place
- Log in once
- Automatic game updates
Well according to them, some guy did his presentation on how to grow !@#$%^-* (the technical aspects of it, however many there are) and actually passed with a great mark.
Oh boy, wtf censor lol. How to grow cann@bis.
Edited by nitrocan#1478 on 10/6/2012 2:25 PM PDT
Posted by nitrocan
Did I just get skull !@#$ed? My entire understanding of WW (which I had thought was pretty advanced) just got raped...
@nitro, idk how interesting this topic would be for people that don't play d3. I mean this is complicated enough to explain to people that have hundreds of hours of game time. Anything with heavy math (even super simplified) would cause glazed over look pretty quickly I'd think. Maybe I'm not understanding the nature of this presentation.
Posted by nitrocan
It's supposed to be presentable to engineering students in their 2nd or 3rd year so being complicated is okay. I'm okay if they don't really understand haha but I'll try my best.
Don't worry, I won't tell them :D
But if you want to do it right, this stuff is really hardcore. Something you also may want to keep in mind: LoH is very easy to calculate, as it just adds up - number of ticks multiplied by healing per tick and you are done.
With CM (and also APoC) it's not so easy, because suddenly you are getting deep into probability theory...
Thinking about it, as soon as you mention critchance, even someone who doesn't play diablo has to realize this. So be careful not to mix things up here.
Posted by nitrocan
Come on over. We need more brains. Also, feel free to verify my testing results. It was a long day with many numbers whirling around in my head and on the screen, so I may have made a mistake somewhere.
The rounding is definitely real, we just haven't figured out how exactly or why it does that. And we also don't know the scope of this phenomenon yet. By now I just tested ET-WW with LoH. Could be different when we look at CM, APoC, and of course any skill other than ET-WW.
Yes I realize that CM is another story lol. I'm just going to assume that if I crit 8 times with WW, I'll proc CM approximately once, but the real probability of lets say being able to cool down frost nova completely in 1 second, that stuff could get messy lol. I don't even remember combinations all that well.
Does anyone know how often meteor hits? I think it was mentioned that it's independent from your attack speed. I calculated my numbers assuming that it hits once when it lands and then once per second for the molten fire duration.
This is what I have so far (except the introduction part)
GROUP OF 3:
Old 1.0.5 Liquefy:
0.125 proc rate
Against a group of 3, there's a 87.5% chance that the molten fire effect will last 8 seconds, 12.5% chance that it will last 3 seconds. This averages to an 7.375 seconds of molten fire effect.
Because of attack speed, we can cast 2 meteors in 1 second. Because the average molten fire effect lasts 7.375 seconds, that means 7.375*2 = 14.75 molten fire effects can occur at the same time. Since there are 3 enemies, we are producing an average of 44.25 hits per second from the molten fire effect. The amount of meteors we can land per second is 2, so 6 more hits from the meteor itself and the grand total is 50.25 hits per second.
Because of liquefy's proc rate, the return coefficient is 6.12825. If we have 30 arcane power gain on critical hit, that means we will 6.12825/2 * 30 (because we have 0.5cc) = 91.92375 AP gained per second. By simply using the Prism rune of Diamond skin, we can infinitely cast meteor on a group of 3. (43 AP per meteor, 86 AP spent per second)
Regarding CM, approximately 25.125 of those hits will be critical hits and each of them will have a 1/8 chance to proc CM. This suggests that on average, CM will be triggered 3.140625 times per second.
0.125 proc rate.
What separates wicked wind from meteor is that wicked wind hits the enemy at a rate of twice your attack speed per second. Also, because you can cast more twisters per second by increasing your attack speed, the return coefficient from wicked wind is quadratically proportional to attack speed.
Against a group of 3, we can assume that the twisters are wide enough to hit 2 enemies at the same time. Since twisters stay on for 6 seconds, we can have 12 twisters up because of our attack speed of 2 aps. Those twisters will hit 4 times per second at 48 attacks per second. However, because each twister is hitting 2 enemies on average, they are producing a total of 96 attacks per second.
Multiplying the number of hits by the proc coefficient gives us a return coefficient of 12. Since twisters have a lower AP cost and a higher return coefficient, sustaining it isn't a problem with a minimal amount of APoC.
Out of those 96 hits, approximately 48 of them will be critical hits and will have a 1/8 chance to proc CM. Using WW, CM will be triggered about 6 times per second.
Old 1.0.5 Liquefy:
Against a single enemy, there's a 50% chance for a 8 second and 50% chance for a 3 second effect, averaging to 5.5 seconds. Multiplying this by our attack speed gives us 11 molten fire effects at the same time. Because we're hitting only 1 enemy, we're producing 2 + 11 (initial meteor hit) = 13 hits per second.
This results in a return coefficient of 1.625. With 30 APoC, this is equal to 1.625/2 (0.5 cc) * 30 = 24.375 AP gained per second. With certain items, meteor can be reduced to cost only 34 AP to cast. Even with the wizard's 10 AP regeneration per second, casting 2 meteors per second requires 68 arcane power which doesn't seem feasible against a single enemy.
TO BE CONTINUED...
Any major discrepancies or incorrect calculations? (assume that my assumptions are correct :P)
I seem to remember assuming meteor was a 0.250 proc rate for a while due to the LoH return. Doesn't that mean its actually a 2x per second tick from DoT, not a 1 per second if the official proc rate is now announced at 0.125?
Also in the above post your only permitted 2 meteor DoTs to stack from a single second of casting, but then comparing it to 12 stacked twisters, or 6 seconds of casting.
I would REALLY try to keep things at a single target. Cause against 3 targets you'd be talking 14 DoTs stacked at max metoer cast rate (87.5% of 16).
Against a single target things will appear much more balanced as well. And less messy by oversimplifying the statistics. Half of the meteors each seconds will crit, giving an 8 second DoT duration. So you'll have 3 stacked from the non-crits and 8 stacked from the crits for 11 DoTs and 2 meteors per second at 'max wind up'.
Then with WW you'll have 12 twisters at 'max wind up'.
The only problem with that is it actually becomes EXACTLY the same. 11 DoTs hitting twice a second = 22 hits, +2 from new meteors. 12 twisters hitting 4 times a second is 24 hits a second. Both at 0.125 proc rate. Idk if you want it to work out perfectly like that, but thats kinda neat.
You can't have meteor only count 1 second worth of casting and WW to count 6 seconds worth of casting however.
Edit; oh nvm, i see how you did it now. Just didn't wrap my head around it quite right on the first read cause it was a little different than my mindset. Its still accurate though as near as I can tell. The point of meteor tick rates still stands though, I'd hop in game and check the LoH return.
Edited by Shandlar#1961 on 10/6/2012 3:36 PM PDT
i think we need more graph and plot to summarize the result in a easy to look matter.
Those would be tremendously useful to me:P
Meteor apparently does tick once per second. Shame haha.
I'm trying to keep it as basic as possible. I'm assuming that in a champion pack of 3, all 3 elites are being hit by every meteor and all stay inside the molten fire for its complete duration. I also initially assume that we have infinite AP and then at the end see if the AP we get from 30 APoC is enough to cast meteors at the rate we assumed.
Casting delay just makes it so that you start getting procs 1.25 seconds after you cast meteor, nothing really changes.
Edited by nitrocan#1478 on 10/6/2012 5:25 PM PDT
I see apo already ask you to come here and share the love : http://eu.battle.net/d3/en/forum/topic/5208511084
Here is a piece of art I made to convince you :
Alright I am not here to do propaganda (or maybe just a little xD). If you did this presentation for your engineering class like I understood, keep in my mind engineers love to find a compromise. Your conclusion could be : "alright energy twister is great but it require attack speed and he did not deal that much burst damage. On the other hand meteor does not require attack speed, deals some damage. But you cannot really stack that many meteor on a target and the landing time force you to use a way to protect you before you finally get some proc. So meteor is kinda more risky. So in that situation we proved that WW is optimal while star pact is optimal on that one."
It is sad but it does not matter if you truly find a compromise or not : teacher will probably focus on the maths/logic part. And compromising show that you are trying to be objective and logical. Students will probably applaud anyway because they want all the prom to have the best mark as possible. At least in France :D... I guess that makes sense when you know that you have to pass several tests (depending on the school you want to have) after the two first years and half of the people that prepare the tests abandon/are fired/loose their mind (rarely) even before trying it :D. When you loose half of your friends : you are kinda attached to everybody that made it through. So everybody is basically and viscerally united against the teachers after that point :) !
Also I did not see anywhere in that thread that meteor get 3 ticks of dot and 1 tick on impact. If a target get hit by the impact he also take the first tick of the dot simultaneously. But I read the thread very quickly so I might have missed it.
I made engineering studies myself... Gosh that kinda of presentation bring back good memories. Even if my subjects were more... classical.
Edited by Guybrush#2456 on 10/6/2012 6:33 PM PDT
Even if they actually will focus on the logic/math part, they are more or less looking for my ability to prepare and present, confidently and professionally, a technical presentation. For example I won't really lose marks if my calculations or logic are a bit off but I will lose marks if I go overtime :P
I really like the compromise part, I'll definitely try to squeeze it in, somehow lol. I really hope I can somehow fit all this in 15 minutes.
Threats of violence. We take these seriously and will alert the proper authorities.
Posts containing personal information about other players. This includes physical addresses, e-mail addresses, phone numbers, and inappropriate photos and/or videos.
Harassing or discriminatory language. This will not be tolerated.