Okuplok?

[Grain of Salt]

5 hours ago, PjSpartacus said:

His Sunder map 6 still bothers me to this day. Don't understand how he can get hit 7-8 time in the opening 2 mins for the absolute min damage every time.

 

Why use tas to get min damage when you could just use tas to dodge the projectiles?

[Paul977]

15 hours ago, loveless said:

  

He's coming back, you'll see.  

  

He went with my dad to go get that pack of cigarettes. 

 

lol^

[Catpho]

Wow, is this speeedrun history deep...

Hmm, its really an odd series of actions Oku did there....

Bah, i dont want to drag this further, i just wanna see that map he made get finished :P

[Bob9001]

Having just went and watch the sunder map 06 demo, I guess it's legit. But a few questions I have: 1. Almost all the baron's died in 4 rockets, now to get that amount of lucky rng is probably possible. 2. The archviles died in 3 SSG blasts??

 

I guess we could argue all day about this, because there is no way to check if it's a TAS without the submitter stating it. Unless I'm wrong??

[GarrettChan]

6 hours ago, Bob9001 said:

Having just went and watch the sunder map 06 demo, I guess it's legit. But a few questions I have: 1. Almost all the baron's died in 4 rockets, now to get that amount of lucky rng is probably possible. 2. The archviles died in 3 SSG blasts??

1. Did I miss something... I think it's a group of mixed monsters at the beginning you're talking about? If it's a group, you need to consider splash damage. (eg: killing 1 HK in 2 rockets is rare, but killing 1 group of 2 HKs in 4 rockets is normal)

2. Indeed you can kill an Arch-vile in 3 SSG blasts.

 

Usually TAS would include some "weird" movements or stuff. At least from the demo, it seems normal to me. For example, Oku didn't do some ridiculous movements during the platform sections or such. Well, just my 2 cents.

[Bob9001]

4 hours ago, GarrettChan said:

1. Did I miss something... I think it's a group of mixed monsters at the beginning you're talking about? If it's a group, you need to consider splash damage. (eg: killing 1 HK in 2 rockets is rare, but killing 1 group of 2 HKs in 4 rockets is normal)

2. Indeed you can kill an Arch-vile in 3 SSG blasts.

 

Usually TAS would include some "weird" movements or stuff. At least from the demo, it seems normal to me. For example, Oku didn't do some ridiculous movements during the platform sections or such. Well, just my 2 cents.

Yes you make some good points, I for one have very rarely killed an archvile with 3 ssg blasts, yeah I did know it was possible

but he killed most of them with 3.

Yeah they must of been grouped. This is an interesting topic.

[GarrettChan]

2 hours ago, Bob9001 said:

killed most of them with 3.

I only saw one, so I guess I definitely missed something.

 

*The following calculation could be $h!t because I totally don't understand how things work.

Spoiler

According to my calculation and the p_rng table, damage of one full blast of SSG ranges from 170 to 255. Let's make this a bell curve, where:

mean = (170 + 255) / 2 = 212.5

SD = (212.5 - 170) / 3 = 14.17 ≈ 14

 

Then, damage of 3 blasts should be a bell curve, where:

mean(1) = mean * 3 = 637.5

SD(1) = SD * 3 = 42

 

Given an Arch-vile 700 HP, in this bell curve, P(x>700)=0.0684=6.84%. With this, it should be rare enough.

 

Edited March 27, 2018 by GarrettChan

[Bob9001]

13 minutes ago, GarrettChan said:

I only saw one, so I guess I definitely missed something.

Yeah you're probably right, most of the others he killed with splash from rockets.

 

Oh well.

[Player177]

  Some comments reminded of probably the first utube doom speedrunner channel with the name Okuplok that I watched along with Armane15, but all I can call to mind is that it was deleted a couple days later which was sadly strange and that its content was interesting at some point. I suppose that I had to watch some Holy Hell just episodically at that time but on Armane15 channel I think.

  Well, of course! I forgot a bit. I watched HH in pieces but not completely and damn, the map5 is a crazy task if you look from the standpoint of those who doesn't know well the capabilities of tasing in Doom.

  I didn't know that there is a UV max for it and the fact that there is, if I get into this statement, leaves a very striking image of a dedicated not less than a month work on every aspect of the demo - theory and practice. Only to make the first exit on the safest route. It is controversial, but the duration of the map is too long and when I say a month I don't mean casual playing or playing for fun. Let it be better two or three in a gross time. Normal first logical goal will be with a promise - don't test fate, do not hurry, take your time - and only with this approach may be, perchance you'll reach the exit. I didn't watch the demo and I don't know how to take it - is it close to, lets call it, reality category or is it close to speed category. Summing up the time of all the HH maps 1-4 UV max demos from dsda will give ~2.20. It most likely shows that the demo is close to reality category but its well prepared and I think that it could not be otherwise if you really know what you're dealing with if aiming for the fantastic map time or just for an optimised run but that's another story.

  Unfortunately I can't accurately perceive a map difficulty not playing it seriously but I can tell that at first, everything is always bad for me. Over time, it becomes easier almost automatically. Afterwards - lack of time, the forgetfulness, the mistakes, the random, the anxiety, the fatigue and other things can become very influential in recording and you will think that's it. But sometimes luck overlaps them all. Is it really, the speedrunners bitter share, was unfamiliar to Okuplok while he was pulling this run? He realized that he has the ability to cope with HH after some significant amount of time, thats probably the most understandable explanation for me. The details of how it was I don't know but I can roughly imagine.

  After a little meditation I'd choose to TAS the map5 apriori, one way or another and I would have known in advance that it won't be a walk in the park. It's too long - main argument and the good reason will be my expirience with such things and me always watching at stopwatch. I plan to do one TAS and it looks like it will be devoted to Okuplok style but there is some unpleasant snag - the demo will be marked as a TAS. 

[Frits]

Little bump, but this seems like a good place to write this now that this thread exists. It's sad to see that he left the dooming/online community for the second time. This is just in case he ever lurks here again. I don't believe Oku ever cheated. I've spend a lot (a really unhealthy amount) of time playing together with him and he did the same things online he did in his demos.

 

So Ku, last I talked to you, you seemed to have your things in life together sorted out pretty well. I hope that's still the case after all this time.
So If you ever read this buddy, which I hope you'll do at one point. Add me on Facebook (again) or send a mail my way. you know my info. (it's also on my this forum account)
Wish you teh best man!

Edited April 28, 2018 by Frits

[42PercentHealth]

On 3/27/2018 at 6:05 PM, GarrettChan said:

  Hide contents

According to my calculation and the p_rng table, damage of one full blast of SSG ranges from 170 to 255. Let's make this a bell curve, where:

mean = (170 + 255) / 2 = 212.5

SD = (212.5 - 170) / 3 = 14.17 ≈ 14

 

Then, damage of 3 blasts should be a bell curve, where:

mean(1) = mean * 3 = 637.5

SD(1) = SD * 3 = 42

 

Given an Arch-vile 700 HP, in this bell curve, P(x>700)=0.0684=6.84%. With this, it should be rare enough.

 

You can tell it's correct because the standard deviation of 3 shots is 42. 😉

 

The mean is correct, but I don't think your calculation of standard deviation is correct.

 

Let's treat all the pellets as independent random variables (they are not independent because of the way the PRNG table is structured, but who cares) with constant distribution. The variance (which is not quite the same as standard deviation) of each one is the sum of all the (u-x)^2*P(x) [Interpretation: each value's distance from the mean, squared, and multiplied by the probability of rolling that value]. For a single pellet, the variance is therefore 10 (Is it a coincidence that this happens to also be the spread of values, 5-15? Probably not, but statistics is not my strong point.).

 

The variance of a sum of *independent* random variables is the sum of the variances. (The formula gets more complicated if they are not independent, which is why I chose to make that assumption.) Since we have 20 pellets, each with a variance of 10, then our variance is 200. Standard deviation is the square root of variance, so we have a final answer of 10*sqrt(2), or approximately 14.14 damage points.

 

The other way to do it, of course, would be to just shoot stuff with the SSG and record damage values, plug these into Sage or Matlab or your calculation engine of choice, and just get a mean and standard deviation of the data. But what's the fun in that? That's how scientists and engineers do stuff... Mathematicians think on a higher plane. 😛

 

EDIT: I just realized I neglected to calculate the standard deviation of 3 SSG shots. The variance of 1 shot was 200, so the variance of 3 would be 600. the square root of 600 is 10*sqrt(6), which is approximately 24.49 damage points.

Edited May 11, 2018 by 42PercentHealth