Firstly: I cannot see how you used calculus to run a simulation. I am passing calculus with an average grade of 85%. Care to enlighten me?
Secondly: How on earth did you get 30% expectation??
Thirdly: You did not see my simulation. I did run it 1000 times. There were 115 attacks which were completely nullified. However, they weren't frequent enough to be that useful. The nullified attacks would be clumped together, but then the clumps would be far apart by a relatively large gap. Not particularly helpful in my book because that ensures that in any one battle you can't expect to have any blows from your opponent nullified or reflected

On your experiences with lions vs clannfear - that seems particularly strange to me. Clannfear had more health. Then again they were more susceptible to the daedra killing shock spells.

Well, I stand corrected on that. The wording of the perk changed. For heavy armor it was resist damage 10%, for light armor it was 10% chance to nullify damage. Wording of the perk has changed the implementation as I understand it. Now they are indeed, about equal.

Light armour, if its anything like it was in oblivion then you will be a lot faster whilst wearing it

Armor? I don't need no stinking armor!

I'll go as I did in Oblivion. Robe, blue suede shoes and whatever gloves and hat looks the best.

I always prefer light armour myself.

I find it looks better.

And since most of my classes incorporate the sneak skill heavy armour is always low on the list.

For sure my friend . Its actually a combination of calculus and statistics. Its not a full simulation but it is just doing some math involving probability. Since you have N number of trials we can make a probability table for the percent chance to nullify 10 blows following the percent chance to nullify 11 blows and so on. I only went up to the chance of nullifying up to 40 blows out of 100 trials. From there I randomly generated a number that corrosponded with each number of hits and came up with the number 5. Which represented the data group 35-40 blows. The weight on the number 5 was balanced tot he weight of the data group and I just happened to get it. It was only a 6% chance but I did draw it. Which means for me I nulled about 40 blows out of 100. or, roughly, 30-40%.

I don't mean you running it a thousand times I mean you running that ANOTHER thousand times. right now you have N = 1,000 for one thousand trials but the more you do it the more it should balance out to roughly 1 out of 10. There really is no reason to do a simulation here since we know the pre programmed probability. I only did my own to show you how much chance influenced your results. Someone could easily have far better results then you did and it wouldn't be considered unusual.

But we already know the probability that you have a 10% chance anyways so, unless your simulations are off, the simulations should represent the actual data. You should be seeing about 1/10 blows nullified. The fact that you are a little skewed tells me you need to run the sim more as we already know the probability so the probability will match up directly with the data. Either way though its already representing about 10% so its good. I'm sure you know what I'm saying by now so i'll move on, if not let me know and i'll try to re-explain.

I DID see the sim. lol. Its pure chance that your procks clumped together like that. It really says nothing about what to expect in the game as someone can have widely different results. The sim was already pretty accurate with 115 procks out of 1 thousand but the more you do it the closer it will get to 1/10.

I could be biased on clannfear vs. lions. I HATED the lions. I was a bow wielding character and it was always easier for me to kill clannfear rather then lions.

Ah that could be where the missunderstanding lies then. Yeah there is only a 10% chance to reflect damage taken off of a normal attack with heavy armor, not to reflect 10% of every attack. They are virtually the same perks except light armor completely nullifies the damage while heavy armor makes the enemy "feel your pain" so to speak.

until you actually do get hit.... then its game over.

lol good luck to you


R.I.P.

Haha just kidding. you may want to at least invest in alteration since it does eventually give you the mage armor perk which gives you a bonus to all protective alteration spells as long as you aren't wearing any armor.

That has absolutely nothing to do with calculus, whatsoever.

And in my opinion, that maths is flawed. The chance of having 40 attacks in 100 nullifying damage is 2.47exp-13%, not 6%. I have a lot of confidence in my method of calculating that..
It already was ~10% of the time. That wasn't the issue. My issue is with frequency. Now that heavy armor perk has changed, it is about even.

Repeated the simulation 21 times, 1000 "attacks" in each trial. And yes, of course the average value was going to 102, which is to all extents and purposes 10%. Its probability.

and no you didn't see the sim - because I didn't post it here until that post. I just posted the results of clumping and jumping. I would have much preferred the constant chance.

Likewise, the simulation is what to expect in game. Both seem like rather unimpressive perks to me, to be honest. Frequency is way too low to be a game changer, the ability to rely on the perk to save our ass is likewise quite low. My feeling is that there are much better places to put this particular perk.

It does. Though only in one equation. I'm not going to go into depth about calculations on a skyrim board though lol. I am taking statistics in college though and there is some calculus involved (not to say I'm some expert or anything).

Sorry about that, yes I saw your RESULTS is what I meant. No I didn't see the actual sim but I assumed we were talking about your results.
What I'm trying to say is we didn't need a sim to know the frequency in the first place. The frequency is 1 out of 10. Or 10 out of a 100. Take it or leave it.

Which 1 out of 10 is high enough for me. Your not supposed to be able to rely on perks to save you, they are just perks. Just little added things to make you that much stronger. I feel those are great master perks. I've been saved by procks with a much lower chance in almost every other RPG.

In DA for instance I only had a .01% chance to insta kill an enemy with a certain dagger. That just so happened to prock on one of the hardest boss fight that I had of the game.

Point is is that 1 out of 10 is high enough to expect it to prock almost every single fight. There are only going to be very easy fights where you are not hit at least 10 times.

So, considering that, we should see this prock relatively often considering how much we will be in harms way. one out of 10 is actually one of the highest probability ratings i'v ever seen for anything in an RPG to prock.

Well I guess light armor is out now.

Nah go on. Which equation? We're on a gaming forum. We can't possibly look any nerdier than we already are.

Simulation was to look at permutations, not combinations. I frankly was surprised when you took that view of it.

Yeeas.. and yet, according to the simulation, it doesn't. It averages out to be that number, yes. But in reality, we can expect to go 20 - 30 hits without a nullified one. I don't know about you, but I don't really intend to be hit that many times in a single fight... Especially in light armor.

My guess is light armor will have better advantages on using it while using stealth and If stuff like speed bonuses,etc will more than likely be on light armor instead of heavy. Heavy armor reduces magicka regeneration more and I want to say stamina as well but no quotes on that, I'm not sure.

Gunna be using heavy armour on my first character.

Light Armor
http://i.imgur.com/JGju6.jpg


Heavy Armor
http://i.imgur.com/SqCiX.jpg


I would say that the heavy armor perks are far better then the light armor perks especially by the time you get to 70 in the skills

light armor perks are on the wiki here http://elderscrolls.wikia.com/wiki/Perks
i think heavy is a bit better though

I'm bit eh on some of the Heavy perk tree structure. The entry is fine, it's gravy. Conditioning is awesome but having to go through the eh Fists of Steel and Cushioned Perks isn't sixy. Tower of strength looks very good though.

On light armor I'll give the % to armor perks a slight nudge because the difference might be more dramatic because of the generally reduced protection of LA. Unhindered doesn't require crappy buy ins so there's a plus. Deft Movement does seem like it's not worth the perk point though.

For me it's a crapshoot since Wind Walker in LA and Tower of Strength in HA are the sixy choices for me.

I am definitely going to use light armor. Why?

1) Heavy armor may become weightless with a perk, but for that I'd have to invest points into two completely useless perks.
2) Since I am going dual wield I will need the +50% samina regeneration from the light armor tree, whereas heavy armor drains stamina faster (confirmed, though I can't find a source atm. I'll post it when I find one.)
3) I want to use spells as a backup weapon. Heavy armor has a higher magicka regeneration penalty than light armor (still need confirmation on this one)

The damage doesn't reflect every hit for the Heavy Armor perk. There is a 10% CHANCE to reflect damage. I'm sure Light Armor will have some sort of advantage. Hopefully it'll be faster and be much better for sneaking or something.

When I first saw the respect HA and LA perks I thought that the HA one was a little better. But after really thinking about it I think post are a waste of a perk. They're too unreliable to really waste a perk point on.

personally i think light armor should get 20% chance to dodge (negate all damage, of any kind including magic/arrows[extra/seperate perks?]) OR alteration spells shield for their full stated value in light armor and a corresponding perk in heavy for 50%, if you could get double bonus in no clothing, full in light, and half in heavy i think that would balance out evenly, because it would still benefit each one, just the ACTUAL AMOUNT of benefit would change. it jsut seems slightly unbalanced to me because the % increase for heavy is still actually more then the % increase for light. lets say iron armor has 5 defence + 100% from perks = 10 defence, fur armor = 3 defence + 100% from perks = 6 defence, the gap just seems to get larger rather than smaller

Keep in mind that if armor affects movement speed then the HA % perks will have more value because you'll more than likely get hit more often. With LA if it's not nearly as important as mobility to avoid attacks all together then you're saving some on perk points since you can have a lot of builds that just get the initial 20% perk to open the rest of the tree.

Apparently there's a magicka regeneration penalty for Heavy and Light armors (with lighter armor having less of a penalty), so there's one reason to pick light over heavy. If you're playing a character that relies on spells more, then you'd probably have an advantage using Light Armor. Plus magic oriented characters can always enchant more AR into their light armor. That's what I'm doing.

Haha I suppose you are right. This is pretty much the penicle of nerd congregations. Well I assumed order didn't really matter in this case so I used combinations. Doesn't matter to me wheather out of 10 hits one prock hits at trial 8 or 9. The only fact I was looking for is the percent chance of finding any number of procks out of a certain trial amount. What I did was to create a probabilities table for n=100. I then calculated each chance of all probabilities for N values. We know the probaility is 10% so all it really takes is to take a scientific calculator and use binomial calculations for each value.

However we could also just use standard binomial equations to find pretty much all useful information in this case. Since this is a binomial probability (the thing eiher procks or not, and it is involving discrete, not continous, data) we can just use N*P*Q = Standard of Deviation^2.

We find that the standard of deviation is 3 with 100 trials and a probability of 10%.

We can then use alpha to find the unlikely chances (or just double the standard of deviation and add and subtract it from the mean).
The mean is, obviously, 10, so adding six and subtracting six means its normal to have 16 procks or 4 procks.

Yet the probability when considering alpha is 6% for it to be outside of that range. When I randomized some number it just so happened to prock the number representing 30-40 times out of a hundred.
That's not to say this is the most reliable way.

Since we are considering order though and you are saying they prock together, when considering permutations I found a less then 5% chance of having 4 sucesses in a row, which in statistics terms is deemed "unusual". It was semi likely to recieve 2 in a row. but mostly orders of just 1 in succession were the highest.


The only calculus used was some minor differential calculus to discover the rate of change in the alpha proportions.

Forgive me I got kind of into it. Somewhat of a math buff, but then again I do love programming.

Feel free to let me know if I did something fishy here though, i'm no expert and since this was just an extra curicular activity I didn't put very much thought or effort into it. Just writing down results from my calculator really.

I never saw much difference between no armor and light armor. Light armor does offer more defense, but not so much that it actually makes a difference.
If you face a tough monster that hits hard, you will die in almost as many hits as if you didn't wear any armor. Light armor probably lets you take a few more hits, that's it.
Light armor lets you move faster than heavy armor, but wouldn't no armor be faster.

So I hope the perks for light armor make it more noticeable.

Sorry - prock? Not familiar with the term.

Wait. My simulation was to look at the order of nullified hits vs non nullified hits. So I didn't use Combinations or Permutations myself.

I.e. attacks 1 - 15 were non-nullified, attacks 16 and 17 were nullified, attacks 18 - 56 were non nullified and so on.

Your method (with combinations etc.) sounds legit. I am not that much of an expert on the subject - I'm finishing off high school this year, but dropped Stats coz with NCEA it becomes more of an english essay, or so I've been lead to believe.

As for my own method, I screwed all the niceties I could only semi remember, after looking for my notes book from last year and failing to find it (probably ritual burning, but come to think of it I didn't use it much for my revision as I knew all the material, so it's probably wandered off into the sunset by now) and instead just plugged it straight into an excel spreadsheet. Much faster for me than doing a calculator method.