Rogue - Assassin Changes

[slarden]

I used SK because I am playing an Int based SK for flavour reasons and I was comparing my current build do what it would be if I did an LR or ER.
Sure If I made an Int Drow instead my DC would go up by 2, still leaving me 4 behind a Dex build but I would also loose a few more HPs and open that gap further, sure not so big a concern and I have often considered drow for the extra DC but as I said, I am playing SK for flavour, I even put up with the Iconic rubbish and no racial traits for it.

Yes my Int build can use any weapon with Int to Hit and damage but I almost never do, I used to before celestia got bugged but now I never use others.

I think I would go with drow simply because of Turbine's history of pendelum swinging. This gives you a race that would work as an int or dex build so despite what Turbine does you won't need to do a full heroic TR.

The benefit of being able to use other weapons with strategic combat 1 and 2 is way overblown. If int to-hit and damage was offered in the assassin tree like dex is I would never take either enhancement. I would do just fine without being able to use the larger selection of weapons and be more than happy for my freed up 4 AP.

[Qhualor]

The bottom line is that the calculation is very simple although there are many things impacting the actual variables

1 - (opponent chance to hit through displacement * opponent chance to hit through AC * opponent chance to hit through ghostly * opponent chance to hit through elusive target * opponent chance to hit through dodge, etc.) = total miss chance

You are focused on the dodge variable and I am focused on the total miss chance.

If you take [change in total miss chance] /[ original total miss chance] it gives you a rough idea of how much less damage you will be taking from the change.

I would look at total change in dodge, total miss chance before/after and % of damage reduction I can expect. Still, a 3% change in dodge doesn't overly excite me much. If it's free I will happily take it.

a good way to know how good that 3% difference is deduct from one of your characters and run around in a quest. than go back and do it again with the 3% added back on. I personally notice a difference, especially the higher I get it with layers of other miss chance defenses.

using myself right now I have 14% dodge with 18% cap and with perma blur from enhancements and feats, no dodge gear. by level 23 I can start using Treads for ghostly. I have a displacement clicky I use for boss fights and oh no moments. lesser perma displacement can be found in game, but im just doing a past life and im not playing a rogue. I know that rogues can get higher than this and other options but for sake of making a point.

I have about a 32% miss chance pre-23 while regularly questing.

I have a 38% chance rounded down at level 23 and up while regularly questing.

I have a 61% miss chance rounded down at level 23 and up when I cast displace.

this is what I like to refer to just to give me a guideline for what I can more or less expect http://ddowiki.com/page/Miss_chance





Miss chance

Dodge

Concealment

Incorporeal


0.00% 0%
5.00% 5%
10.00% 10%
15.00% 15%
20.00% 20%
25.00% 25%
10.00% 0% 10%
14.50% 5% 10%
19.00% 10% 10%
23.50% 15% 10%
28.00% 20% 10%
32.50% 25% 10%
10.00% 0% 10%
14.50% 5% 10%
19.00% 10% 10%
23.50% 15% 10%
28.00% 20% 10%
32.50% 25% 10%
19.00% 0% 10% 10%
23.05% 5% 10% 10%
27.10% 10% 10% 10%
31.15% 15% 10% 10%
35.20% 20% 10% 10%
39.25% 25% 10% 10%
20.00% 0% 20%
24.00% 5% 20%
28.00% 10% 20%
32.00% 15% 20%
36.00% 20% 20%
40.00% 25% 20%
28.00% 0% 20% 10%
31.60% 5% 20% 10%
35.20% 10% 20% 10%
38.80% 15% 20% 10%
42.40% 20% 20% 10%
46.00% 25% 20% 10%
25.00% 0% 25%
28.75% 5% 25%
32.50% 10% 25%
36.25% 15% 25%
40.00% 20% 25%
43.75% 25% 25%
32.50% 0% 25% 10%
35.88% 5% 25% 10%
39.25% 10% 25% 10%
42.63% 15% 25% 10%
46.00% 20% 25% 10%
49.38% 25% 25% 10%
50.00% 0% 50%
52.50% 5% 50%
55.00% 10% 50%
57.50% 15% 50%
60.00% 20% 50%
62.50% 25% 50%
55.00% 0% 50% 10%
57.25% 5% 50% 10%
59.50% 10% 50% 10%
61.75% 15% 50% 10%
64.00% 20% 50% 10%
66.25% 25% 50% 10%
25.00% 0% 25%
28.75% 5% 25%
32.50% 10% 25%
36.25% 15% 25%
40.00% 20% 25%
43.75% 25% 25%
32.50% 0% 10% 25%
35.88% 5% 10% 25%
39.25% 10% 10% 25%
42.63% 15% 10% 25%
46.00% 20% 10% 25%
49.38% 25% 10% 25%
40.00% 0% 20% 25%
43.00% 5% 20% 25%
46.00% 10% 20% 25%
49.00% 15% 20% 25%
52.00% 20% 20% 25%
55.00% 25% 20% 25%
43.75% 0% 25% 25%
46.56% 5% 25% 25%
49.38% 10% 25% 25%
52.19% 15% 25% 25%
55.00% 20% 25% 25%
57.81% 25% 25% 25%
62.50% 0% 50% 25%
64.38% 5% 50% 25%
66.25% 10% 50% 25%
68.13% 15% 50% 25%
70.00% 20% 50% 25%
71.88% 25% 50% 25%

[gwonbush]

The bottom line is that the calculation is very simple although there are many things impacting the actual variables

1 - (opponent chance to hit through displacement * opponent chance to hit through AC * opponent chance to hit through ghostly * opponent chance to hit through elusive target * opponent chance to hit through dodge, etc.) = total miss chance

You are focused on the dodge variable and I am focused on the total miss chance.

If you take [change in total miss chance] /[ original total miss chance] it gives you a rough idea of how much less damage you will be taking from the change.

I would look at total change in dodge, total miss chance before/after and % of damage reduction I can expect. Still, a 3% change in dodge doesn't overly excite me much. If it's free I will happily take it.

I don't think you understand how total miss chance works properly. As each hit chance is multiplied together, it is completely fine to look at any term in isolation to determine how much less you get hit.

The correct way to calculate how much less damage you take isn't [change in total miss chance] /[ original total miss chance], but 1-([New chance of getting hit]/[original chance of getting hit]).

Let's even throw in some numbers:

Assuming displacement, 27% dodge, 10% ghostly and 5% chance to be missed due to AC. In other words you have a chance to be hit of (1-.50)*(1-.27)*(1-.1)*(1-.05) for a total hit chance of 31.2075% chance to get hit.

Bringing dodge up to 32% gives us (1-.50)*(1-.32)*(1-.1)*(1-.05) for 29.07% chance to be hit. If you calculated it using your faulty method it comes out to ~3.1% chance less to getting hit, but that's faulty math. 29.07/31.2075=~.93151, which EXACTLY correlates to the dodge increase I showed in my previous post.

[CThruTheEgo]

That's not really how math works...

Here is the wiki entry on total miss chance for a complete explanation and chart of what slarden is talking about. You are correct that there are multiple layers of miss chance that an attack goes through. Any single attack is going to have to get through them all to land, so it is useful to consider how they "stack" with each other by calculating the total miss chance. When you calculate the chance of an attack getting through the first one, then the second, third, etc., you end up with a total miss chance for that attack. The more miss chance you have from more sources, you will receive a lower return on your total miss chance by increasing any one of them. In this regard, there is a kind of diminishing returns for damage avoidance just like there is for damage mitigation. In the chart you can see that going from 20% to 25% dodge when you have 50% concealment and 25% incorp only yields a 1.88% increase in your total miss chance.

This is one argument against elusive target. When you consider all other sources of miss chance, the additional 5% from it only ends up adding about 1% more total miss chance. You do not simply avoid 5% more attacks.

[Grace_ana]

Here is the wiki entry on total miss chance for a complete explanation and chart of what slarden is talking about. You are correct that there are multiple layers of miss chance that an attack goes through. Any single attack is going to have to get through them all to land, so it is useful to consider how they "stack" with each other by calculating the total miss chance. When you calculate the chance of an attack getting through the first one, then the second, third, etc., you end up with a total miss chance for that attack. The more miss chance you have from more sources, you will receive a lower return on your total miss chance by increasing any one of them. In this regard, there is a kind of diminishing returns for damage avoidance just like there is for damage mitigation. In the chart you can see that going from 20% to 25% dodge when you have 50% concealment and 25% incorp only yields a 1.88% increase in your total miss chance.

This is one argument against elusive target. When you consider all other sources of miss chance, the additional 5% from it only ends up adding about 1% more total miss chance. You do not simply avoid 5% more attacks.

Oh, I understand total miss chance completely. The problem is that it isn't relevant to what depositbox was saying, which is what slarden was commenting on. Depositbox said he doesn't need more dodge because it makes no difference. That was the foundation of the discussion. When you start from that point, using a total miss chance is incredibly misleading.

[gwonbush]

This discussion is really just about two ways of viewing miss chance: relative miss chance (if I took 100 hits before, how many do I take now) and total miss chance (How many times do I get hit in 100 swings).

It's easiest to showcase the difference between these views at the extremes: let's say you get avoid 98 swings of every 100 then manage to raise this to 99 swings. From the view of total miss chance, you have increased the miss chance by 1%. On the other hand, from the view of relative miss chance, you now get hit half as much as you did before (1 hit every 100 swings instead of 2).

[brzytki]

I don't think you understand how total miss chance works properly. As each hit chance is multiplied together, it is completely fine to look at any term in isolation to determine how much less you get hit.

The correct way to calculate how much less damage you take isn't [change in total miss chance] /[ original total miss chance], but 1-([New chance of getting hit]/[original chance of getting hit]).

Let's even throw in some numbers:

Assuming displacement, 27% dodge, 10% ghostly and 5% chance to be missed due to AC. In other words you have a chance to be hit of (1-.50)*(1-.27)*(1-.1)*(1-.05) for a total hit chance of 31.2075% chance to get hit.

Bringing dodge up to 32% gives us (1-.50)*(1-.32)*(1-.1)*(1-.05) for 29.07% chance to be hit. If you calculated it using your faulty method it comes out to ~3.1% chance less to getting hit, but that's faulty math. 29.07/31.2075=~.93151, which EXACTLY correlates to the dodge increase I showed in my previous post.

"Slarden's" method is in no way faulty and it works just like he said. To calculate how many times you'll get hit out of 100 you don't need to calculate a relative increase in miss chance, that's not what it is for. In your own example you increased dodge by 5 percentage points and got a little under 2 percentage points increase in total miss chance. That's what slarden, Cthru and me are talking about. That is enough to tell you that out of 100 hits 31 hits will connect using the lower dodge value and 29 hits when using higher one. Relative difference is a wrong way to look at it as is illustrated in your next post because you take an extreme situation:

This discussion is really just about two ways of viewing miss chance: relative miss chance (if I took 100 hits before, how many do I take now) and total miss chance (How many times do I get hit in 100 swings).

It's easiest to showcase the difference between these views at the extremes: let's say you get avoid 98 swings of every 100 then manage to raise this to 99 swings. From the view of total miss chance, you have increased the miss chance by 1%. On the other hand, from the view of relative miss chance, you now get hit half as much as you did before (1 hit every 100 swings instead of 2).

Do you know, that to reach 98% hit avoidance you'd need 1 source of 98% hit avoidance or 2 sources of 87% hit avoidance? Or 3 sources of 73% hit avoidance or 4 sources of 62% hit avoidance and so on? If you get even one source that is lower, say by half, you'd need (hyperbole here) a dozen that give you 90% hit avoidance.

Let's work on more real numbers:

• 50% concealmeant+25% incorporeality+25% dodge = 71,9% miss chance 1-(0,5*0,75*0,75)
• 50% concealmeant+25% incorporeality+30% dodge = 73,8% miss chance
• 20% concealmeant+25% incorporeality+25% dodge = 55,0% miss chance
• 20% concealmeant+25% incorporeality+30% dodge = 58,0% miss chance

First two rows: increase of 5 percentage points in dodge gives you 2 more percentage points in total miss chance. That means that (assume rounding for argument's sake) you are missed 74 times out of 100 now while you were missed 72 times before. 2/72*100%=2,(7)% relative increase in total miss chance when you added 5 percentage points of dodge.
Last two rows: increase of 5 percentage points of dodge gives you 3 more percentage points in total miss chance. That means that you are missed 58 times out of 100 now while you were missed 55 times before. 3/55*100%=5,(45)% relative increase in total miss chance when you added 5 percentage points of dodge.
And that's not even counting AC which you did in your previous calculations. If i did, then the incraeses both relative and in total miss chance would be even smaller. It's very easy to comprehend: the more hit avoidance sources you have (concealment, incorporeality, dodge, AC, elusive target) and the higher the numbers are, the less of an increase in total miss chance the increase of any of them by 1% gives. That's because of the multiplicative nature of the formula used to determine the total miss chance.

[slarden]

This discussion is really just about two ways of viewing miss chance: relative miss chance (if I took 100 hits before, how many do I take now) and total miss chance (How many times do I get hit in 100 swings).

It's easiest to showcase the difference between these views at the extremes: let's say you get avoid 98 swings of every 100 then manage to raise this to 99 swings. From the view of total miss chance, you have increased the miss chance by 1%. On the other hand, from the view of relative miss chance, you now get hit half as much as you did before (1 hit every 100 swings instead of 2).

Yes there are multiple ways of looking like at it - just like PRR. Each point of PRR gives you the same relative percentage protection against the damage you were taking previously (which keeps going down with each point of PRR). The other side of the argument is that each point of PRR provides less total damage reduction than the point before it. Both are true.

If my total miss chance changes from 20% to 19% I am certainly happy, but since the biggest risk I face is a megacrit and that only happens in 1 in 10 quests or so, I may not even notice the impact of that. The other way of looking at it is that I just reduced incoming damage by 5% (20-19)/20 vs. what I had previously, which is also true. If I find myself having trouble keeping up with heals that 5% could be very helpful.

The same is true with elusive target. It always reduces damage by 5% compared to what you had before the elusive target (assuming no other changes), but it might only change your total miss chance by 1%. I took spell power positive over elusive target which changed my positive spellpower from 284 to 304 (7% more benefit from my cocoon). My reasoning, which others are sure to disagree with, is that while 7% more healing isn't necessarily any better than 5% less incoming damage - it has the added side benefit of reducing the # of times where my cocoon is not good enough. It also theoretically reduces my total spell point usage (by how much I have no idea).

That is the great thing about this game - there are so many ways to play it. That is also unfortunately the reason why the devs didn't bother to balance int and dex. Because while all the people actually playing assassin agreed dex would be far better than int, we disagreed on the specifics of why it wasn't balanced. So in the end int builds were killed by the devs. The non-assassin dex builds that wanted to add assassinate to their build all agreed that it should be added and no balancing was required. The devs bought the non-sense reasons why int was equal - most likely because they wanted to make the change and didn't want to go through a little extra work to balance.

As for my math being flawed/faulty - no it's not.

[brzytki]

And another test from me, this time about helplessness and SA for those who asked. Here's a link to the gallery of images of my tests on the topic.

As you can see i was testing it on a regular dummy (not helpless) while unarmed. I've reset my destiny and enhancements, turned off active past lives so that i didn't have any +damage bonuses, stripped of all my gear (except for Golden Guile for Imp Deception and Flawless Black Dragonhide for more fort bypass). Because there is no easy way to get rid of ship buffs (except for reincarnating) keep in mind that i have 0,25[W] and +5% helpless damage buffs active.

So, here's how it goes:

• unarmed regular hits are 1d3+4, so the range is 5-7
• regular hits with MP and 0,25[W] making the range between 7-11
• SAs with MP in my screenshots are in the range between 59-75
• helpless regular hits with MP are in the range between 10-16 or 10-17 depending on the rounding (10-15 in my SS)
• helpless SAs with MP in my SS are: 42+54 / 54+71 / 60+79 / 58+76 , so:

1. 42+54=(x+54*0,55)+54 --> x=42-29,7 --> x=12 or x=13 (depending on rounding)
2. 54+71=(x+71*0,55)+71 --> x=54-39 --> x=15
3. 60+79=(x+79*0,55)+79 --> x=60-43,45 --> x=16 or x=17 (depending on rounding)
4. 58+76=(x+76*0,55)+76 --> x=58-41,8 --> x=16 or x=17 (depending on rounding)

I think this answers the question.

Edit: Oh, and for those who will catch the difference in HP of the dummy - if you don't kill it, it goes back to full HP after a few seconds without going into helpless state.

[slarden]

With the freed up feat I am trying to figure out the best way to use it.

I was looking at improved feint which is most likely not worth 2 feats, but I was wondering if anyone has any experience using it. Is the animation slow? is an 83 bluff enough in higher level EE content?

[gwonbush]

As for my math being flawed/faulty - no it's not.

Yeah, as I got a better view of the situation, I realized that I was looking at it only from a relative view point and assuming it was superior, when both view points are equally valid.

As for the 98%->99% miss chance, I mostly chose the numbers at the unsustainable extremes to showcase the differences between the two views. While it is reachable on a dodge based build that then activates Improved Uncanny Dodge or Primal travel, it definitely isn't sustainable.

[depositbox]

is an 83 bluff enough in higher level EE content?

A bluff of mid 50s works in ee storm horns. 80s should no fail everything everywhere. never used fients, sorry. bluff was just something i tried out for fun when i saw my bluff was high with fot ring equipped.

[CThruTheEgo]

This discussion is really just about two ways of viewing miss chance: relative miss chance (if I took 100 hits before, how many do I take now) and total miss chance (How many times do I get hit in 100 swings).

It's easiest to showcase the difference between these views at the extremes: let's say you get avoid 98 swings of every 100 then manage to raise this to 99 swings. From the view of total miss chance, you have increased the miss chance by 1%. On the other hand, from the view of relative miss chance, you now get hit half as much as you did before (1 hit every 100 swings instead of 2).

While I understand how the relative perspective works, I don't think it's a useful way of looking at it. Let me give an example. Let's say you have 0 avoidance and get hit 100 times out of 100. Then you increase your total avoidance and get hit 50 times out of 100, so you get hit half as much as you did before. Relatively speaking, you increased your miss chance by the same amount as in your example, but in this case you get hit 49 times more than in the first case, which is a big difference but is unseen by looking at it from a purely relative perspective. So, I don't know, the value of looking at it relatively seems relatively low (hehe) imo. The only situation where this seems useful is in considering the comparative costs (in feats, enhancements, gear, etc.) between different options of improving your miss chance.

[Tinco]

The only situation where this seems useful is in considering the comparative costs (in feats, enhancements, gear, etc.) between different options of improving your miss chance.

It's very useful to determine changes in updates, as it's very valid viewing the pre-update situation as a baseline. If a class that has an old baseline of getting hit 40% of the time and a new of of getting hit 30% of the time, imo it's more valid to speak of a 25% avoidance buff.

[VinoeWhines]

Though the cost always factors in to an equation.
The real "live" chances in combat when you have other "variables" thrown in is what makes a "little more" always better. Sometimes you won't need it in certain skirmishes where you never got down to the third form of save. But since this isn't a static game, there will be times where worst case scenario can happen on multiple levels, with multiple adversaries.
You take getting neg leveled, glancing blows, enemy IPS ranging and AOE's/DOT's, I'll take my 1% any time I can get it.
If your gonna barb Rambo into melee, than you don't really care.
But in EE TOR soloing the first part it's a matter of sticking around to wait for the door to open or clicking "release" as you see your soul stone drop.

[CThruTheEgo]

It's very useful to determine changes in updates, as it's very valid viewing the pre-update situation as a baseline. If a class that has an old baseline of getting hit 40% of the time and a new of of getting hit 30% of the time, imo it's more valid to speak of a 25% avoidance buff.

That makes sense. I can see the value of it in this situation as well.

[bbqzor]

I was wondering if anyone has any experience using it. Is the animation slow?

Its so slow it doesnt need a delorian to go back to 1985, its still there. Havent tested in awhile, but dont recall ever seeing a change in patch notes either.

is an 83 bluff enough in higher level EE content?

Probably, although lately Ive had some mobs seemingly unaffected by a mid 70s. Its probable the extra 10ish you have would get you to a livable point, but I havent tested consistently at that dc value. To be into no-fail, you may need to see 90+ (if they no-fail my 70ish youd need at least 20 more to get to yourself to no-fail, etc).

[slarden]

Its so slow it doesnt need a delorian to go back to 1985, its still there. Havent tested in awhile, but dont recall ever seeing a change in patch notes either.

Haha

I might roll up an iconic shadar-kai to see if it's been improved. I am out of character slots so I have to figure out what to delete first.

[Tinco]

Haha

I might roll up an iconic shadar-kai to see if it's been improved. I am out of character slots so I have to figure out what to delete first.

It's actually pretty short. Well, in theory it's very long, but you can cancel it by simply attacking (no special action needed) and the effect goes through regardless.

[depositbox]

Probably, although lately Ive had some mobs seemingly unaffected by a mid 70s. Its probable the extra 10ish you have would get you to a livable point, but I havent tested consistently at that dc value. To be into no-fail, you may need to see 90+ (if they no-fail my 70ish youd need at least 20 more to get to yourself to no-fail, etc).

interesting. the bluff check for mobs is 10+hit dice+ wis mod+ scaling. quick way to check if the rule is true is to enter ee fire peaks and see if you can bluff the dragon(any). if you pass then the formula is false as they have hit dice something like 1d80k+600k hp.