Stam Stacking vs Avoidance... Math I saw in a forum
Posted: Tue May 12, 2009 4:56 pm
(COPIED FROM MMO-CHAMPION THREAD)
As a rule of thumb, HP/Armor stacking is generally more effective for progression. Why? Because it relies less on RNG and healer skill. You may have AMAZING healers who know how to perfectly time heals through your inconsistent spikes, but MOST healers have knee-jerk spam reactions at best. This is the #1 reason for tank death in a raid, period.
I love the sample that Omen gave (what a charming fellow, btw). It is so accurate. Basically, according to his data, if you stack stamina, you will take 4/5 hits (20% avoidance! awesome). If you stack avoidance, however, you will tak 2/9 hits (77% avoidance! wow. I didn't think that was possible).
Now here's the thing about avoidance.
1. It has a cap
2. It relies on RNG
3. It only really fleshes out in HIGH-YIELD situations
What do I mean? Well our dear friend Omen gave Patchwerk as an example of where avoidance is especially key. Let's take a look at some cold hard facts.
http://wowwebstats.com/mcwqpcihrsd5q?s= ... &a=x25c00e
Not only does this guild fail miserably because they took 6:20 to down Patch (I took the longest duration fight available), but there is a really interesting thing that happened to the main tank.
He took 255/598 hits (57.4% avoidance). Forget the small detail that the average hit was 5200 (not 30k) and focus on the number of total hits for the duration of the fight. Anyone who has taken statistics will tell you that trying to argue a noticeable difference from a 1-5% occurance rate in 598 swings is just ludicrous.
In other words, you're going to get virtually the same results (total avoidance) whether your paper doll says 45% avoidance or 50% avoidance. The RNG does not decree that if you have 50% avoidance, you will take EXACTLY 299/598 swings, and if you have 45% avoidance, you will take EXACTLY 329/598 swings. The only time you will ever come close to seeing a noticeable difference in ~5% avoidance is when the sample size gets to be in the mid thousands.
And we are talking about the worst of the worst here, where simple fights take EONS. Want to see a real pro fight?
http://wowwebstats.com/liar3d4t2clsc?s= ... a=x1e87fa5
1:32 minute fight. 34/72 hits taken by the MT (52.8% avoidance). 72 swings. Am I making sense?
So what next? Hodir? Let's take a look.
http://wowwebstats.com/mnnhz1rrqkhza?s= ... 00013bbdf1
8:30 minute fight (well above hard mode). 73/157 physical hits landed (53.5% avoidance). 157 swings. Even +15% avoidance is completely irrelevant to 157 swings.
Oh, btw, there were also 62 Frost attacks for this parse, but they can't be dodged or parried. Remember when Omen was talking about those 40k hits? Yeah...frost damage...non-avoidable. Woops.
Now, you give me a boss with 1.0 attack speed, or a swing-applied debuff like Kologarn, and I will admit that avoidance is probably a good thing to stack....for THAT fight. However, as my first point indicated...
"As a rule of thumb, HP/Armor stacking is generally more effective for progression."
As a rule of thumb, HP/Armor stacking is generally more effective for progression. Why? Because it relies less on RNG and healer skill. You may have AMAZING healers who know how to perfectly time heals through your inconsistent spikes, but MOST healers have knee-jerk spam reactions at best. This is the #1 reason for tank death in a raid, period.
I love the sample that Omen gave (what a charming fellow, btw). It is so accurate. Basically, according to his data, if you stack stamina, you will take 4/5 hits (20% avoidance! awesome). If you stack avoidance, however, you will tak 2/9 hits (77% avoidance! wow. I didn't think that was possible).
Now here's the thing about avoidance.
1. It has a cap
2. It relies on RNG
3. It only really fleshes out in HIGH-YIELD situations
What do I mean? Well our dear friend Omen gave Patchwerk as an example of where avoidance is especially key. Let's take a look at some cold hard facts.
http://wowwebstats.com/mcwqpcihrsd5q?s= ... &a=x25c00e
Not only does this guild fail miserably because they took 6:20 to down Patch (I took the longest duration fight available), but there is a really interesting thing that happened to the main tank.
He took 255/598 hits (57.4% avoidance). Forget the small detail that the average hit was 5200 (not 30k) and focus on the number of total hits for the duration of the fight. Anyone who has taken statistics will tell you that trying to argue a noticeable difference from a 1-5% occurance rate in 598 swings is just ludicrous.
In other words, you're going to get virtually the same results (total avoidance) whether your paper doll says 45% avoidance or 50% avoidance. The RNG does not decree that if you have 50% avoidance, you will take EXACTLY 299/598 swings, and if you have 45% avoidance, you will take EXACTLY 329/598 swings. The only time you will ever come close to seeing a noticeable difference in ~5% avoidance is when the sample size gets to be in the mid thousands.
And we are talking about the worst of the worst here, where simple fights take EONS. Want to see a real pro fight?
http://wowwebstats.com/liar3d4t2clsc?s= ... a=x1e87fa5
1:32 minute fight. 34/72 hits taken by the MT (52.8% avoidance). 72 swings. Am I making sense?
So what next? Hodir? Let's take a look.
http://wowwebstats.com/mnnhz1rrqkhza?s= ... 00013bbdf1
8:30 minute fight (well above hard mode). 73/157 physical hits landed (53.5% avoidance). 157 swings. Even +15% avoidance is completely irrelevant to 157 swings.
Oh, btw, there were also 62 Frost attacks for this parse, but they can't be dodged or parried. Remember when Omen was talking about those 40k hits? Yeah...frost damage...non-avoidable. Woops.
Now, you give me a boss with 1.0 attack speed, or a swing-applied debuff like Kologarn, and I will admit that avoidance is probably a good thing to stack....for THAT fight. However, as my first point indicated...
"As a rule of thumb, HP/Armor stacking is generally more effective for progression."