L-RANDLE Posted July 23, 2012 Share Posted July 23, 2012 If I flip a coin, and heads gives me a million dollars while tails gives me nothing, the expected value of that coin toss is 500,000. If you opt to take a payout of 300,000 instead of the coin toss, it does not mean you are "right" in any sense, or that 300,000 > 500,000, it just means you're risk averse. Which is fine, that's your choice, but please stop pretending you have some theoretical high ground. Telling me to get off a theortical high ground, while you are standing right next to me on it telling people that WP wins? Why do you believe it? Because someone told you to? LOL ok.. Your "example" also disregards that you add a third "probability" to the equation, but I'll let that one slide. It also assumes one flip. Bad example. Here is a better one, you are in a storm, what is the probability of you being struck by lightning while holding a sand wedge versus holding a 5 iron? Either way ur toast... Link to comment Share on other sites More sharing options...
Aurojiin Posted July 23, 2012 Share Posted July 23, 2012 (edited) Telling me to get off a theortical high ground, while you are standing right next to me on it telling people that WP wins? Why do you believe it? Because someone told you to? LOL ok.. I believe it because someone told me, because I did the math myself, because I've seen and run simulations myself, and because I've seen parses consistently demonstrate it. That's a fair bit more evidence than you're providing. Your "example" also disregards that you add a third "probability" to the equation, but I'll let that one slide. It also assumes one flip. Bad example. Do you even understand what you're talking about? The expected value of a set of outcomes with known probabilities is basic, textbook statistical theory, taught the world over. It's simple probability theory, something you seem to have a very tenuous grasp of. Here is a better one, you are in a storm, what is the probability of you being struck by lightning while holding a sand wedge versus holding a 5 iron? Either way ur toast... Oh, you're deliberately trolling. Touche. Edited July 23, 2012 by Aurojiin Link to comment Share on other sites More sharing options...
Khevar Posted July 24, 2012 Share Posted July 24, 2012 (edited) I discovered that SimC did have a 100,000 iterations test but I got bored after 10 minutes and canceled it. Instead ran a 10,000 iteration version. I used this AMR profile from the 1.3 BiS thread: http://swtor.askmrrobot.com/character/a5189a89-2f15-4a76-aed0-b8fcb08b7cbd 1836 dps I then swapped out all willpower augments for power augments and re-ran the 10,000 iteration http://swtor.askmrrobot.com/character/1106480c-8172-4ea7-949a-1226a9dfc963 1825 dps Mind you, at that vast level of stats, 180 points swapped is sort of a drop in the bucket. I suppose I could spend time trying to swap out every single piece of gear that had +crit for a +power version for a more definitive test, but the above test was good enough for me. In the grand scheme of things, it's not like you're killing your dps by picking power over willpower. I mean, if you've already spent the credits on overkill augs, keep them. But for a true Best in Slot: Willpower > Power Edited July 24, 2012 by JeffKretz Link to comment Share on other sites More sharing options...
Khevar Posted July 24, 2012 Share Posted July 24, 2012 And the silence is deafening. No more witty retorts, L-RANDAL? No Sith comments? Exploded by Lightning Comparisons? Internet Lemming Accusations? Of course, I would be happiest with some sort of response like, "Well, that's interesting data, I'll have to look into that." But I decided not to ask the Magic 8-Ball if that was going to happen, because it would probably laugh in my face. Link to comment Share on other sites More sharing options...
LahLahSr Posted July 25, 2012 Share Posted July 25, 2012 Hi JeffKretz, I've been following this debate for a while as I was curious about the reigning view on this. Just wanna say thanks for make this very apples-to-apples comparison by setting up two fully built characters like that. That was very helpful! What I take away from it is that, on the whole, the difference really is quite marginal and that it's highly unlikely that any difference is going to make or break a team's chance of success in a raid. In other words: if you fumble the rotation or don't activate perfectly efficiently just ONCE in a boss fight, you've probably made more of a difference in the DPS than any decision on the augments you may have made WP/P-wise. After seeing this, I have decided not to get rid of all my power-augments and replace them with willpower-augments. Not only is such an undertaking rather cost-prohibitive, but more importantly: it's not going to make any practical and noticeable difference. Thanks for providing clear, readable data on this matter! Cheers :-) Link to comment Share on other sites More sharing options...
L-RANDLE Posted July 25, 2012 Share Posted July 25, 2012 And the silence is deafening. No more witty retorts, L-RANDAL? No Sith comments? Exploded by Lightning Comparisons? Internet Lemming Accusations? Of course, I would be happiest with some sort of response like, "Well, that's interesting data, I'll have to look into that." But I decided not to ask the Magic 8-Ball if that was going to happen, because it would probably laugh in my face. I shouldn't have to say anything it took 10000 tests to give a .6% dps output. WOW amazing. I think you just proved my point that in the realm of probabilities it 1.2% is meaningless. Go look up "outlier" and try apply it to our class versus another class and *maybe you will understand why your analysis is poor at best. Link to comment Share on other sites More sharing options...
Underpowered Posted July 25, 2012 Share Posted July 25, 2012 (edited) I shouldn't have to say anything it took 10000 tests to give a .6% dps output. WOW amazing. I think you just proved my point that in the realm of probabilities it 1.2% is meaningless. Go look up "outlier" and try apply it to our class versus another class and *maybe you will understand why your analysis is poor at best. You really don't know anything about statistics or probability do you? Please stop pretending like you do. I shouldn't have to say anything it took 10000 tests to give a .6% dps output. That's not the purpose of running 10000 tests. Go look up "outlier" Outlier has nothing to do with anything here. Stop. Just stop. Your freshman-level knowledge of statistics is killing us all. Anyway, for the rest of us that have any semblance of proficiency in mathematics, Willpower > Power. To take Power augs over Willpower augs is to lose DPS, plain and simple. Is the difference small? Yes. If you're fully augmented out in Power, is it worth your time and money to change them all to Willpower? Probably not unless you have nothing better to do with your credits. But if you're starting fresh, then start and end with Willpower. This thread is over. Edited July 25, 2012 by Underpowered Link to comment Share on other sites More sharing options...
Khevar Posted July 25, 2012 Share Posted July 25, 2012 (edited) L-RANDAL, I believe you'll find Aurojin said something similar a few posts up in this thread. My point is, that "Stacking Power is Not Better Than Stacking Willpower". "Better" in this case is minor and they're almost the same. But both math and actual testing show that power doesn't have the edge. However, it's obvious to me you're more interested in trying to "Be Right" than have a meaningful discussion. Feel free to post (or not). I don't believe there's anything else to get out of this. Edited July 25, 2012 by JeffKretz Link to comment Share on other sites More sharing options...
L-RANDLE Posted July 25, 2012 Share Posted July 25, 2012 (edited) Outlier has nothing to do with anything here. Stop. Just stop. Your freshman-level knowledge of statistics is killing us all. So I guess having the ability to overwhelmingly change odds on rolls means about as much as the change in DPS? Outlier has everything to do with this. We have a talent that throws a whammy in the test data... Our class is a "1" more than the "crit number" says (i.e. we crit WAY more than other classes with the "same" crit rating). That test data is good for every class, except ours because we are not static in our crit rating. Our "RL" or "effective" crit rating fluctuates between 90%+ and 30%+, so unless SimC accounts for this, which I have no indication that it does (if I am wrong about that then fine, I give), the value of our actual output is suspect, even past the actual meaninglessnes of .6%. Someone higher than freshman level should understand that.. Stack willpower, and be happy though. Edited July 25, 2012 by L-RANDLE Link to comment Share on other sites More sharing options...
Khevar Posted July 25, 2012 Share Posted July 25, 2012 So I guess having the ability to overwhelmingly change odds on rolls means about as much as the change in DPS? Outlier has everything to do with this. We have a talent that throws a whammy in the test data... Our class is a "1" more than the "crit number" says (i.e. we crit WAY more than other classes with the "same" crit rating). That test data is good for every class, except ours because we are not static in our crit rating. Our "RL" or "effective" crit rating fluctuates between 90%+ and 30%+, so unless SimC accounts for this, which I have no indication that it does (if I am wrong about that then fine, I give), the value of our actual output is suspect, even past the actual meaninglessnes of .6%. Someone higher than freshman level should understand that.. Stack willpower, and be happy though. Hmm. Maybe there is room for discussion. Perhaps I misunderstood your argument. It sounds like you're referring to the Force Potency skill to add +60% to the crit chance. To answer your question about SimC, here is the action list I used that modeled the DPS result. actions=stim,type=exotech_resolve actions+=/mark_of_power actions+=/snapshot_stats actions+=/power_potion actions+=/use_relics actions+=/affliction,if=!ticking actions+=/thundering_blast,if=dot.affliction.remains>cast_time actions+=/recklessness,if=buff.lightning_barrage.react|buff.lightning_storm.react actions+=/force_lightning,if=buff.lightning_barrage.react actions+=/chain_lightning,if=buff.lightning_storm.react actions+=/polarity_shift,moving=0 actions+=/crushing_darkness actions+=/sequence,name=pewpew:lightning_strike:force_lightning actions+=/restart_sequence,name=pewpew actions+=/shock,moving=1 "Recklessness" is the Sorcerer equivalent of Force Potency, so it is in the rotation. Link to comment Share on other sites More sharing options...
L-RANDLE Posted July 25, 2012 Share Posted July 25, 2012 Hmm. Maybe there is room for discussion. Perhaps I misunderstood your argument. It sounds like you're referring to the Force Potency skill to add +60% to the crit chance. To answer your question about SimC, here is the action list I used that modeled the DPS result. "Recklessness" is the Sorcerer equivalent of Force Potency, so it is in the rotation. Glad we can discuss more... I am not a programmer,a this rotation is exactly why I question stuff like this. It looks like Force Lightning is "eating up" both Reckless charges. Its like me using FP to guarantee a crit hit on TK throw. Can you confirm this or can you change the next sequence to "Chain Lightning" and rerun each test? Yes, I am sure my argument was misunderstood. I am accused of "talking in code" alot, but the fact that force potency/recklessness exists makes willpower's contribution moot in the normalized area of the SimC result. But what about the unnormalized portion? Questionable + Moot = Take Burst(POW) Link to comment Share on other sites More sharing options...
Khevar Posted July 25, 2012 Share Posted July 25, 2012 Unfortunately, I'm not familiar enough with the "actions" instructions of SimC to fiddle with it -- this was taken straight off their site as the "Best in Slot" dps test for gear. However, I would encourage you to fiddle with the program. It's primary use is to compare gear changes one to the next. My "theoretical" gunsligner rotation according to SimC was like 1302 as a Fresh 50 with Tionese/Columi and I don't get anywhere near that. But when I tweak my gear, re-run the simulation and see that it went up to 1382, I knew my gear choices were good. Then in game I get a similar increase of dps, even if I'm not actually hitting the same numbers. Link to comment Share on other sites More sharing options...
Aurojiin Posted July 26, 2012 Share Posted July 26, 2012 That instruction list basically says the following: 1. If a Lightning Barrage or Lightning Storm proc is available, and Recklessness is up, use Recklessness. 2. If Lightning Barrage is available, use Force Lightning. If Lightning Storm is available, use Chain Lightning. Consider that the majority of the times the simulation loops through that priority list, it won't use Recklessness/Force Potency because of the cooldown. But when it is available, it checks if the most efficient options for consuming Potency are up then uses it appropriately. In short, the sims fully account for the effect of Force Potency on DPS. Given the hundreds of attacks you perform over a fight, the ability does not significantly devalue the benefits of crit chance. Link to comment Share on other sites More sharing options...
CJNJ Posted October 27, 2012 Share Posted October 27, 2012 Because somebody just made another thread on this I figured I would bump it Link to comment Share on other sites More sharing options...
anthrobeau Posted October 28, 2012 Share Posted October 28, 2012 I read most of this thread, and thought there was something missing from this discussion. I apologize if this has already been addressed. While simulations may prove the relative weight each stat should be given, deciding on actual pieces of gear presents unique choices. For instance, it may be better to go for a mod with 53 WP and 41 Power instead of 61 WP and 11 crit. Since it's theoretically possible to have 30 extra power by sacrificing 8 WP (per mod), is it better to stack power? Link to comment Share on other sites More sharing options...
Aurojiin Posted October 28, 2012 Share Posted October 28, 2012 While simulations may prove the relative weight each stat should be given, deciding on actual pieces of gear presents unique choices. For instance, it may be better to go for a mod with 53 WP and 41 Power instead of 61 WP and 11 crit. Since it's theoretically possible to have 30 extra power by sacrificing 8 WP (per mod), is it better to stack power? That choice is pretty straightforward. Always use unlettered mods, because the lettered variety have a lower stat budget and thus are inferior outright. Link to comment Share on other sites More sharing options...
SoonerJBD Posted October 28, 2012 Share Posted October 28, 2012 What you are talking about is the lettered mods. And the answer is, no, you don't want lettered mods. Willpower is better than power point-for-point, but the lettered mods make you sacrifice about 2 secondary stat points for one point of Willpower. One point of willpower is not worth two points of power. The question about what is better point-for-point is really only relevant when it comes to augments because that is the only time you are making a 1 for 1 trade between main stat and power. If the choice is 1 main stat or 2 power (as it is for lettered mods), take the power. If the choice is 1 main stat or 1 power (as it is for augments), take main stat. Link to comment Share on other sites More sharing options...
Recommended Posts