An exercise in chance: A gambling thread...

A very nice well-wisher sent me a message the other day that said something along the lines of "I may not agree with everything you say, but your posts always make me think."

Well, to return the favor, I'd like to share how much YOU guys in the Kheldian community make me think! Contrary to "popular belief," I actually do pay good attention to every new post in this section of the forums (I rarely go anywhere else on these boards!).

Recently, I made a post about a particular build I made on my second Warshade (Coincidentally named "Worst Warshade Ever"...haha!), based on a post by LordXenite in another thread regarding the "presence" pool...

Well, there are quite a few more people on these boards that I appreciate, and have caused me (more than once...or twice...or 8 times) to "go back to the drawing board" as it were, with VestigeOne's build. Justaris is one. Lord Xenite is another. Princess Darkstar comes to mind, as does Bill Z Bubba. Redlynne is another. Even Obsidian, when he's throwing out stats instead of insults, is a pretty awesome source of information.

I cannot count how many times I tell myself that "THIS time will be the LAST time I respec VestigeOne!"

It never works.

Now, a few people have told me "Dude! Why does a Warshade who gets 'ordered' by the team leader to take a tower by himself while the rest of the team takes the other tower during an iTF need a respec?"

Well... I blame you guys. Based on information contained in posts YOU all have put on these boards, I find myself constantly trying to think of different IO combinations, different set bonuses, and different calculations to figure out how to "max" my human-form build and maintain (or up) his damage capacity, without losing any of his surviveability or having to drop a single toggle (while fighting).

A interesting recent thread is the culprit for my latest respec endeavor with VestigeOne.

Now, let's try to set aside any "tri-form vs. human-form" arguments here, if that's possible. This is purely a "brain exercise" on my part, if you will...

Everyone's posts in that thread were really informative, but specifically Redlynne's post, and particularly this quote

Quote:

Originally Posted by Redlynne The question is ... is that chance for profoundly increased effectiveness "worth it" to you over something more "obvious" and enduring such as a simple Flight IO? For a lot of people, the answer is simply going to be "NO" ... just because they look at the spreadsheet of statistical analysis and figure "it's not worth it" ... without having actually experienced the effect in actual gameplay.

The latter part of the quote (bolded for emphasis) is what got my wheels turning... I am personally guilty of being the sort of person who not only never takes the "chance" IOs in any of the sets, I sometimes in the past have encouraged others not to, "because it's not worth it."

Now, chalk that up to my habit of always "5-slotting" a set (because the 6th set bonus--especially on purple sets--never seems to be worth it) and wanting to get "as much damage as possible" from the power.


But... am I really getting as much damage as I can from a power by avoiding the "chance" IO? Well, technically, according to a "spreadsheet" that never trusts the *reliability* of "chance", yeah... I am. As an added note (that may also have bearing on my choice), I've never been the "gambling" sort of person.

"Always go with the sure thing," I say.

But, what is chance? And is taking the chance worth it?

I was extremely satisfied with the "continuous, never-ending, health/endurance bars always full" build that I had...

However, I thought, "It would be cool to see something pop off like that in my normal attack chains... How would it affect them? If one or two procs popped off at the same time, would it shave 3 or 4 seconds off the *usual* time it takes to take a mob down... randomly? This would be interesting to see!"

First, I decided to check and see what kind of actual numbers I'd be changing on each of the powers, and overall in the "global" numbers...

Let's take (for example) the "Gravity Well" power... With Hetacomb 5-slotted in it (without the proc IO), it did 383.8 points of total damage at any given time. Well... Since I already had it 5 slotted, how much would taking out the Hetacomb: Damage IO hurt it?

Not a heck of a lot. Taking that IO out took the overall damage down to 363.8 total points of damage. Hmph. 20 points is 20 points....But, is it worth a tradeout with a proc that has a 33% chance to do an extra 107.1 points of damage on top of that?

I was convinced. I wound up changing out ALL the straight damage IOs (losing a few overall points here and there) for a LOT of "chance."

On top of that, I even traded out my Gravity Shield for Gravimetric Snare, based off of a recent build by Princess Darkstar (which I was a bit hard on). That not only saved me some slots to put elsewhere, but yet another "chance" for some damage mitigation (the chance for hold proc)...

"Who knows?" I thought... "Maybe one 20% chance for something to proc seems like a little thing... But, how about seven or eight 20-33% chances for something happening working in tandem?" Something's bound to happen, right?

Sure enough! During the next few missions I did, I had to sometimes double-check my buffs just to make sure I wasn't being fulcrum shifted or something...

I'm converted. I'm a believer. I'm gonna have to alter my guide and my guide's builds a bit to reflect this newfound knowledge, thanks to you guys...




...You remember that "you know you play too much CoH when..." thread? Yeah, this is how I know I play CoH too much...

Here's the new build.

So, how do you view "chance?" Has it proved to be worth it to you in-game?

"Alien"

*EDIT*
Using the same "chance" philosophy, I decided to respec my tri-form build as well...

Quote:

Originally Posted by Infernus_Hades The way you play changes your IO slotting..

I basically have three things to say:

1. Congratulations on yet another well-written post.
   
   
2. The [Chance for *] procs wouldn't have been added by the Devs if they were completely worthless.
   
   
3. Knowing however how computer random generation works, there are still some procs I'd generally avoid slotting in some powers.

I only use a proc if, on average, it will do more damage than the alternative. This is often the case for purple sets. It also tends to diversify your damage type which is an added benefit.

Ok, I'll play.

If you assume a 33% chance (purples) then roughly 1 in 3 attacks will proc for 107 points of damage or an improvement per attack of around 35.666666666666666666666666666667.

20% chance being around every 4-6 attacks per proc with substantialy reduced damage.

As always damage procs and procs like it are almost always best used on AOE attacks because the checks happen vs. each enemy in the area of effect with equal chances to affect each.

What is commonly missed and the reason I like the Guassiens proc in toggles, is the damage spike especialy when speaking of an AOE attack or better yet group of attacks.

While over time the improvement in damage can be small (depending on the timespan you measure, it all adds up.) That spike in damage can make a huge difference between killing a bunch of mobs in a single attack or having them almost dead and shooting back at you while you finish them off.

Other notable procs I have seen used to good effect are the chance for 100% recharge placed in an aoe attack able to do KB with a semi-quick recharge like White dwarf foostomp, and the previously mentioned Guassiens chance for build up proc in Nova.

For sure... Like someone else said once, "you can't *depend* on a proc poppin' off, but when it does, it's VERY nice."

I could tell a definite difference between when I didn't have the procs slotted and when I did. I'd see random enemies getting held (whereas they weren't before, meaning it was just that much more damage mitigation), and there'd be a few times where I hit Gravity well on a lvl 52 boss (at 75% health after Sunless Mire), and he'd go down in just that one hit, due to the proc hitting...

Just tonight, I was doing a LGTF, and I happened to get to the AV on the "rescue captives" mission before anyone else did... Now, I realize this is just blind luck, but I hit Gravitic Emination, Gravity Well, and Gravimetric Snare on him, targeted a minion, hit sunless mire, hit Gravity Well on him, followed by Unchain Essence, and then hit the AV with with another Gravitic Emination... and gosh darn it if he didn't walk around stunned for over 10 seconds... On top of that, "chance for" hold wound up popping off once or twice during that attack chain.

Amazing stuff....when it happens.

Any other "testimonials?

"Alien"

Quote:

Originally Posted by Infernus_Hades The way you play changes your IO slotting..

Excellent post.

I would have to say that after further experience with all the procs just today, I'm not sure how I thought before they couldn't benefit me...

The whole approach to mathematical equasions in this game is something I've wondered about for some time now... When a game puts together a "random chance" system, how does it work? I mean, each system may work differently in any game that uses "chance" for drops (say, the Diablo II game, for instance)... So, how does this one work?

Is it as "reliable" as a "constant and reliable damage increase?" I'm sure that over a longer period of time (as Obsidian points out), the numbers are a little more accurate, but exactly how random is it?

If one looks at the Law of large numbers, you'll see the proof that the numbers become a little more reliable and predictable in the long run, but are those numbers from constant, continuous gameplay? If so, then how does that match up to any player in game? Or, better yet, should a "casual player" slot procs, because their gameplay may not be enough to warrant the "possible" extra damage over time?

"Alien"

Quote:

Originally Posted by Infernus_Hades The way you play changes your IO slotting..

I don't have much proc experience on my WS, but on my scrapper they are great. The obliteration proc goes off at least once every time I lotus drop a decent group and the ToD proc is a nice addition to the relatively low damage gambler's cut.

The gaussian's proc is another story. If I didn't like the set bonuses I would drop it, and on my WS build I have. I may have made a mistake somewhere in here, but in a toggle it has a 5% chance of activation every 10 seconds. So it fires roughly once every 200 seconds and lasts for 5.25. During that time it boosts my damage by approx ~45%. So the effect it has on dps over a LONG period of time is 1+(5.25/200)*0.45=1.0118 or a 1.18% increase. And that only goes down as your damage buffs go up. Personally, I'd rather have the chance for +end in there.

Wonderful and informative post and I really have very little to add.

The main point I have noticed is the appearance that people who play less frequently or for shorter periods of time appear to be the ones who view procs as less valuable.

This is not a hard and fast rule obviously as both LX and AlienOne (currently) see the advantages of procs and I believe both do not conform to what i'm talking about, but whenever we begin talking about the advantages of procs in a build I have noticed the ones who are not happy using procs appear to be the sort who will run a few missions then do something else, switch toons or call it a night.

It's simply a pattern I have noticed on the forums, and I have no doubt if you played a character with setted out procs for just a few missions and did not come back to that toon for a day or three it could very well seem they are not going off very often.

I am not a mathmatician, and have only gotten fairly good at running numbers by necessity and dogged determination. My skill is seeing patterns and trying to use math to find out why.

This is just a pattern I "think" I am seeing and would love to see a more skilled mathamatician figure out the why's of.

Could it be that over very short periods of play the straight enhancer comes out ahead of the proc? And if so, why?

While I "get" what your saying I personaly try to leave out likes/dislikes from questions of performance.

Having said that likes and or dislikes of procs do not change their value, I simply lack the mathmatical ability to accurately judge their value on performance as were now getting into what I call meta-math. A single proc has a qualifiable percentage but back to back procs follow a very different probability which is beyond me.

I did once read multiple procs in a single power follow a roughly halved value in succession, meaning a single 33% chance proc in a single power gives 33% chance while 2 in the same power do not give a 66% chance but "around" a 49% cumulative chance of one of the procs firing and 3 in the same power around a 57% chance of one of them firing and not the 99% chance one would think.

But as I said previously were now into mathmatics well beyond my ability.

If I were to guess based soley on my subconcious perception of a pattern I would say the procs do a better job.

As an aside I use the same purple set in Gravity well and left out the end/dam one to keep the damage "capped" in addition to having the proc. I might try running the numbers on that to see how that looks to you.

I have a little more to add to this discussion.

One benefit to procs: They allow you to exceed the ED cap. Three damage SOs will hit the cap, and adding any more will increase the power of an attack very little. Adding damage procs are an option to significantly increase damage output after that cap has been reached.

One problem with procs: They are not affected by damage buffs. That 20 extra points of damage your gravity well does every time you click it will be increased by sunless mire and any team buffs. The damage proc is the damage proc, even fulcrum shifted to the damage cap. On the flip side, they'll still do full damage through a rage crash. Not that warshades have a rage crash, but any -dmg from enemies won't affect your procs. Resistances still play into the equation as normal.

So my conclusion is "Add damage enhancers to the ED cap, then fill the rest with procs, unless it's a fast charging, low damage attack, in which case, fill with damage procs, unless you can reliably get high self damage buffs, in which case add damage enhancers to the ED cap, then fill the rest with procs." That's just a general rule.

Another plus in favour of damage procs are quite a few of them are irresistable.

Okay ... now that I've had a chance to get over my shock at corrupting one of this forum's gurus (while in process of creating a guide no less!) into looking at things in a different light.

Er ... *ahem*

I don't think you're going to be able to say (or at least not with a straight face, anyway) that Procs are always better than the alternative(s). Just like building a Kheldian in Mids' doesn't have a single answer to the proposition, Procs are going to have to be a value judgement by the player depending on a whole lot of variables ... many of which you can't compute on a spreadsheet to accurately predict in-game playing performance. Adding (or rejecting) Procs when slotting simply has to be a value judgement on the part of the player. There's really no other way to think about it.

Each Proc has its advantages and its drawbacks ... however some general guidelines do apply. Just be mindful of the exceptions to the rule.

Fast/Light Damage powers tend to (notice, "tend to") "leverage" the power of Procs more effectively/efficiently than Slower/Heavier Damage powers often will. That's a trend to be aware of, not a dogma to adhere to. This is because faster powers offer more opportunities for Chance to Proc and other considerations. AoE powers (Cones, et al.) also offer substantial opportunities to leverage Procs, sometimes in unusual or innovative ways. Slotting a Perfect Zinger Proc into Tanker attack powers, to leverage the Taunt effect of Gauntlet AoE on Tanker attacks for instance, can result in some peculiar advantages in combat.

The important thing to consider, with Damage Procs, is that when the Proc "hits" does it add enough effect to "feel like" a crit to you or not? With faster, lighter damaging attack powers, it's perfectly possible for a Proc to very nearly double the damage throughput of an attack ... which is very close in performance to Scrapper Criticals.

The most important thing to realize from a spreadsheet point of view is that while Procs can result in minimal/marginal advantages when analyzed over an infinite timespan, they can offer very obvious benefits when analyzed over remarkably short timespans ... hence why I commented about them being "nice" when they go off (and you notice). In this regard, I think that a good rule of thumb concerning slotting procs is the idea that if a proc actually "hits" then does it stand a good chance of "saving" you from needing to make an extra attack with the power that proc is slotted into? This is why thinking of Proc IOs in terms of "crits" can be very useful. You never know "when" you're going to "crit" on something, but when you do ... you get something extra. The key is, is that "something extra" actually powerful enough to save you the effort of making/animating yet another attack against whatever it is that you're fighting.

That's why things like the Lockdown Proc (+2 MAG Hold, 20% chance, not unique) is such a useful thing to put into Hold powers. It gives you a chance to "Overpower" (and thus, Hold) a Boss, who would normally not succumb to your Hold(s). Same deal with the Devastation Proc (+2 MAG Hold, 20% chance, also not unique) is also a big benefit to builds that have at least one Hold power, since the proc can stack with the Hold (no matter who it is coming from). These things can yield very obvious (and therefore, appreciated) tactical advantages in battle ... advantages which cannot be conveniently analyzed by a spreadsheet, but ought to be understood by players.

One thing that players ought to pay attention to, when going out of their way to slot a lot of procs into their build(s) is how you can get multiple procs to stack up with each other. As Alien One said ... a single Proc IO by itself is nothing much, but 7 or 8 of them slotted in a variety of powers can be absolutely game changing. Being able to "predictably" get 3 or more chances to Proc a specific effect per 10 seconds based on your attack chain starts looking like "pretty good odds" for having that particular effect going off "a lot" in your favor. I for instance have slotted up my Warshade with lots and lots of Pacing of the Turtle Procs in multiple attacks. This is because the duration of the Proc is one of the longest (20 seconds!) and it's something that can stack up into being remarkably effective when combined with the Warshade's native Recharge Slowing effect in their attacks. A mere 2 Procs within 20 seconds can easily put a target (or multiple targets!) at a -50% Recharge Debuff disadvantage (or worse!) ... which in PvE is as good as a -34% Damage Debuff (in terms of Foe Damage throughput).

So no ... Procs don't activate every time, but when they do, it's NICE.
I've had fun using Taunt on my Tanker, slotted with a Perfect Zinger Proc, as an occasional "finishing move" on Foes. I hit Taunt, the Proc goes off ... and kills a Foe (or two) because that's all the green bar they had left when I Taunted them. Same deal with Damage Procs in Damage Aura powers like Orbiting Death ... or Icicles, or Chilling Embrace ... where the Procs act like "crits" which can dramatically reduce the number of activation cycles necessary to "burn down" a Foe. PBAoE Stun powers (such as Inky Aspect) can slot the Razzle Dazzle Proc, which is a MAG 2 Immobilize (which I *have* tested and confirmed does work, despite not having any visual cue to indicate the proc has actually "hit").

And so on ...

All of these things can have tactical effects which do not yield (easily) to spreadsheet analysis. You have to PLAY them in order to be able to appreciate these effects.

And some effects are more "helpful" to your playstyle than others, depending on how you like to play ...

It's the end. But the moment has been prepared for ...

First: Excellent topic AlienOne

Second: I have always been in favour of procs, but only once you have at least 80% +damage in the attack (IE: Two damage IO's). At that point they average out better than the last straight damage IO, but they don't average out better than the first 2 (Not even purples). It is very easy to check that in mids because it can automatically calculate averaged damage.

Third: For your guide AlienOne I would put a procs section in the introduction section where you can talk about how on average over a long period of time the damage procs in your builds work out better than a straight damage IO, but that it is up to the readers discretion if they want to, and include a line saying how the effects may be less noticable to those who only play for short periods of time. That way you can make your guide include procs (Where you think they are better) and don't have to worry too much about how people feel about them.

Last: Purple sets are the place the advantage of procs is seen most (Because they do higher damage than most procs, and proc more often), the pure damage IO is completely overkill and is best left out for the proc (The pure damage IO is usually the cheapest for this reason). The only time this isn't true is if you can leave out an acc/dam or a dam/rech in a build where you don't have to care about accuracy or recharge (Or endurance etc). Personally even on a low budget build I will sometimes look out for a purple proc to put in an AoE, because the difference can be incredible.

Quote:

Originally Posted by Aggelakis PrincessDarkstar: "RAAAAAAAAARRRRRRRRRGHHHHHHHHHH SOMEONE IS *WRONG* ON THE INTERNET!"