Sentinel DPS sheet

[Hjarki]

2 hours ago, SwitchFade said:

I'm sorry, but including procs, epics and attack rotations that are irrespective of standardized builds or enhances renders all the data invalid.

 

If you want to compare these sets you would have to remove variables that skew data so badly that it is uncorrectable 

Those variables are essential to evaluating a set. If a set is really strong except for a single hole, but that hole is filled effectively by an epic pool it's not really a 'hole'.

 

Attack rotations are essential to actually evaluating the damage potential of a set. While looking at attacks in isolation can often be useful, you really need to know the rotation you'll be using to see how it will perform as a whole.

 

I do think a bit of a caution needs to be used with respect to the 'proc-ability' of a set. While some procs - notably FF in AE powers - dramatically change the damage output of a set, most procs are merely additive damage. So while I can say with reasonable certainly that a Water Blast Sentinel (or Blaster/Defender/Corruptor) is better at AE damage than a Fire Blast Sentinel due to FF procs, I can't say the same thing about a ST attack chain unless I first assume some level of +damage. If my build's attack @ +200% damage can bridge the ST dps gap with a single proc in each of the attacks, I would need two such procs @ +400% (above Sentinel damage cap, but principle remains the same).

[oldskool]

@Frostweaver

 

Psychic Blast can feel a bit slow for both its animations and its cooldowns.  My Psi/EA has perma-hasten level global recharge and it still has some gaps.  That's partially due to the amount of frankenslotting I've got going on with TK Blast, and Will Dom.  To help counter this somewhat, I run a Force Feedback in TK Blast to give me a chance at some recharge bursts. 

 

At full range I have TK Blast, Will Dom, Scramble Thoughts, and Dominate.  Scramble and Dominate refresh in just over 5.5 seconds.  TK Blast doesn't recharge within Will Dom most of the time but the gap is less than a second.  Plus, FF can iron that out on occasion.  Ageless can also help, but I don't like being fully reliant on it.  

However, that's with purple sets and all that goodness.  What you have is more budget friendly, and I think you can still salvage it.  It doesn't need to be all that crazy either.  

 

Think about dumping Mental Blast.  Think about picking up Psionic Strike.  Yes, Psionic Strike.  Since you're not pushing a procced out Dominate, Psionic Strike isn't a bad filler.  Or more importantly, Psionic Strike is a better filler than Mental Blast.  You're not going to suddenly move mountains but it would be a slight upgrade.  

Whenever you're not in melee range you don't have Mind Probe to fill in gaps.  So when you're attacking from full range you have to run Mental Blast as a filler due to the general animation/recharge time frames.  Psionic Strike has a 2.9 arcanatime animation but it's still better than Mental Blast.  It also does have higher base damage to compensate the slow factor.  That power can help make the ranged attacks feel a bit smoother, and you can probably drop it for Mind Probe when you need.  

I'd take the full 6 slots and set from Mental Blast and shift it to Scramble Thoughts.  May as well ED cap the damage and get 62% recharge into it while you're at it.  You could consider swapping Entropic Chaos with another 5pc Decimation.  Its a little better on recharge mod and accuracy than Entropic while still giving the same global recharge bonus.  I'm not that fond of the heal proc though. 

 

You could go a minimalist slotting route on Psionic strike.  Like 3x Thunderstrike or whatever.  I don't know if I'd really worry too much about how I enhanced it since it is low priority and filler.  Still, I'd want to at least hit harder.  

Everything looks fine though for the most part.  If you ever complete Accolades then you'd be over cap on hit points.  You wouldn't need to slot Dismiss Pain that way anymore.  You could just go with a Heal, Heal/End without the boosters and be about 10 hp over vs 100.  You could move the proc to Health or Fast Healing.  
 

With the way this build is set up, Bio Armor is really only adding damage from Offensive Adaptation.  The mitigation may or may not feel better.  

 

 

[SwitchFade]

2 hours ago, Frostweaver said:

You know, except that procs, epics, and attack rotations are an important part of the game. And direct determiners of actual in-game performance rather than some magical, nonexistent, even keel model.

And you wonder why economists are always wrong.

Fortunately for all of us, that's not how statistics works.

 

Valid data sets are na integral part of analysis. With the data present no aNova analysis could be performed with ANY reasonable rSquare value.

 

Further, when doing comparison analysis, there must be a basis for statistical inference.

 

I understand your feelings about how things work, but that is not statistical fact.

 

Glib comments aside, your comments have little to do with the fact that there's no valid information gleaned from a corrupt data set.

[SwitchFade]

41 minutes ago, underfyre said:

1) Please elaborate on "standardized builds." I have never once seen two people build the same exact way. Pretending that Epic attacks aren't available when comparing between builds is going to skew data more than making them available. I've already said why procs should be present. As for attack chains, a lot of attack chains would rather pause for 0.2 seconds for an ability to refresh than just keep spamming whatever is up. That's a reasonable argument. I, however, don't have the spreadsheet know-how to implement something like that.

 

2-3) You know that everything about the game is based on random numbers, right? You have between a 5% and 95% chance to hit a mob and a dice is rolled for every. single. attack. A spreadsheet like this operates between now and infinity. If I have a 95% chance to hit with an ability doing 100 damage, and you do a parse of 100 attacks, you will more than likely come up with a number around 9500. Will it be exactly 9500? Probably not. But it will be really close. Extend that to 1000 attacks, you'll probably edge closer to 95000.

 

4) The only bias in this sheet is the fact that it's for Sentinels only. There were times when I really wanted to see if modeling something would give me a result that would beat Fire Blast, such as factoring in the -res debuffs from sonic just to find out that that ain't it. Actually, one bias I'm working on is that I have the up-time of Hybrid Assault Core factored into its effect, where I don't have it factored into the Radial, which technically it would be doing half of its damage since it's up for half the time (up 120 seconds, down 120 seconds). It's a "quick" fix that I have to apply to every sheet, which makes nothing quick, but I'll probably be doing it soon.

 

At the end of the day, if you don't like it, don't use it. 

When creating a valid data set, randomization in multiple variables across unstratified fields renders any data invalid, as the R-square value, p score, z score, median etc will be meaningless.

 

If you're attempting to do a comparison analysis, you must have a basis of reference.

 

Standardization is one such measure, among others. Don't cherry pick one example from a Sentence to attempt to refute it.

 

Sorry, but the maelstrom of variables, possible permutations, bias and the lack of common framework renders the data largely unusable from an analytical standpoint.

[Underfyre]

13 minutes ago, SwitchFade said:

When creating a valid data set, randomization in multiple variables across unstratified fields renders any data invalid, as the R-square value, p score, z score, median etc will be meaningless.

 

If you're attempting to do a comparison analysis, you must have a basis of reference.

 

Standardization is one such measure, among others. Don't cherry pick one example from a Sentence to attempt to refute it.

 

Sorry, but the maelstrom of variables, possible permutations, bias and the lack of common framework renders the data largely unusable from an analytical standpoint.

Well rather than come here and tell me how months of my work is wrong, make your own, better spreadsheet since you clearly know more about math, statistics and spreadsheets than I do.

[Frostweaver]

17 minutes ago, oldskool said:

@Frostweaver

 

Psychic Blast can feel a bit slow for both its animations and its cooldowns.  My Psi/EA has perma-hasten level global recharge and it still has some gaps.  That's partially due to the amount of frankenslotting I've got going on with TK Blast, and Will Dom.  To help counter this somewhat, I run a Force Feedback in TK Blast to give me a chance at some recharge bursts. 

 

At full range I have TK Blast, Will Dom, Scramble Thoughts, and Dominate.  Scramble and Dominate refresh in just over 5.5 seconds.  TK Blast doesn't recharge within Will Dom most of the time but the gap is less than a second.  Plus, FF can iron that out on occasion.  Ageless can also help, but I don't like being fully reliant on it.  

However, that's with purple sets and all that goodness.  What you have is more budget friendly, and I think you can still salvage it.  It doesn't need to be all that crazy either.  

 

Think about dumping Mental Blast.  Think about picking up Psionic Strike.  Yes, Psionic Strike.  Since you're not pushing a procced out Dominate, Psionic Strike isn't a bad filler.  Or more importantly, Psionic Strike is a better filler than Mental Blast.  You're not going to suddenly move mountains but it would be a slight upgrade.  

Whenever you're not in melee range you don't have Mind Probe to fill in gaps.  So when you're attacking from full range you have to run Mental Blast as a filler due to the general animation/recharge time frames.  Psionic Strike has a 2.9 arcanatime animation but it's still better than Mental Blast.  It also does have higher base damage to compensate the slow factor.  That power can help make the ranged attacks feel a bit smoother, and you can probably drop it for Mind Probe when you need.  

I'd take the full 6 slots and set from Mental Blast and shift it to Scramble Thoughts.  May as well ED cap the damage and get 62% recharge into it while you're at it.  You could consider swapping Entropic Chaos with another 5pc Decimation.  Its a little better on recharge mod and accuracy than Entropic while still giving the same global recharge bonus.  I'm not that fond of the heal proc though. 

 

You could go a minimalist slotting route on Psionic strike.  Like 3x Thunderstrike or whatever.  I don't know if I'd really worry too much about how I enhanced it since it is low priority and filler.  Still, I'd want to at least hit harder.  

Everything looks fine though for the most part.  If you ever complete Accolades then you'd be over cap on hit points.  You wouldn't need to slot Dismiss Pain that way anymore.  You could just go with a Heal, Heal/End without the boosters and be about 10 hp over vs 100.  You could move the proc to Health or Fast Healing.  
 

With the way this build is set up, Bio Armor is really only adding damage from Offensive Adaptation.  The mitigation may or may not feel better.  

 

 

I think I got the crap animations confused... Will Dom is the one that seems to take a lifetime to actually get to a target, although mental blast is slower than crap, too.

My only concern about losing mental blast is also losing offensive opportunity. I mean, regen... defensive is hardly useful. I do like the idea of losing the extra slots in pain, though... yeah, I lose some set bonuses, but I can shift those slots around to proc a few things and crank my damage up at least 40%.

[Zeraphia]

1 hour ago, underfyre said:

1) Please elaborate on "standardized builds." I have never once seen two people build the same exact way. Pretending that Epic attacks aren't available when comparing between builds is going to skew data more than making them available.

Now I actually agree with this, but there is the fact, and I'm going to bold for emphasis, you gave the Psionic pool to sets that you seemed to favor for one reason or another over other sets and gave those sets a higher ranking. For someone who doesn't understand these principles, who doesn't read all of the forum literature, this is extremely misleading, and a valid criticism. Fire Blast's highest rotation (from what I've done) is Blaze -> Blazing Blast -> Mind Probe -> Dominate -> Fireball (when up) and using Flares only when you have full opportunity. That is significantly higher than having two different parts of the rotation relying on Flares, versus other sets where you literally gave them full reigning advantage of the Psionic pool. That is not just taking into account procs, that is outright giving a huge bias in favor of other sets.

1 hour ago, underfyre said:

4) The only bias in this sheet is the fact that it's for Sentinels only. There were times when I really wanted to see if modeling something would give me a result that would beat Fire Blast, such as factoring in the -res debuffs from sonic just to find out that that ain't it. Actually, one bias I'm working on is that I have the up-time of Hybrid Assault Core factored into its effect, where I don't have it factored into the Radial, which technically it would be doing half of its damage since it's up for half the time (up 120 seconds, down 120 seconds). It's a "quick" fix that I have to apply to every sheet, which makes nothing quick, but I'll probably be doing it soon.

As above, you're looking for something to beat Fire Blast. That is a confirmation bias, and the fact you've used Psionic pools for other sets and use Flares 2x in your rotation gives a pretty hefty bias? Look, if you're going to give one set the Psionic pool and all the procs galore, give it to them all and really compares apples to (at least) fruits. 

1 hour ago, underfyre said:

At the end of the day, if you don't like it, don't use it. 

You posted this, on a forum. Everything you say and everything you do on the internet is open for criticism. Some are going to bash you harder than others, that's just how it operates (though there is a line between bashing a user personally and criticizing). You created a data sheet with your own rules and sets, and you said to even take your own post with a grain of salt, so why are you so defensive over when people point out that it really (and I'm sorry I really don't want to come across as bashing you...) ultimately proves nothing about the relativity in terms of the DPS of the set, and the set rankings are extremely misleading for newer players who can't decide the powersets for themselves upon initial inspection, which they may have been lead to believe by the numbers you gave. That is actually quite harmful to them, because then they may pick up this set, not really look at how the numbers were created, make a character based upon this list and not see you used the psionic pool or a ton of procs and go "wait... why does my damage feel so low if I chose the highest dps that this guy listed?!" 

Edited June 16, 2020 by Zeraphia

[Frostweaver]

I am confused... why exactly is it okay for stalker comparisons to add in the pools that give snipes in a favored position to those sets that gain an advantage from them, but it's not okay for sentinels to compare with pools that are greatly influenced by the primary choice as well?

I mean, the reality is that psionic is one of the best epics for sentinels that can exploit them, and the other epics mostly suck, which actually adds in to the comparison. I took the favored pool on my favorite BR sentinel, even though it's not the best choice for BR specifically, simply because I had free slots and the other pools suck.

Do you simply hate sentinels and despise anyone that dares suggest that there are better builds that let them compare favorably to other AT's?

[GM Miss]

Hey guys! I love discussion (and knowing that I do the weekly discussions - you know I love constructive feedback!) but please be kind to one another. 

[oldskool]

23 hours ago, Frostweaver said:

My only concern about losing mental blast is also losing offensive opportunity.

I really only ever care about Offensive Opportunity in cases where I can actively build to shorten the downtime of Opportunity in general.  I generally want my attack chain to build up Opportunity in less than 15 seconds.  Ideally I want that in 13 seconds or lower.  Psychic Blast can potentially have issues with that.  Some sets are really good about it.  Especially sets with very strong T1 powers (Dual Pistols, Fire Blast, etc).  Some sets that are less reliant on their T1 or T2 make that choice easy as well as I can weave it into my rotation when Opportunity is up vs being largely dependent on it (Archery, Beam Rifle, Ice Blast, Dark Blast, etc.).  If I can't get the kind of damage I want within a 13 second window, or if the T2 is just far better (Psychic Blast, Water Blast, etc.) then I worry less about Offensive Opportunity.  If I am incorporating Dominate/Mind Probe, then I worry about it even less.  Those two powers contribute to letting me prune down other attacks as well as bring strong proc options.  

I'm not sure Offensive Opportunity is doing that much for you.  It can be a positive DPS gain, but the amount of that really does depend on how fast you're cycling Opportunity.  The below is ONLY thinking of single targets.  In large spawns any set can generally build Opportunity quicker because AoE powers contribute more to the meter (its reverse Defiance).  Still, even in builds where I actually include Offensive Opportunity I think my DPS gains are single digits.  I can get a proc to do that to my overall damage.  I always say that Sentinels pick up a lot of pennies though and going with Offensive Opportunity can help there.  

Right now, Scramble Thoughts has a 7.55 cooldown.  So I'll use that as an anchor point.  

 

Will Domination - Mental Blast - TK Blast - slight pause - Will Domination - Mental Blast  is 7.524 seconds without the pause accounted for.   

 

Mental Blast will get you 8 points of Opportunity meter.  Most of the other powers are 13 pts.  I don't recall off hand is Scramble is the same.  Some powers like it have been 14 pts of meter.  For now, we'll assume its lower though.  

 

Over the 7.5+ block of powers you generated 55 meter.  If this was sandwiched by two Scramble (eggs?) Thoughts it would come up to 81 pts.  Total time 13.86 seconds.  You'd need 9 more points to actually activate an Opportunity.  Will Domination will get you there and the following Mental Blast will automatically trigger it [Mental Blast cannot both contribute to 90% meter AND activate the effect.  It is one or the other in my experience].  That's an additional  3.168 seconds (funny enough that is Scramble Thoughts animatione time).  This means it takes around 17 seconds to rebuild opportunity on a single target.  In cases where you're hammering AOE it can be a bit faster.  

By including TK Blast in the mix you run the risk of triggering Defensive Opportunity if your Mental Blast misses.  On the plus side, Mental Blast + TK Blast is an easily repeatable 1-2 punch.  Just Mental Blast sucks.  

And yes, all of this thought I put into how Opportunity works is a very big reason why I think there is a problem with it. 

 

[Edit notes:  I decided to revisit my own Psychic Blast Sent to check on Mental Blast + Offensive Opportunity vs TK Blast (generic sets) vs TK Blast frankenslotted with procs.  TK Blast generic set contributes slightly better DPS than Mental Blast and that's including Offensive Opportunity.  The longer the downtime of Opportunity, the less value Mental Blast contributes because TK Blast spam is that much better.  TK Blast with procs can be better than either, but that only helps when there is enough recharge to make it work that well.  

With lower grades of enhancement slotting there could be room for Mental Blast plus Offensive Opportunity, but that downtime between activation is a real killer to its effect since it just doesn't contribute that much on its own.  Couple that with long animations and recharges in the later powers of Psychic Blast, and well the inherent kind of falters for the set.  The -20% additional resistance debuff is handy and can be nice against resistant enemies.  Thankfully, you can still get that with just TK Blast.] 
 

Edited June 17, 2020 by oldskool

[nihilii]

22 minutes ago, GM Miss said:

Hey guys! I love discussion (and knowing that I do the weekly discussions - you know I love constructive feedback!) but please be kind to one another. 

Yes!! I knew our secret plan to pretend argue with each other would land an orange name on the Sentinel boards some day.

Edited June 16, 2020 by nihilii

[Underfyre]

38 minutes ago, Zeraphia said:

As above, you're looking for something to beat Fire Blast. That is a confirmation bias, and the fact you've used Psionic pools for other sets and use Flares 2x in your rotation gives a pretty hefty bias? Look, if you're going to give one set the Psionic pool and all the procs galore, give it to them all and really compares apples to (at least) fruits. 

Out of curiosity, I changed the Fire Blast set from Fire Mastery to Psionic Mastery and slotted it like it is in the DP build. Then fixed its rotations. It caught up slightly, but it didn't pass it. The other issue is that all the sets are sharing the haste percentages from the Buffs sheet. I have yet to be able to use the same exact set bonuses across power sets and that is definitely something I need to migrate into each sheet separately. But that's also something I said I needed to do on the sheet itself in the welcome page. That's just simply is not a quick thing to fix. I have to change cell references in literally every single power in every single sheet.

 

You all do know that point of this spreadsheet isn't to use my numbers, right? There are literally THOUSANDS of variables that can changed in this sheet. To think that my half-assed slotting on some of the sheets is the end-all-be-all of what that set is capable of is quite a lofty ideal. The idea is for someone to make a copy in their Google Drive and use it as a platform to gauge what their build is theoretically capable of. If they find a much better slotting than what I have in there, they can send it my way and I can put the build in the master sheet so other people don't run into the problem you did thinking that this is some how the only way to build a set and see how little damage it will do.

[Underfyre]

16 minutes ago, nihilii said:

Yes!! I knew our secret plan to pretend argue with each other would land an orange name on the Sentinel boards some day.

My dream was to some day, draw out an orange to pin this thread lol

 

And maybe get Sunsette to come back and update the Guide at the top with all the new reviews since she's left.

Edited June 16, 2020 by underfyre

[Underfyre]

I don't know if anyone cares, which I'm sure you don't, but adding in drop downs for enhancements isn't extremely feasible it seems. I have to point to the available sets, which then points to the individual enhancements in the set, then further grab each bonus for that enhancement. That took an entire day to figure out. In order to expand that to more than just a single ability I would have to recreate the process for every single ability in every single page, which is simply not worth the effort to save people from having to type in some numbers they fetched from Mid's.

 

I could make it so it is just a single page with drop downs for each main instead of separate pages, but then the only difference between this sheet and Mid's would be the ability to see the DPS from a rotation and have the correct numbers used.

 

So now I guess I'll move on to trying to implement Mob resistances.

Edited September 26, 2020 by underfyre

[Marshal_General]

To me the only way to do a valid comparison for attack sets is to start with a base to see how they stand. That would be using level 50 IOs with an acc, end, damx3, and a recharge in each attack. I feel IO sets skew the base data since some AT sets fair much better with slotting them.

 

Once you have the base, then you can make a valid comparison and see how sets and procs change that.

[Underfyre]

On 7/6/2020 at 6:41 AM, Marshal_General said:

To me the only way to do a valid comparison for attack sets is to start with a base to see how they stand. That would be using level 50 IOs with an acc, end, damx3, and a recharge in each attack. I feel IO sets skew the base data since some AT sets fair much better with slotting them.

 

Once you have the base, then you can make a valid comparison and see how sets and procs change that.

Wish granted. No more useful builds. Just Base +5 IOs.

[Underfyre]

Version 1.4 up.

[Underfyre]

I'd also like to point out that @SwitchFade still hasn't provided anything but big words to refute this spreadsheet. Granted it took me months (nearly a year now total), but he's had 3 now. 

Edited March 6, 2021 by underfyre

[Underfyre]

My bad, This is a whole new sheet, so I forgot to change the setting from private. Fixed that.

[Underfyre]

1.6 up.

[Underfyre]

1.6.1 up

 

Minor fix that was majorly broken. Amateur hour over here.

[Underfyre]

2.1 up.

 

Have fun fiddling with theoretical Seismic Blast stuff.

[DarknessEternal]

Why do i get more than 39% accuracy with all powers against +4 opponent with no accuracy, to hit, opportunities, non-accuracy armor set, etc?

[Underfyre]

1 hour ago, DarknessEternal said:

Why do i get more than 39% accuracy with all powers against +4 opponent with no accuracy, to hit, opportunities, non-accuracy armor set, etc?

 

I'm out and about right now so I can't see what you mean, but accuracy is one of the things that hasn't been changed much at all once I had it set right. The only reason it would be higher than the base 39% is if the base accuracy is higher than 1 for the ability. So something like Archery will always be higher than 39% on every ability since they're all 1.05x accuracy.

[Underfyre]

Okay, so as the sheet stands with how it loads with Seismic/Invuln, Invuln has a 10% to hit bonus. Then on top of that, Opportunity gives a 3.75% -defense debuff, which is then reduced by 87% and further by brought down and additional 48% from purple patch to 0.23%. So from the moment you load the sheet up, a Seismic/Invlun Sentinel will start at 49.23% hit chance on a +4 target.