Some time after Armor Up: where do we stand?

[Baktiotha]

Incoming damage is important.

It appears that you and I agree.

DR has no meaningful value in the game for the time type of toughness we are looking at (end game, challenging content).

If after a hit in that end game, challenging content the character has 2 hit points remaining rather than none, is that important? If I am playing the character and he now has the chance to heal -- when being at 0 he would be unconscious and unable to do so -- I would consider DR to have been meaningful.

I agree that as a component in overall defense DR is the least significant. That does not mean it is not meaningful.

The cell formula is already there to incorporate DR so why not use it?

Player DPS is not at this point in the discussion.

If monster attacks are at the rate of ~20 per minute and the probability of a hit is 40% then we expect 8 attacks to hit and 12 to miss. The fraction 12/8 tells how many misses occur for every successful hit. The value 1.5 is used as a multiplier to the speed of the monster's attack to tell us how much time we have to take an action -- in this case 4.5 seconds.

In game the monster is not going to engage in half attacks. It means that damage is coming on average in pulses of 3 seconds and 6 second. This sets a minimum time for (self) heals at 3 seconds.

If mitigation is not sufficient to reduce the incoming damage to less than the character's hit points all of this information is moot. The character has to be alive in order for the 3 seconds of time to have any value.

Next, the amount of healing must be enough to lift the character's hit points to more than the next amount of damage. If it does not then the character dies on the next hit regardless.

The model cannot assume that every form of (self) healing occurs simultaneously. Players must take action sequentially. Neither can it assume that sequential actions take place without delay -- players still must press keys and the electrons must still communicate to the servers.

I do not think that it is probable that any character is capable of (self) healing in end game, challenging content. I think that will be obvious when actual damage numbers from monsters are applied. I think at best (self) healing will be practical for some characters against the preliminary monsters -- the trash mobs.

However, none of the melee builds used as samples will have sufficient healing to recover in a boss fight. In fact at 3000 damage (which was identified as low end LE trash damage) the two rogue builds die immediately upon being hit.

All of this suggests that trying to incorporate (self) healing in the analysis is not a productive next step. It suggests that the only viable healing method is external healing from players running characters designated for that role.

And, if that is the situation then it is better to evaluate DPS than (self) healing since the melee roles in legendary content are those of glass cannons.

[Baktiotha]

A note about my previous post, the success rate of 40% is not chosen at random.

1) Of the 7 sample builds the successful hit rates are 43%, 40%, 39%, 57%, 34%, 38%, 30%. (Edit: average is 40%)

2) At 3000 damage the blitz rogue is at -16 hit points and the shadowdancer rogue is at -132. The heavy armor paladin must heal 245 hit points, the medium armor paladin 324, the ranger 530 and the barbarian 238.

3) Only the S&B paladin can skip a heal but doing so is risky since if hit a second time it would need 687 additional hit points.

And this is from a low end LE mob.

Question, what self healing is the barbarian using that restores 238 hit points in 3 seconds? Paladin lay-on-hands possibly handles the healing until LoH run out, then what? The ranger?

We don't need to worry about the rogues, they are dead.

The mitigation numbers don't change but the avoidance numbers do in boss fights because displacement is routinely neutralized by true seeing.

[Baktiotha]

Before anyone attempts to make accusations of derailing, recovery downtime is very much statistically based on survivability - affected by armor up, which is the thread topic, and therefore its discussion is on topic.

The question asked is, "Some time after Armor Up: where do we stand?"

Every effort to answer that question is on topic and not a derailment even when the effort doesn't match that of other posters. There seems to be at least as much interest in pursuing a model that incorporates the whole of a build as there is in developing one that only includes select elements.

It is no more a derailment to discuss the whole build approach than it is to discuss the partial build approach. They are simply competing concepts and both are wholly on topic.

Proponents of either could start their own threads and more tightly constrain the topic. However, within this thread there is room for both perspectives as well as others that may be completely different from or hybridizations of the two.

[Axeyu]

also against the rules.

I dont think it is.

The entire point is system mastery. This does include the entire system. Purposely leaving out very significant portions of said system doesn't fit into system mastery. In this thread you claimed math is the only way to play, yet are here in the same thread advocating leaving parts of the math out. Many a nerf in the recent, and not so recent past, has occurred due to ignoring portions of the math and focusing on DPS only - resulting in a more homogenized and less balanced game.

Are you also on the "if you don't have the ultimate ddo spreadsheet you cannot tell any useful information" track?
I am not opposed to anyone working out a way to account for recovery time, I just wouldn't use it much myself. I make and have made my spreadsheets to give me specifically the information I am looking for.

again, against the rules to bring past discussions into current discussions. It is a strange rule to be sure, but a rule nonetheless. No need to continue to bait with calls for evidence you know its against the forum guidelines to provide. You can look those up as well as I or anyone else can.

Then you would do well to stop throwing around accusations and assertions that you have no way to back up.

Conclusion: The more repetitive the advocacy to continue ignoring significant portions of the system gets, the more likely there is a back end agenda to base and support positions using purposefully incomplete data with. Due to several 1-2 year old examples of this we now play Gauntlet with better graphics. Green Elf has shot the food!

Not including it in a specific spreadsheet =/= ignoring it.

I have never supported any position on balance based on a spreadsheet alone, they just give very useful data points.
Your constant claims about peoples agendas are frankly ridiculous, less tinfoil and more facts please.

[Baktiotha]

According to the community guidelines --

11. Reposting and Quoting

Do not re-post or quote any materials which violate the Community Guidelines; replies that include quotes of violating material posted by another player may be removed, even if the reply itself contained no offensive content. This includes re-posting a thread or topic which has been previously closed or removed by the Turbine Community Team.

Depending on where the information is located, it may be that reposting would violate the rules.

It seems simple enough to me for players that are interested to do a search of the forums on any particular topic. It certainly seems prudent to not repost unless a poster is certain themself that doing so will not violate the guidelines.

[Chai]

I dont think it is.



Are you also on the "if you don't have the ultimate ddo spreadsheet you cannot tell any useful information" track?
I am not opposed to anyone working out a way to account for recovery time, I just wouldn't use it much myself. I make and have made my spreadsheets to give me specifically the information I am looking for.



Then you would do well to stop throwing around accusations and assertions that you have no way to back up.



Not including it in a specific spreadsheet =/= ignoring it.

I have never supported any position on balance based on a spreadsheet alone, they just give very useful data points.
Your constant claims about peoples agendas are frankly ridiculous, less tinfoil and more facts please.

-Not including it in a specific spreadsheet while telling everyone else its irrelevant over and over again for 15 pages = ignoring it, willingly.
-Demanding nerfs, rebalancing, game changes etc, on one metric and one metric alone, with lack of willingness to acknowledge how other metrics affect that metric, assuming it away as equal when it is most definitely not equal, all the while being willing to include other far less impacting metrics in the calcs = ignoring it.
-both done on purpose.
-resulting in homogenization and less balanced game.
-The data points are not "useful" when purposely tailored to be incorrect in order to support a position, due to willful ignorance of significant data which affects the calc far more than many of the things willfully included in the calc.

Conclusion: Stop dismissing things others want to include outright because you don't want to include it. The more repetitious this dismissal becomes, the more it appears (and rightfully so) that this dismissal is happening purposely and not simply because it makes the calc too complex as claimed (I sure hope we aren't calling something that looks like a 4th grade extra credit math problem too complex by adding 1-2 more lines to it).

[Chai]

According to the community guidelines --

11. Reposting and Quoting

Do not re-post or quote any materials which violate the Community Guidelines; replies that include quotes of violating material posted by another player may be removed, even if the reply itself contained no offensive content. This includes re-posting a thread or topic which has been previously closed or removed by the Turbine Community Team.

Depending on where the information is located, it may be that reposting would violate the rules.

It seems simple enough to me for players that are interested to do a search of the forums on any particular topic. It certainly seems prudent to not repost unless a poster is certain themself that doing so will not violate the guidelines.

Forumites been asked, both publically and privately, not to bring previous disagreements into the current thread. This has happened several times. At this point its baiting, claiming lack of evidence if it isn't pointed out, and will report the minute something from another thread showing the waffle appears. The issue I have of this is if we aren't supposed to bring previous disagreements into current threads, then demanding it to be done as part of the discussion should also be against the rules. Demanding evidence for every single thing stated creates a discussion which is 10-15% actual discussion, and 80-85% coverass, further overbloating threads with overbloating posts which are nothing more than one sentence which states or defends a premise, and three paragraph essays which must address every exception to the rule, quote sources, etc....

Or we could actually have a real discussion with a real desire to master the system, which purposely excluding everything that doesn't fit the narrative does not accomplish. This thread might be 10 pages long if this occurred.

[Axeyu]

-Not including it in a specific spreadsheet while telling everyone else its irrelevant over and over again for 15 pages = ignoring it, willingly.

Cancer research is irrelevant to evaluate the OP, that doesn't mean I never think it should be pursued. What is problematic with taking things one step at a time?

-Demanding nerfs, rebalancing, game changes etc, on one metric and one metric alone, with lack of willingness to acknowledge how other metrics affect that metric, assuming it away as equal when it is most definitely not equal, all the while being willing to include other far less impacting metrics in the calcs = ignoring it.

I have never asked for a change based on one metric and one metric alone. Even if I use a particular calculation as evidence the analysis that lead me to my position is still based on thousand of hours of gameplay and considerations of all aspects I can think of.

You also seem to be unaware of the difference between making an analysis under the condition that all else is equal and believing that all acutally is equal. Google "ceteris paribus".

-resulting in homogenization and less balanced game.

No, it was the reptilians that caused the homogenization and less balanced game.

-The data points are not "useful" when purposely tailored to be incorrect in order to support a position, due to willful ignorance of significant data which affects the calc far more than many of the things willfully included in the calc.

Do you even believe that yourself? What position is that I so deviously support?

Conclusion: Stop dismissing things others want to include outright because you don't want to include it. The more repetitious this dismissal becomes, the more it appears (and rightfully so) that this dismissal is happening purposely and not simply because it makes the calc too complex as claimed (I sure hope we aren't calling something that looks like a 4th grade extra credit math problem too complex by adding 1-2 more lines to it).

The complexity is not an issue for making the calc, its an issue for using it and evaluating the results. The more assumptions you make the more uncertainty in the results. As long as you know the limitations of the values those limitations do not make the results misleading or inaccurate.
From the many years we have been discussing this my take away is that you believe that I take far more from the results of calculations than I acutally do.

[Chai]

Cancer research is irrelevant to evaluate the OP, that doesn't mean I never think it should be pursued. What is problematic with taking things one step at a time?



I have never asked for a change based on one metric and one metric alone. Even if I use a particular calculation as evidence the analysis that lead me to my position is still based on thousand of hours of gameplay and considerations of all aspects I can think of.

You also seem to be unaware of the difference between making an analysis under the condition that all else is equal and believing that all acutally is equal. Google "ceteris paribus".



No, it was the reptilians that caused the homogenization and less balanced game.



Do you even believe that yourself? What position is that I so deviously support?




The complexity is not an issue for making the calc, its an issue for using it and evaluating the results. The more assumptions you make the more uncertainty in the results. As long as you know the limitations of the values those limitations do not make the results misleading or inaccurate.
From the many years we have been discussing this my take away is that you believe that I take far more from the results of calculations than I acutally do.

Somehow you thought I was accusing you?

Purposely excluding data means it is assumed as equal in any calc it is purposely excluded from.

While I do not need to accuse you directly, I do know how discussions which hinge on purposely leaving out relevant information end however. This is why I call it when I see it, so any objective reader understands what they are observing, and also understands that any conclusion which is drawn based on purposely incomplete data should not be taken seriously.

[Baktiotha]

As long as you know the limitations of the values those limitations do not make the results misleading or inaccurate.

It is very probable at some point a person will read the thread title and jump to the end of the discussion to see what the conclusions are. That person will not be aware of the caveats used. They will take what they read at face value. They will not be aware that there were limitations and as a result they will be mislead by what they read.

That does not happen if *all* available data is incorporated into the model. That means including damage. It means including damage reduction. It means including DPS. It means including healing -- both self and party. Of these the DPS and healing parts are hard. The damage and damage reduction parts are easy.

It would be sufficient for the thread to stop if the model could answer two questions: 1) Will my character be alive after being hit and 2) how does that compare to other melee characters?

The obvious data points that will provide the information in a way players will relate to are hits-to-kill and attack-to-kill. Both allow a one-to-one comparison of builds. The everyday players understand hits-to-kill and attacks-to-kill so the best model should present the information in that format.

The model would be more complete if DPS and healing were included. But those are questions that players can answer heuristically and will intuitively understand as not being included if DPS and healing data points are not shown in the model. With respect to making defensive comparisons the model could end with showing just hits-to-kill and attacks-to-kill. If it did so the thread could conclude today.

[Baktiotha]

The issue I have of this is if we aren't supposed to bring previous disagreements into current threads, then demanding it to be done as part of the discussion should also be against the rules.

8. Provoking, Trolling, and “baiting”

Threads or posts that are intended to create a strong negative or emotional reaction, provoke conflict, or are made simply for ‘shock value’ are considered trolling. Debating a topic on the merits or holding a strong opinion are acceptable, however making posts to provoke conflict or incite, bait, or mock others who disagree with you are not.

[Baktiotha]

Did not mean to cause discontent. I understood the comment to be that there was a wish that baiting be against the rules. It is.

[Axeyu]

It is very probable at some point a person will read the thread title and jump to the end of the discussion to see what the conclusions are. That person will not be aware of the caveats used. They will take what they read at face value. They will not be aware that there were limitations and as a result they will be mislead by what they read.

That does not happen if *all* available data is incorporated into the model. That means including damage. It means including damage reduction. It means including DPS. It means including healing -- both self and party. Of these the DPS and healing parts are hard. The damage and damage reduction parts are easy.

It would be sufficient for the thread to stop if the model could answer two questions: 1) Will my character be alive after being hit and 2) how does that compare to other melee characters?

The obvious data points that will provide the information in a way players will relate to are hits-to-kill and attack-to-kill. Both allow a one-to-one comparison of builds. The everyday players understand hits-to-kill and attacks-to-kill so the best model should present the information in that format.

The model would be more complete if DPS and healing were included. But those are questions that players can answer heuristically and will intuitively understand as not being included if DPS and healing data points are not shown in the model. With respect to making defensive comparisons the model could end with showing just hits-to-kill and attacks-to-kill. If it did so the thread could conclude today.

For people who just read the title and the conclusions they are better off the fewer assumptions and the less situational the results are.

Showing a list of HTK and ATK and claim that its a representative comparison of the builds would be misleading. That is because there is no damage number that can be used for those values to acutally mean anything. If they are displayed in a normalized list it would be better to simply jusy use toughness and EHP instead, as demonstrated many pages ago EHP is more useful than HTK.

[Baktiotha]

The thing about a model is that it is not stagnant. That means users can change data values to see what the different results are. That is sort of the whole purpose of creating a model.

We might change dodge percentage or armor class or hit points or concealment -- any value can be changed.

We can change monster damage. In fact, I suspect that to be one of the things players would want to be able to do most often. That is because their builds are generally the same -- once they have "everything" there isn't much change going on to the characters. But what does change is the difficulty of the quests -- and players want to know where the breaking point is for their character. That is represented by the damage the monsters do.

I think players will look intuitively at the damage value and modify it to represent the content they expect. We have been doing it all thread long with values like 1000, 2800, 3000, 4500, 6500. Each of them allows a snapshot of the characters at a particular place. And players will easily grasp that the snapshot is only valid for that place.

To show the challenges with (self) healing I referred to minimum posited damage from LE mobs and to dealing with LE bosses. That is seen rapidly with the current spreadsheet by setting damage to 3000 (LE mobs) or 6500 (LE boss).

I don't think players will be confused by that at all. To the contrary, the three things I think they will look for in the model are incoming damage, hits-to-kill and attacks-to-kill. I think those three elements provide the greatest information to players.

[Baktiotha]

Just as an example, I have a thread elsewhere about a build I call the Dwarven Defender.

I place the build in the spreadsheet alongside the sample builds to see how it stacks up. I move the damage value up and down to see the changes in HTK and ATK. Looking at those two values tells me a lot more about the build than looking at the TV for the build.

To anyone reading, which number or combination of numbers has most meaning in context of 3000 damage: Toughness Value 3.79, Hits to Kill 2.23, Attacks to Kill 9.84, Effective Hit Points 6686.91?

[Chai]

Just as an example, I have a thread elsewhere about a build I call the Dwarven Defender.

I place the build in the spreadsheet alongside the sample builds to see how it stacks up. I move the damage value up and down to see the changes in HTK and ATK. Looking at those two values tells me a lot more about the build than looking at the TV for the build.

To anyone reading, which number or combination of numbers has most meaning in context of 3000 damage: Toughness Value 3.79, Hits to Kill 2.23, Attacks to Kill 9.84, Effective Hit Points 6686.91?

For players it will be HTK. If tanking something that does that damage and the mob misses 3x in a row they keep doing what they were doing previous to those swings. No change in what theyre doing because no damage was processed. If the mob hits once its time to go into recovery mode, or for someone else to go into recovery mode (depending on the arrangement) - since HTK is what changes the player behavior, its most important. Furthermore, if they got hit once on the first swing, theyre not engaging in the "gamblers fallacy" believing that they wont get hit again on the next few hits simply due to a higher STK value - as the mob can still land a blow on the next swing leaving the character at dangerously low HP, and then instead of being in a more comfortable recovery mode, the player has to scramble quickly to get some recovery done before the next damage occurs. Chances are they will have to e less efficient as well with resources at that point, if the character lives at all, as a full on meta'd heal will be needed where a HoT could have helped earlier.

[Baktiotha]

for players it will be htk. If tanking something that does that damage and the mob misses 3x in a row they keep doing what they were doing previous to those swings. No change in what theyre doing because no damage was processed. If the mob hits once its time to go into recovery mode, or for someone else to go into recovery mode (depending on the arrangement) - since htk is what changes the player behavior, its most important. Furthermore, if they got hit once on the first swing, theyre not engaging in the "gamblers fallacy" believing that they wont get hit again on the next few hits simply due to a higher stk value - as the mob can still land a blow on the next swing leaving the character at dangerously low hp, and then instead of being in a more comfortable recovery mode, the player has to scramble quickly to get some recovery done before the next damage occurs. Chances are they will have to e less efficient as well with resources at that point, if the character lives at all, as a full on meta'd heal will be needed where a hot could have helped earlier.

[Axeyu]

The thing about a model is that it is not stagnant. That means users can change data values to see what the different results are. That is sort of the whole purpose of creating a model.

We might change dodge percentage or armor class or hit points or concealment -- any value can be changed.

We can change monster damage. In fact, I suspect that to be one of the things players would want to be able to do most often. That is because their builds are generally the same -- once they have "everything" there isn't much change going on to the characters. But what does change is the difficulty of the quests -- and players want to know where the breaking point is for their character. That is represented by the damage the monsters do.

I think players will look intuitively at the damage value and modify it to represent the content they expect. We have been doing it all thread long with values like 1000, 2800, 3000, 4500, 6500. Each of them allows a snapshot of the characters at a particular place. And players will easily grasp that the snapshot is only valid for that place.

To show the challenges with (self) healing I referred to minimum posited damage from LE mobs and to dealing with LE bosses. That is seen rapidly with the current spreadsheet by setting damage to 3000 (LE mobs) or 6500 (LE boss).

I don't think players will be confused by that at all. To the contrary, the three things I think they will look for in the model are incoming damage, hits-to-kill and attacks-to-kill. I think those three elements provide the greatest information to players.

Now you are back to creating the Ultimate ddo spreadsheet. You keep moving the goal posts.
Is it really unfathomable that I think doing X for the purpose of Z is bad and doing X for the purpose of Y is good?

I also disagree with what values you think provide the greatest information. Attacks to kill seems like a pretty useless number. If you are interested in how many hits you can survive the misses will not matter.

[Baktiotha]

Now you are back to creating the Ultimate ddo spreadsheet. You keep moving the goal posts.
Is it really unfathomable that I think doing X for the purpose of Z is bad and doing X for the purpose of Y is good?

I also disagree with what values you think provide the greatest information. Attacks to kill seems like a pretty useless number. If you are interested in how many hits you can survive the misses will not matter.

I think it could be a fundamental difference in how we view the problem. I view the problem from the perspective of a player and ask, "What information do I want?" You are likely doing the same thing. But our answers are different.

I see the game in a way similar to Chai. The most important piece of information is delivered by the red bar. When it dips I take notice. The bigger the dip the more attention I pay.

That doesn't mean that I don't look at the theories behind all of that. I want to know the targets for AC, dodge, PRR and so on. All of that helps me to build a character that won't suffer a huge dip in the red bar.

So all of the data points have value. It is just a matter of where our emphasis is placed.

To me the hits-to-kill number tells me a lot about any group of characters. But, the attacks to kill also tells me a lot. That is because attacks to kill tells me if I have time to self heal. Two builds with the same round up to whole number HTK and the same round up to whole number ATK are going to be functionally equivalent. It does not matter if the one build is 2.1 w 6.1 and the second 2.9 w 6.9 -- both take 3 hits to kill and have 7 attacks to work it all out so that doesn't happen. (Edit: And, because the hits are taking place inside the attacks I know I don't have time to dawdle before I do something since every other attack is doing damage to my character.)

As I push the damage value up the build on the bottom is going to drop out and the build on top is going to stay in for a bit. That tells me that the two builds are functionally equivalent in some content but not in all content.

And when I push the damage number to the highest expected damage in the game I can tell if the builds are functionally equivalent in the "end game, challenging content."

Many players have advanced degrees but just as many do not. I want the metric to be one that anyone can understand. The three data elements that everyone understands are incoming damage, hits to kill and attacks before dying. Use whatever abbreviations or variables or acronyms seem to fit.

In the long run a model that includes everything would be great. It would probably be a useful tool for Turbine as they could measure the impact of changes. But for concluding about where we are since Armor Up we only need to the spreadsheet with final values showing for incoming damage, hits to kill and attacks before dying.

[Baktiotha]

Let me talk a bit about goal posts.

The goal posts for me have always been down on the far end where we have a good modeling of the whole character. The goal posts for others have been somewhat closer.

I would say that there are three different distances where people have put out goal posts. The closest is to just incorporate avoidance and mitigation and call it a day. The next furthest out is to take that setting but move it a bit further with (self) healing. Then I'd like to have the posts set even further out to accommodate all of those things plus all healing and DPS.

All of those represent different viewpoints on where the goal posts should be set. When I advocate for including additional data points I am not moving the goal posts because I already set them much further out to begin with.

That does not mean I am opposed to getting to the first marking. We treated that as if it were the goal for a great deal of the thread. But, as I observed a few posts back, when we arrived we found it wasn't a destination it was just a way point.

There are some for whom the destination is (self) healing. I'll go along on that journey but for me the objective will still be further out.

If, however, we want to end this thread and stop at the nearest mark then, as I wrote previously, we can do that by using the spreadsheet that includes damage and damage reduction and also reports to us hits to kill and attacks to kill.

We could have closed the thread yesterday but can close it today on that basis. And the reason to use those rather than other metrics is because they represent what the players of the game will want to see and what is most intuitive to them when they compare builds.