PvE Guide - Redux

loroku

Eddiemon's Complete PvE Guide - Redux
==============================================================================
This is a rewording/updating of Eddiemon's PvE Guide. It's been a while since it was updated and several things have changed; additionally the original guide is a bit technical for newbies and can be a bit hard to follow at times. Full props to Eddiemon for the guide; this is just an update.
==============================================================================
Explanation of important terms:
Scaling: The increasing difficulty of your opponents based on personal community success
Rubberbanding: An increase in points multipliers based on how far behind the leader you are
Tanking: Deliberately losing missions to **** or reverse scaling effects
Grinding: Repeatedly completing a mission over and over
Node: Any individual battle (a circle on the map)
Base points: The normal, unadjusted amount of points you get for a node
Shard/bracket: Your sub-group in a PvE event (i.e. your group of 200 players, 1000 players, etc.)
==============================================================================

The points you get for each battle - also called a "node" in this guide - are determined by three things: the base points for that node, whether or not the node is recharging, and how far behind the leader you are.

1. Base points are set by the game.

2. Each time you defeat a node, there is a ~20% point reduction on a ~2.5 hour cooldown (I am unsure of the exact numbers at this time, but these are close). These stack, so if you quickly beat the node twice, you'll have a ~40% reduction on a ~5 hour cooldown. However, these cooldowns work like health: they gradually come back over time. So after beating a node once and waiting about an hour and 15 minutes, the node will have about a 10% point reduction.

3. "The leader" is ambiguous, because it depends on the event. For some events, the leader is always the global leader. In other events, the leader is the leader in your bracket (or shard). Either way: the further behind you are, the more points each node is worth.


Rule 1: Don't Grind
Grinding looks attractive. You are getting more points and you get to pick up all 4 individual rewards. What's not to love? Scaling!
There are two forms of scaling in the game - personal and community. Community scaling increases the level of opponents based on how many times the node is cleared in the event. This is offset by the amount of time other nodes are cleared and the amount of people who fail to clear the node.
Personal scaling is based on how well you perform. This isn't just based on wins and losses, but also looks at how healthy you are entering each match. So if you are healing in the event or even in battles outside of the event your difficulty will continue to scale up.
Community scaling drops between sub-events and events. Your personal scaling also seems to decay somewhat [anecdotal, not confirmed] between events and possibly between sub-events.
That 500 ISO bonus looks attractive and is worth just 1 more go of the node right? Not when you go to the fourth sub event and can't get any of the bonuses because everything is level 230. Basically if the node isn't on your path to winning, let it go.
Rule of thumb: This is based more on intuition than math at this point, but it's probably not a good idea to hit nodes more that about three times at once (unless it's the last couple of hours of an event, in which case go wild).

Rule 2: Manage your rubberbanding bonuses
Rubberbanding is based on your current point value compared to the current points leader. It can get as high as 10x the amount of base points.
If you have little to no bonus, the order you undertake nodes doesn't matter. Otherwise it is better to attempt the highest base value nodes first, because you'll be getting the most bonus for those high-value nodes.
On the other hand it is generally good to do low value nodes when you are at or above the 10x bonus mark, since you can get the most for those low-value nodes without raising your rubberbanding bonus as much.
Rubberbanding is difficult to manage because in order to do well, you will need to be out in front. But the further up you are, the more effort you have to put in to stay there. But everyone will have a rubberbanding bonus at some point because no one can play this game every 2.5 hours for 2.5 days straight.
Rule of thumb: If you are playing a lot, always do the high-value nodes first. If you are coming in late or think you're probably way behind, do the low ones first and switch to the high ones if they start to drop in value.

Rule 3: Usually let a node recharge
Apart from rubberbanding, the other bonus affecting your points is whether or not the node is recharging. Letting each node fully recharge gets you the most points for your effort / medpacks / etc. However, the way this works is radically different than when this guide was originally written, and now you have a sliding scale for how much a node is worth after each defeat, so letting a node fully recharge can be good but sometimes it can also be worth hitting a few high-value nodes several times to get lots of points. More research is needed to give the best advice on point value vs. return, but given Rule 1, it's probably best to let your nodes recharge as much as possible.
Rule of thumb: This is based more on intuition than math at this point, but hitting high-value nodes two or three times at once isn't a bad idea, so long as the scaling won't kill you. It's definitely worth grinding nodes a few times if you know you will be offline long enough for them to fully recharge (like hitting nodes 3 times before you sleep).

Rule 4: Don't rely on boosts
If you rely on them for sustained performance they will hamper your long-term ability to defeat nodes. If you need a boost to beat that opponent that one time for the special bonus mission, or to get past a progression mission, then that is a good use of a boost. Otherwise, thanks to scaling, regular boosts will mean that you come out of missions better than you should. Which means that your personal scaling will increase to allow for the boost. You end up as hard up as you were before you started abusing boosts, but you are also flushing ISO down the drain.
Rule of thumb: Boost are good to unlock nodes and to slam those last few nodes right before the end of an event if scaling has caught up to you. Don't use them otherwise.

Rule 5: Powered up characters are awesome
Obvious, but important. It's worth noting that 1* characters scale better than 2*s who scale better than 3*s and so on. A 3 star at 120 is about on par with a 2* at level 85. Their edge is that they can go to level 141. Even better, a 2* with only 2 abilities is designed to scale even better because their maximum level is lower again that a tricolour 2*. So if you have a level 85 2* with a 30 level bonus it is more powerful than any 3* can achieve. Even better if it is a 40 or 100 level bonus.
Rule of thumb: Always look at the buffed characters and be very very sure that they aren't better than your favorite regular team. Because they probably are.

Comments and clarifications are welcome.

Much more explanation is given below here and here.

loroku

I think the PvE guide is a little more nebulous because it isn't quite the precise matter of timing that it used to be. In fact, it's tough to really make a solid recommendation. If you wait for a 10x rubberbanding bonus, you'll definitely put forth less effort. But it also means you're 10x behind! I think it comes down to whether or not you're trying to get progression rewards, or actually rank. If you just want progression rewards, waiting for a solid rubberbanding bonus seems like the way to go. That will give you the least amount of effort for the most points as quickly as possible. Don't start until a few hours in, let the nodes fully recharge each time, and just take your time.

However, if you want rank, I think it comes down to a matter of effort. If you hit every node every couple of hours, your rubberbanding bonus will be poor, but you're going to be getting a ton of points. If you think about it, rubberbanding is really just a way to make everyone competitive. It rewards those who don't have the amount of time to spend grinding nodes every couple of hours. But I don't know if it's something that can slingshot you ahead of someone who is dedicated in clearing all the nodes every time.

Either way, clearly using the boosted heroes is ridiculously effective. If you're slamming enemies with heroes who have a 40-90 level advantage, you'll be faster and require far less healing. You'll also stay effective longer even as the enemies grow stronger. The effect compounds itself as you're able to recover more quickly and thus clear even more nodes. The not-so-subtle nudge into expanding your roster to be more effective in more events is probably a bigger factor than anything else.

Unknown

loroku wrote:

However, if you want rank, I think it comes down to a matter of effort. If you hit every node every couple of hours, your rubberbanding bonus will be poor, but you're going to be getting a ton of points. If you think about it, rubberbanding is really just a way to make everyone competitive. It rewards those who don't have the amount of time to spend grinding nodes every couple of hours. But I don't know if it's something that can slingshot you ahead of someone who is dedicated in clearing all the nodes every time.

I don't know. I don't have the time to grind, so i play twice a day. After breakfast and before i go to bed. Then I do a three hour grind the end if I can. One time I didn't top 10 (iso 8 brotherhood), but that has do with the fact that I forgot the main doesn't rubberband.

Unknown

I need some more clarification on how to get the top rank as I have only gotten 1-2 rank once (I'm not sure how I did it, but I was ecstatic when it happened). So is it better to hit every node once every 2.5 hours so all of the nodes recharge or just focus on the top nodes like essentials and the last few of them so you can rubberband more often? I usually get the top 150, but I would like to get top 25 so I get more covers more often. I'm just curious how that 10x bonus works.

jim1200

ecir2002 wrote:

I need some more clarification on how to get the top rank as I have only gotten 1-2 rank once (I'm not sure how I did it, but I was ecstatic when it happened). So is it better to hit every node once every 2.5 hours so all of the nodes recharge or just focus on the top nodes like essentials and the last few of them so you can rubberband more often? I usually get the top 150, but I would like to get top 25 so I get more covers more often. I'm just curious how that 10x bonus works.

I have played around with this a little bit and found that if I want a top position, then hitting every node once every 2.5 hours is the way to go. Obviously, no one can hit every node every 2.5 hours, unless you can play MPQ in your sleep. Rubberbanding is made so everyone can be competitive. However, if two people played over the course of 10 hours and one person cleared every node every 2.5 hours, while the other just cleared them once initially and then at the 10 hour mark, the person who cleared each node every 2.5 hours will generally get more overall points within that time frame. While rubberbanding varies in every PVE and can get you close to the top, relying only on rubberbanding will likely not get you a 1-2 position. So someone playing twice a day will probably not get a 1-2 place if they are playing against someone that is clearing the nodes more frequently.

I usually think about what my positional goal for each PVE at the start. If I want a 1-2 top finish, I think about making as many 2.5 hour node clears over the course of the tournament than others in your bracket. I can say that I have easily obtained a 1-2 finish by doing this; however, I wouldn't recommended it for your social life. Obviously scaling is also going to be a huge pain if you chose this route. If I am ok with a top 10, 25, 100 finish then I mix in some rubberbanding with a few refreshes as I can get to them. The more consecutive 2.5 hour refreshes you can get in during the course of the PVE will likely get you placed higher. In the end, it about getting your self into position for the mad scramble of the final hours where everyone is grinding away everything. This is my own experience and I think has worked for me. I hope it helps.

loroku

ecir2002 wrote:

I need some more clarification on how to get the top rank as I have only gotten 1-2 rank once (I'm not sure how I did it, but I was ecstatic when it happened). So is it better to hit every node once every 2.5 hours so all of the nodes recharge or just focus on the top nodes like essentials and the last few of them so you can rubberband more often? I usually get the top 150, but I would like to get top 25 so I get more covers more often. I'm just curious how that 10x bonus works.

I am pretty sure near-constant node clearing is the way to accumulate the highest rank. First, it makes intuitive sense: which should be best rewarded, the person who grinds like mad or the person who comes in late and just does a little? But to be sure, let's try an example:

Example 1
Let's say an event has 5 nodes at 100 base points each, or 500 points to clear all 5 nodes at the beginning. Let's assume person A - who is also the leader for this event - does this each 2.5 hours to maximize the refresh time. Since person A is the leader, their rubberbanding bonus (RBB from now on) is 0 each time. After 22.5 hours (10 clears at maximum each time), person A would have 10 x 500 = 5000 points. Now let's assume person B waited 22.5 hours and then cleared each node. Thanks to person A, person B's RBB should be 10x. So each node should be worth 1000. Meaning one clear would net person B 5000 points as well, for 1/10th the effort - right?

Except there is a flaw in that calculation! Person B's RBB is 10x for the first node cleared, but after that it is lower than 10x. After the first node, person B will have 1000 points and their RBB should be lowered to 8x. That means each node after the first one is now worth 800 instead of 1000. After the second node cleared, person B will have 1800 points, and their RBB will be lowered to 6.4x (or something similar; I don't know how it is rounded), meaning each node will now be worth 640 points (or something similar). Continuing using those assumptions: 2440 points + 4th node valued at 512 = 2952 + 5th node valued at 410 = 3362. So, person B's final total will actually be 3362, NOT 5000 like person A.

Granted, person B got 3362 points in one clear, which is a tremendous value for the effort compared to poor person A's 500 points each clear and doing it 10 times. But after 22.5 hours, person A will still be ahead. (However, another important factor is that person A will have defeated each node 10 times to person B's one, which - thanks to scaling - means person A will have a much tougher time going forward.)

There is no amount of time person B can wait and still catch up to person A, or any combination of node skipping that would help, unless an event had exactly one node (I think 10 is the fewest I've seen). So, generally speaking, we can say that clearing nodes more often will get you better results (for MUCH more effort and at the cost of higher scaling) than clearing nodes while trying to use a RBB.

loroku

Now let's try a second example that is much more complex to determine if grinding nodes is a good idea:

Example 2
Person C decides to grind each node 5 times each, gaining 100, 80, 60, 40, and then 20 points for each node at the beginning. Because person C is playing more, they will be the leader and thus have an RBB of zero compared to person A. That's 300 points per node - 1500 total - BUT it will take 12.5 hours for the nodes to recharge to full. After 12.5 hours, person A has cleared the nodes 6 times (first clear + 5 recharges) for a total of 3000 points. EXCEPT they didn't get 3000 points, because up to 1500 points, person A now has an RBB greater than zero. At their second clear, person A should have gotten 667 (1167), 222 (1389), 174 (1563), 100 (1663), 100 for each node, or 1763 for their final total instead of 1000. So after 12.5 hours, person A should have 3763 points instead of 3000 thanks to person C giving them a RBB for a short time.

Person C then plays again at 12.5 hours, and plays each node once before grinding them all 5 times again. The first node is worth 601 (2101) thanks to RBB, then 442 (2543), 324 (2867), 238 (3105), 175, for a final total of 3280. I'm not exactly sure how the 20% factors into RBBs (or if it's exactly 20%), but assuming it's still linear and 20% each time, that means that grinding each node down will get: 102 (3382), 88 (3470), 86 (3556), 84 (3640), 83 (3723). The next round would be 61, but then all 60s again since person C has caught up point-wise, and then the 40s and 20s would be normal, for a final total of 4324.

So what do we learn from this? After 12.5 hours, person C has played twice and ground each node down to 0, and come out ahead points-wise compared to person A who played each 2.5 hours but only did each node one time. HOWEVER, it's worth noting that during the majority of that time, person A was ahead, and it was only by grinding each node at least 3 times the second time that person C was able to catch up. If the event had ended after only 12 hours, person A would have won.

Also worth noting: person A will have defeated each node 6 times, but person C will have defeated each node 10 times. That means that thanks to scaling, person A will still have slightly easier enemies for each node than person C (which affects healing pack use and other factors not measured here).


What would also be worth discovering is what happens when person C grinds each node to zero each 2.5 hours. Unfortunately I am tired of doing all this math right now. I suspect, however, that grinding each node at 20% of base value, especially with a small RBB, will not come out ahead. So, that makes it seem like there is an optimal amount of time (that is less than 2.5 hours) to let nodes recharge that will maximize your point gain in this narrow example. HOWEVER this example is extremely limited and only has 3 players; in real life, there are hundreds of people all playing at random times based on their own schedules.


So what does all this mean? There is a balance between grinding and waiting. It makes sense; more effort is valued more than less effort TO A POINT, but they set it up to discourage people from playing the game 24/7 in order to win. In fact, I would suspect that grinding the nodes 24/7 would actually hurt your score, not to mention the scaling would eventually destroy you. So this means there is some sort of optimal point between grinding each node and waiting a bit between playing that yields maximum results. HOWEVER this point is impossible to predict since it is constantly changing based on how each other player is doing in the event! Also each time you beat a node, you are increasing future difficulty thanks to the scaling, potentially shutting yourself out of an event at some point.


Ultimately, your choices are probably most realistically determined by the number of boosted heroes you have for an event. If you have a full boosted roster then you can tend to ignore scaling to a larger extent, and hit nodes as much as possible and grind them down at least two or maybe three times each time you hit them. If you don't have a boosted roster and scaling will become an issue, you should not grind nodes and probably even let your rubberbanding bonus increase a bit to keep your point total high with the least amount of scaling you can get away with.

Regardless, I think it's clear from the above examples that there is no reason not to grind nodes down to the ground toward the end of an event. If you have less than 2.5 hours to go, recharge time is meaningless and so grinding each node to pulp will get you the best results - although again, this advice is pretty intuitive and everyone pretty much does this already.

loroku

loroku wrote:

Ultimately, your choices are probably most realistically determined by the number of boosted heroes you have for an event.

As this most recent event reminds me, sometimes your ability to even compete is determined by whether or not you own a specific hero that is required to tackle all the nodes. Nodes with hero-specific requirements are always an awesome source of points, since they tend to have high base points and the scaling is always low thanks to fewer people farming the node ("community" scaling is low).


Another non-insignificant factor: random chance! Specifically, which bracket (or shard) you end up in can make a HUGE difference in whether or not you are able to place. People either aren't sure or don't seem to be sharing what causes you to be in which bracket, but the prevailing theory seems to be it's just sort of randomly determined based on the time you join. There may also be a bit of a hidden PvE MMR involved, and it may also be a bit (or completely) random. In any case, this can make just as big of a difference as anything else, and you have (presumably) no control over it whatsoever.


Edit: And as I am learning, the devs apparently change the rules all the time - so note that some things said in this thread may or may not apply to a given event! In some events there is extremely little / no rubberbanding, or scaling - in others the scaling is apparently global instead of shard-based so it's much worse. It's tough to have good rules of thumb when the ground is constantly changing but generally you just have to adapt your approach, and the golden rule is still: the one who plays most will usually win.