Optimal sub-end PVE grinding strategies

Vhailorx

inside of the last 8 hours, it is obviously impossible to recharge any given node back to a full stack, but what is the optimal strategy for grinding down to zero?

is the rate of points recharged flat (i.e. X points regained per hour no matter how many times the node has been cleared)? if so, then it shouldn't really matter much what order you grind the nodes down, or when you do so. If not (i.e. X points per hour after once clear, but a lower number of points per hour after 2 or more clears), then it makes sense to clear nodes in sequence (one go each time so the first node has a few more minutes to recharge at the higher rate).

Similarly, the ideal timing would be to grind the last node down to 1 as the clock runs out, but that's clearly not possible in the real world. If the recharge rate is flat, than clearing early is less of a problem because any points you miss will just recharge anyway and can be collected at the very end if there is time.

Has anyone done the math on this?

Overall, these strategies won't add up to more than a couple hundreds points. But without community scaling, and with the super long end-of-sub grinds, scores can be fairly compressed at the top, especially in shorter events.

Unknown

The math's probably been done around here before, but I'll take a crack at it anyways.

The points regenerate at a constant rate of 1/6 max points for the node every 8 hours - that is, every minute, you get back 1/2880 of max points.

However, you only get the points currently "charged up" when you complete. Thus, you want to let the points recharge before clearing again.

Let's take a simple example: say you have a node worth 600 points on a 48-hour sub (this gives us time to fully recharge). We'll assume you can clear nodes instantly, to simplify calculations.

Method 1:
Clearing a full stack at the beginning, then again at the end:

600 + 5/6 * 600 + 4/6 * 600 + 3/6 * 600 + 2/6 * 600 + 1/6 * 600 = 2100 for the first clear

Then multiply by 2, since we fully recharge our points after 48 hours: 4200 through Method 1.

Method 2:
Clearing every 8 hours, then a full clear at the end:

600 + 600 + 600 + 600 + 600 + 600 + 2100 = 5700, clearly better than clearing the full stack at the beginning and end.

So you see, even with the same number of clears, the points gained can differ drastically depending upon time.

Furthermore, especially with the more difficult nodes, it becomes uneconomical to repeatedly clear. When you need 2 healthpacks a run, it's not worth it to go back for another 50 points when you could get them by doing the same amount of work half an hour later. In addition, if you finish early enough for X points to recharge by the end, you lose out on X points per clear, coming out 5X points short of where you would have clearing later.

In essence, every clear reduces the "base" amount of points for future clears. Obviously, over shorter times and less differences between clearing methods, the effects will be smaller but still present and like you said, every little bit counts.

Let's try to take into account difficulty/time spent to clear.

Say you have easy node E, which takes you X minutes to clear and is worth Y (max) points.
And hard node H, which takes Z minutes to clear for W (max) points.

Let's say we start 6X+6Z minutes from the end, just enough time to fully clear both stacks.

Method 1: Easy, then Hard

We first look at E:
The nth clear is worth (1 - (n - 1)/6 + (n - 1) * X/2880) * Y points.

Then for H:
The nth clear is worth (1 - (n - 1)/6 + (6X + (n - 1) * Z)/2880 ) * W points.

Method 2: Hard, then Easy

Looking first at H:
The nth clear is worth (1 - (n - 1)/6 + (n - 1) * Z/2880) * W points.

Then for E:
The nth clear is worth (1 - (n - 1)/6 + (6Z + (n - 1) * X)/2880) * Y points.

Comparison
We subtract our total for Method 1 from total for Method 2, which tells us the number of points we gain by executing Method 2 rather than Method 1. Since the point totals are sums with corresponding terms, we can simply subtract term-wise to obtain our final result.

For the Easy nodes:

(1 - (n - 1)/6 + (6Z + (n - 1) * X)/2880) * Y - (1 - (n - 1)/6 + (n - 1) * X/2880) * Y =

(1 - (n - 1)/6 + (6Z + (n - 1) * X)/2880 - (1 - (n - 1)/6 + (n - 1) * X/2880) * Y =

(6Z / 2880) * Y, or Y * Z / 480.

Note that this value is the benefit observed during *ANY* of our 6 clears (the n cancels out). At this point, I'll multiply by 6 to get the total value saved as Y * Z / 80. (You could work completely term-wise, but I'll reduce it because smaller numbers are less scary).

Likewise, the the point difference for the hard node is - X * W / 80. Note the negative sign, since we lose points on the hard nodes by clearing them first.

The total point difference is therefore

(Y * Z - X * W) / 80.

Which is positive (ie. Method 2 is better) when

Y * Z - X * W > 0

Y * Z > X * W

Y / W > X / Z

So if the hard node is worth N times as many points as the easy node, you want to clear the hard node first only if you can clear it in less than N times the amount of time it takes you to clear the easy node.

For example, let's say in today's Juggernaut Heroic, the first node is worth 200 points and the last node is worth 600. Let's say you can clear the easy node in maybe 1 minute. Then you should only clear the last node before the first if it takes you fewer than 3 minutes; otherwise you lose out on points. While this difference may be small (remember, we are dividing by 80 in the end), it can help in close situations.

Of course, this analysis doesn't take into account scaling, health, rosters, etc. I'll work on creating something more comprehensive in the future, but this will at least allow you to have some basic guidelines for clearing strategies.

Unknown

Vhailorx wrote:

Overall, these strategies won't add up to more than a couple hundreds points. But without community scaling, and with the super long end-of-sub grinds, scores can be fairly compressed at the top, especially in shorter events.

This is only true if the top few have very similar clear times and using the same clearing strategy. If there is a significant time difference in clearing a node by players, you will notice a much larger gap. Even if two players do an eight hour refresh and grind down in the final few hours. The player that can grind down in a shorter time span and closer to the end of the event will have more points, as the nodes will have refreshed more points than the guy clearing the nodes on the same strategy but slower at doing it.

I found the first time time brackets, the top placed guys tend to have a much tighter group of scores. However it's almost impossible for players to have identical scores because of the clearing time discrepancy between players, coupled with the fact the players may have different clearing strategies. Some people grind the node right down as soon as a sub is in the last eight hours, some people grind down in the last hour of a sub and others somewhere in between. Then Heal pack management also plays into it too, unless you spend HP on hero packs if you run out, this can especially at the end of a sub and on a much smaller roster impact when and how quickly you do a grind down.

Ryz-aus

Optimal strategy is to wait as long as possible to start the final grind, while allowing time to grind every node to 1. If you can accurately predict your timing and can handle the scaling, it's also better to start your grind on nodes worth the least points.

Points refresh linearly (1/48 of max for node per hour), so leaving higher point nodes to clear later gives you slightly more points. Unfortunately, I think it also will result in them having more difficult scaling - but if you're looking for the highest score possible that's the method to use.

Overall these are minor differences. The more important strategy is if you can't make your clears every 8 hours to clear early and not late. If you clear late your final grind will have had less time to generate pts and be worth less - because you are clearing each node 5-6 times the difference is significant. Clearing an hour early is superior to being 15 minutes late up for any refresh until the final grind.

Vhailorx

Thanks for the math qkumbre, but I was just asking about inside the last 8 hours of an event. I completely agree with you that clearing every 8 hours (and clearing as fast as possible) is the best strategy for pve.

The only real question I needed answer was whether points recharge linearly for all nodes regardless of how many times they have been cleared.

Unknown

Might be a bit hard to think about it this way, but points "regenerate" proportionately to how many more times you can clear them. Let's say that 600 point node has been cleared once, and it is at 500. If it regenerated for 8 hours to 600 and you clear it 5 times you get 100 more per clear. For you this node is regenerating at 500 per 8 hours. If you only can do 1 more clear, the same node is considered regenerating at 100 per 8 hours.

In the grand scheme of things, every "clear" remaining on the map is regenerating points, so it doesn't matter too much which order you do those, but if you want to grasp at some extra (single digit?) points, you can save the nodes with the highest points per clear time for last. Note that the node with the highest full pts is usually not the node with the highest pts per clear time.

To illustrate this, imagine there's two nodes with 5 clears left, an easy one that you can clear 5x instantly (imagine zero seconds) that is max 600 pts (regen at 100 per 8 hrs per clear left), and a hard one that will take 8 hours to 5x clear that has double the pts of the easy one.
- If you do the easy first then the hard, after you clear the hard node for 8 hours, the easy node regenerated 100 pts (0 -> 100) and you can do 1 clear of the easy node and gain 100 regenerated pts. The total points gained from the easy node is 500 + 400 + 300 + 200 + 100 + 100 = 1600 pts
- If you do the hard first, after you clear the hard node for 8 hours, the easy node regenerated 100 pts (500 -> 600) and you can do 6 clears of the easy node and gain 600 regenerated pts. The total points gained from the easy node is 600 + 500 + 400 + 300 + 200 + 100 = 2100 pts
- Since the easy node is instant clear, the order you do these two nodes does not affect the pts regenerated for the hard node. The pts gained from the easy node, however, differed by 500 pts depending on the order.

From this extreme case, you can see that even though the hard node has more pts, saving it for later resulted in fewer points because it takes too long to clear. In real life situations, pts per clear time are much closer between the different nodes, so the amount of pts that you can optimize is limited.

Unknown

one thing you also need to consider is the content of the node. Certain combinations are going to be harder to clear and/or take more time to clear. YOu can't just look at the points value of a node. Some players I notice will always save a relative easy high point node for the last few minutes of a sub... This is especially important if you are on the borderlines between getting additional character cover and/or reward or not. I seen a number of times a player in the tenth position ten minutes before the end of a full event replaced by some other person who managed to outscore them in the last few minutes of a sub/event. Heck I've even done it to one person during the event to win 2 hulkbuster covers instead of one. I suspect the very same thing happens with similar positions further down the reward structure.

Take enemy of the state, those events often have wave nodes, and each of those nodes especially if they have a lot of waves, can easily take 10-15 minutes to successfully complete. These are probably much better to complete early on rather than late because of the amount of time they can take and you definitely don't want these scaling out of hand!

If you also have an extensive roster, it's always a good idea to suicide one character or two in a match, this will actually slow down your personal scaling. I seen nodes drop levels because even though I successfully completed it, having a character or two defeated actually lowered scaling. You also probably want to avoid characters that use protect tiles because of how personal scaling works, you never want to walk out of a match taking zero or very little damage. Although this allows you to do nodes more often without using health packs, it actually slows you down because nodes have got harder and thus more time consuming to complete.

Additionally you want to avoid situations like using a fully levelled hulkbuster/ironfist combo on a lvl 40 goon node. This would do your personal scaling no favors what-so-ever! Try to use your roster characters relative to the level of node. i.e. don't use capped 4 stars on nodes lower than lvl 140, or capped 3 stars on nodes lower than lvl 100 because this will ultimately make your personal scaling rocket up even faster.

In the end the strategies you use will depend very heavily on how you access the situation and consider your resources you have available to you.

puppychow

Ryz-aus wrote:

Points refresh linearly (1/48 of max for node per hour), so leaving higher point nodes to clear later gives you slightly more points. Unfortunately, I think it also will result in them having more difficult scaling - but if you're looking for the highest score possible that's the method to use.

ding-ding-ding! Scaling indeed becomes a factor at the end. Remember that node with 4* deadpool, dino, and hulk? While I was grinding it, every clear added another 1,000 health to these guys! Each guy was fielding 12k health, then 13k, 14k. Finally it wasn't worth the points to grind anymore, and I moved on.

I was really lucky AI didn't get a red cascade or else my team would be wiped asap.

puppychow

skydivearcher wrote:

one thing you also need to consider is the content of the node. Certain combinations are going to be harder to clear and/or take more time to clear. YOu can't just look at the points value of a node. Some players I notice will always save a relative easy high point node for the last few minutes of a sub... This is especially important if you are on the borderlines between getting additional character cover and/or reward or not. I seen a number of times a player in the tenth position ten minutes before the end of a full event replaced by some other person who managed to outscore them in the last few minutes of a sub/event. Heck I've even done it to one person during the event to win 2 hulkbuster covers instead of one. I suspect the very same thing happens with similar positions further down the reward structure.

Take enemy of the state, those events often have wave nodes, and each of those nodes especially if they have a lot of waves, can easily take 10-15 minutes to successfully complete. These are probably much better to complete early on rather than late because of the amount of time they can take and you definitely don't want these scaling out of hand!

If you also have an extensive roster, it's always a good idea to suicide one character or two in a match, this will actually slow down your personal scaling. I seen nodes drop levels because even though I successfully completed it, having a character or two defeated actually lowered scaling. You also probably want to avoid characters that use protect tiles because of how personal scaling works, you never want to walk out of a match taking zero or very little damage. Although this allows you to do nodes more often without using health packs, it actually slows you down because nodes have got harder and thus more time consuming to complete.

Shhhh! Stop giving away secret strategies to top 10 placement!

GrumpySmurf1002

Vhailorx wrote:

Thanks for the math qkumbre, but I was just asking about inside the last 8 hours of an event. I completely agree with you that clearing every 8 hours (and clearing as fast as possible) is the best strategy for pve.

The only real question I needed answer was whether points recharge linearly for all nodes regardless of how many times they have been cleared.

Example:
Node A worth 240 points. Each clear reduces by 40, refreshes 5 per hour
Node B worth 480 points. Each clear reduces by 80, refreshes 10 per hour.

Starting with 4 hours left, so nodes are refreshed to 220, 440. Lets make the unrealistic assumption that it takes 0:00 to clear the nodes, but an hour to finish. (Makes the math easier)

Clear A then B - (220+180+...+60+20) + (450+370+...+130+50) = 2220
Clear B then A - (440+360+...+120+40) + (225+185+...+65+25) = 2190

Clearing low to high is the answer.

Unknown

Clearing low to high as long as you're not #1 (in that case you can wait unless you're ahead enough to simply finish off the competition) is optimal but the value of the node is not just its points. Rather it's points * difficulty. An Ultron + 2 guys who can move the board node worth 10000 points is not very valuable for the simple fact that there's like a 99% chance you'll lose to such a node without mega whales. Take Enemy of the State, a node with Gorgon + 2 goons is much lower expected value than even the 'easy' survival node because in all but 1 of the survival nodes you start with goons only which gives you a ton of time to setup while you'll be eating Debiliating Slashes/Awaken the Hands almost immediately when the opponent is just Gorgon + 2 goons. In this case the Gorgon node should be attempted earlier even if it's a higher value compared to the generic 5 waves of Ninja + Gorgon at the end which is trivially handled because the Gorgon node has a lower expected value (because you can lose!). Deadpool vs MPQ has a node featuring Iron Fist, who I'm sure is still one of the toughest opponent in the game. That node should be considered as a 'low' value node due to its high risk. And if you can beat Iron Fist trivially then you can obviously completely steamroll anything and wouldn't need much of a strategy to begin with.

wymtime

I just have to say everyone who grinds all these nodes down to 1 regardless of when are completely insane. I bow down to your insanity and will gladly give you the covers you have earned. I prefer my sanity over the new 4*.

Bow down to the insanity

Unknown

wymtime wrote:

I just have to say everyone who grinds all these nodes down to 1 regardless of when are completely insane. I bow down to your insanity and will gladly give you the covers you have earned. I prefer my sanity over the new 4*.

Bow down to the insanity

I completely agree. I stopped playing PVE competitively ages ago. Nowadays, i play the Late Join Lottery and watch the brackets for a chance to jump in a score new covers with as short of a grind as possible.

However with the release of 4* DeadPool IN a DeadPool event, i (along with the rest of the normally "PVE Optional" Tacos) decided to go all in and grind for our Namesake.

Aside from one double tap of nodes that was necessary bc i wouldn't have been able to hit a refresh otherwise, I played "optimally" and ground every single one of those Fuggers to 1 (finished 7th). In a 4 day, 48 hour sub event, it was doable. No way i would ever do that in a 7 day 24 hour sub event. Not for 5* Deadpool.

Respect to all of you crazy PVE grinders.

easy....marc

Vhailorx

djsquillz wrote:

wymtime wrote:

I just have to say everyone who grinds all these nodes down to 1 regardless of when are completely insane. I bow down to your insanity and will gladly give you the covers you have earned. I prefer my sanity over the new 4*.

Bow down to the insanity

I completely agree. I stopped playing PVE competitively ages ago. Nowadays, i play the Late Join Lottery and watch the brackets for a chance to jump in a score new covers with as short of a grind as possible.

However with the release of 4* DeadPool IN a DeadPool event, i (along with the rest of the normally "PVE Optional" Tacos) decided to go all in and grind for our Namesake.

Aside from one double tap of nodes that was necessary bc i wouldn't have been able to hit a refresh otherwise, I played "optimally" and ground every single one of those Fuggers to 1 (finished 7th). In a 4 day, 48 hour sub event, it was doable. No way i would ever do that in a 7 day 24 hour sub event. Not for 5* Deadpool.

Respect to all of you crazy PVE grinders.

easy....marc

Exactly this. Deadx seemed fun, the event was short and had longer subs. So I gave it a shot, ground everything to 150 or less and finished 10th.

Haven't done top 10 in a new character release since kingpin and don't plan on doing it again any time soon. It's just too much of a time commitment and turns an otherwise fun puzzle game into a tedious chore.

Late-entry lottery ftw!

Cymmina

If you're in a position where you can take advantage of rubberbanding, clearing the most valuable nodes first is going to be the most optimal, isn't it?

When *does* rubberbanding kick in? I've heard 10k difference from the leader, but I'm certain I've seen it kick in sooner than that.

Buret0

Rubberbanding?

Is that like how you could throw pinpoint 60 yard passes in NFL blitz when you were losing?