Speculation on rubberbanding

Unknown

So recently I've noticed whatever they did to rubberbanding, it's making the event relatively insensitive to when you played your last refresh, and I don't see how it could possibly work with just tweaking the formula, so here's my theory. As soon as the last refresh starts, the leader is replaced by a fictional guy who has points set at a level D3 thinks is about right to hit whatever progression goal they intended to hand out. This guy has points that is generally far above anyone can possibly reach (since we know from past experience nobody was ever close to hitting the levels you'd need to get the top progression), and the rubberband is set so that if you grinded all your mission down to 1 you'll barely surpass this hypothetical guy.

Therefore because nobody can realistically beat this fictional bracket leader, the timing of your last refresh becomes irrelevent because this fictional guy is the bracket leader the moment the last refresh started, unless someone grinded ultra hardcore to actually surpass this fictional leader. From this, D3 can then set the progression reward to be achieveable by someone who got say, 90% of the fictional leader's points (or any other % based on whatever metrics they use).

Or it can just be that we're overdue for progression rewards that are actually reachable, but the reachability of Daken cover seems way too well planned to be an accident. Looking at the rate, the top scorers will comfortably hit it, but from there on it's far less certain. Based on history, usually we have the progression rewards that are totally trivial to get (I think we had top score that are twice of the max progression reward in the last Hulk), or totally impossible to get (quite a few cases where top score wasn't even half of what's needed).

Maybe someone from D3 can confirm if they indeed installed mechanism to make the last minute push less important. I'm sure it'd help people enjoy the game more if it is indeed the case that you no longer have to worry about playing at the exact last 2 hours of a PvE event finish. Note that for the no rubberband brackets, there are obviously no timing issue at all. The more you can win, the better off you'll be.

Unknown

As far as the last pair of subs in the current PvE go I think it's more a case of them pretty much removing the rubber banding and making it a grind fest. When I first checked after the unlock nodes opened they were worth, IIRC, 191. By the time I started doing nodes for my last run through they had increased to 200 points.

I then started clearing from the bottom scoring nodes and by the time I got round to the last of the new nodes it was worth 150. I hadn't reached first in my sub let alone neared a fictional, higher rated pacing player. I know people who were a fair way behind me in score and their nodes barely broke 200.

I think, at least for those subs, they decided to replace having to be there at the end with having to be there every refresh and grind like hell. Not sure I see it as an improvement.

Unknown

bonfire01 wrote:

As far as the last pair of subs in the current PvE go I think it's more a case of them pretty much removing the rubber banding and making it a grind fest. When I first checked after the unlock nodes opened they were worth, IIRC, 191. By the time I started doing nodes for my last run through they had increased to 200 points.

I then started clearing from the bottom scoring nodes and by the time I got round to the last of the new nodes it was worth 150. I hadn't reached first in my sub let alone neared a fictional, higher rated pacing player. I know people who were a fair way behind me in score and their nodes barely broke 200.

I think, at least for those subs, they decided to replace having to be there at the end with having to be there every refresh and grind like hell. Not sure I see it as an improvement.

LA, Wakanda, Washington DC, and Hell's Kitchen have no rubberbanding so of course there's no advantage to timing and no need for a pacing player either.

But the events before that definitely had rubberbanding, and these brackets I saw a consistent pattern of some guy with a very high score well before the last 2 hours of the final refresh, and that guy's score held as #1 without being increased. I don't think that's even possible with the old rubberbanding, so the only way I can think of this can work is if there's a fictional pacemaker, and in that case as long as nobody surpasses the fictional pacemaker on the final refresh, then the timing wouldn't matter because everyone would work their rubberband off the fictional pacing guy.

Unknown

If that's not how it works, it probably should. Sounds like an elegant way to make progression rewards better calculated.

Unknown

I thought D3 was gonna throw us a cookie to get Lazy Daken, but the last 4 sub-event proved otherwise. D3 just doesn't give **** about goodwill anymore.

daveomite

What I had noticed in Wakanda, at least towards the final 12 hours... was the change to how points are dealt with.

In previous events, like the Hunt, etc. If you were to have 3 nodes, all three at 800, then play one of those nodes, the other two would reduce...say, both would then be 768.

But, at least from what I was seeing, was that I had a node at 195, another at 195, and another at 209, for instance. I played the 209 one, then noticed the other two remained at 195 each. To test it, I then ground that 209 node all the way to 1. All of the other nodes were still at the same point value as before.

That is decided different than before, and yes, indicates there is no rubber banding controlling the pool of points. This lends more to, as Phantron stated, you basically are left with little option but to grind the nodes, but doing so increases scaling...leading to spending (D3 hopes), and you still seem to wind up behind no matter how many times you hit the nodes.

When I started Washington DC, it was literally when the node was first released, just after the previous subs closed. I did the first 4 nodes in it, had 393 points I believe. Looked at the leaderboard, and the leader already somehow had over 1400 points. That just didn't seem logical as its not like I spent an hour clearing 4 nodes, I had done them immediately.

So that leads me to think, it's possible that some people, maybe newer players, are getting more points with at least some nodes. That perhaps their stack in one node equals to a greater number than a stack of one of my nodes.

Right now, there seems to be no timing really needed. If you grind all the nodes to 1, then at some point unknown, they restack. If you miss a stack, possibly the points diminish to some small degree. Just seems like from what I was seeing, one node I had at 1, restacked well within a 12 hour time span after getting it down to 1.

I think we're all still trying to figure it out, but I'm thinking nodes stack independently at some unknown time frame, regardless of the previously established 8 or 12 hour refresh windows. Of course, given I'm also in PvP's, it's not like I can continually watch a node like that for an extended period.

But, it does at least, in my mind, provide at least one possible answer to how, even if you were to clear all nodes to 1, some people are still continuing to climb in points, even though you just have 1 point nodes.

Spoit

daveomite wrote:

What I had noticed in Wakanda, at least towards the final 12 hours... was the change to how points are dealt with.

In previous events, like the Hunt, etc. If you were to have 3 nodes, all three at 800, then play one of those nodes, the other two would reduce...say, both would then be 768.

But, at least from what I was seeing, was that I had a node at 195, another at 195, and another at 209, for instance. I played the 209 one, then noticed the other two remained at 195 each. To test it, I then ground that 209 node all the way to 1. All of the other nodes were still at the same point value as before.

That's because they ate up some of the rubberbanding, reducing the amount of RB points you got. These subs don't have that, so it's set to the base node value, and doing the others doesn't effect it

Unknown

LA/Wakanda/Washington DC/Hell's Kitchen has no rubberbanding. If a mission is worth 100 points it's always worth 100 points the first time, then 80 the second time, and so on. Now these missions had some weird values like 79/63 so it might feel like it's rubberband but in reality they just had weird base value. Obviously with no rubberbanding the timing is irrelevent, and you can clearly see that the guys who grinded more, or managed to get an extra refresh, had score higher than what you could have possibly gotten even if you grinded everything to 1. I think the Mayhem missions had rubberbanding, but I wasn't able to collect any data on that on time (and it could easily have been that those have weird base value too).

But the brackets before that still had rubberbanding, and I was pretty surprised at how the top score in the sub bracket appear to have never changed even though the high was established more than 2 hours before the final refresh, which is not possible in the old system, so I'm wondering if something has changed.

JamieMadrox

I have to agree with the no rubberbanding comments. I opened the nodes on one account right when they were available and again on another hours later and they were worth the same on both. There is no rubberbanding. It's just grinding until the next refresh.

Unknown

I think there is RB, its just muted because the multiplier is a lot lower and the points are lower so you won't notice it much. For example, if you hit a low node 1 or twice during last 2 hrs, if you don't hit any more, it has increased in points for me. Bc RB is so low, you probably don't realize it bc you move on and hit the next ones and reduce RB effects.

Clearly, the best bet this pve is to hit the nodes each refresh for optimal play. That's too much for me so I have done 2 playthrus and that's been good enough for top 5 but not enough to overtake top2 who are grinding every refresh

Unknown

In the first 2 brackets I was doing my usual strategy of keep pace (around top 20) before last refresh and grind hard at the end, and the guy at the top consistently looks like he has roughly the points of what I would've gotten if I grinded everything down to 1. I'm sure it's a little bit less than that, but of course I didn't grind everything down to 1 either. I know I was on at least the 3rd pass of all nodes and some nodes I did more than that and I still never caught the leader who remained static for the last 2 hours. Here there is definitely rubberbanding (if it was possible to grind all nodes down to 1, I'd still have caught up) but it's also unlike anything in the past. I've never had a case where I was unable to overtake a static leader by the time I was on my 3rd pass on my missions. Remember that I keep pace until the last refresh (usually within top 20) so it's not like I'm trying to make up a huge lead at the very end.

Now, the newer brackets I'm pretty sure there is no rubberbanding at all. I've been writing down the points of the missions since the beginning of the day and they have not changed at all. Also, even though I was #6 in Wakanda, the #1 had like 2K lead which is far more than the total points that are left when I was doing them, and the only way that can happen is if there is no rubberbanding (I certainly wasn't trying to get grind every refresh to the max). Now, there seems to be a an issue where the points you gain is not the same as what the node reports, like it might be the node says 79 and you get 83. I have no idea what's causing that, but the reported points on the node never goes up (or down) on their own.

Unknown

MikeyMan wrote:

I thought D3 was gonna throw us a cookie to get Lazy Daken, but the last 4 sub-event proved otherwise. D3 just doesn't give **** about goodwill anymore.

I saw people over 40k already, so daken can be taken alright. It's your call whether the grinding worth the cover or not.

daveomite

pasa_ wrote:

MikeyMan wrote:

I thought D3 was gonna throw us a cookie to get Lazy Daken, but the last 4 sub-event proved otherwise. D3 just doesn't give **** about goodwill anymore.

I saw people over 40k already, so daken can be taken alright. It's your call whether the grinding worth the cover or not.

the leader of my main bracket right now has over 40k, but only has 2 85's on their roster. Seems odd.

and less than 7 hours ago, our alliance was first in DC, and the second alliances top scorer had 2,700 points. At that point, all my nodes were at 1 each.

When I got back into it today, I saw that alliance was now in 1st place, which is expected, but now their top scorer has over 4,100 or so. Yet, all of my nodes are still at 1. So, they jumped from 2700 to 4100 within 6 hours.

something is definitely off. Or at least, certainly seems to be.

Unknown

In the no rubberband brackets persistence pays off a lot more than anything else. After all if you got 1 more refresh than other people while the points are flat, there's almost no way they'll catch up to you unless there's some unbelievable disparity in scaling.

Usually the guys with the huge score in these sub events are the guys who started late to the event, so they have very little scaling to begin with and can just grind their way. In general, people have to remember it is possible for someone who joined the event 5 days later to be the leader of your sub bracket, and in fact that's a very easy way to win a sub bracket if you join an event 5 days late because your scaling is going to be way lower than anyone else's (but not exactly a good way to get any progression prizes and risky on main bracket placements).

Spoit

Phantron wrote:

In the no rubberband brackets persistence pays off a lot more than anything else. After all if you got 1 more refresh than other people while the points are flat, there's almost no way they'll catch up to you unless there's some unbelievable disparity in scaling.

Usually the guys with the huge score in these sub events are the guys who started late to the event, so they have very little scaling to begin with and can just grind their way. In general, people have to remember it is possible for someone who joined the event 5 days later to be the leader of your sub bracket, and in fact that's a very easy way to win a sub bracket if you join an event 5 days late because your scaling is going to be way lower than anyone else's (but not exactly a good way to get any progression prizes and risky on main bracket placements).

We'll, that and all the leaders in my subs also are only halfway in their 2* transition, which means their base scaling is crazy low to begin with