rewards for rubberbanding exploiting need reducing.

Unknown

because i have manipulated the mmr system to have enemies that are ridiculously low level and thus i slaughter through them and thus spend a quarter of the time achieving the same thing the other rubber banders are doing.

thus if i was forced to play and win more , which BTW i do not have to do because the huge rubber-banding rewards not playing so much, my mmr would go up and the difficulty of fights would go up.

we have all seen the legit player with barely any roster dominate a bracket because their up leveled guy stomps through super low levles.... now imagine if you could manipulate the system so you not only had a good roster, but the super low level enemies, thats what im doing.

and i can do it only because of how heavy the rubber-banding is.

I can not be the only person to figure this out, confess!!!

I just did not want to oust my self as a doer of the thing that i don't think is cool to be done.

Sandmaker

Malkavian wrote:

when the enemies i face have 1/4th the hit points yours have and i never have to stun lock or heal because they do no considerable damage, i get through them 4 times faster, so... yeah

My average match is about 3-4 minutes... I highly doubt you're consistently knocking out nodes at 1 minute a piece.

But lets say you are, then that just means next sub-bracket against you, I will have to start 3 hours before the end, or 4 hours before the end. Are you going to tell me that you can grind 6-8 times faster than me now? At some point, I will beat you out if you consistently start at 30 minutes before the end.

Unknown

Malkavian wrote:

because i have manipulated the mmr system to have enemies that are ridiculously low level and thus i slaughter through them and thus spend a quarter of the time achieving the same thing the other rubber banders are doing.

thus if i was forced to play and win more , which BTW i do not have to do because the huge rubber-banding rewards not playing so much, my mmr would go up and the difficulty of fights would go up.

we have all seen the legit player with barely any roster dominate a bracket because their up leveled guy stomps through super low levles.... now imagine if you could manipulate the system so you not only had a good roster, but the super low level enemies, thats what im doing.

and i can do it only because of how heavy the rubber-banding is.

I can not be the only person to figure this out, confess!!!

Yeah I'm pretty sure you're full of it at this point.

Unknown

Malkavian wrote:

because i have manipulated the mmr system to have enemies that are ridiculously low level and thus i slaughter through them and thus spend a quarter of the time achieving the same thing the other rubber banders are doing.

thus if i was forced to play and win more , which BTW i do not have to do because the huge rubber-banding rewards not playing so much, my mmr would go up and the difficulty of fights would go up.

we have all seen the legit player with barely any roster dominate a bracket because their up leveled guy stomps through super low levles.... now imagine if you could manipulate the system so you not only had a good roster, but the super low level enemies, thats what im doing.

and i can do it only because of how heavy the rubber-banding is.

I can not be the only person to figure this out, confess!!!

I just did not want to oust my self as a doer of the thing that i don't think is cool to be done.

Thanks, now we all can do this.

Eddiemon

jozier wrote:

The person who plays the most will win, every time. There's a difference between playing the most and playing optimally however.

Grinding nodes for 1 point a battle is bad. Playing the nodes when they have a full stack is good. Rubberbanding has no bearing on that. In fact, if you play optimally, you would never really notice rubberbanding since every time you did your allotment of missions you'd end up back at the top.

That's a commonly held perception here, but it isn't actually true.

Assume we have a field with 3,000 base points available from missions. You complete it 4 times with no rubberbanding in play. You have 12,000 points.

I come along and start. I have a 10x multiplier until I get 4000 points behind you. That means I use 800 base points to generate 8,000 points.
Now I am at the edge of the multiplier decline. At 4,000 points behind I have a 10x Multiplier and at 400 points I will have a 1x multiplier. So over those 3600 points I will have an average 5.5x multiplier. It takes me 700 base points to generate the next 3600 points.

I have so far used up 1500 base points and I am only 400 points behind you. I end up cruising past to a 900 point lead if I clear the board.

Granted it is an unrealistic bracket with just the two of us in it, and in a real bracket you would be enjoying rubberbanding effects too as proplr jostled for position at the top, but you'd probably never hit the 10x multiplier stage. While I would just need to clear a few 10x missions to keep pace with you until that final stretch play comes into effect.

Eddiemon

beezer37_84 wrote:

I feel like at some point the argument went from "higher level teams exploit rubberbanding" to "lower level teams exploit scaling", but what do I know...

I feel we should do the Futurama thing at this point and turn on the neutrals....

Nemek

Eddiemon wrote:

jozier wrote:

The person who plays the most will win, every time. There's a difference between playing the most and playing optimally however.

Grinding nodes for 1 point a battle is bad. Playing the nodes when they have a full stack is good. Rubberbanding has no bearing on that. In fact, if you play optimally, you would never really notice rubberbanding since every time you did your allotment of missions you'd end up back at the top.

That's a commonly held perception here, but it isn't actually true.

Assume we have a field with 3,000 base points available from missions. You complete it 4 times with no rubberbanding in play. You have 12,000 points.

I come along and start. I have a 10x multiplier until I get 4000 points behind you. That means I use 800 base points to generate 8,000 points.
Now I am at the edge of the multiplier decline. At 4,000 points behind I have a 10x Multiplier and at 400 points I will have a 1x multiplier. So over those 3600 points I will have an average 5.5x multiplier. It takes me 700 base points to generate the next 3600 points.

I have so far used up 1500 base points and I am only 400 points behind you. I end up cruising past to a 900 point lead if I clear the board.

Granted it is an unrealistic bracket with just the two of us in it, and in a real bracket you would be enjoying rubberbanding effects too as proplr jostled for position at the top, but you'd probably never hit the 10x multiplier stage. While I would just need to clear a few 10x missions to keep pace with you until that final stretch play comes into effect.

It's true in the sense that if player A had a 100 (or any) point lead on player B. And then both player A and player B both played the exact same amount of missions at the same time, player B will never catch player A through rubberbanding.

Unknown

You can't beat a guy who can play more than you unless that guy is somehow unable to play in the last refresh. Let's assume the guy who plays more than you is the overall sub bracket leader to make calculations easier, though it doesn't actually matter. Let's say that guy plays twice as much as you can, so he jumps to a large lead early on and you try to take advantage of rubberbanding. Well, going into the final refresh, that guy is of course #1. At this point, he has no reason to do his missions, so you're forced to do your most valuable missions first to try to close the gap, but whenever you do, he just does his most valuable mission (which is at 100%) while you probably already depleted all your 100% missions just to catch up. If you somehow still caught up, the same logic still applies, as that guy will have his missions at 80% and you probably are forced to go down to 40% range to just overtake him.

Yes, if the #1 guy for some reason decides to just do all his missions very early, you might miraculously come back from behind. It's usually not even possible (I've seen quite a few bracket where nobody ever caught the sub bracket leader), but even if it is, that assumes the leader has some kind of time constraints. Given the time it takes to do all these missions, it'd basically mean that guy is forced to do his last refresh at pretty much the earliest time possible, because if that guy waited until even 4 hours to go, it'd likely take you 4 hours to grind every mission down to 1 which is what you'd need to overtake him. In the TaT sub brackets where people actually were doing every mission down to 1 (prior to the boost nerfs), you will see people fall in within 2-3 points of each other no matter when they did their missions because that's how rubberbanding works, and that's 2 people doing all the missions and they're still only separated by 2-3 points (2-3 missions at value of 1). Remember that even the lowest point mission in TaT is worth 5, so it means if you missed even one mission, instead of being +3 you'd be -2 and still lose, and that's 1 mission. Someone who plays a lot more than you, given the current scaling, is certain to do a lot more than one mission than you can.

Eddiemon

Nemek wrote:

It's true in the sense that if player A had a 100 (or any) point lead on player B. And then both player A and player B both played the exact same amount of missions at the same time, player B will never catch player A through rubberbanding.

Yes that holds true. But that turns into a timing is more important than effort message again. Only when you have timing right does effort matter.

We just seemed to be telling people that if your grind everything you can't be beaten. Which isn't the case.

Nemek

Eddiemon wrote:

Nemek wrote:

It's true in the sense that if player A had a 100 (or any) point lead on player B. And then both player A and player B both played the exact same amount of missions at the same time, player B will never catch player A through rubberbanding.

Yes that holds true. But that turns into a timing is more important than effort message again. Only when you have timing right does effort matter.

We just seemed to be telling people that if your grind everything you can't be beaten. Which isn't the case.

I'd agree with that.

Unknown

Eddiemon wrote:

Nemek wrote:

It's true in the sense that if player A had a 100 (or any) point lead on player B. And then both player A and player B both played the exact same amount of missions at the same time, player B will never catch player A through rubberbanding.

Yes that holds true. But that turns into a timing is more important than effort message again. Only when you have timing right does effort matter.

We just seemed to be telling people that if your grind everything you can't be beaten. Which isn't the case.

Sure you can win by just grinding. Nobody has come close to grinding everything down to 1 since the boost nerf except in some very select cases (like the initial part of The Hulk), and in those maps if you waited you'd never be able to overtake someone who's been doing 1 point mission the whole time. Given the difficulty and the length of the missions there's no reason to believe there are very many people doing more than two passes at every mission. I've had time where my missions are not going up in points at all for at least an hour, which means the sub bracket leader is not changing. As long as the sub bracket doesn't change, the timing is actually irrelevent, because both you and anyone else is based on that guy's score. Given you can win a sub bracket while still getting significant rubberbanding bonus, there's no reason to believe the overall sub bracket leader is constantly changing (why would you grind more to overtake that guy when you can already easily get #1?), and timing is irrelevent as long as the overall sub bracket leader doesn't change.

Sandmaker

Eddiemon wrote:

That's a commonly held perception here, but it isn't actually true.

Assume we have a field with 3,000 base points available from missions. You complete it 4 times with no rubberbanding in play. You have 12,000 points.

I come along and start. I have a 10x multiplier until I get 4000 points behind you. That means I use 800 base points to generate 8,000 points.
Now I am at the edge of the multiplier decline. At 4,000 points behind I have a 10x Multiplier and at 400 points I will have a 1x multiplier. So over those 3600 points I will have an average 5.5x multiplier. It takes me 700 base points to generate the next 3600 points.

I have so far used up 1500 base points and I am only 400 points behind you. I end up cruising past to a 900 point lead if I clear the board.

Granted it is an unrealistic bracket with just the two of us in it, and in a real bracket you would be enjoying rubberbanding effects too as proplr jostled for position at the top, but you'd probably never hit the 10x multiplier stage. While I would just need to clear a few 10x missions to keep pace with you until that final stretch play comes into effect.

Edit- Nvm! Didn't read it right.

Unknown

Eddiemon wrote:

beezer37_84 wrote:

I feel like at some point the argument went from "higher level teams exploit rubberbanding" to "lower level teams exploit scaling", but what do I know...

I feel we should do the Futurama thing at this point and turn on the neutrals....

If I don't make it, tell my wife I said hello.

Eddiemon

Sandmaker wrote:

Eddiemon wrote:

That's a commonly held perception here, but it isn't actually true.

Assume we have a field with 3,000 base points available from missions. You complete it 4 times with no rubberbanding in play. You have 12,000 points.

I come along and start. I have a 10x multiplier until I get 4000 points behind you. That means I use 800 base points to generate 8,000 points.
Now I am at the edge of the multiplier decline. At 4,000 points behind I have a 10x Multiplier and at 400 points I will have a 1x multiplier. So over those 3600 points I will have an average 5.5x multiplier. It takes me 700 base points to generate the next 3600 points.

I have so far used up 1500 base points and I am only 400 points behind you. I end up cruising past to a 900 point lead if I clear the board.

Granted it is an unrealistic bracket with just the two of us in it, and in a real bracket you would be enjoying rubberbanding effects too as proplr jostled for position at the top, but you'd probably never hit the 10x multiplier stage. While I would just need to clear a few 10x missions to keep pace with you until that final stretch play comes into effect.

I don't think this is right. Your example is making the assumption that the leading player went through 4 refreshes, without you going through any, but then you still have 12k base points left. That is incorrect. What will actually happen in your example is that in the last refresh you'd use 800 base points to get 8000, but then you'll only have 3200 points left, which means the most points you can end with is 11200 for second place.

No, you missed a step.
The first 800 netted me 8000 points at a rubberbanding multiplier of flat x10
The next 700 are spent at a rubberbanding multiplier of an average of 5.5 for 3600 points. (It's actually a higher non round number but 3600 is good enough for this model)
I have 1500 left with no multiplier.

You missed that middle multiplication step I think.

Of course I missed that 1500-400 is 1100 instead of 900, but whats a bit of addition amongst friends.

Eddiemon

Phantron wrote:

Sure you can win by just grinding. Nobody has come close to grinding everything down to 1 since the boost nerf except in some very select cases (like the initial part of The Hulk), and in those maps if you waited you'd never be able to overtake someone who's been doing 1 point mission the whole time. Given the difficulty and the length of the missions there's no reason to believe there are very many people doing more than two passes at every mission. I've had time where my missions are not going up in points at all for at least an hour, which means the sub bracket leader is not changing. As long as the sub bracket doesn't change, the timing is actually irrelevent, because both you and anyone else is based on that guy's score. Given you can win a sub bracket while still getting significant rubberbanding bonus, there's no reason to believe the overall sub bracket leader is constantly changing (why would you grind more to overtake that guy when you can already easily get #1?), and timing is irrelevent as long as the overall sub bracket leader doesn't change.

No, you need to use the actual math models. In the model given the first player had completely ground down every mission every time to end up with 12k points.
The second player passed him with 1100 base points to spare (I know I blew subtraction in the model. I blame Daredevil)
On a 3000 point total board, The 100% missions are worth 1000 base points, the 80 and 60% missions are worth 1400 base points and the 40 and 20% missions are worth 600 base points. If I don't do any 40 or 20% missions then I still win by 500 points.

In order for the first person to win purely by grinding he would have to additionally have ground over 10,000 points worth of 1 point missions.

Sandmaker

Eddiemon wrote:

No, you missed a step.
The first 800 netted me 8000 points at a rubberbanding multiplier of flat x10
The next 700 are spent at a rubberbanding multiplier of an average of 5.5 for 3600 points. (It's actually a higher non round number but 3600 is good enough for this model)
I have 1500 left with no multiplier.

You missed that middle multiplication step I think.

Of course I missed that 1500-400 is 1100 instead of 900, but whats a bit of addition amongst friends.

Yeah, I realized I didn't read it right, and edited it, but not quick enough apparently.

Edit- Actually.. I still didn't read it right, haha. Let me just remove that one and start again.

Unknown

Eddiemon wrote:

Phantron wrote:

Sure you can win by just grinding. Nobody has come close to grinding everything down to 1 since the boost nerf except in some very select cases (like the initial part of The Hulk), and in those maps if you waited you'd never be able to overtake someone who's been doing 1 point mission the whole time. Given the difficulty and the length of the missions there's no reason to believe there are very many people doing more than two passes at every mission. I've had time where my missions are not going up in points at all for at least an hour, which means the sub bracket leader is not changing. As long as the sub bracket doesn't change, the timing is actually irrelevent, because both you and anyone else is based on that guy's score. Given you can win a sub bracket while still getting significant rubberbanding bonus, there's no reason to believe the overall sub bracket leader is constantly changing (why would you grind more to overtake that guy when you can already easily get #1?), and timing is irrelevent as long as the overall sub bracket leader doesn't change.

No, you need to use the actual math models. In the model given the first player had completely ground down every mission every time to end up with 12k points.
The second player passed him with 1100 base points to spare (I know I blew subtraction in the model. I blame Daredevil)
On a 3000 point total board, The 100% missions are worth 1000 base points, the 80 and 60% missions are worth 1400 base points and the 40 and 20% missions are worth 600 base points. If I don't do any 40 or 20% missions then I still win by 500 points.

In order for the first person to win purely by grinding he would have to additionally have ground over 10,000 points worth of 1 point missions.

Your model is wrong because events where both people grinded to full actually happened in TaT and it didn't work like how you claim it would. I was in several sub bracket where you can watch the top 3 points going up one at a time and they're separated by 2-3 points the entire time until someone gave up, despite the fact that those 3 guys sure didn't all play at the same time. That 2-3 points is less than the value of the lowest mission, which means it was not possible to catch up to the overall leader without doing at least as many missions as he did on the final refresh.

Nobody is grinding anywhere close to 'down to 1' after the AP boost nerf. You simply thought you had a large lead but you likely didn't even do every mission twice (since that is usually enough to win a bracket) and when someone did every mission twice + one more, they pass you up like they're supposed to. By the way, the guy who grinds the most always gets to dictate the timing, because he's #1 by definition and has no reason to start grinding any earlier until people start getting close. Unless that guy has some other time constraints, the guy who is ahead can always make the last move compared to the guys trying to catch him because he started out ahead.

Nemek

I think the model is only "wrong" in the sense that the first player completely flubbed the last refresh in terms of timing and the point being made is that timing the last refresh correctly is more important than being in the lead going into the last refresh.