The new bounty and spree system

[Wupti]

Alright, Beta 2 is up. Hosting will begin soon™.

Regarding the bounty and spree changes, lets discuss this in a seperate thread.

I'm quite amazed no one has done this yet

I believe the idea is to make feeders less of an influence, correct me if im wrong, but am I the only one that find this new system heavily unbalanced?

To Put it bluntly mid-part of the game is ruined. The only way you see a game ending before 1 hour is duo to 3 reasons.
a)leaver
b)dc
c)kick


I see teams dominate heavily early/midgame but thx to the new bounty system they gain nothing from it unless they completely obliterate a team and take a CP. duo to the fact the spawn timers are fairly short and its almost impossible to get an egde this is near impossible and can only be done by huge missplays from either team.
This is all personal but to me a good game isnt always a long boring titan war.

Unfortunately there is no stat in league named average game time, so I cant compare my old acc with my new which could have been interesting. If any admin is able to do that i would very much apprecciate the result posted here.

The new bounty system is a great idea but imo need to get a lot more balanced.

Furthermore I think It's a problem the only real objective to get in btank is the checkpoints. I'd like to see for instance to towers to be spread throughtout team instead all to the lasthitter.

An idea could be

before: player1 400g player2 0 player3 0 player4 0 player5 0

New: player1 200g player2 100 player3 100 player4 100 player5 100.

feel free to give input


Food for thought

(last 10 ranked games)
t1 ~ Team 1
t2 ~ Team 2
Length:
57m9s
1h47m50s (t1 dc)
1h1m8s (t1 afk and dc)
1h4m57s
58m40s (t1 dc, t2 kick)
24m27s (t2 dc dc)
11m58s (t1 dc)
1h11m20s (t2 dc)
1h51m16s
1h19m15s (t1 dc)

Here are some example values with a tank worth of 10k:

Code:

old bounty:
125 + tank worth / 65 (262g)

new lower boundary:
125 + tank worth / 100 (225g)

new upper boundary:
125 + tank worth / 33 (428g)

Assuming all players are worth 10k and 0 kills

player 1 kill player 6 (262 + 2 assists)

player 2 kill player 7 (262 + 2 assists)

player 3 kill player 8 (262 + 2 assists)

Player 1 kill player 6 (225 + 2 assist)

Player 2 kill player 7 (225+ 1 assist)

Player 3 kill player 8 (225 + 1 assist)

Net gold kills: 1491g

then lets assume they all die once

1491-3x428

Net Gold kills: 207g

Team: Kills 6 deaths 3

Since i dont know the boundary its hard to give a concrete example and these estimations should be some what off, but you get the idea. havent included assists since they should even out somewhat aswell

[Exodus]

Alright, moved and edited your post.

To be honest, I'm also suprised noone gave any feedback on this topic. Especially in this case it's very important.
It clearly wasn't my goal drag out games and I'm also not sure if that's what really happened, so it really might be interesting how the average game length developed after the last version.

So you are saying, that the new system is too forgiving for player mistakes? I guess you get this from your own feeling of the games you play, but not the facts that you stated here, since they seem to base on false assumptions. Because of that, I'm again trying to explain what those boundaries mean and how the bounty works. Bascially, your own bounty increases everytime you receive gold, except from the passive income, force gold or selling items. Your own tank or your items have nothing to do with that and buying and selling won't change your bounty anymore.

Rather, your bounty increases when you gain gold and it decreases, when you die. But to prevent the bounty from going too low or too high, I implemented the boundaries. These boundaries depend on your tank worth (which is why I stated that in my example above). In other words: the range of your bounty is limited by your own tank worth, but it is not determined by it.
What I wanted to show you with my example was the following: in a normal situation, an average player should have about the same bounty as before. Better players will have a higher bounty and others will have a lower one. Just to give you an idea on how high these differences are, I showed an example of the old bounty (the 262 gold) and the minimum bounty a tank with the new system could have, with the same tank worth (the 225 gold). With the new system, you can't go below 225 gold, after dying several times. This miminum bounty obviously increases as you get better items / tanks. The same is true for the upper limit, with the difference that this limit is farther away from the old bounty, meaning that players with such bounty are far more lucrative targets as they were before.

Taking your example and assuming everybody really starts with 262 gold, it would actually play out like this:
player 1 kills player 6 (+262g)
player 2 kills player 7 (+262g)
player 3 kills player 8 (+262g)

The bounties of the players 1, 2 and 3 now increases by 5% of the gained bounty, so by 13 gold. Their new bounty is 275g
The bounties of the players 6, 7 and 8 decrease by 12%, so they now have a bounty of 231g, just above the limit of the 225g.

player 1 kills player 6 (+231g)
player 2 kills player 7 (+231g)
player 3 kills player 8 (+231g)

The bounties of the players 1, 2 and 3 now increases by 5% of the gained bounty again, this time by 12 gold. Their new bounty is 287g

Net gold kill: 1479g

I guess you substracted the bounty of the dark force players next, right?

In that case you would have:
1479 - 3*287 = 618g

As you can see, this number is about 3 times as high, as you calculated and not at all out of proportions. Just to see how it would turn out, lets see what it would look like if everyone from the dark force would die again:
After their first death, they'll now would have a bounty of 253g.

1479 - 3*287 - 3*253 = -141g

As you would expect with this new system, the light force would come out on top, by about 140g. This is an advantage of 47 gold per tank, so about two creep kills


Just to be clear again, I didn't show you this to tell you, that the system works perfectly as it is right now. I just wanted to make sure, that everyone actually understands how it works correctly, so people don't start making false assumptions.

[stibi-]

from my experience i can tell that the games last longer since there is no suuuuuuuuuuuuuper fed pro anymore thanks to double/triple/mulitkill
so there are 2 sides of one coin

[Wupti]

First of all nice, with a fast response I hope more players will join this discussion because I think it's a very important topic.

Before answering i want to get 1 thing straight when you say

I guess you get this from your own feeling of the games you play, but not the facts that you stated here, since they seem to base on false assumptions.

Yes I said my calculations would be wrong (note the last comment of the post). Since I only knew the boundaries, and not how the bounty scale(5% obv) it wasnt possible for me to make exact estimations which i also pointed out.
I still think it gave an idea of my point though.

So you are saying, that the new system is too forgiving for player mistakes?

Yes I do. Unless they missplay heavily and lose a CP I find it impossible to gain a real egde now. The answer for this you actually wrote yourself

Bascially, your own bounty increases everytime you receive gold, except from the passive income, force gold or selling items.

The only way of getting an egde (gold without getting a disadvantage), is to:
a) increase your passive income (leaver, dc, kick)
b) forcegold, which as far as I know only can be done by taking or destroying a cp (correct me if im wrong)

Since spawn times early and midgame wise are so short, this would mean enemy team has to make a huge missplay by either.
a) get obliterated completely
b) getting caught heavily out of position

Which against a decent team just isnt possible

What I wanted to show you with my example was the following: in a normal situation, an average player should have about the same bounty as before. Better players will have a higher bounty and others will have a lower one.

***********************************************************************************************************
This is where the main problem lies in my opinion and maybe I'm the only one.
What you do here is an effort to balance the games. You forget that the games are already balanced somewhat with elo already.
This creates a very bad scenario for high elo players since they already get the "worse" team AND The new system also works against them?
***********************************************************************************************************

Just to give you an idea on how high these differences are, I showed an example of the old bounty (the 262 gold) and the minimum bounty a tank with the new system could have, with the same tank worth (the 225 gold). With the new system, you can't go below 225 gold, after dying several times. This miminum bounty obviously increases as you get better items / tanks. The same is true for the upper limit, with the difference that this limit is farther away from the old bounty, meaning that players with such bounty are far more lucrative targets as they were before.

I understood all this, but as I pointed out i couldnt make these calculations without knowing the scaling and made a very simplified example.

Taking your example and assuming everybody really starts with 262 gold, it would actually play out like this:
player 1 kills player 6 (+262g)
player 2 kills player 7 (+262g)
player 3 kills player 8 (+262g)

The bounties of the players 1, 2 and 3 now increases by 5% of the gained bounty, so by 13 gold. Their new bounty is 275g
The bounties of the players 6, 7 and 8 decrease by 12%, so they now have a bounty of 231g, just above the limit of the 225g.

player 1 kills player 6 (+231g)
player 2 kills player 7 (+231g)
player 3 kills player 8 (+231g)

The bounties of the players 1, 2 and 3 now increases by 5% of the gained bounty again, this time by 12 gold. Their new bounty is 287g

Net gold kill: 1479g

I guess you substracted the bounty of the dark force players next, right?

In that case you would have:
1479 - 3*287 = 618g

As you can see, this number is about 3 times as high, as you calculated and not at all out of proportions. Just to see how it would turn out, lets see what it would look like if everyone from the dark force would die again:
After their first death, they'll now would have a bounty of 253g.

1479 - 3*287 - 3*253 = -141g

As you would expect with this new system, the light force would come out on top, by about 140g. This is an advantage of 47 gold per tank, so about two creep kills

Great Example and nice explanation
I'm very curious about what a underlined in your post. Can you explain why the difference between the increase and decrease is so huge.

Now let me extend it even further to show my point on how the gold will balance out resulting in any early/midgame egdes mostly are lost.


player 1 kills player 6 (+262g)
player 2 kills player 7 (+262g)
player 3 kills player 8 (+262g)

The bounties of the players 1, 2 and 3 now increases by 5% of the gained bounty, so by 13 gold. Their new bounty is 275g
The bounties of the players 6, 7 and 8 decrease by 12%, so they now have a bounty of 231g, just above the limit of the 225g.

Lets assume this happens 6 times

262x3+231x3+225x3+225x3+225x3+225x3 = 3504g

Teams 1's new bounty should now be:
262x0.05+231x0.05+225x0.05+225x0.05 = 13.1+11.55+11.25+11.25+11.25+11.25 = 69.65

Original bounty + added bounty = new bounty

262 + 70 = 332

Now Team 2 kill team 1 half the amount 3 times.

Bounties decreases
after first death
332x0.12 = 39.94
332-40=292
after second death
292x0.12 = 35.04
292-35=257

332x3+292x3+257x3 = 2643g

3504 - 2640 = 861

this is without taking assist into account since I can't find any formula and forgot how much % it is. But I assume it should be close to 1000g when added assist
Team 1 12 Kills
Team 2 6 Kills

Compared to old system

where it should be 3504:2 = 1752g

+ double amount of assists which I assume make it close to 2000g

Difference without assists

1752-861 = 891

With assist (Roughly guess)
2000-1000 = 1000

Regardless whether my guess is wrong or not compared to old system you gain around half the egde you originally would (in this scenario).

Just to be clear again, I didn't show you this to tell you, that the system works perfectly as it is right now. I just wanted to make sure, that everyone actually understands how it works correctly, so people don't start making false assumptions.

I know this.
I'm only here to discuss, give feedback and maybe from the POV of an higher elo player since they are too lazy to post



This post is most likely full of flaws since it's so long i don't bother check

[stibi-]

just a general thought to reduce the impact of massive feeders - could you lower the lower boundary?

[Prog]

Lowering the lower boundary makes it even more difficult to snowball advantages (which is exactly Wupti's criticism).



It's a problem that every time the bounty system changes comes up. Last time it resulted into the split up system with league being kinda unforgiving, and other modes having high stat influence and low boundary. However, only league was played so that solution wasn't great either.

It's just very difficult to find a middle way between what is right from a competitive aspect (as low of an stat influence as possible) and for a fun casual gameplay experience (keeping the influence of single feeders low). There are very good reasons for both sides. Personally I still don't know enough about games in the new bounty setting to have a justified belief about it, so I withhold judgement for now.

[Exodus]

Wupti, if you are interested in the specific numbers, just take a look into the Mechanics thread.

You get at least 25% additional gold from assists, so it would be more than you put there.

I knew about the fact, that this system would more or less close the gap between players somewhat. But I felt it was necessary with the addition of the killing sprees. The additional gold you can get from that is nothing to sniff at. It's a shift in the gold balance and on how you have to approach the game. As I'm sure you understand, it's kinda hard to get the balancing between such systems right, especially on the first try.
Bascially, the idea was to make comebacks somewhat more likely (and thus potentially increase game length), while also rewarding good play more (in form of more gold for killing and assist sprees). Since this latter one is a usually one-sided bonus, it would favor a shorter game length. My hope was, that these two mechanics would balance each other out in the long run.
If you feel they don't do that yet, we just have to figure out, where we have to adjust things to move them in the right direction.


The values for increasing and decreasing the bounty are purely based on empirical testing. I wanted to have a steadily increasing bounty on every tank over the course of the game and those values came closest to the pace we had with the old bounty.

[Wupti]

Lowering the lower boundary makes it even more difficult to snowball advantages (which is exactly Wupti's criticism)

Partly yes. I feel like you can play a veeery sloopy early and midgame without it resulting in anything and then win the lategame.
One may ask if you are so great then shouldnt you shine even brighter late game?
The answer is no.
What imo is the difference between a good player and a great player is he take those small advantages early on. Now these edges have become even smaller and almost marginal. So yes you might be 5k ahead early game. But as the game progresses 5k difference becomes less and less significant.
Yes with slightly inferior mechanics you should still win the game. However duo to the fact that the elo system put the best and the worst players in the same team to balance (ideally speaking) I think this put high elo players into an even harder position
in a titan battle for instance
1pro + 1 noob
vs
2regulars
Who would you favor?
I get that it is hard to balance the new system and every beginning is hard.

I just feel like if a new btt were to come and we played with the same rules every game would end a draw. (if you exclude black suns)

Wupti, if you are interested in the specific numbers, just take a look into the Mechanics thread.

You get at least 25% additional gold from assists, so it would be more than you put there.

ah thank you.
So to end my example assuming every single kill has 2 assists

Team 1: 18 kills
Team 2 9 kills

New system:
Team 1:
3504+3504x0.25+3504x0.25 = 5256

Team 2:
2643+2643x0.25+2643x0.25 = 3965

Differential: 5256-3965 = 1291

Old system

Team 1:
262x3x6 = 4716 (no idea why i picked 3504 before since its the new system)

4716+4716x0.25+4716x0.25 = 7074

Team 2:

262x3x3 = 2358

2358+2358x0.25+2358x0.25 = 3537

Differential: 7074-3537 = 3537

I think the numbers speak for themselves.

I knew about the fact, that this system would more or less close the gap between players somewhat. But I felt it was necessary with the addition of the killing sprees. The additional gold you can get from that is nothing to sniff at.

Yes the sprees indeed is nothing to sniff at, however you also balanced this somewhat by removing multikill bonuses? besides the spree end after 3 minutes without a kill, so you will have to play exceptionally well or the opposing team must be on tilt to take advantage on this. I mean Just to reach the spree without dying is hard enough, but you also have to play almost suicidal to keep it going.
Therefore when you write:

Bascially, the idea was to make comebacks somewhat more likely (and thus potentially increase game length), while also rewarding good play more (in form of more gold for killing and assist sprees). Since this latter one is a usually one-sided bonus, it would favor a shorter game length.

I think the huge problem is the later part. remember you have to hit 3 kills before the spree even starts. at least if you find me a player with at least 10 games and a k/d of more than 3 let me know.

[Max]

Hi,

Just to react on this, I want to say that, in the few games I've played I kinda like the new system :

- The new spree system give opportunities for comebacks (which was +/- impossible before).
- It gives some more balance : as you say Wupti when you play with a ~1350 ELO player, when he feeds he gives less gold than before, and the more skilled players will give more bounty when get killed. I don't think it's a bad thing. When we see your stats for example : you don't die a lot of times every game (you have an average ~16 which is low), so I'm not shocked that you give more bounty when you die (you get more gold, you give more).

I'm telling that about strategy : now to win a game, you'll have to fight every player in enemy team to get enough gold to upgrade your tank enough. If you focus one guy he will give less gold, so it will be less worth it, and games will be more about killing every players in enemy team than taking advantage of a feeder in a team. And a team who get a feeder will be less handicapped by him.

Well my message isn't good structured but I think you got my point. Again I didn't play a lot of games on this version, so I may not have a full idea of this change, but that doesn't change my point.

[Wupti]

Good to see more faces in this thread

Hi,

Just to react on this, I want to say that, in the few games I've played I kinda like the new system :

- The new spree system give opportunities for comebacks (which was +/- impossible before).

Not sure about others, but I often made comeback victories in earlier versions

- It gives some more balance : as you say Wupti when you play with a ~1350 ELO player, when he feeds he gives less gold than before, and the more skilled players will give more bounty when get killed. I don't think it's a bad thing. When we see your stats for example : you don't die a lot of times every game (you have an average ~16 which is low), so I'm not shocked that you give more bounty when you die (you get more gold, you give more).

So we should punish the players who plays well and benefit the players who play poorly.
I'm sorry but this isn't logical to me

I'm telling that about strategy : now to win a game, you'll have to fight every player in enemy team to get enough gold to upgrade your tank enough. If you focus one guy he will give less gold, so it will be less worth it, and games will be more about killing every players in enemy team than taking advantage of a feeder in a team. And a team who get a feeder will be less handicapped by him.

The exact opposite of this is what's happening to the game.
Since people now can stop care about stats and no real egdes are gained from fighting, what happens to the game is it appeal to backdoors.
This is why you see more guys with adv troop and tp behind enemy lines or sky forts suiciding on factories.
Personally this is not the direction i want battle tanks to go. What I like about battle tanks is the big team fights that can turn the games, and not some solo backdoor plays.

[Ch4oSxD]

in earlier version was a comeback possible. but 1-3 Hour games are to long, just because its possible to prolong a game.

[Max]

- It gives some more balance : as you say Wupti when you play with a ~1350 ELO player, when he feeds he gives less gold than before, and the more skilled players will give more bounty when get killed. I don't think it's a bad thing. When we see your stats for example : you don't die a lot of times every game (you have an average ~16 which is low), so I'm not shocked that you give more bounty when you die (you get more gold, you give more).

So we should punish the players who plays well and benefit the players who play poorly.
I'm sorry but this isn't logical to me

Well I don't get this change is about to punish good players and benefit the bad players, but to balance games more. Some times you have a "bad balance", play with bad players vs 2-3 really good players. A game almost impossible to win. With this change games are balanced with ELO and there's an in-game balance.
That's how I understand this and how I see this change...

I'm telling that about strategy : now to win a game, you'll have to fight every player in enemy team to get enough gold to upgrade your tank enough. If you focus one guy he will give less gold, so it will be less worth it, and games will be more about killing every players in enemy team than taking advantage of a feeder in a team. And a team who get a feeder will be less handicapped by him.

The exact opposite of this is what's happening to the game.
Since people now can stop care about stats and no real egdes are gained from fighting, what happens to the game is it appeal to backdoors.
This is why you see more guys with adv troop and tp behind enemy lines or sky forts suiciding on factories.
Personally this is not the direction i want battle tanks to go. What I like about battle tanks is the big team fights that can turn the games, and not some solo backdoor plays.

Well for this point as I said I didn't play a lot of games. It's my "theoretical" opinion about how the new bounty system should change the gameplay. I say theoretical because I have played ~5 games since the new version is released so my feeling about this might change after some other games.

My opinion about what you are talking about - camping and solo actions - is because of the new spree system (as I dreaded when the map was still in beta) more than the whole new bounty system. The killing spree that stops when you die promote bullshit camping tp thing in my opinion. Maybe not the new bounty system.

PS : sky fort suiciding on factories : it's the best way to use this tank

[LIoOoOoIL]

atm i feel more like:
Probably there is just a need for playing around with thus values , to balance a way which still benefit good playing in a constant line through the whole game.
Hm, thinking about something like, with an higher Bounty, u earn percentag(%) more gold through creeping and specially killing Opponents. But since this gives an advantage for passiv play/camping to focus on creeping, how about to lowering the raising bounty through creeping in generell like it does atm? Im not sure about this, since it still benefit passiv play, but in the other hand it also benefit to kill other Tanks, which more goes in the way to get an advantage about them.

[Wupti]

Well I don't get this change is about to punish good players and benefit the bad players, but to balance games more. Some times you have a "bad balance", play with bad players vs 2-3 really good players. A game almost impossible to win. With this change games are balanced with ELO and there's an in-game balance.
That's how I understand this and how I see this change...

Ofcause its not made with the intention of that, but thats kinda how it turns out to work. Yes it might work perfectly fine in games with bad balance, BECAUSE its so hard to gain an egde. In the rest of games though its just..
You might be unlucky sometimes with the elo balance yes, but trying to fix this by making an ingame balance is just ridicilous. People should just learn to deal with variance.
If I should relate it to poker it would be like people whining about losing with aces all in preflop (variance) and therefore make a change so you still win if you win, but chop if you lose, lmao this would be ridicilous aswell and ruin the game

PS : sky fort suiciding on factories : it's the best way to use this tank

Thanks to the changes yes

[Exodus]

Possible options for change:
In general I'd say, I could try and strengthen the sprees a bit and lessen the effect of personal performance on the bounty. This could be achieved in the following ways:

Killing Spree: I could remove the condition that the sprees end on death and just leave the time constraint (and maybe lower it a little bit). That way it should be a bit easier to maintain, as well as promote more agressive play.

Bounty: I could reduce both the bounty increase and decrease, which would effectively reduce spikes in the bounty progression (or at least delay them), but leave the general trend intact (this is my hope anyway).

[GodStrongArms]

About the differential of value-coef between good players and feeders, I also think it's a good thing. It's not a punishement to be worth more, but a simple reflexion of the reality into a game. Wupti you're one of the league major players. I feel better about myself to kill a high value target like you ingame than I do killing guangzhou or petere... Glory has a price. The prized target is hard to kill, but worth more than a supercreep. So imho this feature must be kept.
Removing the end of death, however, may be a mistake. You want to encourage aggressive play or create kamikazes ? Instead of that, indeed Exodus, lower (a bit) the spree time constraint, and reduce (a bit) the bounty increase and decrease may be a more localized fix to.

Overall, I'm very pleased by this spree system. Psychopaths and mass murderers in game have to be rewarded for their behavior ! And THIS time, the good players, or those who are on a "spree", get a interresting :discussed: advantage on their victims.

About towers, I strongly support to share tower Bounty for whole team, with a bonus to the last-hitter. Or at least to grant an assist (with a few minutes timer) to those who help to destroy it. The current system is not satisfactory at all. It incitates to greed (stealing, senseless push on lane when action is mid), and goes against the teamplay this game is / should be based on.