Dynamic balancing by bidding
matso
Master of Patches Join Date: 2002-11-05 Member: 7000Members, Forum Moderators, NS2 Developer, Constellation, NS2 Playtester, Squad Five Blue, Squad Five Silver, Squad Five Gold, Reinforced - Shadow, NS2 Community Developer
Master of Patches Join Date: 2002-11-05 Member: 7000Members, Forum Moderators, NS2 Developer, Constellation, NS2 Playtester, Squad Five Blue, Squad Five Silver, Squad Five Gold, Reinforced - Shadow, NS2 Community Developer
I have always been found of the idea that teams should bid for the sides rather than playing alternates.
The idea would be that both teams places a secret bid for how many extra tres they want as marines. The lowest bid get to play marines with that many extra tres (negative bids means extra alien tres rather than less than normal marine tres).
So if you bid 20 tres for marines and the other team bids 10, they get to play marines with an extra 10 tres. If you bid -20 tres and they bid 10, you would get to play marines with normal tres vs an alien team with an extra 20 tres.
There are several advantages to this - not the least is that it would give you information just how balanced the game is thought to be; you get an actual tres measurement of it.
And of course, if the game balances differently at different skill levels, that can be compensated for by the teams; if they feel marines are OP, they can bid less to get marines.
Likewise, if certain maps are seen as biased, bids can change to compensate.
It WOULD screw with fixed build orders, as you would not know just how much tres you would start with. Some may see this as a bad thing, others may see it as a good thing... it would require commanders to adjust the gameplan dynamically.
And of course, you would no longer be certain to play both sides in a scrim.
The idea would be that both teams places a secret bid for how many extra tres they want as marines. The lowest bid get to play marines with that many extra tres (negative bids means extra alien tres rather than less than normal marine tres).
So if you bid 20 tres for marines and the other team bids 10, they get to play marines with an extra 10 tres. If you bid -20 tres and they bid 10, you would get to play marines with normal tres vs an alien team with an extra 20 tres.
There are several advantages to this - not the least is that it would give you information just how balanced the game is thought to be; you get an actual tres measurement of it.
And of course, if the game balances differently at different skill levels, that can be compensated for by the teams; if they feel marines are OP, they can bid less to get marines.
Likewise, if certain maps are seen as biased, bids can change to compensate.
It WOULD screw with fixed build orders, as you would not know just how much tres you would start with. Some may see this as a bad thing, others may see it as a good thing... it would require commanders to adjust the gameplan dynamically.
And of course, you would no longer be certain to play both sides in a scrim.
Comments
But this.. I don't even.
Maybe if someone explained in their words what they thought I suggested?
Sorry about that.
Right. Starting over...
NS2 is very difficult to balance. Not only is it inherently difficult to balance, as you have two intentionally asymetrical sides, but it also tries to cater to public play as well as comp play. To make matters even worse, NS2 balance changes depending on the skill levels of the teams playing - top level comp games plays very differently from division 3.
So you can end up with things like lots of best-of-5 scrims ending in 3-2 wins, with all wins belonging to the aliens side (or marine). If that happens to often, you might just want to save your time and flip the coin for who gets to play the winning side in the last game (and skip playing completly...).
But the very same game in the hands of teams of a different skill level could be very balanced - or uneven the other way. Or completly unplayable in public play.
It would seem to make it really, really hard to balance the game for everyone.
By introducing bidding (for comp games only - pub games are too chaotic), teams would have a chance of balancing each game to their own taste (ie, their own belief in where the balance is).
Now, the bid is how much extra tres you want for playing marines (or if you give a negative bid, how much tres you are willing to give the aliens if you get to play marines).
Implementation wise, the commanders would give a console command before the game started, something like "bid -20" (if you really want to play alien) or "bid 25" if you want play marines with an extra 25 tres. Once both bids are in, they are revealed and players moved to the correct sides.
Whoever bids THE LEAST gets to play marine, with that much extra tres. If both bid the same, sides are randomized.
Let's do an example.
---
Assuming team Arsekickers meat BeatsToAPulp in a best of five. They haven't played each other before, but at their skill level, aliens are generally considered to be easier to win with, and we assume that they are reasonably even, skillwise.
First game, Arsekickers bids (secretly) 30 points and BeatsToAPulp bids (secretly) 20 points, so BeatsToAPulp gets to play marines with a 20 points advantage. And they loose. Badly. They realize that they need more of an advantage if they are supposed to win as marines.
Second game. BeatsToAPulp bids 40 points, while Arsekickers bids 30. Arsekickers plays marines with 30 pts, and gets beaten to a pulp.
1-1, both alien wins. For the third game, Arsekickers bids 50 pts, but BeatsToAPulp, realizing that Arsekickers can't aim, bids 60 pts, letting Arsekickers play marines with an extra 50 tres. This time, with the help of fast upgrades, Arsekickers kicks some and manage to win a marine game.
... etc.
---
For the next game, you would expect BeatsToAPulp to drop their bid to below 50 - as that was obviously too much - while Arsekickers might realize the same and drop their bidding down a bit.
So you would end up with a bidding trending towards something that both sides thinks are fair.
Uneven teams .. well, the better team can just bid zero and be certain to not play at a tres disadvantage.
Your bids may be influenced by which map you are playing on, or if you know that the other team has a stronger marine than alien game, or you know your own strengths... but that's just fine. Knowledge pays. And if you don't, you get a chance to adjust your bids for the next game.
I feel you bro =D
The proper response in this case is to balance the game. Adding some arbitrary res betting system is covering up bad balance instead of fixing bad balance. The comp mod itself is attempting to make this game properly balanced and that should be where the future of NS 2 is, not some bettering system.
Getting a balanced game is not incompatible with using a bidding system. In fact, having a bidding system can be a very large help, as it will tell you what everyone actually belives the balance is. No need to argue theoretical pros and cons - just check what the bids are.
If everyone bids very close to zero, you know that the balance is pretty good.
If div 3 games bids 20pts more for marines than div 1 games do, then you know how the balance changes with skill level.
If a certain map have lower bids than other maps, you know that map is known to give aliens an advantage - and you can see exactly how much.
And the fun thing is - you can get balanced, even games even BEFORE the game is perfectly balanced!
Blerg, I am just not impressed. The system introduces too much uncertainty in something that should be carefully planned. No one is going to want to try your suggestion. Joshy already said what 90% of the comp players are going to say.
I'm trying to think of something clever to say but I can't think of a single thing. The idea is insane and even a 10 res difference is huge.
Yea, that's better. Simpler to explain it - just auction out the right to choose team. Winning team gets to choose side, loosing team gets the tres.
There might be a tediousness issue as there is no reason to bid more than 1 more than the opposing team; though I guess that could be solved by requiring a minimum bid increase of a reasonable size (3-5 or so?).
Though you would have to do it for each game - yea, originally when I first thought of this, I was thinking of only doing it for the deciding game. But I realized that it would be kinda meaningless if you only do it for the deciding round; if the game is sufficiently unbalanced to need a tres adjustment, the non-adjusted games are pointless. If it sufficiently balanced not to need the tres adjustment, no need for the extra complexity.
Again though - adding in bidding will TELL you if the game is seen as balanced. If teams figure that they have an equal chance to win, they will just let the other team bid up and cash in the tres.
Yea, that's the source of the idea.
Why would you bid that much? Are you really feeling that unconfident that you need 100 tres more to win?
I have nothing.
To be fair, we lost some marine rounds through pure and simple bad play, and there should definitely be no balancing for bad aim or stupidity, but it's gone a bit too far towards aliens at present, I think. At least with bidding you could try to even things out a bit...
The reason the higher lifeforms can come back a bit quicker is to help aliens with a high amount of RTs end the game faster.
Again, we made some mistakes tactically, and I'm trying not to bias my opinion too much based on that, because I'm well aware of the specific issues we had in our recent marine rounds and they should not be balanced for.
But as it stands, the comp mod, while being very welcome and bringing lots of good changes, does leave the lower divs with some serious issues in its current form.
Yea.
Though in practice, it might be unlikely - if you loose the bid and get stomped, you are likely to bid higher for the next game, until you finally get to play the team of your choosing.
Or the other teams wins the bidding and you stomp THEM, in which case they are likely to bid less the next game, allowing you to choose team.
Bw, an additional advantage to having a bidding system is that you are freer to introduce changes to the comp game without worrying too much about balance, as minor balance changes can be handled by the bidding system.
On top of that, "betting" outside the game (and the I-think-he-thinks that goes with it) isn't an NS2 related skill and shouldn't be applied to it. Otherwise, why not play rock-paper-scissors or a game of monopoly to determine who gets what side? Let risky strategies be implemented in game, not pre-game.
Betting works in a game like Spades because risk analysis and risk taking is 100% of the skill in the game. In NS2, it isn't. And shifting weight away from the current skill mechanics to a risk evaluation mechanic is not something that I would like to see.
Those are all good points.
And there is no reason to introduce bidding to NS2 for its own sake.
The question is simply if the problems solved by bidding is worth the problems introduced by it.
But do note that the disadvantages you point out would only come into play if the game IS unbalanced - if the game is balanced, the winning strategy would be to bid zero and get whatever advantage the other team is willing to give you.
So bidding will only matter if the game is unbalanced - at which time it is probably nice to have a way of getting even games while you wait for the next patch to fix the balance.