Pub team evener

[Kolar]

With less people you get less organized teams, or at least less people who are basers showing people how to play the game and where to go. If people don't have direction in this game they'll sit and spawn and besides the amount of people per public is fine as it is (there's enough people basing and enough people spawn/not playing the game). A bot could do what I said and the balancing thing from the other zone without risking a major effect to the population and stablilty of the zone.

[Troll King]

Lower it from 36 to 30 and you'll still have at least 10 per default pub freq. That's still more than you see in "organized" base matches. What effect would this have to population? Would you leave the zone if there were only 30 people in pubs? How does it make it less stable? As I said at the end of my last post, TW's pub arena cap was at 30 for years. It worked well then and though other factors are involved, you could make a case for pubs being worse now than it was then.

[Sarien]

I like Troll King's idea. I think it makes sense on the face of it, and is backed up by the numbers. A drop in arena population of a bit not only keeps things from being terribly lopsided it gives more arenas for choice, which means there are more likely to be more base games going in some. More choice is almost always a good thing.

I think Kolar has his heart in the right place but he is forgetting something: You can't force people to base. You can only make it worth their while to do so. You can't push people in a direction and not expect them to push back. I'm in favor of a lot of unhealthy hindrances to basing (leviathans) being changed or removed, but that's because they're so extreme as to kill a game that's been started, let alone keep one from starting.

[Kuukunen]

And here's the kicker: with 5 or 6 fewer people in an arena, you can be fairly sure that there'd be one less levi per arena. That's just from pure mathematics.

Not one, but 0.5 or 0.7. And talking about PURE mathematics, if there's no space in pure pub, I don't go to pub.

and is backed up by the numbers

*sigh*
You like numbers then? Let's do some math.

Let's think about two arenas, other with 10 vs 10 and other with 20 vs 20. Let's say probability for player to leave after a round is 10%, making probability for him to stay 90%. So let's count the probabilities for 10v10 turning to 10v6 and 20v20 turning to 20v12. (Which, imho, pretty much just as unbalanced, ratio of people is 5:3)

So 10v6 situation first. Probability for 4 guys to leave is 0.1^4, probability for them to leave AND other 6 to stay is 0.1^4 * 0.9^6. But there is 10C4 = 210 ways to pick 4 people out of 10. The group leaving could be any of those 210 possible groups. So the whole thing is 10C4 * 0.1^4 * 0.9^6.

That's only halfway there, probability for the whole other team to stay is 0.9^10, making the whole expression (210 * 0.1^4 * 0.9^6) * 0.9^10 which is 0.0038913424 so probability for 10v10 becoming 10v6 is 0.389%.

Let's do 20v12 then. 20C8 = 125970. (There's lots of ways to pick 8 people out of 20.) (125970 * 0.1^8 * 0.9^12) * 0.9^20 = 0.000043254115084 = 0.00433%

But you have to remember it might not be exactly 8, with 20 people there's two times more possibilities in the number of people leaving, so the actual number is a bit bigger. Let's do scenario where 7 people leave from the other freq: (77520 * 0.1^7 * 0.9^13) * 0.9^20 = 0.0002395613 = 0.024%.

Probabilities of the scenarios of 7 and 8 people leaving one freq put together makes 0.0002828154 = 0.0283% which is A LOT more unlikely than 0.389% of 10v10 becoming 10v6.

There's your numbers, happy?

Of course in real life, those kind of situations happen a bit more often. Mostly because people on losing team have a higher probability to leave. (Doesn't change the 20v20 against 10v10 though.) Also if there's some kind of group that leaves together, teams might get more unbalanced. (This would be in favour of 20v20 though, because a group of 4 is a lot more likely than a group of 8, no matter what the arena cap is.)

It's been a long time since I've done math and I'm a bit tired right now, so it's very likely I made an error, please point it out if you notice something. But I think I got the basic thing right, which is: Less people in arena DOES NOT mean more balanced teams.

[Sarien]

Well first it's not 20 vs 20.
It's 15 vs 15 (max team size) with 5 on another freq or freqs (probably LTing or spawn warbirding)

With a drop to 30 people max would be 15 vs 15 (max team size)
Which is what I've seen TK suggest before.

Secondly you're doing the math from inside one arena as if you're trapped there. Can you take into account an average number of public arenas with people in them with a 36 limit, vs what that would mean with 30? Then determine the likelihood that a basing game would be going on in each, vs complete chaos.

[Kuukunen]

Well first it's not 20 vs 20.
It's 15 vs 15 (max team size) with 5 on another freq or freqs (probably LTing or spawn warbirding)

With a drop to 30 people max would be 15 vs 15 (max team size)
Which is what I've seen TK suggest before.

I know the numbers.
Just that 15 and 10 don't go well together on those calculations, they would get messy. They were just for making a point that reducing team size actually makes it more likely to have unbalanced teams.

Secondly you're doing the math from inside one arena as if you're trapped there. Can you take into account an average number of public arenas with people in them with a 36 limit, vs what that would mean with 30? Then determine the likelihood that a basing game would be going on in each, vs complete chaos.

I think you missed my point here. I'm talking about teams of ONE arena, the average number of pubs has nothing to do with it. And you still want numbers? So let's "determine the likelihood that a basing game would be going on in each". It's easy to compare 30 vs 36. More pubs = less likely for all to have base game going on. Why? Let x be a probability of basing going on in one pub. x^5 > x^6.

I wasn't talking about basing, I was just talking about team sizes. If we take basing into account, it's a different story. Often there isn't that active basing in even the timed arena. What about people who try to base in the other arenas then? I think I've never seen too many people basing in a pub.

[Theif of Time]

Gonna need to do some research... tabulate some pub base probability values... interview players... Lagout statistics... ave. Session times... usages... tolerance of idiot factors... take into account conversations on politics, religion, science, philosophy that force people to spec... Probability of hosted events... probability of local servers (ISP'S) being down... probability of going to automated events... probability of nuclear strike in europe taking down number of players... probability of govt censorship of free speech in any number of countries... population density, time of day... probability of specific people being on (social factors)... probability that player quit entirely due to RL...

Kuuk, remember that basing can be effected by social factors: Ie if 1 terr and a shark start to base there is a chance the rest of the team follows suit.

[Evul Fly]

limit javs to 3k or more usage and ull have the fair teams!

[Kuukunen]

Kuuk, remember that basing can be effected by social factors: Ie if 1 terr and a shark start to base there is a chance the rest of the team follows suit.

I do know that it isn't all numbers. But I can do numbers on that too.
Smaller teams = less likely for the team to have a willing terr and a willing shark.
Actually: With more pubs, not only levs are spread around the pubs more, but also those people who are willing to shark and terr. (Can you say there's too many good sharks in pubs competing with each others of who gets to shark for the freq?)

Just reminded me: one good terr and shark is all you need http://koti.mbnet.fi/kuukunen/kuu_and_malla.jpg (priv freqqing with malla way back in wr)

Anyways, why all the numbers? I just got tired of things like

The same thing can happen in current pub arenas, and that will change 12-8 teams to 12-4. It's the same problem, but you still have a better chance when it's 10-6 than if it were 12-4. It doesn't get rid of the problem entirely, but it doesn't make it as bad.

I like Troll King's idea. I think it makes sense on the face of it, and is backed up by the numbers.

[Sarien]

And I stand by the numbers that say, with 15 member team limits per freq, when you have a left over amount of people, 15vs15 + 5, then they're almost automatically going to default to detract from the basing game. Why does that matter? That is why people split, that is why they go off to hunt the troublemaker, or that is why they give up, because it ends up being not worth it. That leads to a good number of your uneven teams.

And While if it were the opposite of the way it actually is, if there were a basing game going on in almost every arena, increasing the number of arenas inhabited by players would decrease the chances for a base game. However the number of public arenas where a basing game is either going on or about to kick into action is in the minority, now.

It's all tangled. The way pub is now isn't like a set of numbers and stats that you can look at and tinker with and get a formulaic response. It's like a cotton ball, when you start tearing shit apart and you see how other things are affected. Case in point: Recent Weasel Controversy.

[Kuukunen]

And I stand by the numbers that say, with 15 member team limits per freq, when you have a left over amount of people, 15vs15 + 5, then they're almost automatically going to default to detract from the basing game. Why does that matter? That is why people split, that is why they go off to hunt the troublemaker, or that is why they give up, because it ends up being not worth it. That leads to a good number of your uneven teams.

If that's what you think, let's make the teams bigger.

[Troll King]

Here's the flaw to your logic: you misread my statement that "you'd have a better chance with 10-6 than with 12-4". I'm not talking about the chances of those occuring, I'm talking about your chances of being competitive down 10-6 than you would down 12-4. Look at the wording again and specifically at the word "with". With 10-6 teams, you can still be very competitive. The other flaw in your logic is that you're only looking at the odds of ONE team losing players while the other keeps all of them. The odds of one team losing 4 players are the same as the other team losing 4 as well. You also forget about incoming players. Your model works if no new people come in, and because those new players are from outside the system, probabilities fail us.

The urge to compare teams usually makes us think of the differences in terms of percentages, as in, if they have 12 and we have 9, then they have 33% more people than we do. The problem is that with numbers under 20, percentages aren't as effective as looking at the actual values. Instead of thinking of percentages and multiplication and division, we should be thinking in terms of difference, or addition and subtraction. Even if we use your numbers, you'll see that it's almost as likely for an arena of 10-ship teams to lose 4 (0.00389) as it is for a team of 20 to lose 5 (0.00388). Based on percentages, you'd say that 20-15 is better than being down 10-6. On the other hand, being down 10-6, you're only down 4 people instead of 5. When the next person comes into the arena, you're down just 3. With the 20-man teams, if you're down 5 and someone enters, you're now down 4. Smaller teams will correct themselves and balance out faster than larger team arenas.

All these numbers are conjectures anyways. Don't forget, I'm also presenting anecdotal evidence. For years, TW ran well with a pub cap of 30 per arena. It worked fine at that level and though other changes have been made since then, pub basing was better then than it is now. As for your argument that the number of sharks and terrs are now spread out over more arenas like the levis, you're forgetting something else too. This affects both teams in an arena, not just one. Both teams would be under the same handicap, so in balance, there's no overall difference. Levis are another matter because they don't work within a team concept. It makes no difference whether you're on one team or the other if you're a levi. In fact, you'll often find they are on private freqs.

[Kolar]

Just put it through TWPS, I don't think it will pass but it's still a better method then posting on the forums. As for "forcing people to base", no, publics will never be like that. If you have less players in the arena that means you're gonna have less people doing everything (spawning, basing, dueling ect..), I would say due to the majority of people basing, you would find more of a reduction in people doing it and a reduction in people showing players how to play publics, for each public arena. That's all I meant. Using a bot would not risk this effect or anything else to the population and quility of the zone.

[Troll King]

But that's the thing; a majority of pubbers AREN'T basing.

[Kuukunen]

Here's the flaw to your logic: you misread my statement that "you'd have a better chance with 10-6 than with 12-4". I'm not talking about the chances of those occuring, I'm talking about your chances of being competitive down 10-6 than you would down 12-4.

I didn't misread your statement, I failed to see the logic behind it. Why on earth would you compare the two?

Look at the wording again and specifically at the word "with". With 10-6 teams, you can still be very competitive. The other flaw in your logic is that you're only looking at the odds of ONE team losing players while the other keeps all of them. The odds of one team losing 4 players are the same as the other team losing 4 as well. You also forget about incoming players. Your model works if no new people come in, and because those new players are from outside the system, probabilities fail us.

I don't know if you missed it, but I did include to the probabilities that the other team doesn't lose any. It's the "* 0.9^20" thing there.

And about incoming players. Let's say x is the probability of at least one player entering the arena in the time slice. With more pubs, that x is lower, meaning people don't enter as fast, so even though with less people, you'd need less people to balance the teams, less people will come in, because there is more pubs they can go to.

The urge to compare teams usually makes us think of the differences in terms of percentages, as in, if they have 12 and we have 9, then they have 33% more people than we do. The problem is that with numbers under 20, percentages aren't as effective as looking at the actual values. Instead of thinking of percentages and multiplication and division, we should be thinking in terms of difference, or addition and subtraction. Even if we use your numbers, you'll see that it's almost as likely for an arena of 10-ship teams to lose 4 (0.00389) as it is for a team of 20 to lose 5 (0.00388). Based on percentages, you'd say that 20-15 is better than being down 10-6. On the other hand, being down 10-6, you're only down 4 people instead of 5. When the next person comes into the arena, you're down just 3. With the 20-man teams, if you're down 5 and someone enters, you're now down 4. Smaller teams will correct themselves and balance out faster than larger team arenas.

Of course 20v15 is better than 10v6. With 10v6, let's say they both teams have a terr and a shark, making it 8v4 for the other ships. (And if they don't have shark or terr, they're screwed anyways.) I'd say the percents are pretty good for approximating the power balance, because it basically tells you how many ships you have to kill to keep it even. As in, if it's 8v4 for fighters, you have to kill 2 ships for your every death. And if it's 18v13, the number is 1.38. Of course sharks and terrs kill some but not that much.

And that balancing thing I already went though, more pubs, less people coming to the arena, meaning they actually might not balance faster.

All these numbers are conjectures anyways. Don't forget, I'm also presenting anecdotal evidence. For years, TW ran well with a pub cap of 30 per arena. It worked fine at that level and though other changes have been made since then, pub basing was better then than it is now. As for your argument that the number of sharks and terrs are now spread out over more arenas like the levis, you're forgetting something else too. This affects both teams in an arena, not just one. Both teams would be under the same handicap, so in balance, there's no overall difference. Levis are another matter because they don't work within a team concept. It makes no difference whether you're on one team or the other if you're a levi. In fact, you'll often find they are on private freqs.

Yea, but it got changed. I don't know why, but I guess they had some idea behind it. I wouldn't blame pub size caps for the state of the pubs either. And have you ever noticed how so many people always think "Oh the good old days, things were better then." And I'm not talking just about subspace here.

I'd say basing without a terr isn't basing. I'd also say that if the other freq has a shark and the other doesn't, it isn't basing.