The Bot World

[Kestrel]

There's not enough grind in this game, the noobs are having it too easy.
We should definitely increase the amount of logs needed for Forest Bow to atleast 200, that quest is too easy.
Oh and the quest rewards should be more random so they would have to roll more alts to get everything, oh wait.

[Len]

I assume you’re being sarcastic, In any event it’s not that the game has too much grinding, it’s just that in TMW it’s particularly boring and painful (due to the dull combat and quest design). Which makes players more aware of it than they would be in another game with just as much grinding if not more, like RO.?

[Hello=)]

We should definitely increase the amount of logs needed for Forest Bow to atleast 200,

IIRC it's purely random if particular log is okay, so theoretically you can get to 200, if you're (un)lucky enough. Though it's "low probability" event. And probability could be really funny thing, especially in some arcane theoretic cases.

[Crush]

Hello Kestrel

It seems like you made it your personal goal to lobby for making TMW a less grind-focused game. I completely support you in that regard. But unfortunately your critique is completely destructive. You wrote that the grind-factor in TMW should be removed, but you have so far made no suggestion how. It would of course be possible to make everything much quicker to achieve, but the sad truth is that grind is the main feature of TMW at the moment. When you take that away, there would be nothing left.

So what features should be added in your opinion to make TMW a better game so the grind-factor can be reduced?

Also remember that TMW is a collaborative effort by volunteers. Nothing on TMW happens unless volunteers make it happen. So I would also like you to write which parts of the implementation of your ideas could be made by yourself and with which parts you would need help from an experienced developer.

[Hello=)]

Crush, do you really want to hear what to change in game mechanics to make it more fun, less grind and so on? I can propose changes. The only problem is that devs would want to kill me as it requires to rework half of game. Players would want to kill me because these are rather drastic changes. And botters would want to kill me because their bots would suck .

[SriNitayanda]

As long as you dont actually make these changes, they wont happen I see nothing wrong with posting them. But as you can see it got very minor influence posting them.

[Nard]

Hello Kestrel

It seems like you made it your personal goal to lobby for making TMW a less grind-focused game. I completely support you in that regard. But unfortunately your critique is completely destructive. You wrote that the grind-factor in TMW should be removed, but you have so far made no suggestion how. It would of course be possible to make everything much quicker to achieve, but the sad truth is that grind is the main feature of TMW at the moment. When you take that away, there would be nothing left.

So what features should be added in your opinion to make TMW a better game so the grind-factor can be reduced?

Also remember that TMW is a collaborative effort by volunteers. Nothing on TMW happens unless volunteers make it happen. So I would also like you to write which parts of the implementation of your ideas could be made by yourself and with which parts you would need help from an experienced developer.

Many suggestions have already been made, some by your servant. Replace the random geometric trial by an hypergeometric one for example. Other also suggested just to bound the number of items required to a maximum of your choice as there is none at the moment. I even remember to have suggested to write some math extensions of eA script in order to simplify non specialist developer's job. Finally I asked for the math extension in wiki, with, in my mind, the goal to make a crash course of probabilities (and maybe other maths and physics) for game developers... This was at least a year and a half ago. I still have only an answer from Platyna: "yes the extensions you ask for are interesting"... Conclude what you want.

edit1:

Code: Select all

function HypergeometricSuccessTrial(Nmax, Nsuccess, k)
%% This function simulates an Hypergeometric first success trial,i.e. trial in a set 
%% containing Nmax possibilities among which are Nsuccess success possibilities.
%% k is the order of the trial an has to be incremented by 1 at each trial, 
%% Nfails = Nmax-Nsuccess
%% Nsuccess < Nmax 
%% k < Nfails, as soon as it equals success is sure.
%% Initial probability of success: p0 = Nsuccess/Nmax
%% Probability of success after k-1 trial pk = Nsuccess/(Nmax-k+1)
%% It returns success=1 if the trial was a success, 0 if not.
%% You have to update the Nsuccess variable by ±1 depending of the result; typically
%% Nsuccess:=Nsuccess-success.
success:=0;
trial:= Rand(Nmax-k+1);
if trial ≤ Nsuccess then success := 1; 
return success;
endfunction;

pseudo code
edit 2:
Wikipedia: Hypergeometric distribution.
http://en.wikipedia.org/wiki/Hypergeome ... stribution


edit 3: A standard player is not supposed to know a single thing in game programming, or in computer drawing/pixart, or in music. But his feedback should be welcome. A programmer is supposed to be aware of possible solutions. If he does not feel skilled enough, he is supposed to be able to write the question with the appropriate language (analysis), ask it to some skilled person (that may wander around), and if (when) he gets an answer, be able to transcribe it into machine understandable stuff (programming).

[Hello=)]

As long as you dont actually make these changes, they wont happen I see nothing wrong with posting them. But as you can see it got very minor influence posting them.

Then I would briefly outline what's wrong and what IMO to change. But it's way too much job on reworking and balance to be realistic.

TMW haves several key issues:
1) There is way too much grinding and low drop rates. This forces players to spend too much time on monotonous boring crap, which isn't epic. So it provokes botting. Machines are better at monotonous routine work which isn't dangerous at all.
2) On other hand, crafting and mining is absent. Yet, crafting and mining can be nearly main source of income instead of grind. Mining can also be turned into (dangerous) adventures in horrific caves.
3) Economy does not assumes items sinking, only items sourcing. This reduces value of existing items produced by players and would not coexist well with crafting.
4) XP rewards also not really well-balanced.

Solution? In general, to have good reward players should be forced to do relatively dangerous and challenging actions where outcome isn't 100% predefined so player haves some chance to die if playing poorly and doing too much mistakes. That's where bots suck.
Obvious things to admit are:
1) Killing weak monsters should not give any noticeable rewards in terms of XP. The best approach I've seen is taking (Target level - Maximum attacker's level) as "difficulty base" and computing XP reward based on this difference. This way only challenging actions are rewarded well. In principle, tweaking XP curve can achieve something like this without drastic changes, though mentioned approach leads to fact botting for XP getting nearly 100% pointless thing as you have re-target bot for different monsters and it would die a lot if it expected to get reasonable rewards.
2) If easy monsters drop too much of valuable stuff, it would attract bots which are good at safe routine works. Monsters should drop worthless crap most of time. You can sell it for few GPs, but that's all. At most, worst monsters could drop something valuable (e.g. needed by crafters, etc). But then there should be serious chance to die to keep bots away. For humans it can add some fun and team action(this is MMORPG after all?!) . For botmakers it would mean awful headache since bots suck when it comes to decision-making in conditions where battle outcome isn't immediately obvious.
3) Obviously, crafting items should give most income. Crafting good items should take some efforts or prices to get right materials. But action itself could be not too time consuming and not very repetitive so there is no need to bot.
4) More random, dammit. If you roam in caves and stab some crap for a while, you can awake some powerful horrors roaming around. Then you better to run for your life. If you can run fast enough, of course. Or try to fight back... if you came with friends and can make a stand against real horrors. Needless to say it can turn botting and grinding to epic expeditions and adventures, attracting many players around to have fun (and some profit) together. And that's what MMORPGs for. In TMW there isn't too much of multi-player action, actually. Which is odd for MMORPG...

[Nard]

Designing with hypergeometric trials

Code: Select all

function HypergeometricSuccessTrial(Nmax, Nsuccess, k)
%% This function simulates an Hypergeometric first success trial,i.e. trial in a set 
%% containing Nmax possibilities among which are Nsuccess success possibilities.
%% k is the order of the trial an has to be incremented by 1 at each trial, 
%% Nfails = Nmax-Nsuccess
%% Nsuccess < Nmax 
%% k < Nfails, as soon as it equals success is sure.
%% Initial probability of success: p0 = Nsuccess/Nmax
%% Probability of success after k-1 trial pk = Nsuccess/(Nmax-k+1)
%% It returns success=1 if the trial was a success, 0 if not.
%% You have to update the Nsuccess variable by ±1 depending of the result; typically
%% Nsuccess:=Nsuccess-success.
success:=0;
trial:= Rand(Nmax-k+1);
if trial ≤ Nsuccess then success := 1; 
return success;
endfunction;

(pseudo code)

1. You can first set the success to zero until some value when you want a minimum items to be given. (it would be a less frustrating soluiton than the one used in TMWbr), then begin the random trials.
2. Choose Nfails: it is the maximum number of (additional) items that you have to give.
3. Choose Nmax:compute: Nsuccess = Nmax-Nfails, p0= Nsuccess/Nmax is the initial probability of success.

The probability to have k fails is constant and equals to Nfails/Nmax, thus the probability to have a number of fails lower than a given value K increases proportionally with K. As a result, the number of players which have a given value of fails should be uniformly scattered between the bounds.
Ref.:Wikipedia: Hypergeometric distribution. http://en.wikipedia.org/wiki/Hypergeome ... stribution

[Crush]

Discussion about MathML extension for the wiki moved to viewtopic.php?f=11&t=18043

[Nard]

Discussion split from The Bot World --Crush

Finally I asked for the math extension in wiki, with, in my mind, the goal to make a crash course of probabilities (and maybe other maths and physics) for game developers... This was at least a year and a half ago.

I tried installling MediaWiki Extension:Math, but ended up rejecting it because it pulls in over 900 MB of TeX dependencies.

My reply has been moved there: viewtopic.php?p=141051#p141051

I had warned about this dependency. Do you have so few memory and hd room that you cannot afford TeX? btw on my computer, with the tug installer; Texlive, with comfortable install occupies a bit more than 100Mb. I doubt that you need the complete bunch of big fonts for mediawiki. Anyway Thank you so much for your quick answer.

It has it's place here as it is not a discussion about the wiki but an information about the care that some members of the development team care with (some) contributions.