Post Reply 
 
Thread Rating:
  • 0 Votes - 0 Average
  • 1
  • 2
  • 3
  • 4
  • 5
Brilliant Researchers
08-17-2013, 12:15 AM (This post was last modified: 08-17-2013 12:15 AM by Diebo.)
Post: #21
RE: Brilliant Researchers
(08-16-2013 11:42 PM)rocco Wrote:  The only scenario i can imagine anyone taking clumsy scientists is in pre-made team games.

I was thinking that a genius might take it. They crank out so much research, it could allow them to also get freethinking and amazing industry. There are so few negative traits. And maybe 25% is too high - but I can see taking that risk (and hoping hoping hoping that it fails only on cheap techs!).

That said, I didn't think about team play - of course people would take that for free points. This is reason enough to not do it.

(08-16-2013 11:42 PM)rocco Wrote:  So overall I still favor PlanetaryScientists idea as a good medium between the two; it retains the feeling of BR as a gambler's trait, limits the chances of striking the jackpot better than the current 3x cost implementation, and it helps late game when techs roll over (higher chance of getting the cheaper techs).

I like it too. Very nice compromise. And if the formula is right (i.e., strongly favors the cheaper techs), you could remove the 3x cap.

I like the idea of being able to leapfrog a tech, too, (i.e., get shield 2, bonus shield 3) but I imagine that would add a bit of coding headache.
Find all posts by this user
Quote this message in a reply
08-18-2013, 01:11 AM
Post: #22
RE: Brilliant Researchers
(I'm putting together a full design statement as a proposal.)
Find all posts by this user
Quote this message in a reply
08-18-2013, 02:13 AM
Post: #23
RE: Brilliant Researchers
Design Statement for Brilliant Researcher bonus...

Assume there is a tech ladder, so relationships in the tech ladder can be determined.
Assume the current owned techs are known
Assume the lists and ranks (red,yellow,green) of available techs are known. (Question, is this list before or after the researched tech?)

All possible techs for the bonus must be either in the list of techs and/or in the next tech in the tree of the currently researched tech.

Definitions
R = cost of just-researched tech.
B = cost of a bonus tech
T = the tech type of B
c = the tech type of R
W = Weight of a given T
N = cost of next-cheapest possible tech after current ®
F = cheap penalty factor
S = sum of weights of possible techs
U = random numer generator universe size, say 1000 for now
P = normalized weight of penalty factor
B = bin size in Random array U for a given bonus choice (i.e., normalized probability)

All the possible techs (B's) are weighted (W) as follows...

W = (R/B)**2

In addition, the weights are adjusted as follows...

if T is in the next tech tree of C, W = W * 2, otherwise... (maybe times 3, we'll tune?)
if T is red, then W = W * 1.25
if T is green, then W = W * 0.75
if T is in the tech tree of ANY of the current techs, then W = W * 1.25

In addition, if the current tech is the cheapest in the available tech tree, a cheapness penalty is computed by...

F = (N - R)/N

The various weights are summed into S.

The cheap penalty weight is computed (if appropriate)

P = S * F

The various techs (and penalty) are then given bin ranges sizes B in U via...

B = W * ( U / (S + P) )

...and the leftover bins are the penalty ones (no bonus)

(The above could be transformed to avoid floating point calcs, if needed.)
Find all posts by this user
Quote this message in a reply
08-18-2013, 04:29 AM (This post was last modified: 08-18-2013 04:46 AM by Sentar.)
Post: #24
RE: Brilliant Researchers
This didn't post so it's behind the discussion now quite a bit -- fwiw:

(08-16-2013 06:32 AM)Diebo Wrote:  
(08-16-2013 04:57 AM)Sentar Wrote:  1) The trait doesn't stand up to "believability" -- a researcher doesn't discover a malaria vaccine and stumble onto high-speed pneumatic tube trains on accident in the process -- it's just weird to get a "bonus" tech, nevermind the problem of equal value, I just think its weird to end up with Colony Security when I was aimed at e.g. L3

Oh, I dunno. It is a six-point trait, so you race is REALLY good at riffing off ideas and turning them into something else. Maybe you are looking into L3, and one of the brains realizes that they could improve colony security with a modification of laser technology... Seems plausible. And as a racial trait, there is probably full support for an idea being turned into a product.

Sure, I didn't mean to say it is consistently unbelievable -- agreed that on occasion a given line of research could result in a breakthrough even highly tangential -- but per my example and as happens in practice, we are getting totally unrelated techs -- as if. My real point goes beyond this finite concern.


(08-16-2013 04:57 AM)Sentar Wrote:  3) There's no equivalent build tech (and I wouldn't want one, but to make my point), you don't set out to build a frigate and get ACP built as a bonus -- but that's what BR suggests should happen in a sense -- i.e. it's nonsensical and doesn't scale across the logical categories.

Quote:Again - it is a racial trait/ability. And just imagine, as I said above, that the race fosters this brilliant development into parallel development. If you want realism... Why are we only studying one thing, and it takes years to research? Shouldn't we be investing a little bit into everything? It is made simple.

Now that RPs spill over I think we are essentially researching multiple things at once (?) --in one of three main categories anyway. However, perfect realism is not what I'm trying to suggest here or we'd have to take that to its logical extreme and end up debating this in space, that could take a few years.

Maybe I wasn't clear above -- I understand that it's a trait/ability, what I am saying is merely to add a bit to the overall argument against it, is that it's an odd duck. You don't have "bountiful harvest" double the food on random turns and you don't have "production explosion" create two builds in a single turn -- and again, I am not trying to suggest we should have those abilities -- rather that BR is curiously unique as a grossly obtuse trait that more often than not makes no real sense. The two examples I used here for food & prod are much more believable, you can draw a line between those happenings and a plausible reality even thousands of years ago. We're not talking about a Magic Research Machine that occasionally spits out a two-fer ---or are we? That's the issue I have w BR... Remember, my larger point remains articulated best below, the above is just to tee it up, but picks on BR because its the best example of what I'm arguing against, and is currently at issue.


(08-16-2013 04:57 AM)Sentar Wrote:  4) Just a quick statement I could go on about: There's good random and there's bad random -- good random is a trait like Combat Pilots or Smugglers... traits w either/both subtlety or unpredictable impact/effect are what give SO its "texture" and "depth" -- the most interesting matches are through these dynamics that are potentially applied in a "novel" way, or even countered in a manner you didn't expect. BR doesn't pass the "good random" test, and it hits a high mark on "bad random" (see #1) test.
Quote:There is nothing random whatsoever about combat pilots. You move and turn faster. Very predictable.
Nonsense, the *outcome* as a result of having that trait, whether in a single battle or over the course of an entire match is wholly unpredictable when taken in context with all else never being completely equal.

Quote:If you mean that spending 2 points on pilots vs, say, a rich home world adds strategic depth (you are a better fighter, but you are worse at cranking out ships, so you may have a smaller fleet), then I would say that brilliant certainly does that. It is a second path to research.

Sorry I wasn't more clear, no that's not what I'm saying, rich home world makes good sense just as combat pilots -- it's easy to imagine a race well built hand-eye coord or what have you just as it is a population that has its genesis on a rich home-world relatively speaking.

Quote:In research, you can go genius, or brilliant (or neither, or both).

Genius: Pro, you get techs faster. Con, you miss techs.
Brilliant: Pro, you get techs slower. Con, you don't miss very many techs.
Add in broad field, narrow minded, and you have some options.
And you can take both, but at 12 points, you need to give up on industry, or food, or population growth, or... well, a lot of things.

I find brilliant to be very strategic. Not the best choice for a quick blitz, but for a longer game, it can be a nice path to follow. And more strategic (if less plausible) with the buy one, get the second of equal or lesser value free. It is kinda cool - do you go with Research Center 2, knowing you'll get colony infrastructure, or go with cloning, which gives you the chance of getting RC2.

A key to this game is research, and Roc has provided numerous paths to that end. The main thing I'd like to see is the second pass costing more so it isn't just one tech/turn once you research the first pass.

I get that. I believe I do understand what you're saying -- we may just disagree fundamentally, and I'm just taking up the position here. I see a Research Build/Focus as a strategic choice, BR merely a tactic to that end. I'm sure that's what you mean (very strategic ~ powerful tactic). A tactic that I and others find powerful, attractive and almost priceless (esp. as a two-fer)? Heck yeah --but that doesn't mean I like it as part of the game (as is), and seeing it go away or change significantly doesn't necessarily have to mean one cannot adopt a Research Strategy. I don't want that either.

So my feeling is that there's room for improvement -- the only thing I am arguing be taken away, and replaced with something better, is the entirely "simplistic" factor BR represents. It's boring, ultimately, because in the matches I've been playing (admittedly < a dozen) its not just a "nice path to follow" -- rather I'm seeing everyone adopt BR -- rendering it effectively a global R&D accelerant, and a trait you "have to" take just to level the playing field. This applies less now that it is "equal or lesser value" but I'm skeptical that tweak is going to play out well.

To be clear I'm not saying just remove it and be done with the matter -- I think these things need to be carefully considered, not knee-jerk anymore than a production feature ought to be sacrosanct.


(08-16-2013 04:57 AM)Sentar Wrote:  I want to post some ideas about other traits and primarily just modifications to existing ones that would go to deepen the game as I allude to in #4 -- but there a better tread(s) for that --however, merely to reinforce my points above a single example of what I mean: Leaders to be permanently on the auction block -- in recognizing some degree of loyalty perhaps it takes a significant bribe (and a raise) to get them to defect and join you, but all Leaders would be "in play" at any given time technically. If a Leaders gets passed up by everyone at first, they shouldn't vaporize, in "reality" of course they are at home sitting by the phone waiting for the phone to ring (albeit watching GoT and mowing down bags of Valurian Meat Chips) -- one could hire them at any point they are needed. This would add some cool decision factor to,the game where people are trying to build just the right team but not too soon.... I would keep the leader turns to fully join @ 5 and announce globally when they get hired --possibly including to whom (gossip travels faster than light). In terms of max #of Leaders maybe that's strictly limited only by budget -- in that sense we let galactic capitalism work its magic.

Quote:The problem is that people drop everything to get leaders, and it just isn't fun if capitalism runs its course. RE: Availability. I just think those leaders are off in other galaxies, and swing by to check things out. They aren't just sitting around the phone. They are probably in the cloning centers in Alpha Centuri, building up clones for when they get shot up

Adding "assassinate" as a spy option would be fun.


Yeah, except that if they can get to other galaxies and back why can't we -- but I get what you mean, and I can be down with that also. It's not the driver behind my idea. Capitalism running its course is the most powerful instrument of change mankind has ever created -- it's anything but "not fun" (as in not interesting) -- does it have a certain stigma? Sure, but leaving that out of it, purely letting the economics drive choices, seeing it work in and around "true power" (as Kevin Spacey would say) makes for all sorts of fun.

Assassinate -- now that's an interesting idea. :-)

I hope I did the quote thing right, I almost never post to forums.[/php]


Planetary I applaud your well crafted effort -- and I will look closely and think about it before weighing in one way or the other and climbing back into my hole. In the meantime I just want to caution the complexity -- if you want this game to continue to grow (broader audience) you need to keep things readily understandable to the casual gamer market. Maybe there's a way a copyrighter could do that with what you've proposed by that's a consideration you may want to make Rocco.

My suggestion is to just consider what the world would look like if BR wasn't there at all, a bottom-up exercise just to see what you get...
Find all posts by this user
Quote this message in a reply
08-20-2013, 12:34 AM
Post: #25
RE: Brilliant Researchers
First, before I respond to Sentar's comments, I wanted to get this thought out there:

What if brilliant (using Planetary Scientist's ideas) took from the entire list (within astro, military, or civil) of available traits? So the brilliant trait wouldn't necessarily be one of the red, yellow, or green traits, and would be more heavily weighted to pull from one of the lower-costing techs? This would minimize the strategic side of it (I have a 50% chance of getting this or that tech, because only three are available), make it more believable, and weaken it some (by mostly getting lower techs)? Something to ponder, anyway.

(08-18-2013 04:29 AM)Sentar Wrote:  Maybe I wasn't clear above -- I understand that it's a trait/ability, what I am saying is merely to add a bit to the overall argument against it, is that it's an odd duck. You don't have "bountiful harvest" double the food on random turns and you don't have "production explosion" create two builds in a single turn -- and again, I am not trying to suggest we should have those abilities -- rather that BR is curiously unique as a grossly obtuse trait that more often than not makes no real sense. The two examples I used here for food & prod are much more believable, you can draw a line between those happenings and a plausible reality even thousands of years ago.

I'd disagree here. With production, you are limited by your racial ability to produce and the aid from machines. Same with farming. With researching prototypes, I can easily imagine that 50% of the work is coming up with a workable innovation/idea, and 50% is figuring out how to implement it. If you are a slow scientist but brilliant, you might be really good at churning out very workable ideas, but technically you are slow about figuring out how to test and implement them. Alternatively, a genius race that isn't brilliant might be really good about figuring out how to implement a specific idea, once they get that idea, but not so good about coming up with the ideas in the first place. In the end, you don't actually produce anything but that workable prototype. You research Titan, but you don't actually get any until you have you people build them. So a brilliant race has more Einsteins than a genius race. It works for me.

The 100 percent chance... well I can see that being tweaked to make it more realistic. But I can also deal with how it is currently implemented. If it is OP, change its point value (or what the bonus is; see my top paragraph).

(08-18-2013 04:29 AM)Sentar Wrote:  I get that. I believe I do understand what you're saying -- we may just disagree fundamentally, and I'm just taking up the position here. I see a Research Build/Focus as a strategic choice, BR merely a tactic to that end. I'm sure that's what you mean (very strategic ~ powerful tactic).

We are on the same page here.

(08-18-2013 04:29 AM)Sentar Wrote:  It's boring, ultimately, because in the matches I've been playing (admittedly < a dozen) its not just a "nice path to follow" -- rather I'm seeing everyone adopt BR -- rendering it effectively a global R&D accelerant, and a trait you "have to" take just to level the playing field. This applies less now that it is "equal or lesser value" but I'm skeptical that tweak is going to play out well.

What you are saying is that it is OP right now. I see 4 ways to change it:

1. Nerf it (so you get mostly cheaper traits) while remaining 100 percent success. This is currently being worked on.
2. Increase the trait point cost. If everyone takes it at 6 points, would they at 8 points or 9 points? Use player data to tweak and determine the break-even point.
3. Nerf it (by reducing the % chance from 100 to something lower).
4. Get rid of it, just allow research in one way.

I like it as a trait - I do see it as one path to research that is different, and interesting, from the genius route. And I think items 1-3 are all probably being considered right now.

(08-18-2013 04:29 AM)Sentar Wrote:  My suggestion is to just consider what the world would look like if BR wasn't there at all, a bottom-up exercise just to see what you get...

BR didn't use to be there, or at least, what it was before wasn't often taken. 1.2.1 introduced both the 100 percent success AND the need to use scientists to produce science. Before 1.2.0 it was a not-frequently used trait.

My two cents: tweak and balance it, but keep it.
Find all posts by this user
Quote this message in a reply
08-20-2013, 11:33 AM
Post: #26
RE: Brilliant Researchers
Thanks PlanetaryScientist.

I have implemented this change for the next beta build (I did not implement it exactly as PS laid out).

I can tweak the values if needed, but for the next build the results look like this (in this example, I researched barracks for 200)

8.400000% : Laser Beam II (1000)
42.399998% : Nuclear Missile I (300)
27.799999% : Pacifist Overhaul I (400)
21.400000% : Heavy Armor Plating I (500)
Find all posts by this user
Quote this message in a reply
08-21-2013, 07:01 AM
Post: #27
RE: Brilliant Researchers
(08-20-2013 11:33 AM)rocco Wrote:  Thanks PlanetaryScientist.

I have implemented this change for the next beta build (I did not implement it exactly as PS laid out).

I can tweak the values if needed, but for the next build the results look like this (in this example, I researched barracks for 200)

8.400000% : Laser Beam II (1000)
42.399998% : Nuclear Missile I (300)
27.799999% : Pacifist Overhaul I (400)
21.400000% : Heavy Armor Plating I (500)

..and similar to the above example, what would the odds be for researching Pacifist Overhaul (including the possible enhanced chances for Pacifist Overhaul II)?
Find all posts by this user
Quote this message in a reply
08-21-2013, 07:25 AM
Post: #28
RE: Brilliant Researchers
(08-20-2013 11:33 AM)rocco Wrote:  8.400000% : Laser Beam II (1000)
42.399998% : Nuclear Missile I (300)
27.799999% : Pacifist Overhaul I (400)
21.400000% : Heavy Armor Plating I (500)

This looks great for one part of the OP nature of Brilliant, and makes it more interesting, too. You just don't know what you'll get for sure, but you have a pretty good idea.

But for 200 research, you are getting on average 429.4 bonus research. That is, you get 629.4 worth of research for the cost of 200. If you have a normal race with RC2, that would take 40 scientists worth of effort (40 x 3+RC2 = 200) to accomplish (say 10 over 4 turns).

40 genius scientists (at 6 points for the trait) with RC2 would be able to research 320 points (40 x 6+RC2), or only 75% of the brilliant.

Maybe barracks isn't the best comparison. What if the tech cost 300, but the above table stayed the same? That would take 60 scientists for Brilliant to get 729 worth of tech, compared to 480 for 60 geniuses. You're still better with Brilliant.

You could boost it up to an 8-point trait. Or use probability to equal it out. So if the trait is 25% better in terms of overall research points, maybe have it succeed only 75% of the time? You could pretty easily calculate parity this way.

Just some ramblings...
Find all posts by this user
Quote this message in a reply
08-21-2013, 07:31 AM
Post: #29
RE: Brilliant Researchers
(08-21-2013 07:25 AM)Diebo Wrote:  
(08-20-2013 11:33 AM)rocco Wrote:  8.400000% : Laser Beam II (1000)
42.399998% : Nuclear Missile I (300)
27.799999% : Pacifist Overhaul I (400)
21.400000% : Heavy Armor Plating I (500)

This looks great for one part of the OP nature of Brilliant, and makes it more interesting, too. You just don't know what you'll get for sure, but you have a pretty good idea.

But for 200 research, you are getting on average 429.4 bonus research. That is, you get 629.4 worth of research for the cost of 200. If you have a normal race with RC2, that would take 40 scientists worth of effort (40 x 3+RC2 = 200) to accomplish (say 10 over 4 turns).

40 genius scientists (at 6 points for the trait) with RC2 would be able to research 320 points (40 x 6+RC2), or only 75% of the brilliant.

Maybe barracks isn't the best comparison. What if the tech cost 300, but the above table stayed the same? That would take 60 scientists for Brilliant to get 729 worth of tech, compared to 480 for 60 geniuses. You're still better with Brilliant.

You could boost it up to an 8-point trait. Or use probability to equal it out. So if the trait is 25% better in terms of overall research points, maybe have it succeed only 75% of the time? You could pretty easily calculate parity this way.

Just some ramblings...

If it's 25% better, it'd actually even out at 80% success rate, not 75 (though with the % chance of success, it could perform better or worse obviously). Say 100 RP for Genius and 125 for brilliant (so 25% better), 80% of 125 would be 100. 75% would only be 93.75. Though I'm not opposed to a true research build being better than brilliant...

GC ID - VanderLegion, GMT-9. Sandbox GC ID (Beta) - VanderLegion
Find all posts by this user
Quote this message in a reply
08-21-2013, 07:34 AM
Post: #30
RE: Brilliant Researchers
(08-21-2013 07:01 AM)PlanetaryScientist Wrote:  ..and similar to the above example, what would the odds be for researching Pacifist Overhaul

2013-08-20 17:28:22.635 BETA[52100:c07] [0.400000] 24 4.800000% [0] : Laser Beam II (1000)
2013-08-20 17:28:22.635 BETA[52100:c07] [2.000000] 272 54.400002% [1] : Marine Barracks II (200)
2013-08-20 17:28:22.635 BETA[52100:c07] [1.333333] 125 25.000000% [2] : Nuclear Missile I (300)
2013-08-20 17:28:22.635 BETA[52100:c07] [0.800000] 79 15.800001% [3] : Heavy Armor Plating I (500)


(08-21-2013 07:01 AM)PlanetaryScientist Wrote:  (including the possible enhanced chances for Pacifist Overhaul II)?

I am not convinced this is a good idea, so the trait currently acts as it always has and just pulls from the other available research items.
Find all posts by this user
Quote this message in a reply
Post Reply 


Forum Jump:


User(s) browsing this thread: 1 Guest(s)

Contact Us | Starbase Orion | Return to Top | Return to Content | Lite (Archive) Mode | RSS Syndication