This is the energy/metal relation that was coded in r7342 for testing. The purpose is to allow an equal sharing of Overdrive(OD) metal and to give incentive to people who contribute the most Energy. Here's the illustration:  ^This is what it do when there's growth in E & OD metal The following is what it do when there's dip in E & dip in OD metal:  The graph was created using Excel and the value can be changed to show different output, here's link to the file: https://skydrive.live.com/?cid=E7F297F4F70F2EF0&id=E7F297F4F70F2EF0%21121&sc=documents -> Open with MsOffice or OpenOffice, and edit the yellow portion to test different condition http://code.google.com/p/zero-k/source/detail?r=7342
+0 / -0
|
Wow, wow, wow. Just wow! This must be the best reason for why energy based OD distribution sucks, good job. They look so confusing, I don't want to understand these graphs.
+0 / -0
|
I was hoping the change wasn't reverted... that's why I go to great lenght to make this presentation :(
+0 / -0
|
This doesn't help explain it to players. What does this mean, actually, in terms of return on investment?
+0 / -0
|
Dang, that must have taken some work. I have to admit it is a little hard to follow, perhaps if you explained your proposition with text along with the graphs, as my only issue is that I don't know the system well enough to know what has changed. So far it looks like you want overdrive to be split based on overall contribution to the grid, so changing that composition will give you more metal, but less so as the total pool of E expands. See if you can explain more precisely what your proposition will do and how it will work with text to make the graphs easier to interpret.
+0 / -0
|
quote:
[by norm0616] "So far it looks like you want overdrive to be split based on overall contribution to the grid, so changing that composition will give you more metal, but less so as the total pool of E expands." |
Yes! :D you are absolutely correct. It split OD based on overall contribution to the grid, so that changing the contribution will give more/less metal, and [the increase is] less so as the total pool of E expand. - But I also changed so that every 1 minute the split is reset, so that all player has same OD split regardless of their E, and then the cycle repeat (to give more OD to people who increased the OD further and so on). In present OD scheme the split is parmanently favour person who own the E. So it is not fun because in late-game you want everyone to have as much metal as possible to fight, and not just more OD for the one who hold the E. -so that's different. quote:
[by SaktothIVL]"What does this mean, actually, in terms of return on investment?" |
The return of investment is not guaranteed. People who contributes to the increase only get ~1 minute OD boost, and its value depend on whether it is possible to substract from ally. -If you are in mex rich map the OD is big, but in metal poor map the OD is small, and the return is small or big depending on OD... * the return time can be altered to whatever value in the code.
+0 / -0
|
Example situation: * Player1: contributes 40% of average in team1* Player2: contributes 160% of avg in team1* The whole rest: contributes 100% of avg in team1Past fails of 2 contrasting OD-code-trials: * p1 is at disadvantage when: * * p2 sits in the back of p1 and p1 can't make much E because of 2v1 * * p1 powers some more shields (which are linked to the team shield grid) and thereby contributes in another way. * But p2 complains when: * * The team supports a p1 being behind a p2-top-ten player who additionally fight 2v1... Who deserves the overdrive-metal? * Only players without huge excess deserve metal - or not? * So for the 2v1 player, each of his enemies spend about the same as he spend (assuming both teams have an equal eco)In this situation, the player is a gain for the team (and this should be rewarded) if: * Method A: od-energy / team.od-energy.avg-per-player > 1.0 * Method B: mexes / team.mexes.avg-per-player > 1.0 * Method C: [metal] killed / [metal] lost > 1.0 * Method D: reclaim and rez / wreckages lost > 1.0 ... And each factor should affect his personal overdrive-metal gain! * Some factors can be smoothed or roughened.
+0 / -0
|
To: SinKitty Ok I understand... You still favour the old OD split because you believe the ECO player can still contribute much to the team. You said they might be a PRO (top 10 player), or they might be front player (2v1), or they might be feeding the shield grid (indirect contribution). -this all might be true, but I am assuming everyone is average. You might want to OD split based on ELO? Also, you might want OD split based on enemy-kill, mex-building, mex-efficiency, energy-difference?? This is all possible if there's enough attention span to study, document, modify & debug the OD system... -its really hard.
+0 / -0
|
I have no opinion on this... could we get a graph of a simulated run, with x-axis time and y1-axis metal (instant or integrated) and y2-axis = energy contributed (instant or integrated) ? I'm sure it would be easier to understand.
+0 / -0
|
Certainly!... Luckily I have a replay of r7349. I can make graph.
+0 / -0
|
Just so you know, nobody has any idea what is going on and we fear to let that stand with such a vital gadget.
+0 / -0
|
Anything mentioned in this thread seems just terribly and unnecessarly complicated. The easy, and natural, solution would be to just bring back communism as 90% of all people ive talked about it with want anyway. This will strip overdrive-enenergy contriubuting players of their returns on their investment, but actually that doesnt matter at all because its a team game, and if you team wins, you win. youre making powers? youre contributing to victory, thats an incencitive. youre waiting for your team mates to build powers, because you dont wanna make the investment? you need an additional incencitive to do something that will already increase your chance of winning? then you just suck. its really the same thing as if you would give a front player returns for building units, as it may also increase the teams overall metal production (or make it win), but the front player will never get the metal for it back! so i wanna have an incencitive to build units, i cant stand my teammates not paying for my units and ill not build units again until i get paid for every single unit. even makes sense if you look at it the same way od is looked on. as an intermediate solution, why not increase the shared percentage on od to bring us closer to communsim, while maintaining that those people building the power will eventually get their investment back?
+0 / -0
|
This is a reply to [V]Sheep: The following graph came from a game of Chicken VS versus666 & mastergandor ( http://zero-k.info/Battles/Detail/91922 , ZK r7349). The explaination & caption is drawn in the graph itself. -I'm happy to elaborate if asked to do so. Graph of- Integrated value (cumulative value) of OD-income & Energy-feed-into-OD (as reported by mex_overdrive Gadget):  Graph of- a close up of Integrated OD-income:  Graph of- Absolute value of OD-Income & Energy-feed-into-OD, (raw data):  Graph of- Mean (averages) of metal distributed by OD in r7342 (distribution pattern):  Others: Raw graph of total E-income and OD-income (for reference):  Raw graph of OD-distribution pattern in r7342:  Raw data+ processed variables from chicken battle with versus666&mastergandor: https://skydrive.live.com/?cid=E7F297F4F70F2EF0&id=E7F297F4F70F2EF0%21121*Data processed using SPSS & captured using a modified widget (dbg_fpslog.lua+gui_chili_resoure_bar.lua)
+0 / -0
|
Awards prints economy data to a file with the right widget enabled. Automatically, it's in the current stable.
+0 / -0
|
This all seems unnecessary. What am I missing that prevents overdrive energy from being collected from the players, and splitting the overdrive metal proportionally every resource tick? Why have reset times and rolling contribution periods? This is all, of course, assuming the naivety of the system with regards to where the energy is actually coming from - the physical energy structures and the mexes. When I got my head around the overdrive system via reading the manual and viewing grid behavior ingame, it seemed obvious to me that the system would have full knowledge of which structures were producing energy and supplying it to which mexes. Based on the way people talk, it seems to me this is not done and a player's excess energy is available to the entire system globally, leading to a strange disconnect between the interactions of energy structures as energy production, and energy structures as grid-creating structures. However, the tooltips in game seem to support the idea of multiple isolated grids having maximum energy pools allowed to them, which I think led me to the original notion. Implemented naively, this still might allow surplus energy to sneak its way in via team energy sharing. I can think of ways the OD system should work given the concept and documentation provided, I have ways I think the OD system works given the in-game presentation and player behavior, and I have ways I think the OD system works given the way people talk about it on the forums. None of these three seem to coexist. In summary: the fuck? Question: is there an engine/performance limitation on implementing "perfect" overdrive, considering all aspects of structure placement, physical grid connection, and player energy usage/contribution? If not, why all this roundabout; if so, what are the considerations and why not just dial down the tick rate/sequencing of overdrive calculation until happy?
+0 / -0
|
Honestly, this is getting increasingly baroque. I'm partially bothered by the fact that "spam sing" is helpful *at all*. Players do it because it really does cause a much larger output of metal... teamgame or otherwise. Are we sure that's a good thing? Either way, my opinion, either: a) Revert the 50% personal-OD thing, or b) Re-design the OD/Energy system itself into a more localized form so that we can coherently tie metal-output to each structure.
+0 / -0
|
Klon's story is the only one I understand. And I agree. Maybe you can make it like this to prevent other players totally leeching your energy: - first use own energy for building/repairing/rezzing etc. - 50 % excess energy is put into OD and shared to whole team equally - remaining 50 % is shared to teammates energy income, if they too excess energy it get into OD So you always get more metalincome if you build energy.
+0 / -0
|
Acceptable solution is to return investment to owner - this will of course take variable time but it is possible to repay metal cost first without damaging others. Proposal is described in that other thread. Btw the graphs look impressive and cool, it suddenly feel like a business strategy presentation :)
+0 / -0
|
You have your Y axis labeled as 'Value'. What is this in units of!? Is that total metal/energy amount? The first one ranges from 0 to 800,000: 800,000 what? Not energy production, the numbers are too high. Not total energy or the curve would be exponential rather than flatlining at the start. Looking at the first graph, you have one player making 0 E until 45m in, and then making a ton, while if you look at the second graph, the curves are almost IDENTICAL. How can the investing player possibly make back their income when the curves are so close? Even the second, per minute graph, the lines track eachother almost exactly (only they fluctuate constantly for no apparent reason). Why should they do this when in the first part of the first graph, only one player has E, and for the giant hump, it's the second player who has made all the E: Yet the curves track almost exactly. Seriously, this system is so close to 100% communism that you might as well go with that and avoid the complication.
+0 / -0
|
quote:
[by Shadark] "In summary: the fuck? Question: is there an engine/performance limitation on implementing "perfect" overdrive, considering all aspects of structure placement, physical grid connection, and player energy usage/contribution? If not, why all this roundabout; if so, what are the considerations and why not just dial down the tick rate/sequencing of overdrive calculation until happy?" |
Its just that editing an existing system can lead to more bugs because it is made by different author.I think making extension is easier & safer. If you code a system to track all wind,solar,fusion,and geothermals' age & owner then it would require more work. Its really hard to keep r7349 commit for more testing unless you made an amazing Powerpoint presentation to keep it. ZK dev doesn't want a ridiculous commit. quote:
[by SaktothIVL]"You have your Y axis labeled as 'Value'. What is this in units of!? Is that total metal/energy amount? The first one ranges from 0 to 800,000: 800,000 what? Not energy production, the numbers are too high. Not total energy or the curve would be exponential rather than flatlining at the start." |
-I don't make 2 saperate graph because it is a waste of space. -Its not exponential because the players are not spamming E, and that big hump is Singularities which dwarf all Energy source (make them look tiny). quote:
[by SaktothIVL]"Looking at the first graph, you have one player making 0 E until 45m in, and then making a ton, while if you look at the second graph, the curves are almost IDENTICAL. How can the investing player possibly make back their income when the curves are so close? |
-I keep reseting the OD-distribution every minute because I want to keep communism. It made them equal. *I want to test differently if given chance* quote:
[SaktothIVL]"Even the second, per minute graph, the lines track eachother almost exactly (only they fluctuate constantly for no apparent reason). Why should they do this when in the first part of the first graph, only one player has E, and for the giant hump, it's the second player who has made all the E: Yet the curves track almost exactly." |
- OD is added/substracted from the player given by how much he/she added/substract E from the grid. - I don't know why mex_overdrive is eating an irregular amount of E from player, which was causing such random up/down responses.
+0 / -0
|
