Free Allegiance Forums

Posted: **Fri Aug 18, 2006 3:18 am**

Noir, I like the idea of playing "Classic" Allegiance, but then if the game gets big and SY would be helpful in ending a stalemate there is no SY. =(

Posted: **Fri Aug 18, 2006 4:25 am**

mcwarren4 wrote:QUOTE (mcwarren4 @ Aug 17 2006, 11:18 PM) Noir, I like the idea of playing "Classic" Allegiance, but then if the game gets big and SY would be helpful in ending a stalemate there is no SY. =(

A seperate "classic" switch would be nice.

Posted: **Fri Aug 18, 2006 7:07 am**

Noir wrote:QUOTE (Noir @ Aug 18 2006, 02:48 AM) Grimm, you are certainly entitled to your opinion, as anyone is, however I do NOT and will NEVER balance DN on just complaints alone. The data (statistics) MUST back it up. It is the ONLY way I can insure that the changes DO in fact need to occur and that I am not just making knee jerk changes. In my opinion, it is the only responsible way to balance the game. I'd hardly call a ~10% win% variance for the majority of the factions over the past 8 months huge and HARDLY "unplayable". Can it be tightened up? We'll see. I have serious doubts given all the other variables in the game that <5% variance is maintainable, but DN gets tighter with every release.

Please tell me where you find statistics on what techpath was used. I haven't heard of such data being available. I imagine there may be some data on what techbases got built, but that does not signify usage of that techpath (like getting a tac when enemy has galvs, just so your ints can launch from a forward base). When analyzing game logs, I don't think there is a distinction between stealthbombers or regular bombers destroying a base. Or galvs/ptbombers for that matter.

That Tac change you made, since you clearly claim it's not following your 'need statistics to back up a change', is nothing sort of a "¤#&&¤% you people, you whine at my core and you deserve to be ¤/%/%/d". Sorry.

Also, a small request... While everyone has access to ICE and can find out all sorts of interesting quirks about your core, most people don't opt to do so... There are plenty of 'undocumented' (or documented last in DN 3.04 or so.. forgotten to the majority anyway) features in the core, could we get some lists?

Like TF 20% mass reduction.
Or IC ints having a bigger thrust value than other interceptors.
Or Dregh stations not being affected by the factions sensor nerf (other factions stations do get affected by sensor nerfs/perks).
Rix and Nix figs having 10mps more base speed.
Rix and TF basic sfs being able to drop Teleport Probes (among other things).
GT Devastator being able to mount heavy cloak.

That sort of stuff.

We need that list of weird stuff so we can play the core as it was meant to be played, for surely the stuff is in the core for a good reason.

Posted: **Fri Aug 18, 2006 10:47 am**

Noir wrote:QUOTE (Noir @ Aug 18 2006, 01:48 AM) The data (statistics) MUST back it up. It is the ONLY way I can insure that the changes DO in fact need to occur and that I am not just making knee jerk changes.

Real win percentages:

Code: Select all

Faction A
-----
Games vs Faction B   |   Total number   |   Won   |   Lost   |   Drawn   |   Percentage
Games vs Faction C   |   Total number   |   Won   |   Lost   |   Drawn   |   Percentage
Games vs Faction D   |   Total number   |   Won   |   Lost   |   Drawn   |   Percentage

Faction B
-----
Games vs Faction A   |   Total number   |   Won   |   Lost   |   Drawn   |   Percentage
Games vs Faction C   |   Total number   |   Won   |   Lost   |   Drawn   |   Percentage
Games vs Faction D   |   Total number   |   Won   |   Lost   |   Drawn   |   Percentage

Faction C
-----
Games vs Faction A   |   Total number   |   Won   |   Lost   |   Drawn   |   Percentage
Games vs Faction B   |   Total number   |   Won   |   Lost   |   Drawn   |   Percentage
Games vs Faction D   |   Total number   |   Won   |   Lost   |   Drawn   |   Percentage

Faction D
-----
Games vs Faction A   |   Total number   |   Won   |   Lost   |   Drawn   |   Percentage
Games vs Faction B   |   Total number   |   Won   |   Lost   |   Drawn   |   Percentage
Games vs Faction C   |   Total number   |   Won   |   Lost   |   Drawn   |   Percentage

Bogus win percentages:

Code: Select all

Faction A  |  Total number  |  Won  |  Lost  |  Drawn  |  Percentage
Faction B  |  Total number  |  Won  |  Lost  |  Drawn  |  Percentage
Faction C  |  Total number  |  Won  |  Lost  |  Drawn  |  Percentage
Faction D  |  Total number  |  Won  |  Lost  |  Drawn  |  Percentage

'nuff said.

Posted: **Fri Aug 18, 2006 12:04 pm**

"Do not put your faith in what statistics say until you have carefully considered what they do not say"

Posted: **Fri Aug 18, 2006 12:50 pm**

Denga, you would also need columns for Win % with ranges of team probability of winning using ELO. The biggest factor in any game is the strength of the team. You can only get reasonable statistics where teams are even. Even Bios was winning before the update when they had the players.

Posted: **Fri Aug 18, 2006 1:46 pm**

mcwarren4 wrote:QUOTE (mcwarren4 @ Aug 18 2006, 08:50 AM) Denga, you would also need columns for Win % with ranges of team probability of winning using ELO. The biggest factor in any game is the strength of the team. You can only get reasonable statistics where teams are even. Even Bios was winning before the update when they had the players.

Agreed. Basing core changes on uncontrolled stats is not totally reliable. There are too many factors in games to have the stats be reliable. Perhaps a certain tech rock wasn't avilable to go Belters Tac, or folks got sick of it so it was not played as much. Those can also be factors that skew data.

Posted: **Fri Aug 18, 2006 1:48 pm**

mcwarren4 wrote:QUOTE (mcwarren4 @ Aug 18 2006, 02:50 PM) Denga, you would also need columns for Win % with ranges of team probability of winning using ELO. The biggest factor in any game is the strength of the team. You can only get reasonable statistics where teams are even. Even Bios was winning before the update when they had the players.

McW, if you read my original post where I was explaining the problem in greater detail, you'll notice that I mention ELO splits and settings as the next refinement. However, at the initial stage it's sufficiently representative by only considering games not counted as "excessively imbalanced" (even though that threshold could sure use moving down to the 65/35 mark).

But this whole discussion is moot, as apparently noone realizes that aggregating win totals by faction obfuscates problems that would be otherwise shown.

Posted: **Fri Aug 18, 2006 2:13 pm**

ELO itself will probably become more useful data after the balance patch has been in effect for a while, if my understanding is correct and only the balance-enforced games will contribute to ELO. However it's probably safe to assume a bunch of stackers on one team makes for a stacked team /smile.gif" style="vertical-align:middle" emoid=":)" border="0" alt="smile.gif" />

Equally important for core balance purposes is the primary tech used by one side as you can't ignore the tech choices in these things. Retrieving this data isn't easy. (I know you've already covered all this Deng)

Posted: **Fri Aug 18, 2006 2:16 pm**

Now throw in one last variable. What was the rank/rating/skill level of the commander?

Free Allegiance Forums

"Classic" Allegiance on DN