Free Allegiance Forums

As I understand it you want to measure experience. You choose to base this of unique days logged in. Now I would suggest that this is not the way to measure experience. Rather a system that takes games played (which Helo does) into account would be better.

Well, the idea of games played works to an extent. The thing of it is is there a noticeable or worthwhile difference in say, playing 54 games as opposed to 55 games, the benefit of AllegAge is it takes blocks of time (relatively). Now I had suggested in the past that AllegAge should in fact be based on number of games played (outcome immaterial) but taken as blocks.

Instead of Unique LogIns it would be replaced by something like 5 games played (just an example), so you'd be looking at 1125 games for vet1. Or if it was based on 2 games, 450. I read over TrueSkill in awhile though I don't recall it taking something akin to experience in it. I 'generally' like TrueSkill as it is one of the nicer systems out there... not perfect but better then Helo. It seem that it would be best if we can lay down this whole rating system for a long time, as it always is divisive. Of course to do that we need something more solid.

My rank is what, 15 today, the thing occilates from 12-23... seriously that's insane. Last night a vet2 didn't understand why launching from base in a bbr without missiles armed was a good idea, yet another inter8 didn't know the use of a nan, and a vet ran into mines, in his ng bbr, in the middle of open space, spidey is what an inter4?... the list goes on and on.

You have 770 games played. TB has 100 games played. The situation will probably be similar with regard to unique log-in days. And yet there can not be a shadow of a doubt as to which one of you is the better player.

(An example of why games played aren't indicative)

Why thank you Deng. /mrgreen.gif" style="vertical-align:middle" emoid=":D" border="0" alt="mrgreen.gif" />

Personally I think TB is pretty darn good too. /smile.gif" style="vertical-align:middle" emoid=":)" border="0" alt="smile.gif" />

Anyway, all that aside, TB easily has more logins I would wager. Also if ratings were used since TB started it's really hard to say what would be the situation.


Big difference 7 over 3 years of data collection.

But we don't have 7 years of data. That's part of why age won't work, because anyone of our time and older will be rated equally by that part of the system. Which is claerly not right even if we think that measuring experience can be done.

We have more data from AllegAge that often shows more accuracy then with game data showing wins and losses.

In a nutshell AllegAge is less radical though slighly less accurate overall then a rating system, which while slightly more accurate is notiecably more radical.

Even more still, AllegAge is more stable.

jgbaxter wrote:QUOTE (jgbaxter @ Dec 13 2007, 12:39 AM) We have more data from AllegAge that often shows more accuracy then with game data showing wins and losses.

In a nutshell AllegAge is less radical though slighly less accurate overall then a rating system, which while slightly more accurate is notiecably more radical.

Even more still, AllegAge is more stable.

Erm... what do you mean by radical? Wouldn't by radical you mean extreme??? How can something be more accurate but more radical??? (Rating System Calculates while doing consecutive 360's in an Aston Martin Vanquish) (Why a vanquish??? cuz its so pretty!!)

Now if you mean by radical as in extreme in the sense of an anomalous piece of data of ridiculous proportion then its a bit silly when saying its more accurate.

All alleg-age will do is seperate people into two tiers, those who are newbies(or returning vets) and those who play a lot of Alleg. Neither is a seperation of skill and you might as well have ranks be 0 and 1 where you only get to 1 after playing for 2 months (or some other ridiculous time like last time).

AllegAge will be stable in the way that it will more consistantly be inaccurate rather then the current system which actually has sort of accurate (give it a 5 rank leeway) system.

Unless you guys have all of our data from logins back during BETA for those of us who have been here that long, something ranking us based by age alone is useless.

HELO works about as good as anything is going to work. Besides, I know I'm not alone when I say that I never even look at somebody's rank - I look at the name.

This thread is full of people who think age = skill.

God, get the nuke! And make sure to write "Darwins Avenger" on it too.

HSharp wrote:QUOTE (HSharp @ Dec 12 2007, 06:47 PM) How can something be more accurate but more radical???

For example;

18/20 AllegAge is accurate withing ±2 ranks.
19/20 Helo is accurate within ±6 ranks.

Get it? /smile.gif" style="vertical-align:middle" emoid=":)" border="0" alt="smile.gif" />

Bacon wrote:QUOTE (Bacon @ Dec 12 2007, 06:53 PM) Unless you guys have all of our data from logins back during BETA for those of us who have been here that long, something ranking us based by age alone is useless.

I disagree. You make a point though, and that's the same for the other ranking systems, the more data they have, the longer the data is collected, the better the system. Also it likely is better to have the AllegAge/AllegSkill averaged then taking either one on it's own. AllegAge alone is probably not as good as this AllegSkill alone, neither are as good seperate then apart.

Bacon wrote:QUOTE (Bacon @ Dec 12 2007, 06:53 PM) Besides, I know I'm not alone when I say that I never even look at somebody's rank - I look at the name.

Strongly agree. However, if we had a more reliable system, you'd more likely look at someones rank? Even now, you're very likely to look at someones rank whom you have no familiarity with? /smile.gif" style="vertical-align:middle" emoid=":)" border="0" alt="smile.gif" />