Yea, I like this idea too, and maybe total battles rather than just wins, like Deestan suggested.

But I'm not picky, Maelstrom's done an awesome job with this ranking in the first place. I've played a lot of folks in the top 50 and they're all damn good.

Hmm. I *really* like this scoring over the current one. Though I'm not sure if it's tweaked right, though I didn't have the patience to understand the system as good as I should.

My ranking from my first win was with my friend, EightBit, who lost that game (rematch!), and it was our first ever game of M.A.N.

Our ranking is:
Rank Player Rating RD #battles
293 Fraggler 1590 337 1
752 EightBit 1410 337 1

It seems the Glicko system favors a person with their first win as I dont think I played well enough to be ranked so high.
It's weird to see people at the bottom of the chart with so many games played. It makes me wonder if they lost most/all those games?
It would be nice, but not needed, to list the wins and losses, for statistical purposes.

I remember a long time ago I played a chess tournament regulated by US Chess Federation, and my official rating, having not won a single game, was I think 600-800. Was that score based on Glicko?

Good work on this ranking system! I would like it to be integrated into the Massive Assault official ranking, though I'd imagine people who are ranked General and such, wouldnt like to be dropped down in rank due to their position in the charts. It could be integrated with the Conscript/General type naming scheme by their position in the Glicko rank. If you're at the top you're the highest title, the bottom would be the lowest title.

The reason you're so high up is because there's a TON of people who lost one game and never came back, or else won one game against somebody really low on the charts....and never came back. A lot of the guys down on the bottom have probably long since quit playing.

As to chess, I think chess uses the Elo system. The Glicko system is based on the Elo system, but in such a way that Elo is actually a special-case scenario of Glicko...or something like that. Maelstrom has some links in the original post that might be helpful.

Yeah, from what I understand, Glicko is based on Elo, but includes the RD which represents the possible variation in the person's official score (one standard deviation, based on frequency and number of games played).

In both Elo and Glicko, everyone starts out with a base score, 1500 in both systems I believe, and then with each game played, your score goes up or down depending on if it's a win or loss. The amount your score goes up or down depends on the score of the person you're playing against, and Glicko also takes into account the opponent's RD.

I'm no math wiz, but that's the basic idea, hopefully that helped.

As for integrating it with the game interface, there was a big discussion about the scoring system somewhere on the boards a while back, so maybe check that out. I think the current in-game system is fine the way it is, it's really meant to measure experience, rather than raw skill, and it does a fine job at it. The only thing I can see being a problem is that eventually everyone who plays enough games will become a General or Marshall, and it won't be particularly special anymore, except to noobs I guess. But we can cross that bridge when we get to it I spose.

Just look at your RD though!

Great posts guys! Makes more sense now.

I'll repeat my previous call:

I don't know much about the php code that was originally written by Qohnil. I'm a veteran C++ programmer, but I'm new to PHP, and don't really have enough time to figure it out.

If there is anyone with expertise in PHP that has some spare time to help out, we could get some additional sorting methods (established players only, etc) and additional data displayed (number of wins/losses). If you can help, go ahead and post here, or contact me.

Well, I got really bored and decided to go ahead and create a system that would make it show only those players with an RD below 100....and by "create a system", I mean "scroll down the list looking for 2-digit RDs".

http://legacy.rooms.rubberrooms.net/MAstats2.txt

If I missed anybody, just say so.

Nifty, thanks for that, though impractical to maintain I suspect.

this time it looks a little bit better...but still i repeat my statement - any scoring system which not takes in account the planet size is meaningless.
one victory on New Paradize against strong opponent worths 2 wins on Anubis or Rust and probably 5 wins on small planets.

I fully agree! I'd love to be able to take into account planet size!

It'd also be nice to see who is the best on what planet, as there really are different strategies for each planet, and some pick those up better than others.

I agree that there are different strategies and styles of play for different maps and different SA layouts, and some do pick up on them faster than others. So having rankings by planet would be cool.

I don't think we should be assigning weights to different planets tho. A map is a map, and each requires it's own style of play. If a player can only win on a certain type of map, either big or small, that suggests the player is limited to that niche. This happens with the Non-Pay accounts. Since they can only play on Emerald, they get very very good at it, but if you take them out of Emerald, their unfamiliarity with the terrain and long-term strategies will destroy them. A really good player will be able to adapt to any situation and do well on any map, even one they've never played on before.

It's kind of like the Olympic Decathalon. You can't say that the 100m dash is any more or less a measure of an athlete's ability than the hurdles or pole vault. Each event requires certain talents, but the athlete who receives the gold medal is the one who is the best overall.

I'm #1 now?! LMAO!!!

Ok, that's gonna change real fast, cuz I'm about to lose like 5-6 games in a row here.

While I agree with you on many points, I do think that the small maps definately require a lot less strategic expertise to win, and so a player that excells in tactics can beat one that is good all around, with a little luck.

The larger maps do require long term thinking, front management, as well as unit tactics to be able to come up on top. Additionally, naval units are more important, so that goes into the mix of things you have to consider. With more decisions to make, the geniouses seperate themselves from the masses. And for that, they should be recognized.

I would prepose grouping maps of similar size together in order to weight battles, but don't think its possible to really give an exact weight to each map (for the reasons Redfox stated). So a map would either be small or large in affecting the rating.

While this is certainly true, don't forget that the opposite applies as well. On a smaller map, a good tactician can easy counter the one or two extra units that an all-around player or a strategist can bring to the front, but I've found that on a large map, a strategist can accumulate enough extra troops to smash an all-around player or tactician under wave after wave of his unit of choice, with no real need for tactics in the deciding engagements.

This is why I personally think that the Medium maps are a better test of overall skill than the others (except Crateus, which I haven't played enough to judge yet). On a medium map, if one player slices the enemy's forces to shreds in every firefight, then it will certainly affect the course of the war, but if that same player diddles around in some backwater for too long while his enemy is taking all the neutrals he can find, he's going to find himself outnumbered and outgunned pretty badly when the endgame comes. Also, medium maps (again, except Crateus) have enough water to factor into the equation.

This is also why I agree with map weighting, but for reasons of time, instead of skill. If I can finish six games on Emerald in the time it takes to finish one on New Paradise, I should get six times as many points for New Paradise, regardless of whether it takes more skill or not.

Ahhh, well this is no longer a moot point! Morda just added the planet to each game for me, so soon we will be able to do all sorts of planet-based stats!

Well as soon as I can considering I have about half an hour to work on it each day after all other obligations and don't know what I'm doing

Ahhhh! Got a first look. So far, they aren't very pretty, and I plan to later have good links between them, but here they are in their raw form.

These are the ratings for individual planets:
Antarticus
http://www.massiveassault.com/clans/nwo ... atings.php

Anubis
http://www.massiveassault.com/clans/nwo ... atings.php

Bizarria
http://www.massiveassault.com/clans/nwo ... atings.php
(No wonder Redfox has a high rating, he's played a bajillion games on this map )

Brimstone
http://www.massiveassault.com/clans/nwo ... atings.php

Craetus
http://www.massiveassault.com/clans/nwo ... atings.php

Emerald
http://www.massiveassault.com/clans/nwo ... atings.php

New Paradise
http://www.massiveassault.com/clans/nwo ... atings.php

Noble Rust
http://www.massiveassault.com/clans/nwo ... atings.php
(watch out for Tiger here!)

Wasserland
http://www.massiveassault.com/clans/nwo ... atings.php

He is always saying he's king of Biz, now we have proof, lol.

and again it is noncence i have won more battles on New Paradize that ANYBODY and still am i on 10th place?and somehow i am placed higher than Drakon on Emerald ...

phew screw this lame glicko ....

Sweeeeeeeet