Ranks broken?


(radiantSquid) #1

I placed gold 5, I’ve won about 20 straight with one loss… and no rank up. The others in my party only ranked up once from it… Whats going on? If it is a score thing I play proxy so I always have good score per minute or whatever.


(Humbug) #2

It’s because you play against people without rank and you only help them to calibrate their rank i guess.
Should still rank up to get at least the chance to get better enemies though.


(Lagenberg) #3

Have you been top scoring or being average? If average, your rank up might be slower. I got to G5 and am now G2 after ~12 wins? Going to be Elite soon. Imo it works slightly better than CSGO ranking.


(radiantSquid) #4

I’m usually 30+ kills, I’m global in cs go and to get global it took less wins to get up from gold 5 here haha.


(Serious Sam) #5

Well, it is mentioned that there’s a lot of math behind the scenes for comp. Maybe something you’re not doing is heavily weighted, like support xp or objective xp. No actual clue tho, just a guess.


(radiantSquid) #6

I top frag while doing all the objectives… I play proxy hahah. My buddy ranked up and he just shoots people. I just won another two games… nothing.


(radiantSquid) #7

I’m going to try and play healer to see if it support XP that’s the key to ranking up.


(radiantSquid) #8

I just ranked up after playing a healer and second fragging… ummm… the key to ranking up? or what?


(Lagenberg) #9

I think the value most weighed is support xp. Then comes objective and combat. I’d like to say it’s calculated via ratio of those in comparison to other players in specific games or the mean value of all the players to score in competitive matches, but haven’t really time to look into it yet.

My friend actually is one rank higher than me and he is playing my dedicated medic like back in ET while I am doing objective’s and fragging.


(radiantSquid) #10

None of my party has ranked up… we haven’t lost. We chain together tons of wins and no rank ups. We have ranked up before, but not now… why is this?


(Gi.Am) #11

Full Disclaimer I have no knowledge of how the ranking system actually works in DB what follows is solely guesswork based on my experiences in other games.

There are two things to consider.

  1. Ranks are most propably groups. most games use a non linear distribution of the whole gaming population across those groups. I.e. the highest ranks contains the best 1% of the players. While the lowest rank may contain 50 % of the whole playerbase so depending on where you are you have to beat alot of people to get into a higher group.

  2. The ranking itself. Pretty much any ranking system nowadays is based on the ELO system which is used to rank chessplayers (you can read up more about it on wikipedia including the actual math). Developers usually have their own spin on the system (for starters you have to modify it for team based games).
    But the gist of it is, that your ELO number (Your actual rank not your rank group) indicates how likely it is to win a match against someone else. The higher the number the more likely it is to win against someone with a lower number. Because of that you gain less points for winning those matches compaired to games where the outcome wasn’t that obvious(and the looser will loose less points). On the other hand if you loose a match that the system deemed as easy for you. You will loose alot of points (and the winner gains alot of points).

What does this mean in the context of Dirty Bomb?
It means that for matches where you curfstomped the opposing team you will see hardly any improvements.

Also right now alot of people are new to the game and just hit lvl 5 therefore you will play alot of those matches (there are not enough players atm. that are ranked at the same lvl. as you). This should improve overtime.

PS. one thing that is important with the ELO system is that it will more and more reflect your true gaming strength.
So while you might experience alot of rank gains in the beginning, the system will sort you into a place sooner or later, where matches against others have a 50/50 chance of going your way or the opposite. Locking you into place unless you truely stepup your game.


(Lagenberg) #12

@Gi.Am Well, as a chess player I’d like to emphasize that ELO is only comprehensive to games if played correctly can only end up as a draw like chess. ELO doesn’t reflect your skill at all, but your ability to win and in my opinion it alone shouldn’t be used in ranking of games like Dota 2 or Dirty Bomb. And which obviously isn’t the case.

Microsoft TrueSkill has been quite good and there’s been an awesome algorithm from an inhouse Dota league (Dota-GC) which actually reflected players’ skill in Dota. The hardest part is to create an algorithm that takes everything necessary to the equation. In Dirty Bomb we have different roles and their accomplishments differ a lot. Medics may frag alot or support (heal) alot. Field Ops may frag insanely if they have a medic supporting them or just support (ammo and airstrikes). Engineers do most of the objective as they are fast at it, but Bushwhacker is a pretty viable fragger also. Sniper/Fragger/Rhino, the assault/fraggers mostly frag, but the rating should make a difference to a high value kill and to a low value kill (eg. engineer constructing/fixing, medic and at which time of the spawn wave it happens).

Rank groups should be always gaussian imo. If it isn’t well have oversaturated top and bottom.


(Gi.Am) #13

Interesting points. tnx’s for the insights. Most likely true about ELO not the true mechanic behind this game. But my major point was that wether you use something similar or something more elaborate, the ranking up or down will still roughly follow something along the line of “higher rank, destroying lower rank, not much gain in rank”. Therefore winning 100 games against someone might have not much of an impact depending on the teams you play against.

I would also argue that the main constributing factor between teams is who wins or not.
Now for ranking single player skill therefore ranking them across a team (i.e. analogues to trueskill) the anwser could actually be right before our eyes and in the open (just speculating).

The XP system.

It is fairly balanced (sure not 100% perfect but what system is). It allows every merc to reach a top spot in the team. Mercs that have a special role will mostly score one type of XP (i.e Fragger getting mostly Combat XP, Aura mostly Support XP, Proxy Game XP). While mercs that are more multi purpose get a combination of it (i.e. Bushwacker gets less game XP but more combat XP, similar with sawbonez and the fieldops).
A Aura that is reviving and healing like crazy can very well be the number one on the team. Even if the rest has higher K/D’s.

If one player does more work than the rest, you can usually see that reflected in total XP. Him having way more than the next best player. On the other side a team where anybody works hard for the team and fullfills their role ususally ends up with fairly similar XP lvls.

Now on their own XP doesn’t say much about wether a team used their skills in a good way only that they were used. But we can bring the Win/Loss back into this.

Consider a Vassili scoring a crazy K/D. If he scored those kills by sitting back and shooting weak enemies after they went through his team. They will lose. If he moves up with the Team and scores those crucial kills before the enemy is even engaging his team reaching the objective they will win.
So even tho the Vassili player might end up the top score quite often in his team. He would lose ranks since he is the main factor why his team loses and therefore he would be in alot of lost matches.


(KangaJoo) #14

Based on the fact that I pretty much won every game I played before receiving a rank, carried hard in most of those games, and still only got silver 1 I think the ranking system is somewhat based on xp/time played. I’ve also heard from a few competitive players that you start at a pretty low rank no matter what and need to grind your way up.


(activeClip) #15

Placed at Gold4, now Gold1 after about 50 games with a good 90% win rate. Gold2-1 took a good 25+ alone


(Sephios) #16

It took me 17 wins from G5 to G4 so meh