• Please make sure you are familiar with the forum rules. You can find them here: https://forums.tripwireinteractive.com/index.php?threads/forum-rules.2334636/

ROLadder announces some of its features.

Well regarding your suggestion i gave my opinion in your thread you can read it when roladder is back up :p

but the general chat system can be used without being on the website,its not a proprietary program or whatever. Afaik it is something existing but cleverly incorporated within the website as well. (Again don't hold me to it as I'm not up to snuff on the chat part).

In general ESL's ladder is not really ideal for RO as the default ELO system they use in the advertised configuration doesn't converge fast enough for instance. Making the Elo system converge faster isn't a fix to all issues but its a step in the right direction, but ESL doesn't allow game Admins to modify the needed settings to do that. Although with leagues there is not much that can go wrong so in that sense it would be nice to have ESL pro gaming leagues.

ROLadder is a big undertaking for everybody as well, if we could get away with just signing ourselves up as Admins on ESL or Clanbase we would have probably done that and save ourselves a boat load of work. But both ESL and Clanbase lack both some technical functionality for, and is especially lacking in community features. Staying alone allows us to tailor make the site to this specific community and fully focus on optimizing things for this community.

The advantage of ESL and Clanbase lies in its marketing prowess due to a big client base and with that a famous name. The advantage of ROLadder lies more in its ability to offer a place where we can custom bring things the community asks of us or necessities towards the game. And that all our resources and time will go to this one game. A lot of the current design plans come based on an actual survey we held a while ago under clan leaders, and generic suggestions on the forums.
 
Last edited:
Upvote 0
Yea the ESL ladder is indeed in few words described as "bad" :p. If the goal is to get a high rank then it is all about avoiding playing against teams that are more skilled than their rating, and try to play against new teams that are not good and get a 100% win and score **** load of points :p. Ideally the roladder system would not only get teams to their true rating faster, but also removing other disadvantages when playing against newly signed up skilled teams =). Don't know if it can be done, but it is interesting things to think about as it includes math ^^.

I am more of a league person though. I find ladders quite boring as if I lose a match, I can just win another and there is no harm done. Everything comes to it's edge when playing that important league (or cup ofc ^^) match that determines the future. Just like in real life sports!

ESL does not really care for the games either. They couldn't care less that they killed INS, and wouldn't work for RO2 but rather try to get RO2 to work for them ^^.

Yea the chat thing sounds cool (trying my best to not assume stuff about anything though). I assume it should be easier to get people to install and use compared to mIRC ^^. In general I'm sorry if I come off as a bit rude, I'm just interested in user-friendly systems and competitive RO2 =).
 
Last edited:
Upvote 0
Although the module set-up of the site allows for us to use different rating criteria of teams, we still need to restrict ourselves in not offering too many systems. So all ladders will most likely show one type of rating system.

A full rung based system won't happen as rung ladders while super quick in converging are highly inaccurate for a ladders overall ranking. As in if you win a lucky match at one time one teams score immediately plummets and the other teams immediately rises. Basically you need to look at past performances as well to somewhat dampen the change of scoring, and at that point you get into a scoring system.

But we might do something like a rung based top 3 so in order to obtain an award for a nr1,2,3 spot in a tier you need to actually beat the previous champions. Even though an accurate score system indeed shows who is really the best, it just feels not so nice when the top team gets kicked off without even playing the one taking the place.


In a sense parameters the ranking system should take notice of when its fully finished.

- how many matches a team played and how long ago they were played, this helps in knowing how accurate someone's score is. If a completely new team plays an old team you know that the new team is likely to have an incorrect score. In a case like that its mostly the new team that should get major changes in it's score and not the old team should probably hardly be affected if at all. The values to start with will be obtained from the current ROLadder website, if for instance the time delay is super important in the ranking system(in the sense that a match played a week ago is like 10 times or more inaccurate than a match played today) then you essentially end up with a rung based ladder.

- And a big basis for the scoring system is that generally the scoring system predicts an outcome based on the teams scores. If you play against a weak team the system will then read from the score difference that you need to win over for instance 80% to actually play better than the prediction. Meaning that even when playing against weak teams or against strong teams in order to "win" you do not need a definite win per say, but need to beat the prediction of the outcome. This makes it impossible for strong teams to just beat on weaker teams to gain a quick route to fame.

- how balanced the maps are. Any prediction on the outcome of a match will always give a chance of success in a percentage. Ideally you want a match to be played in a way that you can say that someone won with for instance 60% or 70% rather than only saying someone won 100% or 0%. As more accurate scoring outcomes allows you to increase the speed of the rating system. So when you say report matches as a 3 rounds won 1 lost type (playing for instance 2 maps on 2 sides). Then you need to take into account map balance, because on a map like Arad or Kaukasus there is a big chance that the allies will win and thus the map ending in a draw (as you play both sides of a map).

----------

Ladders and Leagues both have their advantages and disadvantages. Ladders are more about knowing how strong you are and finding opponents close to you to get better. Together with that in ladders you don't have to play at a strict schedule of like 1 match every 1 or 2 weeks.

In a league its about in a group playing every team so you know for sure what teams are the best, but you are constrained with activity. Often with traditional ladder systems like elo, actual matches are played in leagues, and the results of leagues are put into ladders, and then ladders are used to determine what leagues you can then participate in.
 
Last edited:
Upvote 0
as rung ladders while super quick in converging are highly inaccurate for a ladders overall ranking.

well that's more a matter of opinion....i don't feel like writing out a novel, but just a couple quick notes about rungs:

rung ladders are very simple and to the point. you win, you move up. you lose, you move down. because of that, as you said play is quicker but this style can also encourage more competitive challenging. additionally, issuing rank drops for penalties (no shows, forfeits, inactivity, challenge dodging etc...) are much simpler to do with rungs as opposed to points.

yes a points based system would be good for longterm ranking as it's more detailed like you described, but when were talking teams needing at least a month to accumulate enough matches to calculate "accurate" statistics.....is it really anymore accurate than rungs? ELO and points based rankings are mostly used for competitions where teams are playing hundreds of matches. let's face it, how many teams at roladder played 100 matches at rol over the course of its existence? 2, 3 maybe 4? (i know Core, maybe cc// at least)

also with points you bring into the issue of points resetting. rol, from what i understand never reset points at the completion of seasons, or even at the completion of every other season. that seemed to bring up a lot of drama and controversy. with rungs, there is no controversy. you win and move up, or you lose and move down. imo rung variations would be the best for 2-3 month ladder seasons. then perhaps having some type of points based ranking to comprise an "overall" or "career" ranking.

if people want to try it, why wouldn't rol be willing to experiment?
 
  • Like
Reactions: SgtH3nry3
Upvote 0
I don't see how a rung ladder would make things more competitive. If anything it would make things less competitive because who could take it seriously. Everything would comedown to the end of season matches instead of performance throughout the season.

Anyway I look forward to playing on Roladder. I cant wait for there to be an NA clan scene again.
 
Upvote 0
well that's more a matter of opinion....i don't feel like writing out a novel, but just a couple quick notes about rungs:

rung ladders are very simple and to the point. you win, you move up. you lose, you move down. because of that, as you said play is quicker but this style can also encourage more competitive challenging. additionally, issuing rank drops for penalties (no shows, forfeits, inactivity, challenge dodging etc...) are much simpler to do with rungs as opposed to points.

yes a points based system would be good for longterm ranking as it's more detailed like you described, but when were talking teams needing at least a month to accumulate enough matches to calculate "accurate" statistics.....is it really anymore accurate than rungs? ELO and points based rankings are mostly used for competitions where teams are playing hundreds of matches. let's face it, how many teams at roladder played 100 matches at rol over the course of its existence? 2, 3 maybe 4? (i know Core, maybe cc// at least)

also with points you bring into the issue of points resetting. rol, from what i understand never reset points at the completion of seasons, or even at the completion of every other season. that seemed to bring up a lot of drama and controversy. with rungs, there is no controversy. you win and move up, or you lose and move down. imo rung variations would be the best for 2-3 month ladder seasons. then perhaps having some type of points based ranking to comprise an "overall" or "career" ranking.

if people want to try it, why wouldn't rol be willing to experiment?

Lets first say that the positive and negative properties of a rating system are not a matter of opinion, its a matter of maths.

The reason of not resetting points is that top teams with immediately attack weak teams to gain a higher ranking. Even if you reset points teams do not suddenly get weaker or stronger. Point systems indicate not just an order of teams it actually represents a point amount that dictates how strong a team is. And with a rung based ladder when you reset it, by random chance one team will automatically start as the top of the ladder and one at the bottom, a situation you do not want to have.

A good point system will let you rise in score when you got stronger than your formal self. When a season changes you generally do not get stronger or weaker. In rung when you get challenged by a team just one spot below you, you can only loose and not gain anything. Which is why everybody will try to just get themselves booked full with teams that are higher than them, leading to a odd game play scenario. In a point situation you just need to play better than a prediction, allowing you to gain points even if a system predicted you would win with 75%, by simply winning with more than 75%.

Well at first, ROLadder will likely not use ELO but an in house developed score system (which actually converges faster than rung based ;) ). But even if you look at the current ROLadder website it got a tweaked version of ELO, adapted and modified specifically to Red Orchestra. A big issue with rung based is simply that any performance in the past doesn't matter, your position is solely determined by the last match you played.

Its not so much that we're not willing to experiment, and we could perhaps at some time let the same ladder run multiple rating systems so people could see for them selves how different systems work differently. (Although logically one system would be the one ROLadder uses for handing out awards etc). But running say 3 separate ladders with as only difference the ladder rating system would simply unnecessarily split the community.
 
Last edited:
Upvote 0
I don't see how a rung ladder would make things more competitive.

well it's not really a difference specifically in making things more competitive but in simplicity. however, a rung system actually can encourage more play and activity, especially in using position swap, due the ability to advance up the ladder more quickly. if you are a #16 rank and you can challenge up 5 ranks, that not only gives you at least 5 teams you can challenge but also allows for the potential to move further up in the rankings in just one match than you could with certain point rankings. if you're doing position swap, you could win and move up 5 ranks, whereas in a point system, you may only take 15 points and thus only move up 1 or 2 ranks. even with a halfway up/down system, you can still potentially have more movement than with points. that type of mentality makes each match vital as there is more at stake. higher teams will care more about defending their rung and lower teams will care more about moving up as they realize that if they can squeak out a win, then it will be very rewarding. also, say in this situation your team is ranked #14 and a #16 challenges and beats a #11....plus with a halfway up/down or swap system, you as the #14 will move down despite not even playing. that would in turn encourage your team to make a challenge to regain your lost spot.

a rung system also prevents abusive inactivity because a team cannot build up such an immense amount of points and just sit at the top of the ladder and remain "unchallengable" because they are outside a specified points range.....or that it ends up being the same top 4 or 5 teams only playing each other where losses don't result in much movement. also like i said before, it's also easier to assess inactivity penalties.....inactive for 2 weeks, drop x amount of ranks as opposed to using points, where assessing a 15 point penalty might not move a team's rank at all (i.e. rendering it essentially a worthless penalty).

Zets: with any system (unless teams start unranked until they've play x amount of matches) teams will enter the rankings at a certain point. the difference between rungs and points is that by starting at the bottom of a rung system, you rightfully have to fight your way up. now if say using ELO, you join and start with 1000 points.....well what if in fact your team isn't anywhere near as good as a team with 925 points? it could take 3-4 matches before the rankings pan out to what would be considered "accurate". also without a points reset, you're basing a team's hypothetical "strength" not off of the current team, but of a previous team's performance. my team could have a roster of allstars and work up to a #2 ranking with 1100 points, but 4 months later at the start of the next season, my roster changes and i lose 3 of my better players.....the team at that point is not anywhere near as good as it previously was, yet they are to be treated as if they are simply because of their built up points. in a points system, it could take a significantly longer time for the change in ability to show due to the fact that it could take 4-5 matches for a team to experience significant drops in ranks. on the flip side using rungs, a team could move from #2 to #10 in 2 matches. each match matters and strength is determined on the field, not in "maths".

again, i'm not saying that ELO or a points system doesn't work, because it does work.....just not in this setting. we're talking gaming not chess, tennis or another sport where the majority of teams average playing 20 matches each month. even more important is the fact that our form of gaming is a team oriented competition. ELO works best with individual sports because it's the same consistent variable as team rosters/lineups never change. however with team gaming, ELO/points systems don't truly produce accurate assessments of team strength or match predictions because you're dealing with rosters that can vary on a daily basis. the predictions are based off teamX's previous performance despite the fact that teamX in week 12 might have a lineup with different players than teamX did in week 4, thus potentially resulting in 2 completely different teams in terms of "strength". the team on the field should be the one that produces movement in rankings, not the hypothetical "team" that is represented in the rankings table.

sure ELO/points can result in more accurate longterm ratings if the conditions are right, but you're banking that teams will be playing 3-4 matches per week over the course of 6-12 months and that rosters will not have significant changes during that time. while i don't doubt that could happen, as it most likely will in some cases, the conservative route would estimate that the majority of teams will be "semi-serious" teams and play maybe 1 or 2 matches per week with potentially different lineups each match......in those cases, the complexity of an ELO based system just wouldn't be worth it. i think that's what forced many clans away from competing in ROOST because the rankings were just seen as too "serious". perhaps for say premier divisions or leagues a more complex ELO/points based ranking system would work better, but if you want to pull in the most amount of clans, you have to understand that a large portion of the former RO community, and what will be the RO2 community, will be more relaxed in their attitudes towards competing.....for those clans, the simplicity of a rung system would be much more inviting.

i guess the real question is, what type of atmosphere do you want to create? do you want to cater to the super competitive, super serious types only, or do you want to offer competition for all levels of teams? i'd imagine if there is to be one "source" for RO competition, then that organization should cater to all and not just some. using a rung system would be perfect to draw in the majority while ELO would work well for the highly competitive bunch. mixing and matching and offering a lot of options is what will get the most involvement.
 
Last edited:
Upvote 0
Ok let me first get some things straight.

Before I begin do you have any understanding of how ELO functions, as your post brought me with great doubt. (And remember that ROLadder will us an in house developed system or a heavily modified version of ELO with many additions). http://en.wikipedia.org/wiki/Elo_rating_system

As like with any formula you can change loads of parameters to suit your need without changing anything to the formula. Logically people won't hand everything on a golden or silver platter but the basic formula is quite simple. That you need say 3-4 matches per week and a period of 6-12 months for instance would mean that you would use a super slow system, much slower than even the default ELO rating system. So I wonder where you got your statistics for that.

For instance the starting score of something is just a number that you pick, if you want people to start at the bottom then you can just let people start at the same score of the lowest teams. But you can decide to test a team first and inject him in a position that is close to his approximate strength, as he will then quickly converge to about the strength that he really got.

Or how quickly teams can gain and raise in their point amount is configurable as well, if a team is weaker than their ELO rating gives out. Everybody will then challenge said team as that means they can easily play better than the ELO prediction. Meaning that any team that suddenly lost loads of strong players will quickly get back to the score they should have.

While a Rung bases system encourages activity towards challenging teams above you, as you cannot loose anything by losing. Similarly it absolutely discourages playing teams below you as you cannot gain anything. This requires you to force people to accept matches, which will result in most teams trying to book as many matches against teams above them so they cannot be challenged by teams below them.

When resetting a point system everybody will start out equal, where as with a rung based system people will randomly be assigned to be the best or the worst team.

If you want to offer competition to all levels of teams then some things are important. For one strong teams shouldn't be able to exploit weak teams. If you take a rung based ladder and reset it every 3 months, then the previous nr1 team can be at the bottom and the previous last team can be at the top.

Every team and their grand mother will try to beat up in such cases the lower teams that randomly got at the absolute top of the league. And as said before they will be required and forced to play those beatings. That's not something that less strong teams generally enjoy doing.

Point systems are in place to give a good prediction how strong you are relative to other teams. Which means that you can make rules that protect weaker teams allowing them for instance to decline matches against teams that are too strong.

In rung based systems you need to build in some viscosity as otherwise teams continue bouncing around in rankings without ending up at some stable point. In the end going up in score means you were better than the team you played, but your ranking isn't a good indication of how strong you are or your opponents are. That brings a situation that is not advantageous for anyone.
 
Last edited:
Upvote 0
I never knew there were a word for such systems ^^. Somehow it feels like it shouldn't be that difficult to make some modification to basic rung to make it a little more "exciting". Like if you have won 3 games in a row, if you lose next match you will drop 3 less in ranking (never lower than -1 in rank though). If you have lost 3 matches in a row, if you win next you will win 3 less in rank (never lower than +1 in rank though ofc). That should encourage lower ranked clans to not take on clans that they cannot win against, and not discourage high ranked clans to play agaisnt low ranked clans. Modifications of a rung based system is reasonable just as modifications to ELO based system is resonable ;).

Couldn't it work to give out one award for rating-rank in the ladder and one for rung-rank? If that would be something that people like, and if the ROLadder has a rank display for both to begin with (to just try it out).

I just noticed that on TWL the rung ladders seems more common than rating-based ones :D. I could never have guessed that ^^.

Personally I am against any ladder reset, both rung and rating. The ladder is supposed to be an ongoing update towards accurate rank, not reseting to screw the rank up ^^. Leagues are to be played in seasons, ladders not imho :D.
 
Upvote 0
There are plenty of methods you can use to remove some of the biggest issues with rung based systems.

The easiest one is splitting competition up in relatively small tiers of say 10 teams per tier. As most issues with rung based ladders appear when they are compromised of a large group of people. That allows you to do a ladder reset every few months as well (although it would still be better to not do that :p). Aka essentially grouping people together of about but slightly different skill, so you evade most issues about skill differences as well.

Next to that rather than using a full or half swap method, when you win you should be placed 1 spot above the person that you beat. While the person that you beat should not should not get your old spot but rather ends up 1 spot below you.

But you cannot use a rung based system to make any predictions about a players skill. A rung based system is not a predictor. If you make it a predictor its not rung based any more. Which is the key difference between a point based ladder.

In a way rung based systems share a lot of properties with leagues, and there are advantages with it. The key issue is that for rungs to work well they need to be utilized in a specifically adapted situation.

You cannot run a rung based system next to a point based ladder, although you can run a ladder next to rung based system.

I hope that a lot of rating systems will make it into RO:Ladder so different ones can be used. So although perhaps not initially eventually multiple rating systems can end up being available.
 
Last edited:
Upvote 0
Back in the day, I had such a bad experiences with organized comptetitive fps play that it just never was worth the effort or discomfort for me. The immaturity, the lack of respect between players/teams of all ages, referees who couldn't make a decision, etc. The competitions seemed to bring out the worst in people and no one 'in charge' ever had the balls to much.... .....the list is long and not the point of my post....

I enjoy a good friendly game between organized teams, but I don't really care for 'being the best' or vying for top honors. I just enjoy a good fight within the parameters of the normal game. It would be nice to find a place where casual players/teams could congregate to set something up. This may already be a feature of ROLadder. I've not read completely through the information. If it is and is mentioned, my apologies for my indolence.
 
Upvote 0
One of the main reasons to go for an accurate scoring system is that you can help weaker teams.

Predictive scoring systems as the name implies allow you beforehand to get a pretty good idea what the outcome of a match will be. This allows you to choose opponents that are as strong as you are. And allows rules to be created based on strength difference rather than rank difference.

The best and most fun competitive matches aren't played against teams stronger or weaker than your self, but matches played against an even team that end up pretty close. Point based ladders give people the option to decide to try and climb the ladder or just play matches against teams of similar strength.
 
Last edited:
Upvote 0
Which is imho where leageus beat ladders in how fun they are. They group ideally equal enough teams together for an epic battle for the #1. In higher divisions the assholes roam, but in lower divisions there are way less of them and more casual clan players =).

Actually it is a weak point of leagues and rung based ladders. Its just that if you split up people into many different small leagues based on skill that the issue gets evaded, as then you can never play someone largely stronger than you are as all people in your tier are of about equal skill to begin with.

But to get successful results in tiers with likely skilled people, you still need good methods of creating tiers with equally skilled people.

1) Which can be done either by a point based ladder like in tennis, where all league results are entered in an elo system, and different leagues have min and max elo rating participation rule. Similarly you could use a point system to decide in what tier someone is going to play at the next season (glicko 2 is generally a good point rating system for something like this).

2) Or you could use for instance use an automatic transfer system where the top 3 of a 10 player league go a tier higher and the bottom 3 go a tier lower. But utilizing this with a lot of tiers could make it a very long road for a new super team to actually reach to the top as you can only climb 1 tier a season at a time. Which is why method 1 gains preference.

3) Any more ideas ? I know worluk got some special idea for advancing tiers but I do not know that one :p

Which is why if we would end up using a rung based system (in the form of beating a team you get moved just above the team that you beat without the losing team being moved down), it would still happen in small tiers and likely in conjunction with a point system.
 
Last edited:
Upvote 0
There should indeed be a system to put newly signed up good teams into higher divisions. Winner of a division could move up two divs and the 2nd team one div etc. More teams than just bottom two moves down etc etc etc. A less static system and as you said, some help from the ladder and perhaps other ways of determining which team that get's to play in a higher than the default league for new teams.

It has to be in some way dependent on last season in a league for it to be fun, otherwise it is just a ladder split up. Leagues are more fun than ladders so it should be worth the trouble :D. Or just leave that to other gaming sites ^^.
 
Upvote 0
  • A quickly converging ladder system, which takes into account the activity and skill difference of teams.
  • A fixed slot league system, to offer a different flavor of competition beside ladders.

We will have both, but whether we will or not use tiers and how people transfer from tiers depends on how many people will sign up and participate.

When you have a small amount of tiers then its better to go with the system of moving the top x players of a tier to a tier higher, and the bottom x to a lower tier. Although this system in pretty much all ways is worse than a point based system, people can easily grasp and understand the system.

But if a lot of teams sign up and you get loads of tiers then its much more advantageous to use a point ladder system to decide in what tier someone should play, as otherwise it could seriously take years for a new constructed super team to rise to the top. Remember that when using a league with tiers the next tiers will depend on the last season, as only the games of the last season modified the set up of those tiers.

And I know Worluk had some great ideas in store, but do not actually know them, so it might be something completely different.
 
Last edited:
Upvote 0
Zets, yes i know how the basics of how ELO works. but you're avoiding the heart of the points i made about how that system doesn't accurately assess a teams' skill, at least not for typical online competitive gaming. teams can potentially have varying rosters. there is no guarantee that teamX will have the same lineup during any given match that they did during a previous match. to base a prediction of an upcoming match on the results of a previous match, without being able to keep variables such as match participants constant, you're not really assessing the skill of said team that is involved in the match. again, that's why ELO based systems are more traditionally designed for individual sports and NOT team competition. and before you list team sports using variations of ELO, aside from the BCS, none of those rankings impact actual standings for those sports......also with the BCS system in NCAA football: 90% of americans do not approved of the BCS ranking system and over the years, it's many failings have been exposed.

now if you say, well x multiplayer gaming sites use an ELO based ranking system, we have to realize that the majority of the venues of these "team" competitions are more often then not having teams using the exact same lineups every match......that's because they're playing games where matches are typically no more than 5v5. teams don't need to carry a roster of 10 guys so that they can always guarantee having 8 players for standard 8v8 matches. with RO and standard 8v8 matches, the majority of teams will play different lineups quite often. again, because of that, the variables you're making predictions on are not constant.....that's more guesstimation than accurate prediction.

back to the rung system for a minute, the challenged team will also lose something by losing. the two standard forms of rung systems is that if you are the challenged team, if you lose you either swap positions with the challenger, or you move halfway down and the challenger moves halfway up. that is why typically teams are only allowed to challenge up (or at least matches are only forced to be accepted when challenging up). sure, you can be the lower team challenging up, lose and not drop in rank.....but that's the whole point to the system. as a lower ranked team, there IS a prediction: the lower ranked team is expected to lose because they are lower ranked. if they do lose, then that was the expected, so why should they be penalized by dropping in rank? they shouldn't. on the other hand, the higher team was expected to win, so if they lose, then there SHOULD be movement in rankings. you make the point that this would only lead to many teams trying to challenge the same opponent (the highest ranked that is challengable), but that's why you have challenge limits. example, team can only receive x challenges per day, can only have scheduled matches within x amount of time period, teams can only play the same team x amount of times within a specific period etc.... this ensures that teams don't get overbooked or that people will always end up challenging the same team. most every professional script i've seen that offers a rung system has built in features to address this issue, so idk if it's just a limitation of whatever system you plan on using or that you just haven't seen that it exists.

also, in case i didn't make my stance clear on whether or not there should be resets with a rung system, imo there shouldn't be resets. with points however, i have to believe that there should be resets at some point, again due to the fact that lineups vary over time. that is the least you could do to take into account the hole in the system that results from inaccurate prediction making. maybe not a reset after every 4 month season, but at least after every other season. i mean lets take an example from ROL.....how about fullmj: look how many roster changes there have been since said clan has been active. their current ranking and proceeding match predictions are based off previous "team" matches where there were significant roster differences, which were based off of more previous matches of teams with significant roster differences, and so on and so forth. so technically, their current rank is in part based off of the original fullmj team that began participating in ROL in 2004, despite the fact that the current team roster is almost nothing close to the original roster. imo there should never be that level of emphasis placed on past records, let alone the discrepancies in the hypothetical "strength" of teams. the battlefield should determine that strength.

if you are lower team and beat a higher team, you move up. if a lower team beats you, you move down. it's that simple. there is no need for predictions; no need for complexity. if you are truly working to have "less organizing and more playing", then simplicity should be the focus. teams should not have to think about "well we have to play this team and based on their points we must win 6:2 for them to lose x amount of points so that we can potentially move up to x rank if this other team loses their match 8:0." systems that use ELO always come off to me as super serious just because of how in-depth everything is, whereas rungs are more along the lines of casual play. idk, maybe that's just what ROL is going for, a more highly competitive, super-serious environment. i guess if that's what people want then that's the way to go.
 
  • Like
Reactions: {Core}Craig
Upvote 0
With football you use different tactics and a different roster as well, or is football or as you Americans call it soccer not a team sport? Heck even American Football in the NFL seems to utilize a modified Elo system.

You do not need to keep all variables constant, as you can take into account inaccuracies in a teams ranking.

Score predictions change over time, you do not need to reset the scores of teams, so they get a new ranking. If a team changes in skill so does their score, that is the basic principle of a ladder in the first place.

Of course the challenged can loose something as I said before the challenged can only loose everything. And challenging up you can only gain everything. You do not really need to make a rule that you can only challenge up that much, as you cannot gain anything by challenging down in the first place.
 
Upvote 0