ELO ratings should decrease over time if you are inactive

When the new set of maps arrived, I went from playing 20-25 MP matches a day, to playing almost no MP at all. I've fought human player only twice, while concentrating more on SP and making custom maps. One thing I've noticed is that even though I'm very inactive in MP, my ELO rating staying exactly the same (around 1750), which I think is unfair towards those who have been active. In order to stay at the top of the ratings (or at any level), you should be forced to play frequently to earn that rating. Currently it would be way too easy to just protect your rating by not playing.

I'm not an expert on how ELO ratings work, and I don't know if there is a "right" way of making it decrease over time, but I'll go ahead and post some of my thoughts.

Making ELO decrease 1 point for each day of inactivity, would not help much because you could still cheat by fighting one match against AI each day. I think a better way would be to have a quota that you need to meet in order to avoid rating penalty.

Lets say we had a quota of 20 MP matches per week (based on amount of wins you need to get that purple chest). If the player fights 20 MP matches (win or lose) during the week, his or her rating does not suffer any penalties. But if the player plays for example only 12 matches, his or her ELO rating will drop 8 points at the end of the week (20-12=8). Maybe there could also be a rule that all those 20 matches have to be against human players.

A system like this wouldn't make a huge difference to ratings but at least it would help a bit. Common sense tells us that some kind of system like this has to be implemented.

Anyone has any better ideas?

 

I hate it when games want constant attention. Not everyone is in high school and has the time to play all the time. Some people have jobs and families.

What you could do is hide inactive players in the rankings (maybe after 1 month of inactivity). However, if you log in and play an MP game, your score goes back on the board.

 

I find myself agreeing with both of you...

 

Yes, but I don't think having to fight 20 matches a week would constitute as "constant attention", especially when talking about players who are supposed to be the current top-10 Card Hunter PvP players in the world.

Also I don't think having a job or a family should give a person any special treatment. It would be silly to have a system where inactive people would be able to stay on top of ratings for months, just because they need to change a few diapers. Active people are always rewarded in online gaming, just like they are rewarded in everything else in life, which is how it should be. I mean is it too much to ask that the top players would also be active players? Game keeps evolving, and if you do not play, you do not earn your spot in the ratings.

What you suggest there wouldn't solve the problem at all. I could just sit doing nothing for 29 days and then fight Mom once. Rinse and repeat. We need something that cannot be exploited or avoided.

 

If elo decayed a bit regardless of play time, that would address the problem, as not playing would make you lose points on average, and playing mom would give you a ridiculously low amount of points. The problem would be setting the decay rate to a convenient value (it would not create elo deflation if it makes up for new players losing a few elo points to more experienced ones,before their elo stabilizes to its correct value : just make sure that new players elo rating is at the target elo average).
Another way to do this would be to add decay to scores over 1000, (and reverse decay to scores under 1000?) to stabilize the total elo around a target value of 1000 for instance

 

Personally I think this is only an issue for the very very top of the ladder. If you were at 1400 rating and stayed there inactive for 6 months, no one would notice or really care.

I would break this down into 2 sub-issues:
1) Infrequent PVPing means that your skills, to some extent, would degrade over time. Someone who has lost their keen competetive edge after taking a break for a few weeks, or someone who is not up to date with the most recent counters/meta shifts would be noticably less skilled than when they were active. This reduction in skill is not reflected in the current system.

2) The top positions on the ladder can be considered positions of prestige, and players should (?) be required to fight hard to stay at those positions. The top of the ladder should theoretically display who is the best at the game at that particular point of time. Perhaps the top should display a hybrid of overall rating plus the total delta of Elo change (both plus and minus) they have gone through in the past 24/72/168 hours?

If 2 players both have a rating of exactly 1750, but one of them has played 100 matches in the past week and the other has only played 2, the player who has more matches is more deserving of the top spot, in my opinion.

Of course, these issues are part of the larger issue that there is no real reward/incentive to hold a top spot on the boards...

 

@Scared Little Girl: C'mon, this is a flash game, not the 2016 Summer Olympics... I think you're taking this a bit too serious. I think it's fine to take inactive players off the ranking table after a while, but keep it simple please.

 

I think it's hard to define 'active' satisfactorily. Someone can play 20 games in one day, the other six days of the week don't play at all - is that any more or less active than the guy who played just three matches a day, especially if one of those three happens to be against the AI (which -will- happen at higher levels?)

I personally have no problems with the current system, because - as what was stated earlier - there's no tangible reason for keeping your pvp level higher other than personal satisfaction or perceived prestige.

 

That's an interesting suggestion. I'm not sure if "punishing" everyone is the best way to go though. I'd like a system where active players are not penalized. Makes more sense to me that way.

Yes, it's definitely only visible in around the top-10. Then again if ELO rating would decrease over time, I think those should rules should apply to everyone, even those not in the top-10. The fact that they are more invisible, does not mean rating exploiting cannot happen.

But it would probably be a good idea to have some kind of minimum rating that you need to have before it starts to decrease. I did some research and looks like for example League of Legends has a system where ELO decreases 25 points per week after a month of inactivity but only if you are above 1400. Similar system could work with Card Hunter as well, although it would be easy to bypass by fighting single matches against Mom every now and then. I personally would go with something more aggressive.

This is a very cool idea. It would be a great way to see who the rising stars are. And I think it should be 1 week (168 hours) like you suggested.

"This is not the Olympics" is a really weak argument. I could use that argument against anything, because how important can a flash game be? I mean why do we even bother talking about card balance because this is not the Olympics?

I personally couldn't care less about the ratings. I would be happy if BM decided to get rid of them altogether because then I would get a wider range of opponents and not have to fight the same 10 guys over and over again. But the fact is that Card Hunter does uses ELO ratings, and because it uses ELO ratings, it should use them so that they make sense. Protecting your rating might not be a big issue now, but in the future, when ratings could potentially have a greater meaning (extra loot for top 10 etc.), this could be a bigger problem. That is why it wouldn't hurt to fix it now.

And it's not like this is some weird idea I came up with. It's something that many other games use because it makes sense and is fair. Just Google it and see for yourself.

True. Defining "active" wouldn't be easy but it's not that big of a deal what the actual number is. For example that 20 wins needed for a purple chest is not based on anything big (other than the number of fingers and toes you have), it's just a number that the devs came up with. They could do the same with activity and just decide on a number, maybe choosing a bit lower just to be safe. Even if the number was 1, it would still be better than the current system.

 

I still fail to see a problem that needs to be solved. And elo decay has a huge potential in discouraging players from returning to the game. Two months ago I first took rank1 with an elo of 1497. Now I am at 1800 with a few people nearby in the rankings. Where is the problem if the general elo level keeps inflating? If you do not play at all you will get overtaken by the active players. I can see a potential problem in the future if the playerbase dwindles, but as I said back in beta the proper solution is something like the Spectromancer system. That one adds slight point inflation to the system (wins give +1 point compared to loser's losses).

As for matchmaking there are simply very few people that play at your level Scared Little Girl. Accept that .

 

Really don't understand why this is necessary. It's like chess ratings. Even if a master takes a few months off they're still a master. If they get rusty when they come back, they'll drop points pretty quickly to other people.

If inactive people with good ratings were somehow hogging something- like there was a free chest handed out every week to the top 100 rated people- then I'd agree something needed to be done to ensure only active players were getting it. But rating degradation? Nah.

 

You are right indeed, Turinturamba : Elo inflation is enough to turn this into a non problem

 

The problem is that is happens too slowly. Someone could be inactive for months and we wouldn't see much change. What I suggest wouldn't make any huge rating differences overnight, it would only slightly speed up the natural process.

And I disagree that elo decay would somehow discourage players from returning to the game. I mean who would take a 6 month break from the game, and then get angry because their rating has changed during that time? Nobody.

This is a perfect way of saying it. And it's not how the system currently works.

Chess is different because when was the last time a chess piece was nerfed?

Card Hunter gets monthly patches that change the metagame. A player could increase their rating using some overpowered or even a broken build. In a few weeks that player could be #1, and if that player stops playing after the devs fix the problem, he could stay in the top-10 for a long long time. You cannot do this in chess.

So you agree that there is a problem, but you just don't want to fix it because potential exploiters do not get enough rewards? If the system is flawed, it should be fixed whether or not there are some huge rewards involved. Some people like the idea of being in the top-10 and they work hard for it. I don't see why we should cater for inactive players over active ones.

Instead of asking "why should we do this?", we should be asking "why not do this?". I mean what really is the downside here? People say ratings don't matter, but when we talk about decreasing ratings for inactivity (something that is perfectly logical and used by other games), ratings suddenly do matter? Losing some ratings because you haven't played for a month does not mean you lose your soul

 

Well said.

Less well said

As in my equivocating first post in this thread, I find myself divided. This doesn't seem like something that would hurt anything, or like it would offend too many people. It also just doesn't seem like something that anyone needs. I am not sure what value it adds. I'm totally down with people hitting 1500+ and then being like "I hit this rating milestone, I'm done"—I almost did that at 1337, just 'cause. It "means less" the longer you go without playing, but honestly your rating already means nothing if you stop playing after a winning streak. While I agree that the chess rating scene is different due to the reason outlined above, chess still has high-rated players who can't play anymore (age, disease, etc.), and they don't lose their rating when they don't play, even if they're not as good as their rating would suggest. What need is there for people to get taken down on the ladder? That's what I'm missing.

 

An alternate way to accomplish something similar would be to do add MP seasons based on map sets or builds that balance MP regularly (ie once a month). Also, if you could see a player's highest rating ever and their current rating, that'd allow people to have the e-peen factor, plus motivate players to actually shoot for top rank each build/season.

 

One thing about chess; there is no balancing that occurs. Essentially, the game never changes as the rules remain constant. That's one reason why an elo decay in chess shouldn't matter in theory. This game has new items being added, removed, nerfed, and abilities suffering the same fates. New maps are added which change the nature of a person's playstyle and what works and doesn't work is in flux.

 

Uh, yeah, SLG said that and I agreed But I was just pointing out that in chess, you still have players whose ratings don't reflect their actual skill level for OTHER reasons besides rebalancing. Anyway your suggestions in your previous post all sound good.

 

haha, that queen is getting OP indeed.
Seriously though I agree, just because inflation may have a balancing effect this is a simple matter of what's right even if it's not an important issue right now.
Also - chess players who stop playing for a few months can't be equated to CH players because chess ranked matches aren't a daily fixture. CH is constantly open for play so some off time has to be set as unacceptable and at first it can be a long one, no need to immediately punish anyone who's absent for a week.
In beta, although a smaller player base for sure, the leader for the last 2 months all the way to the finish (you can check the beta leaderboards on the main site) was someone who didn't play a single match in that period and in reality the second ranked ruled the meta game and should have been first, a thread at the time was opened on the same subject but nothing was done. would be interesting to know where the devs stand on this.

 

I'd also like to point out that this is in many ways an issue which should be taken seriously: if this isn't fixed in some way, the competitive scene will suffer. As the people who PVP are able to get much more out of the game than the people who just play through the campaign and quit, a lot of the longevity of the game will depend on the competitive scene's health.

CH is just a game. But any game where people put in hundreds of hours of work deserves to have its issues taken seriously.

 

I could not agree more.