"Ranking Degredation" for song scores

**justin_ator** · 07-3-2016, 10:03 AM

Re: "Ranking Degredation" for song scores

As someone who deals with this struggle I agree something like this would be neat. I play semi casually here and there and have lost a significant amount of skill over the past 3+ years. I think it would be really encouraging to those considering returning (especially during the official timeframes) if you weren't just plopped right back where you were at for skill 3+ years ago. I think you're on to something, but the right implementation will be the challenge.

**Dinglesberry** · 07-3-2016, 10:06 AM

Re: "Ranking Degredation" for song scores

justin made a really good point I didn't think about, a huge part of this game (especially for me) is progressing and improving my scores + my overall rank. This would allow for old players to come back and go through that improvement again, I feel now people do get discouraged when they go into a tournament and get eliminated 1st round because they used to be good years ago.

I used to play quake 3 cpma duels at a high level back in the day, I tried going back to it recently and got stomped and quit, I exactly said to myself "I used to be way better, there's no point of playing and being bad anymore".

**Walrusizer** · 07-3-2016, 10:19 AM

Re: "Ranking Degredation" for song scores

Ranking degeneration more like sandbag regeneration

**MikeShinoda12345** · 07-3-2016, 10:20 AM

Re: "Ranking Degredation" for song scores

Tournaments aren't happening all the time, though. Wouldn't it be nice to quit knowing that your rank can still stand as a testament to how good you got? Or if you're an active player, that when you surpass someone in the Skill Rankings that you're actually improving and not just benefitting off a superior player's perceived rust?

**Xenorosth** · 07-3-2016, 10:33 AM

Re: "Ranking Degredation" for song scores

Originally posted by Dinglesberry

I have a suggestion:

I've noticed in the tournament, there are people in divisions who do not score very well, and if you look at their profiles, they are high rank, but when you look at their top scores, most of them say things like "7 years ago" or "3 years ago" or whatnot. Its not my intention to point at any names, but for example you can see people in division 5 who dropped 30k or less raw score on the round 1 song, when the cutoff was 81k or so~

My suggestion is: For the top 50 scores formula, implement some sort of AAA equivalency degradation.

You wouldn't have to remove the scores from the leaderboards, what I mean is that logically speaking if a player is progressing, they should keep getting higher top 50 scores and improving their ranks.

For my implementation, I propose they should make it so if you don't get a new top 15 score for a couple months (or even a year), the weighting of the scores start to degrade. Once you get a new score in the top 15 songs, the full weighting could come back.

Once the inactivity has degraded a bit, perhaps your rank could be based on the top 16-30 songs, etc? Could perhaps be tier based, I'm unsure exactly what would work best. Maybe just lowering your rank in general by % amount based on inactivity could work too.

This would prevent old players who have lost their skill from having an artificially high rank, but would allow for people returning who have managed to gain their skill back to make all their old scores count again.

Anyone think this is a good idea/have any ideas of how to implement it better?

EDIT: Also, I feel like the rank it shows on your profile is supposed to be reflective of how good you currently are and not your peak rank, that's why its a ranking system to me it seems like that is the FFR developers intention. Because of this fact, it makes sense that scores from 7 years ago shouldn't hold up as well today and should have to be "re-proven". They could also add a "Peak Rank" stat or something for people, but I'd rather the rank just be more reflective of current skill.

My main idea is to simply add a second database of scores. The overall database will stay the same, but the idea would be that it would limit the scores that are currently held to "scores for the past year" instead of of all time. It'd be near sandbag proof because in order to sandbag, you would have to not play for a full year. And if there aren't valid scores that are 1-2 months old, you simply reject them. It would also give the ability to flag inactive accounts for "soft deletion". Moving it to a database of inactive users so they aren't pulled up in any user profile search, but can still be accessed if they are on a friends list, or if the username is known.

**Dinglesberry** · 07-3-2016, 10:46 AM

Re: "Ranking Degredation" for song scores

Originally posted by MikeShinoda12345

Tournaments aren't happening all the time, though. Wouldn't it be nice to quit knowing that your rank can still stand as a testament to how good you got? Or if you're an active player, that when you surpass someone in the Skill Rankings that you're actually improving and not just benefitting off a superior player's perceived rust?

Yeah, I see your argument for sure - scores don't change just because they are old, and people should be ranked for their peak skill. Ranking up isn't just for tournaments and divisions though, isn't the whole goal of this game to improve and improve your scores on harder songs? Ranks are just a side effect of that, but it's supposed to provide an accurate representation of where you stand relative to other people, at the current moment.

You are right though, there would have to be some way to get around this. That's why I was thinking something like getting a new top 15 score undoes this degradation - if you can prove you can still score as well as you used to, all the scores should be valid.

It's not a matter of benefiting off peoples rust, I just want the system to be more accurate. The ranking system is amazing, don't get me wrong. That being said, regardless of how the ranking system is, you can still get to top 100 or 1000 or whatnot if you are skilled, regardless of how good people were 7 years ago. You still need to be able to hit the notes, this would just make the progression more enjoyable and make your real rank more accurate, I feel.

EDIT: I guess at this point, the argument is should rank reflect peak skill or should rank reflect current skill.

**inDheart** · 07-3-2016, 10:50 AM

Re: "Ranking Degredation" for song scores

Originally posted by MikeShinoda12345

Tournaments aren't happening all the time, though. Wouldn't it be nice to quit knowing that your rank can still stand as a testament to how good you got? Or if you're an active player, that when you surpass someone in the Skill Rankings that you're actually improving and not just benefitting off a superior player's perceived rust?

this is basically my stance on this too. for tournaments, it's already required to have (iirc) 250 plays with 75 full song plays on your account but if you want to place people more accurately, you could also stipulate that x of those plays have to be from the past month or so and on a particular set of songs (say, roll your skill rating and play songs within 5 points of that as a benchmark)

plus older players already rust naturally a bit by not being around to play new song releases.

**MikeShinoda12345** · 07-3-2016, 11:12 AM

Re: "Ranking Degredation" for song scores

Originally posted by Dinglesberry

I guess at this point, the argument is should rank reflect peak skill or should rank reflect current skill.

That's a good way of putting it.

Introducing a ranking degradation variable to the current skill rating system is also kind of unprecedented, I think, and adds an expectation of rust to the metric. It sort of announces that the FFR community is prone to being inactive, if something like this has to be implemented.

Sorry if it feels like I'm just tearing into your idea! Something feels off to me about it, though, and I think I'm trying to pinpoint exactly why.

**psychoangel691** · 07-3-2016, 11:17 AM

Re: "Ranking Degredation" for song scores

I'll just note that for official tournaments we've held a strong stance on going by peak skill for everyone. There's no way to predict who will or will not boost back to their original skill and if we do it for one person we have to do it for all. It would cause all sorts of issues when someone does boost back and people start screaming about how well they had x scores why didn't you place them here yada yada yada.

So basically, this is not something that would ever be used for Official Tournaments.

**TC_Halogen** · 07-3-2016, 11:24 AM

Re: "Ranking Degredation" for song scores

Disagree with this because it promotes users having multiple accounts to where they can let one "degrade" and then use it as a point of entry to mask a higher account (and if we checked the stronger account as opposed to the weaker one, then this concept of degradation would be useless).

**Wayward Vagabond** · 07-3-2016, 11:35 AM

Re: "Ranking Degredation" for song scores

You guys can already IP check accounts and have linked accounts in the past so...

**TC_Halogen** · 07-3-2016, 11:49 AM

Re: "Ranking Degredation" for song scores

Originally posted by Wayward Vagabond

You guys can already IP check accounts and have linked accounts in the past so...

That's not quite the same though. Think about it for a moment: if a user has two accounts and is aware of this ranking degradation concept, they can show an older account they have - we'll arbitrarily say that this account is account A, and D5 based off of their level ranks, but gets dropped a division based off of not playing for a while... or so we thought. So, with this degrade, they're showing D4.

Again, these values are arbitrary.

They enter the tournament under a secondary account, account B, in hopes of not being detected/linked up at all with that previous account, because it shows better scores, and they get placed in D3. An IP check, along with other methods of verification, confirms that they are the same as account A, and they get placed back in D4, despite showing D5 stats due to having "degraded" level ranks.

...whoops, they've been playing offline engines and other offsite content to keep their skills sharp and are simply waiting until the tournament progresses to showcase the skill that never lost -- i.e. they're a D5 player, sitting in D4, and are playing other content continuously, only recording relevant scores that would otherwise keep them in a D4 status.

Anything even remotely close to this promotes sandbagging and will be absolute hell to accommodate for. It's shitty to have to accommodate for past skill, I get that - but, you run the risk of ruining the enjoyment of the tournament for a much larger group by inserting someone who later turns out to be a massive sandbagger. We all remember what happened when NeoMasterPie got put in D1 during the 7th Official Tournament - and yes, it's not exactly sandbagging in that case as much as it was sheer stupidity on the part of the tournament host, but... it's the same concept of skill not being effectively measured.

**Dinglesberry** · 07-3-2016, 11:49 AM

Re: "Ranking Degredation" for song scores

Originally posted by TC_Halogen

Disagree with this because it promotes users having multiple accounts to where they can let one "degrade" and then use it as a point of entry to mask a higher account (and if we checked the stronger account as opposed to the weaker one, then this concept of degradation would be useless).

I don't really see why someone would make an alt account. I mean, don't get me wrong, I see them all the time, saw one that was instantly promoted to division 5 after 20 games played, but it just seems so useless.

Do people smurf in the tournaments for an $80 prize pool (just based on the last tournaments prizes)? I mean, I really appreciate the tournaments even having prizes, that's amazing and definitely helps bring competition, but I mean, the amount of time and practice me and others put into FFR... I could make far more than $80 just spending that time working.

Perhaps its so they can say "I won the divison 2 FFR tournament in 1st place, I'm so great?" I just don't see how people making multiple fake accounts to hide themselves ranking up is that big of a deal in a game where the end result is your own personal skill, and that's the only thing you gain.

**TC_Halogen** · 07-3-2016, 11:51 AM

Re: "Ranking Degredation" for song scores

Originally posted by Dinglesberry

Do people smurf in the tournaments for an $80 prize pool (just based on the last tournaments prizes)? I mean, I really appreciate the tournaments even having prizes, that's amazing and definitely helps bring competition, but I mean, the amount of time and practice me and others put into FFR... I could make far more than $80 just spending that time working.

Unfortunately, yes. In the previous tournament, I had to disqualify a user playing under two accounts (one in D3, the other in D2). There's also been instances of players having stronger players play on their accounts in an effort to progress: Starlight-Pwn3 did this in D1 last tournament and was disqualified - marthalolz had Summerschool play on their account to beat Sei-Triplex in D1 on the 5th official tournament, Go_Oilers_Go had... someone play on his account in D3 of the 5th official tournament... the list definitely goes on.

"Ranking Degredation" for song scores

"Ranking Degredation" for song scores

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment