Should Raw Scoring be used to calculate Grandtotal?

[CarrotCake94]

Originally Posted by tosh

Grandtotal can still be based off of the combo scoring formula for all I care, but credits definitely should be based off of raw score imo.

Same. I agree. Credits given should definitely be based on raw score. Don't really mind what happens to Grandtotal, but in my opinion, should stay as comboscore. Grandtotal is something that has been based off of combo scoring since the beginning of each players account. Suddenly changing that it to raw score might throw off the players trying to unlock Vet, or try to hit certain number of combo scores in a day. Grandtotal should be kept around as a relic if it is replaced.

I am open to the idea of having the grandtotal be based of raw score, but it'd be nice to keep it seperate and to have it start fresh with a different counter, like a raw grandtotal.

[scoper49]

Kind of taking suggestions from everyone so far. Raw scoring was implemented as a necessity to accurately define skill. Grand total doesn't affect anything skill-wise and the rank isn't even displayed on the front page's player card.

Combo scoring for GT should just be left alone since it would only negatively affect the current tokens and those who have/have not achieved them.

Finally combo scoring for credits should also remain the same because changing the speed at which you acquire credits whether slower or faster affects anyone trying to get them.

[SK8R43]

Originally Posted by rushyrulz

would pretty much void the value of a full combo. Aside from the raw score benefits on the higher-accuracy level

FC's already dont really mean anything anymore. Id rather get a really good raw score than an OK fc, anyday.

[blindreper1179]

Keep combo grand total, change credits to raw score.

[gold stinger]

if grandtotal is determined by raw scoring

rip Hakulyte's 100,000,000,000 GT.

voted no fyi. I like combo scoring for GT. If it was swapped for raw scoring, not only would it be tons harder to get veteran status, but it would mean that no one would be able to get into the top 10 for the highest daily GT earned. Correct me if I'm wrong about that 2nd point and it's already in effect.

[ilikexd]

Originally Posted by gold stinger

if grandtotal is determined by raw scoring

rip Hakulyte's 100,000,000,000 GT.

voted no fyi. I like combo scoring for GT. If it was swapped for raw scoring, not only would it be tons harder to get veteran status, but it would mean that no one would be able to get into the top 10 for the highest daily GT earned. Correct me if I'm wrong about that 2nd point and it's already in effect.

The first post explains why none of this is relevant or correct.

[gold stinger]

right.

Still voted no though.

It would mean zero reason to keep track or tally total score at all then unless there was a community-driven tournament that relied on it, but no one makes combo scoring tournaments anymore.
Keeping it at least for just credit distribution gives reason for people to exercise its use. Otherwise, it's literally just misused space on the R^3 results screen.

I don't have much of a ground for argument on it, but I think a lot of the support stemming for this is because FFR started with combo scoring, and they don't want to see it changed for the GT aspect of things. Credits I could understand though, because the whole shaboom is calculated under raw scoring now.

[Hakulyte]

Originally Posted by noname219

I have an idea : if possible, implement both scoring models and let the system chose the best score every time (to add to the grandtotal).

I'm going toward this idea too, it would give the insensitive to keep playing even after missing in the middle of the song. I would use the same idea for credits too, like if you missed in the middle of a song it uses the raw scoring formula to calculate credits, if your combo is really high and not the accuracy, you get the combo-scoring credits. This would leave some variety to the player in how to play and it's a win/win in terms of fixing combo-scoring multiplier value.

Need Option 3: Using both values and keeping the highest one. In theory, it should make farming GT slightly faster, but not enough to be a big deal.

[noname219]

Originally Posted by ThunderFlip

With some rough calculations here: The average quad mash on VS Boss Battle would probably go from something like 1.1-1.2m with combo scoring down to 900-950k (did a few test runs and calculated the raw total score). That brings the required time spent from ~14-15 hrs to ~17-19 hrs... which I think would make it much more difficult than previously, considering the amount of leeway time you are cutting. You'd have to really focus on PA the entire time or you'd just have to go for longer and take less breaks, either of which are a lot of strain after you've already been going for that long. Still quite possible, just a lot more of a pain in the butt.

I've just done 6 consecutive 10 minutes mini-session, each testing the two different scoring methods with 3 songs (Yoshi's Cookie, VS Boss Battle and Exciting Hyper Highspeed Star). Note that I tried different playstyle : cleaner runs and playing the whole file everytime for raw scoring and getting the biggest fc for combo scoring (only mashing when I felt was needed).

Yoshi's Cookie (could easily FC and get good PA)
Combo
7,594,095 points in 10 minutes
~816,569 points /minute (removing 2 seconds for writing score and restarting song) or 1,175,859,871 in 24 hours
Raw
238,245 x 31 = 7,385,595 points in 10 minutes
791,314 points/minute or 1,139,491,800 in 24 hours

Verdict : combo scoring is better by ~3%

VS Boss Battle (couldn't FC consistently and PA was average, I'm rusty)
Combo
10,861,655 points in 10 minutes
~1,189,232 points/minute or 1,712,494,511 in 24 hours
Raw
409,330 x 31 = 12,689,230 points in 10 minutes
~1,312,678 points/minute or 1,890,257,710 in 24 hours

Verdict : raw scoring is better by ~10%

Exciting Hyper Highspeed Star (did worse FC and PA-wise than VS Boss Battle, was mashing at parts on both tests)
Combo
8,372,865 points in 10 minutes
~906,808 points/minute or 1,305,804,217 in 24 hours
Raw
352,420 x 31 = 10,925,020 points in 10 minutes
~1,114,797 points/minute or 1,605,309,061 in 24 hours

Verdict : raw scoring is better by ~23%

Conclusion : if you manage to consistantly get a relatively clean high combo on the same song for 24 hours, combo scoring will be better for you, but not by much. The less clean your runs are, the better raw scoring is. Both scoring methods have their advantages but I would definitely chose raw scoring over combo scoring for a 24 hour session. RS, surprisingly, requires less concentration than anticipated and the fact that you can miss certain notes is satisfying - I was really doing messy in EHHS and the results are way better than I thought. Setting an autofail for combo scoring is more stressful, but you can take breaks by mashing certain parts or the entire song. Take those numbers with a grain of salt, a 10 minute playthrough do not scale very well to a 24 hour session.

Should change my vote.

[Silvuh]

To compare, it helps to multiply out the Raw Scoring model:

Combo: 550P + 275G +  55A - 310M -  20B + 1000C
 Raw: 1550P + 775G + 155A - 310M - 155B

and also to think in terms of how much each hit would take away from max GT:

Combo: -275G -  495A -  860M -  20B
 Raw : -775G - 1395A - 1860M - 155B

So perfects, goods, and averages would become worth 2.82 (31/11) times more (which means goods and averages hurt that much more), and a miss would hurt 2.16 times as much by itself (but a miss in the middle in combo scoring does more damage overall), and each boo hurts 7.75 times as much.

Players who trade their PA for combo by mashing would be hurt a lot, but also people who usually get FCs without mashing would be hurt a bit, because their near-perfect scores wouldn't be worth quite as much. (Basically, what lurker said.) So, changing to raw GT would help all the people who don't usually FC what they play.

And as noname219 pointed out, that because GT and combo are more measures of activity rather than skill, there's no need to switch them to raw scoring. But that alone doesn't inherently make the switch a bad idea.

Also as he suggested, the win-win of calculating both and then choosing the greater one sounds good. I'd think the only way it could hurt people is if they think it's unfair to them that people who don't play as well could be increasing their GT/credits just as much.

A few people said they'd want to keep the grand total the same but change credits to raw score, but it would be interesting to hear some arguments about why GT and credits should be treated differently.

Would be cool for the poll to cover all the options: combo GT / raw credits, raw GT / combo credits, and choosing the max of the two. (And perhaps a "neutral" option.)


Oh, also, about the session that noname just did: For the same reason that combo scoring was better for you on Yoshi's Cookie, anyone who can consistently FC and get a good PA on VS Boss Battle (over a long period of time) would prefer to have their GT and credits be based on combo scoring for that. But if they can't play consistently and they start to get more misses after a while, then that makes taking the max of the two methods even better.

[llyair]

Hmm so honestly I'm fine with any option, since personally I'm not too attached to my GT. I'm reading this thread to try understanding the pros/cons for each option, though. Also, it seems like the purpose of proposing this poll was to try making GT more relevant, or more reflective of, skill/PA?

If so, then I'm wondering if GT is necessarily irrelevant (strong word, just trying to say "something that should be phased out")? Already pointed out, but vRofl and Vet both depend on the current GT formula, so how would those requirements change if GT formula changed, if at all?

Just considering, since the strategy for racking up GT would change depending on a user's skill level (it might be better to play the densest song you can PA rather than quadmashing Boss Battle)? I got similar results to ThunderFlip here.

Given a PA spread for maybe D3? of something like (159-17-0-28), the combo score is 1,356,250, whereas the raw score is 1,140,070. Getting to 1billion would take the combo scorer 12.3 hours vs. 14.63 hours for the raw scorer. It becomes even harder for the second PA set for maybe D2? with (300-109-0-75). With combo scoring, it would take 13.12 hours to reach 1billion, but 21.28 hours with raw scoring. (Ehh alright, made lots of assumptions on PA spread and not taking any breaks).

Also, if the GT formula changed, what would the implication be for people who have already reached huge GT milestones? The very meaning of GT would also change somewhat (skill vs. effort), so would that make past achievements seem any less significant, since they would be compared to a different definition in the future? Or, depending on the GT amounts in using combo vs. raw scores, would top ranked GT users need to play even more to stay ahead (which would be the case if it becomes easier to get GT points)?

Given a 1000-note song, raw scoring would give 7500 fewer points than combo scoring if the PA is 6g clean. For any score with good PA basically, combo scoring would give more points. On the other hand, this is where raw scoring gives more points:

In your same 1000-note song, having 1 miss at note 900, all other notes perfects, would already give 99,300 points less using combo rather than raw scoring. Missing at note 500 would give nearly 500,000 points more using raw scoring, out of a total of ~1,550,000. Not that everyone is always missing right in the middle, but ~500,000 out of ~1,550,000 seems significant enough to consider? Especially if the "third" option is possible, how much easier would it be to get GTs this way?

Anyways, just agreeing with most things said already haha, it's understandable where they're coming from, and just restating some questions I had to consider, even though they might not be hugely concerning or be all that applicable.

[Red Blaster]

Here's the question everyone should be asking: How easily is this implemented? I know preventative measures have been taken since the last fiasco, but look what happened last time something combo score related was changed

[PrawnSkunk]

Originally Posted by Red Blaster

Here's the question everyone should be asking: How easily is this implemented? I know preventative measures have been taken since the last fiasco, but look what happened last time something combo score related was changed

The effort to switch or combine a mixture of scoring systems is extremely trivial. Velocity's mistake was attempting to delete scores where a trainer was used to generate impossible totals without a proper testing environment. This should not influence your vote for any future updates. Zageron has deployed a development server which all developers have access to, in addition to our weekly backups.

[TC_Halogen]

Originally Posted by PrawnSkunk

The effort to switch or combine a mixture of scoring systems is extremely trivial. Velocity's mistake was attempting to delete scores where a trainer was used to generate impossible totals without a proper testing environment. This should not influence your vote for any future updates. Zageron has deployed a development server which all developers have access to, in addition to our weekly backups.

!!!!

Praise an omniscient being of some sort! That's great news.

[Red Blaster]

Originally Posted by PrawnSkunk

The effort to switch or combine a mixture of scoring systems is extremely trivial. Velocity's mistake was attempting to delete scores where a trainer was used to generate impossible totals without a proper testing environment. This should not influence your vote for any future updates. Zageron has deployed a development server which all developers have access to, in addition to our weekly backups.

I was just curious more so than anything else if it was akin to merely flipping a lightswitch between combo or raw or if it was something more in-depth. As I said I'm aware that preventative measures have been taken to maintain site stability if a worst case scenario occurred.

[TheSaxRunner05]

What if the multiplier was slightly higher so it was slightly easier to get credits than before? Like 35x?

Someone did some mash math earlier which vastly underestimated the amount of boos mashing achieves.

Honestly, I think it should be left how it is, but I don't mind if there was a change. It makes logical sense that credits and score would be based off the same measure.

It would also further de-incentivize mashing. I may have improved more quickly had it not been for combo scoring in itself, I've mashed a lot of shit in my day, including a WWE FC.

After reading everyone's posts I'm conflicted.

[emulord]

Pure mashing gives you something like 90% of max from Combo scoring, 65% from raw scoring.
I like the idea of mashing your way to a billion in a day. It makes it more accessible to lower level players, they just have to want it really badly. Raw high scores have already de-emphasized mashing to the point that no one will actually do it unless they want HereWeGo.

The "bonus credits from FC" is fine honestly. Credits dont mean much and it gives a nice happiness boost from FCs, which is good game design.

[FF_rules]

Did anyone mention that in the new scoring system [Boos] are significantly more taxing on your Score than the combo system?
Solution for raw scoring formula: (Change the values of Good/Average/Miss so they aren't bringing down your total score compared to a standard FC) The problem here is simple; The formula was made to equalize a AAA in Raw scoring to Combo scoring, leaving everything else to be scaled in the raw scoring method with 0 regards to a combo formula. The 75%/50% can be changed; it's just a random number that sounds reasonable.
Change (Good x 25) to (Good x [1000/31]32.258...) Increases raw scoring of good up to 75% of a FC combo score variant
Change (Average x 5) to (Average x [555/31]17.9032...) Increases raw scoring of average up to 50% of a FC combo score variant
change (Boo x 5) to (Boo x [20/31]0.64516...) Normalize Boo scoring to be equal to the FC combo score variant

31 x ( Perfect x 50 ) + ( Good x 25 ) + ( Average x 5 ) - ( Miss x 10 ) - ( Boo x 5 )

1550 - 550 = 1000 (You don't gain points for getting a perfect)
775 - 275 = 500 (You lose 500 points in raw system vs combo system)
155 - 55 = 100 (You lose 900 points in raw system vs combo system)
310 - 310 = 0 (You gain X points in raw system vs combo system)
155 - 20 = 135 (You lose 135 points per boo you get assuming a FC)

vs

(Max Combo x 1000) + ( Perfect x 550 ) + ( Good x 275 ) + ( Average x 55 ) - ( Miss x 310 ) - ( Boo x 20 )

550
275
55
310
20

Revised formula: 31 x ( Perfect x 1550/31 ) + ( Good x 1275/31) + ( Average x 1055/31 ) - ( Miss x 10 ) - ( Boo x 20/31 )
This formula assumes 100% comparison from Raw scoring to Combo scoring for Perfects/Goods/Averages/Boo. The only thing different in the formula would be how Miss changes the overall score. In this scenario; Revised scoring would give you Combo scoring totals even if you missed a few notes (so basically a buff to low PA scorers). Explanation on the numbers; Goods/Averages had their points changed so they're worth as much as they are in a full combo score variant.

If we want to go one step further and create a miss multiplier that will simulate combo scoring; the lowest number we can go for would be 1310/31. In this scenario; it assumes you miss the final note on a FC. I personally feel we shouldn't be making boo's worth more than they are currently though (Reason: MashMashRevolution). And through powers of basic algebra; we can change the formula to...
Revised Formula v2: ( Perfect x 1550 ) + ( Good x 1275) + ( Average x 1055 ) - ( Miss x 1310 ) - ( Boo x 20 )
-And now it looks like the old combo scoring formula *cough*
~I'd probably agree with Revised Formula v2 so long as the miss multiplier got exceptionally heavier, and if we agree that mashing was not the intended method of playing this game; maybe we can increase the boo multiplier too.

Edit: If we give a Miss equal value to X perfects; we can balance out a raw scoring formula that still caters to lower PA scores and still maintain a comparable FC score vs Raw score.