About this feature...The first time I saw it I was amazed. The resources it would take to run something like that and save all that information...astounding. And then I noticed something. It's just what percent you are to the highest score. That my friends is false advertising. If I wanted to figure out what percentage number I was to the highest score, I'd do it myself. Unless the feature actually starts tallying what it's supposed to, I suggest it be taken out.
-
0yes0%0no0%0Shirtless Adventures. Yea. -
Re: I did better then X% of people?! Oh really?
No it isn't, it's no more astounding than this message board. Just different.Originally posted by Kman3626
It's not trying to sell you something. Though, I do suggest that it be reworded, I was confused too.And then I noticed something. It's just what percent you are to the highest score. That my friends is false advertising. If I wanted to figure out what percentage number I was to the highest score, I'd do it myself. Unless the feature actually starts tallying what it's supposed to, I suggest it be taken out.
209 -
it just say how many people your better than... if you got 5th out of the 10 people who played a song it would say " you did better than 50%of the people"
RAVEnHEXa: Lip ring is because I want to be a professional piercer.
87x: more like.. professional goth.Comment
-
No, no that's not what it's saying.Originally posted by 87x
Let's say 10 people play a song, and here are their scores:
3k, 4k, 5k, 6k, 7k, 8k, 9k, 10k, 200k, 700k
(wow those lower ones are some really crappy scores.....assume that a newbie tries a hard song. but anyway)
So the high score is 700k, and you score 100k on the song. Even though you did better than 8 people, it's only going to say you did better than 14% of the people. Why? Because 100k divided by 700k is .14 (roughly)
What's inaccurate about this is that scores are not evenly spread out among all people who play the game. If they were, this would work. But they aren't, so it doesn't.Comment
-
Whoopee... it doesn't really matter anyway, considering ffr isn't at a competitive level yetComment
-
well... its saying you did better than tha percent of people not the percent of their scores...
RAVEnHEXa: Lip ring is because I want to be a professional piercer.
87x: more like.. professional goth.Comment
-
well, me, being a n00b, tried one of those OMFG IMPOSSIBLE classical songs, and......... got -1% better than anyone ever!
One of two things coulda caused that... either, i got a ton less than anyone else in existance (possible, to got a -1750)
OR
the system is buggy, but, WTH! i have better things to do than complain about such a cool game! BUHAHAHA just get on top 50 on some other song than tutorial (number 3! yipee!)Comment
-
Thank you Chardish for explaining things so that I didn't have to. You see, the system will say you did better then a negative % of people if you get a negative score, and that's just impossible. I'm not saying it should be scrapped, just reworded to say "You're this close to the top score on this song" or something.Shirtless Adventures. Yea.Comment
-
WHOA! cool! so, if i managed to stay allive long enough, than i could get better than -100% of all people?!?!?! that shoudl DEFINATELY go on high *cough* low scores
Comment
-
Grrrr... False advertising? On free game that is in beta? I understand your frustration but you were a little harsh. The feature will start to eventually compare and average scores as time goes on so that it will be accurate. For now I had to make it do what it does in order to atleast get it off the ground.
Stick around for beta .52 or .53 it will be working correctly by one of those.
Cheers,
SynthlightComment
-
I posted this several days ago on the comments of the .51 beta post, in response to nestlekwik's comment that searching through 70,000 records takes too long:
Better option:
Have the server, every 6 hours or whatever, tabulate the 70,000 scores for each song, and then sort them into groups of 700 each. Then retain the top score and the low score in each group.
This way you only have 200 numbers involved, and the game just tells the player which range of numbers his score falls into.
(I hope this makes sense?)
Synthlight, sorry if we seem harsh....we understand that it's free and in beta, but we're here because we love the game, and we love seeing it get better. Thanks for putting up with us
-ChardishComment
-
You wouldn't have to tabulate the percent every round, mang. Given that whomever is in charge of this outfit seems to have an inclination to computer science and data manipulation, something like this is prolly in the works, but:
At the beginning of each day calculate 100 values for every percentile (or 10 for every 10th, 20 for every 5th, etc). Given the number of people that play the game, these values won't change hour to hour by any significant margin, and you'll have accurate values for every percentile every day with only ONE tabulation per day per song. i doubt it'd eat up more than a few seconds of processing time.
As you go, store the values in a one dimensional array of size 100 (because you know there are only 100 percentiles, you don't have to use dynamic memory allocation). And since the memory can be searched arithmetically, you can do an interpolation search to find the percentile for any player.
Total cost:
-calculation of a value for 100 percentiles for each song once a day
-array of size 100 for every song
-interpolation search at the end of every game played
how efficient, and how stupid i feel for wondering why no matter how good i did on ff7 i couldn't change my percent much at all.Comment
-
Which is exactly what I said, you just put it in more technical terms and made it a daily thing instead of a 4-times-a-day thing.Comment
-
Yea, I didn't read your post :/Comment
-
I just don't like this feature because it makes me feel like I'm really bad at this game. Not saying that I'm not, but I think I'm better than more than 30%Comment
Comment