Metaverse Decay: what you want to know

from Stardock Forums
There have been a number of posts here recently regarding the metaverse score decaying and what it means. Since T-Man was kind enough to post the exact decay in the sticky on metaverse submission problems, I wrote a little program to let me simulate the decay. I ran the simulations with the assumption that a given player would submit his scores regularly over a period of years and that for simplicity's sake he would always make the same score. I ran the simulation with a number of average scores, and a number of submission frequencies.

Here's what I found:
The overall decay curve is constant. A person who submits higher average scores and/or at a faster average rate will see higher peaks and a sharper slope but when the metaverse score peaks, decays, and increases again is strictly time dependent.

Your score will initially peak during the 10th month of submissions.
It will decay gradually over time until the 21st month, decreasing to about 93% of your peak score.
It will then gradually increase over time, finally surpassing your 10 month peak in the 47th month of submissions.
If for some reason you're still playing this game after 4 years time, it will continue to increase from that point out.

So, the decay algorithm used for the metaverse scores does not favor the newbie as was feared, nor does it favor the long timer, either. Assuming any two players of equal skill and rate of play, the newbie will find that his score approaches the old timer's in about 6 months time. Depending on just how long the old timer has been playing, the newbie might slightly surpass the old timer for a brief while, but the decay algorithm will catch up to the newbie shortly as well and keep things on an even keel.

*****************************************

Now, what happens to your glory when you shelve this game?

Your score will decay for 10 months from the time you quit submitting, bottoming out by the 11th month (makes sense given that the maximum decay occurs for scores greater than 290 days in age).
That final, bottomed out score is obviously dependent on how long you were submitting prior to quitting.
If you're looking to see that bottomed out, effort free score resembles your initial 10 month peak from your glory days of active playing then I hope you're in this for the long haul: you'll need to keep submitting games at the same rate for about 8 years to have your score level out there.

*****************************************

If anyone wants to play with program (it's just a simple console app that generates a log file with the data) you can get it here: http://filebox.vt.edu/users/channum/files/metaverse_simulator.zip



[Message Edited]
1,149 views 12 replies
Reply #2 from Stardock Forums Top
Whoops, discovered a minor glitch in my calculation for the decay after you stopped submitting, the zip's been updated now.
Siege of Centauri
Reply #4 from Stardock Forums Top
This does not count the "Version decay" of the scores, however which is far more formidable.
End of quote


There is no official version decay of scores. They allegedly recalc'd old scores under the new formula - some people think they just halved things, some people think they penalised, some people just don't know or care. Going by FB's comments, there is unlikely to be much in the way of any adjustments to the scoring formula, so it's all moot anyhow for a formula that takes 10 months to come to fruition on a set of scores that's a little over 2 months old at the maximum.
Reply #5 from Stardock Forums Top
Nice work, CM! :HOT:
Thanks for the clarification! I was wondering what the heck was going on.

~SDC~
Reply #6 from Stardock Forums Top
Oh ho. Thanks for the clarification. I had speculated that after each new update was released, there would be a recalculation of the scores from prior versions. Glad to hear that that is not the case.
Reply #7 from Stardock Forums Top
When I read T-Man's post, I had a different understanding. He said that your old scores decay ONLY IF you aren't an active player. If you remain active, your older scores don't decay. Are you sure that old scores always decay?
Reply #8 from Stardock Forums Top
Yes, in fact, I'm wondering how you missed T-man correcting people with that impression.

Posted by T-Man in the Metaverse Submission Problems thread:
I guess my explantion of the scoring system could have been worded better. Scores will age regardless of whether you have submitted anything recently or not. But in order to stay ranked in the Metaverse you will obviously need to continue to submit new games or your scores will eventually begin to fade.
End of quote

Your metaverse total score won't fade *much* if you continue to be an active player (nor will it increase much beyond a certain point without a LOT of effort on your part), but individual scores all decay regardless of whether you're active or not.

[Message Edited]
Reply #9 from Stardock Forums Top
I think the Metaverse Score Fading feature will favor those that have more time to play. That seems very obvious to me. It hurts the players who have other thinks in life to deal with.

Reply #10 from Stardock Forums Top
Apollo43, the flipside is how long are any given one of us going to be around playing this game actively? Assuming a more or less even submission rate, your score is going to accumulate for *10* months before it plateaus (If I get particularly bored, I'll rewrite my calculator to do some fancier simulations, such as declining submission frequency, to see what that does to the progression curve).

That's 10 months of your score increasing before the decay really catches up to and most people play a game for one, two months tops. The formula ensures that when someone picks up the game 9 months down the road, they're able to fight their way into the metaverse.

Now, I do think the 20% bottom is much too low because it puts you in a position of having to increase your mean score and/or submission frequency just to stay even for *three* years, and that, assuming somebody remains active with the game, is a bit ridiculous.

Putting the basement at 30% still causes your score to level off by 10 months, but the decay only causes it to dip for 2 months versus 3 years before it gradually increases forever until you stop submitting (it also lets your final, fully decayed score look a little more impressive ;p )

E.g., a player submitting games once a week at 1000 points sees his score do the 10 month peak @ 4119 before its decline whether the basement is 20% or 30%.

But, with a 30% basement, it regains that 4119 in the 13th month and continues to increase from there out but it's a very limited rate of increase: whereas it only takes 10 months to get to 4100, it takes almost an additional 3 years to get another 1000 points.

Compare that to the 20% basement which makes the player play for three years just to get back up to his 4100 score and play for another six years to get that whole 1000 extra points.
Reply #11 from Stardock Forums Top
I see now - I didn't read every post in that thread; I just looked for T-Man's first description.

There still needs to be some way to eliminate poor scores from your overall rating. Someone who plays 5 months, stops, and then plays another month should not be at a disadvantage compared to someone who just stared and has been playing for a month.
Reply #12 from Stardock Forums Top
Well, they're not at a disadvantage, they're still at a huge advantage over the 1 month player.

All things being equal, it takes *6* months for the new player to catch up with an active "old timer".

If you take a break from submissions, that's your deal. Significant decay doesn't kick in until the 3 month point, but as a 5 month player, most of your scores are already on their way to having gray hair.

What should be changed, though, is the decay basement of 20% to 30% and T-Man said he'd pass that on to Brad (hopefully he won't mention the source ;) )