User Pass
Home Sign Up Contact Log In
Forum > Goal Line Blitz > Effective Levels - everybody above actual level "paradox"
Page:
 
Fumanchuchu
fonky
offline
Link
 
http://goallineblitz.com/game/roster.pl?team_id=1337

Plenty of down level players here.

I think there are more than enough heavily declined dots in service out there to account for the recent EL surge. Players can only be up by so much, but they can be WAY down from decline.

Including decline in the calculations is probably not the best thing.

 
Deathblade
offline
Link
 
Originally posted by Fumanchuchu
http://goallineblitz.com/game/roster.pl?team_id=1337

Plenty of down level players here.

I think there are more than enough heavily declined dots in service out there to account for the recent EL surge. Players can only be up by so much, but they can be WAY down from decline.

Including decline in the calculations is probably not the best thing.



which makes it a good thing that decline isn't included in the "calculations" of the unchanging numbers
 
Mightyhalo
offline
Link
 
Originally posted by Deathblade
which makes it a good thing that decline isn't included in the "calculations" of the unchanging numbers


So your saying they took a snapshot of the builds in season 12 and EL is continually based off those numbers?
 
Dbride59
offline
Link
 
All my dots are looking great because I am the best
 
Firenze
offline
Link
 
Originally posted by Dbride59
All my dots are looking great because I am the best


lolmulti
 
TigerBloodJoe
offline
Link
 
http://goallineblitz.com/game/player.pl?player_id=1623186

not so effective level
 
jdbolick
offline
Link
 
Originally posted by Darkstrand
it does indeed change and frequently every time your player value changes

If the EL calculation itself was continuously updated then a dot's number would change even if his attributes did not (between levels & no training rollover). That doesn't happen. It seems extremely likely that EL is based on a formula developed around S1/S2/S3 dots who weren't built nearly as well as their current counterparts.


Originally posted by Mightyhalo
So your saying they took a snapshot of the builds in season 12 and EL is continually based off those numbers?

Deathblade is correct. If that wasn't the case, then Bort's decision to include decline in individual EL calculation would have obliterated the overall formula.
Edited by jdbolick on Nov 28, 2010 12:27:18
 
Darkstrand
offline
Link
 
Originally posted by jdbolick
Originally posted by Darkstrand

it does indeed change and frequently every time your player value changes

If the EL calculation itself was continuously updated then a dot's number would change even if his attributes did not (between levels & no training rollover). That doesn't happen. It seems extremely likely that EL is based on a formula developed around S1/S2/S3 dots who weren't built nearly as well as their current counterparts.


it's kids like jdbolick that make me fear for the American education system.

yeah theres no way a system could ever update one thing when it changes but not another thing until something else changes.
Edited by Darkstrand on Nov 28, 2010 13:11:54
Edited by Darkstrand on Nov 28, 2010 13:11:03
 
jdbolick
offline
Link
 
Originally posted by Darkstrand
it's kids like jdbolick that make me fear for the American education system. yeah theres no way a system could ever update one thing when it changes but not another thing until something else changes.

If EL is measured by comparing SP value of the build versus a continuously updating "average" at that position, then the individual's EL should change based on that average even if that individual doesn't level or roll over any attributes through training. That doesn't happen. Your EL stays constant unless you level or you roll over an attribute through training.
 
Darkstrand
offline
Link
 
Originally posted by jdbolick
Originally posted by Darkstrand

it's kids like jdbolick that make me fear for the American education system. yeah theres no way a system could ever update one thing when it changes but not another thing until something else changes.

If EL is measured by comparing SP value of the build versus a continuously updating "average" at that position, then the individual's EL should change based on that average even if that individual doesn't level or roll over any attributes through training. That doesn't happen. Your EL stays constant unless you level or you roll over an attribute through training.


ahahhaha "should" I think it's cute when kids apply random axioms they make up to disprove things
 
TheGreatPuma
offline
Link
 
Originally posted by jdbolick

If EL is measured by comparing SP value of the build versus a continuously updating "average" at that position, then the individual's EL should change based on that average even if that individual doesn't level or roll over any attributes through training. That doesn't happen. Your EL stays constant unless you level or you roll over an attribute through training.


Not if the EL is stored in a field in the database instead of calculated. Perhaps when attributes change, it triggers a re-calc of the effective level field. We really have no idea how it works so speculation seems pointless.
 
mandyross
offline
Link
 
Originally posted by jdbolick

If the EL calculation itself was continuously updated then a dot's number would change even if his attributes did not (between levels & no training rollover). That doesn't happen. It seems extremely likely that EL is based on a formula developed around S1/S2/S3 dots who weren't built nearly as well as their current counterparts.


It's possible that the continuous updating gives changes that are too small and slow-moving to notice with a limited set of players.

I was monitoring the limit for effective level 75 on a guard last season, to see if my guy could get there at lvl 58 or not. And I could have sworn that the cut off was 997.5 or so. Now a quick search shows the cut off to be at least 998.86

http://goallineblitz.com/game/player.pl?player_id=1487581

Of course, my hazy memory is hardly conclusive proof. For proper proof, it would be necessary to find an effective lvl 74 player with a SP value of his build higher than those mentioned in ryan grant's "fastest to 75" thread:

http://goallineblitz.com/game/forum_thread.pl?thread_id=3892630&page=1

That's a lot of searching through players though. It's probably easier to confirm this by asking in one of those Q&A threads exactly how effective level is calculated, and how often the script is run to update it.

As for the initial post, I'd say that all current active players have high effective levels as their builds are much better now than the first few seasons, and they are still being compared to all players, including poorly built ones of agents who subsequently quit. Of course a few positions like C are anomalous due to the introduction of archetypes, but that just goes to show how ridiculous their ALGs were originally.
Edited by mandyross on Nov 28, 2010 13:34:42
 
BigRatUno
offline
Link
 
I wasn't around when this got implemented so I'm not sure it's purpose, but it seems to be an attempt at ranking the quantity of SP in a build at a given level and a given position.
It seems to me that the most direct way to have done this would have been to constantly calculate both the average SP in all players of a certain position at a specific level and the standard deviation for that population thereby giving the ability to calculate a z-ratio (Players SP - Average SP in all players of that position at that current level)/Standard Deviation of that group. That zratio could then be interpreted on a probability scale of 0 to 1 corresponding to the normal distribution.
The scale would then tell you what percentage of other players were above or below your player with .5 being the average.
 
jdbolick
offline
Link
 
Originally posted by Darkstrand
ahahhaha "should" I think it's cute when kids apply random axioms they make up to disprove things

I understand why you're defensive, but it's not my fault that you embarrassed yourself. In the future when you say something dumb, it's far better to simply acknowledge the mistake and move on. By trying to make excuses, you just caused yourself to look worse.


Originally posted by TheGreatPuma
Not if the EL is stored in a field in the database instead of calculated. Perhaps when attributes change, it triggers a re-calc of the effective level field. We really have no idea how it works so speculation seems pointless.

Actually Deathblade's point settled the issue once and for all, since any recalculation including decline would have produced an enormous change in every active player's EL, yet we know that didn't happen.
 
Darkstrand
offline
Link
 
Originally posted by jdbolick
Originally posted by Darkstrand

ahahhaha "should" I think it's cute when kids apply random axioms they make up to disprove things

I understand why you're defensive, but it's not my fault that you embarrassed yourself. In the future when you say something dumb, it's far better to simply acknowledge the mistake and move on. By trying to make excuses, you just caused yourself to look worse.


Originally posted by TheGreatPuma

Not if the EL is stored in a field in the database instead of calculated. Perhaps when attributes change, it triggers a re-calc of the effective level field. We really have no idea how it works so speculation seems pointless.

Actually Deathblade's point settled the issue once and for all, since any recalculation including decline would have produced an enormous change in every active player's EL, yet we know that didn't happen.


hahaah stick to rhetoric kid. While you're average there it's pretty clear you'll never be getting an A (or even a B+) in any of your math classes.

"since any recalculation including decline would have" ahahahah invent some more baseless axioms... theres no possible way they could have possibly made the system so it wouldn't factor in decline for overall effective levels...

When you get older and more experienced and maybe even someday get that B in a math class (provided you get a tutor), you just might learn you can't go around randomly making up axioms for logical deduction.
 
Page:
 


You are not logged in. Please log in if you want to post a reply.