The new system is nice but as said DEWO at the beginning of the thread, i think it would be nice doing something using different systems at the same time. Something like (degressive order of priority):
- A sort is done to see last created/edited build by the "24 hours" - "7 days" - "30 days" - "All time" menu
- 5 stars note. Eventually, can be ponderated with (number of votes)/(min_v) if there are less than min_v votes. The min_v number should be a minimum votes to reach before considering the build has been seen enough to consider it "decently" rated with equity. This supposes a different min_v for each range of time, for instance : min_v=5 for 24 hours, min_v=10 for a week, min_v=30 for a month, min_v=100 for all time.
- List by number of votes
- List by number of views
The first sort allows no to see only "old" builds and already exists. The second one is the 5 stars note and is obvious to understand. The ponderation allows builds with massive votes not to be unfairly relied to the end of the queue because someone voted 5 stars alone on a new build. The hard cap (min_v votes) allows all builds with more than min_v votes to keep entirely their notes.
Once this filters applied, in a same note, posts are listed by number of votes, and then by views.
This system allows new good builds to appear easily in a short time ranking (24 hours, 7 days) but also to get the "best ever rated" (considering notes and number of votes) builds while going with "30 days" or "All time".
I must have forgotten some situations or cases when this system would fail or have to be adapted, but it seems fair to me when i look at it right now :)
- A sort is done to see last created/edited build by the "24 hours" - "7 days" - "30 days" - "All time" menu
- 5 stars note. Eventually, can be ponderated with (number of votes)/(min_v) if there are less than min_v votes. The min_v number should be a minimum votes to reach before considering the build has been seen enough to consider it "decently" rated with equity. This supposes a different min_v for each range of time, for instance : min_v=5 for 24 hours, min_v=10 for a week, min_v=30 for a month, min_v=100 for all time.
- List by number of votes
- List by number of views
The first sort allows no to see only "old" builds and already exists. The second one is the 5 stars note and is obvious to understand. The ponderation allows builds with massive votes not to be unfairly relied to the end of the queue because someone voted 5 stars alone on a new build. The hard cap (min_v votes) allows all builds with more than min_v votes to keep entirely their notes.
Once this filters applied, in a same note, posts are listed by number of votes, and then by views.
This system allows new good builds to appear easily in a short time ranking (24 hours, 7 days) but also to get the "best ever rated" (considering notes and number of votes) builds while going with "30 days" or "All time".
I must have forgotten some situations or cases when this system would fail or have to be adapted, but it seems fair to me when i look at it right now :)
Even if other people make ****, that does not make your actions better !
Attack on Alahric, August 4th - 5th
"NO NEW COMMENTS!"
Annie - from 93% to 90%
Ryze - from 85% to 82%
Veigar - from 85% to 80%
Mordekaiser - 79% to 77%
Taric - 74% to 70%
Discuss, all over 1 day, all of my guides lost a great deal of percentage!
I need to know why, see my problem? We need to enforce comments, all I did between those days was:
Ripping a Mordekaiser guide to shreds for it's lack of description, why Warmog's is a bad item selection, how optaining "immortality" or being "unkillable" is immpossible, and stating the fact that Mordekaiser doesn't compensate for anything etc. It had to be done, someone had to say it: TAKE CAUTION IF YOU ARE LINKALERGIC, NOT MY GUIDE!
Or.. well the other conclusion is that I updated my pictures to show tooltips, but that's irrational ->
I mean it's an improvement, however the same thing happenened when I added pictures.
"NO NEW COMMENTS!"
Annie - from 93% to 90%
Ryze - from 85% to 82%
Veigar - from 85% to 80%
Mordekaiser - 79% to 77%
Taric - 74% to 70%
Discuss, all over 1 day, all of my guides lost a great deal of percentage!
I need to know why, see my problem? We need to enforce comments, all I did between those days was:
Ripping a Mordekaiser guide to shreds for it's lack of description, why Warmog's is a bad item selection, how optaining "immortality" or being "unkillable" is immpossible, and stating the fact that Mordekaiser doesn't compensate for anything etc. It had to be done, someone had to say it: TAKE CAUTION IF YOU ARE LINKALERGIC, NOT MY GUIDE!
Or.. well the other conclusion is that I updated my pictures to show tooltips, but that's irrational ->
I mean it's an improvement, however the same thing happenened when I added pictures.
Use Edit option :) dont double post :PPP
Well it is a change to the rating system not the negative votes system Alahric.
After the change my Trynda Guide was 6/6 and had 100%, it is at 73% atm. Negative votes will still happen.
But i had my 1st succes. Som1 DID COMMENT his -1 vote (on my Jungle WW guide).
The explanation was **** imho (mainly said that ppl that use this guide sucks when he plays with them.) but still...
Well it is a change to the rating system not the negative votes system Alahric.
After the change my Trynda Guide was 6/6 and had 100%, it is at 73% atm. Negative votes will still happen.
But i had my 1st succes. Som1 DID COMMENT his -1 vote (on my Jungle WW guide).
The explanation was **** imho (mainly said that ppl that use this guide sucks when he plays with them.) but still...
DEWO wrote:
Use Edit option :) dont double post :PPP
Well it is a change to the rating system not the negative votes system Alahric.
After the change my Trynda Guide was 6/6 and had 100%, it is at 73% atm. Negative votes will still happen.
But i had my 1st succes. Som1 DID COMMENT his -1 vote (on my Jungle WW guide).
The explanation was **** imho (mainly said that ppl that use this guide sucks when he plays with them.) but still...
Anyway, all of my guides from the night between 4th and 5th August got heavily downrated! This isn't just -1 to a few guides, this was at a large scale I mean 3% on my Annie Guide is alot! You need more than 4 votes to do that. I wouldn't be suspicious if this happened gradually, but this seems to be more like hate negative votes which didn't really have anything to do with the guides. That's the real issue. Mobafire users seems to be just spewing jealousy at competators, that's what happened to your Tryndamere guide. The truth is that if they were to list reasons for negavite voting, they wouldn't actually find a single good arguement for it.. so in the end nobody but spammers deserve -1.
So if we fix this issue which we have to, then we are back at where popularity declears the best guide which is unfair for new users. This is why 5-stars rating, on everything like: Buildsetup, writing, layout etc, eould be more appropriate. This way everyone who sees a guide will be encouraged to vote aswell. I suggest enforced comments with listed reasons for negative voting or a feedback folder for a guide, if these are all false or unvalid one should be able to report it.
Is it still unrelated?
Btw: Obvious troll is obvious, every ******** post you make DEWO..
The annie guide has come down by 2% to be precise, however this is not due to heavy down voting, it is due to the new scoring formula. With this formula all votes are now weighted. This means it is no longer a straight up percent, its a slope on a curve.
With the new formula its much more difficult to get to 100% or 0% rating than before. Your build was reduced by 2% due to the new formula, other builds were reduced by more. But the new scoring system displays every build more fairly now. You will see that all those 2/2 6/6 builds are way below you on the list. This is because your build is considered "more accurate" due to the amount of votes. So although a 6/6 might technically be rated 100%, it instead will be rated much much lower due to the inaccuracy of so few votes.
This formula has been tested and used by very large sites. There is not, nor will ever be a perfect scoring system but this one is about as fair as it can get.
With the new formula it is pretty much impossible to get 100% but keep an eye on the list, you will find its going to be hard for other builds to rise above you. I am confident if they do, and they stay there, it will be deserved. Until then, enjoy your perch on top and don't sweat 2%, in relation to other builds, you all dropped in score fairly, or rose in score fairly.
With the new formula its much more difficult to get to 100% or 0% rating than before. Your build was reduced by 2% due to the new formula, other builds were reduced by more. But the new scoring system displays every build more fairly now. You will see that all those 2/2 6/6 builds are way below you on the list. This is because your build is considered "more accurate" due to the amount of votes. So although a 6/6 might technically be rated 100%, it instead will be rated much much lower due to the inaccuracy of so few votes.
This formula has been tested and used by very large sites. There is not, nor will ever be a perfect scoring system but this one is about as fair as it can get.
With the new formula it is pretty much impossible to get 100% but keep an eye on the list, you will find its going to be hard for other builds to rise above you. I am confident if they do, and they stay there, it will be deserved. Until then, enjoy your perch on top and don't sweat 2%, in relation to other builds, you all dropped in score fairly, or rose in score fairly.
"It doesn't matter if you win by an inch or a mile, winnings winning."
I would also like to add that unless there are any serious flaws found in the next few days with this formula it is likely the one we are going to settle on. From this point on your scores should not change anymore due to scoring formulas.
I would like to take this opportunity to apologize for all the score changing. We are pretty new to running sites like this and the score formula was on oversight on our part when we first started the site. We should have worked it out in the beginning but we simply didn't know any better. We learn as we grow and appreciate the understanding. Everyone has been awesome through all this, thanks for your support.
Regards
FlashJ
I would like to take this opportunity to apologize for all the score changing. We are pretty new to running sites like this and the score formula was on oversight on our part when we first started the site. We should have worked it out in the beginning but we simply didn't know any better. We learn as we grow and appreciate the understanding. Everyone has been awesome through all this, thanks for your support.
Regards
FlashJ
"It doesn't matter if you win by an inch or a mile, winnings winning."
@FlashJ
Thanks, I wasn't informed but so far I like that change because the system is a little more decent now :)
I mean when someone with 1 tip, not even about the guide, in their description has 100%.. it's just not right. I "overreacted" because it seemed like a conspiracy, again.. just a theory but it still happened right? But since you explained why I guess it's only fair ^ ^,
However, it doesn't change the supect about hate negative votes..
The only thing that I consider valid now is fixing that issue.
Thanks, I wasn't informed but so far I like that change because the system is a little more decent now :)
I mean when someone with 1 tip, not even about the guide, in their description has 100%.. it's just not right. I "overreacted" because it seemed like a conspiracy, again.. just a theory but it still happened right? But since you explained why I guess it's only fair ^ ^,
However, it doesn't change the supect about hate negative votes..
The only thing that I consider valid now is fixing that issue.
I do hope everyone knows we are hearing the complaints about the no comment negative voting and have discussed it at length. However this presents some unique challenges to overcome.
For one, simply by forcing the comment we are strongly discouraging negative votes while not equally discouraging positive votes. This will mean far less negative votes but then what does that do to the scores of all the builds? We are now back to the popularity contest rating as whoever has the most positive votes wins (since negative votes will be sooo few). So again the older established guides stay on top regardless of quality while younger upcoming builds have no chance to catch them in pure number of votes.
Also, as most authors are aware, for the most part votes are hard to come by on the site, discouraging voting will amplify this problem. We want to find ways to encourage more voting, not less.
We have yet to find a site that implements any kind of forced comment on down vote, examples like youtube don't force comments and their system works pretty good.
We have tossed this one around and around just like the score and have been unable to find a solution we are more happy with than what we have now. With the new formula our hope is a few negative votes won't hurt your score to nearly the same degree as it used to.
If anyone comes up with a unique way to make this work we are happy to listen and implement. But for now, a simple forced comment to negative vote will have an over all negative effect on the site as a whole, scoring and rating quality.
FlashJ
ps: I don't want this to sound like the option is off the table, its not. I just want everyone to understand why we haven't jumped on this. Its more complex than it seems on the surface.
For one, simply by forcing the comment we are strongly discouraging negative votes while not equally discouraging positive votes. This will mean far less negative votes but then what does that do to the scores of all the builds? We are now back to the popularity contest rating as whoever has the most positive votes wins (since negative votes will be sooo few). So again the older established guides stay on top regardless of quality while younger upcoming builds have no chance to catch them in pure number of votes.
Also, as most authors are aware, for the most part votes are hard to come by on the site, discouraging voting will amplify this problem. We want to find ways to encourage more voting, not less.
We have yet to find a site that implements any kind of forced comment on down vote, examples like youtube don't force comments and their system works pretty good.
We have tossed this one around and around just like the score and have been unable to find a solution we are more happy with than what we have now. With the new formula our hope is a few negative votes won't hurt your score to nearly the same degree as it used to.
If anyone comes up with a unique way to make this work we are happy to listen and implement. But for now, a simple forced comment to negative vote will have an over all negative effect on the site as a whole, scoring and rating quality.
FlashJ
ps: I don't want this to sound like the option is off the table, its not. I just want everyone to understand why we haven't jumped on this. Its more complex than it seems on the surface.
"Sooner or later you're going to realize just as I did that there's a difference between knowing the path and walking the path"
You need to log in before commenting.
<Administrator>