First flaw: Ballot stuffing. Some mapmakers vote countless times to increase their map's rating. Some people vote again and again with multiple accounts to spoil the rating of a mapmaker they dislike, as if the rating system were a popularity contest; and if the mapmaker notices this taking place, his only recourse is to stuff the ballot too, to get the average vote back to normal.

One vote per IP address might solve this and would at least curb such behavior. Or, maybe one vote per IP bundle: determine all the IP's and multiple accounts used by a player and only allow one of the accounts to vote from one of the IP addresses.

Second flaw: Map versions change, votes don't. If a mapmaker's 1.0 version is quite bad but the 1.1 version is quite good, the map's overall rating isn't based on the 1.1 version entirely, as it should be. Ladder games expire. Map votes should expire too somehow.

And since map votes don't change or expire as new, improved/updated map versions go public, this could lead to mapmakers trying to make a similar map go public twice. The old version stays the same. The updated version becomes a new map instead of a new version of an old map.

Third flaw: Comments. Some comments are abusive (for example, this review of Troll's Europe map: "This map is SO GODDAMN FUCKING COOOOOL ASS BITCHY DICKASS PICKLE NOOB"). Some are unrelated to current versions and could mislead players. A way for the voting system to delete comments automatically could be useful. The above review has 1 like and 28 dislikes. That should be enough to delete it.

Fourth flaw: Hard to change a vote. If you change your mind about a map or a new version comes out, the only way for you to change your vote is by playing the map again. Going to the map page to change your vote would be so much simpler.

I've written a lil' JS to add a "Rate this map" link to the map page, may share it if you will.

|> Some mapmakers vote countless times to increase their map's rating. Some people vote again and again with multiple accounts to spoil the rating of a mapmaker they dislike

Do you **know** this happens, or do you **suspect it might** happen?

|> one vote per IP bundle: determine all the IP's and multiple accounts used by a player and only allow one of the accounts to vote from one of the IP addresses.

Not possible. Player A logs in from IP 1 and 2, Player B logs in from IP 2 and 3. Player A votes on IP 1. Is Player B still allowed to vote? And if so, only from IP 3 or also from IP 2? Same questions if Player A had voted from IP 2.

|> Map votes should expire too somehow.

Most definitely not. You might want to consider a system were votes for previous versions somehow expire, but it makes absolutely no sense whatsoever to expire votes cast for the current version.

|> Third flaw: Comments. Some comments are abusive (..). Some are unrelated to current versions and could mislead players. A way for the voting system to delete comments automatically could be useful. The above review has 1 like and 28 dislikes. That should be enough to delete it.

Personally I'd strongly prefer a system where all current comments automatically get marked "This comment pertains to a previous version of this map". Outright deleting comments (especially considering an update might be very small) seems rather excessive. Similarly, I'd say you should hide downvoted comments, not remove them.

|> Fourth flaw: Hard to change a vote. If you change your mind about a map or a new version comes out, the only way for you to change your vote is by playing the map again. Going to the map page to change your vote would be so much simpler.

When you manage to cast a vote, close the rating page, then change your mind, you have a point. But when a new version comes out I sure hope people actually play the new version before changing their vote!

|> I've written a lil' JS to add a "Rate this map" link to the map page, may share it if you will.

Of course, my opinion is only my own, but may I ask you carefully consider the reason why Fizzer didn't add that option himself, before sharing such a tool with the world?

I have no idea why Fizzer didn't put that, except for users to not rate maps they never played, but that could've been done different (like put a link to rate if you played on that map).

@RvW

Ballot stuffing: I know it happens. Two players have told me they do this for maps by mapmakers they dislike. Another mapmaker told me he voted for his map a lot. A third mapmaker told me he voted for his map a few extra times after a guy with a grudge threatened to cast a lot of 1 votes for his map and the mapmaker noticed a significant drop in his map's rating. After calculating the drop, he told me that it was only possible if a handful of 1 votes were cast. He has a pretty good map, so those 1 votes in a short period of time are abnormal.

IP bundles: Maybe it is possible. Fizzer and Warlight's IP man would know better than us.

Expiring votes: Agreed, only for past versions.

Comments: Good idea. Version-specific comments would be best.

Vote changes: You have a point about new versions and not having played them. But for experienced players, having played the old version and simply seeing the new version is enough to know how the map has changed. But since not all players are able to do that, your idea is better.

Note that ballot-stuffing is impossible without multiple accounts. One player can't simply repeatedly vote on a map using the same account. So if people want to be such trolls as to go through the hassle of making extra accounts solely to vote on maps then there is no good way to stop them from doing so.

The whole IP concept is fatally flawed and such a system could never be instituted.

As for the complaints about outdated comments on previous versions, the mapmaker is allowed to respond to all comments so it is easy enough to counter such outdated information.

Vote changes don't require a second game, merely a return to a previous game. If going through your history is too much effort then you can simply create a single player game on the map and immediately forfeit, making the rating tab available. The requirement that a map be played prior to rating is a good one and shouldn't be removed.

I would suggest that the only reasonable course of action in regards to ballot stuffing is to notify Fizzer directly (fizzer@warlight.net) and explain what you think is happening and why you think it.

If he cares about someone getting their rating destroyed or built up artificially, then he can delete the offending votes and potentially suspend/ban the accounts involved. Chances are he won't unless it's egregious.

Otherwise, in the long term regular votes from regular players will eventually far outnumber the efforts of anything but the most dedicated trolls. If you get 200 people to play and vote about a map, 5-10 "stuffed" votes will be pretty meaningless.

1. It's IMPOSSIBLE to vote for any map a multiple times. If I could vote Whirlpool 10 times 5 stars believe me, I'd do it. And yes, I rate all of my maps 5 stars, but not more than once. That rating just means I'm satisfied with the map, as any map maker would be.
2. How about a system that lets votes for a previous version be worth half as much? E.g. if a map gets five ratings of 3 and five of 4, it's rated 3.5. But, when 1.1 comes out, and then 10 people rate it 4, the total rating should be (3.5+4+4)/3=3.8333 instead of 3.75.

One solution might be to ignore the votes from players that have played less than, say, 40 or 50 games.

Seeing as ratings are currently recorded for each player, the collation of the ratings could be very easily modified to include a condition such as the one above.

lets not do any limiting voters and ip blocking after all there might be more than one warlight player in the ip address.
Ballot stuffing is not much of an issue and i can see no better system for rating maps believe me there was a time when map makers didn't know what was felt about their map and i was around that time now they do let using the best available system.
Also if you obviously have the time to vote more than once , first making multi accounts then playing a game winning or surrendering to vote ,obviously has the time to waste and his very pathetic.

Iron, agreed on the IP aspect. That concept was pure folly. However, if it is determined that multi-account voting is a legitimate problem (I don't believe it actually is) then a game limit of even 10 or 20 MP games should greatly reduce the frequency of such a scenario.

@Incognito:

---

Just to be clear, my question about ballot stuffing was NOT intended as "I don't believe you", just trying to make sure we don't fix problems which don't exist. A sort-of solution could be to only allow people to vote on a map once their account is at least one day old (or one week or whatever). Sure, there'll be some people who still do it, but there should be far fewer (and it might even be possible to mass-ban accounts people have "on standby", plus their regular account, which should take the fun out pretty quickly).

---

What I meant is that your idea of "IP bundles" is fundamentally impossible; I don't care what info Fizzer has, it'll never work (or have such severe side-effects the cure is worse than the disease). Just try it yourself, on the example I provided.

To make matters even worse, if multiple people play at the same house (siblings, friends coming over or an entire school being on a single IP address), it's perfectly legitimate for multiple ratings to come in from a single IP. Even worse, because it's entirely plausible people on the same IP will be playing against each other, those ratings could be for the same map and be approximately the same. You can't even block half a dozen one-star ratings, for the same map, from the same IP, within seconds of each other, without potentially blocking legitimate votes.

I don't know about all that but being able to use half stars when you rate would be nice.

i.e 4.5 stars.

Rather than the plain arithmetic mean, the rating should be some combination of 50% mean and 50% median value.
Also, there should be a counter for how often a map has been played over a (to be defined) period of time, and a way to sort maps by that "popularity" value.

sadly, the best method to prevent ballot stuffing would be to make map voting a members only feature.. it would at least force someone to pay 30$ a vote...

map votes expiring makes no sense..

If I played a map 6 months ago and I liked the map, do I suddenly no longer like the map, without a new version being released..?

I don't know exactly how it works, but I can't imagine fizzer would of kept the old votes when new versions are released.. if that is indeed the case, Imho this needs to be changed..

fourth flaw, how can you have a valid opinion about the changes in the map without having played the map again in the new version..?

Perrin, I can confirm that old votes holdover through version updates. I updated two of my maps and the vote totals did not reset to zero despite major changes being made to one of the maps.

http://warlight.uservoice.com/forums/77051-warlight-features/suggestions/2878123-only-count-map-votes-from-people-who-have-played-a