International logo contest/Vote on voting method

historical pages 2003 Wikipedia international logo contest (vote on voting method)
This was an international contest held from July 20 to August 27 2003, gathering 150 proposals. After the early proposal by Chuck Smith on October 12, 2002, the contest was first proposed on June 14, 2003 by Erik Moeller, who argued that the logo (adopted in January 2002 from the Logo suggestions) was unaesthetic, not international, and portrayed a text-only Wikipedia.

This vote is now closed. The winner is average voting with 21 votes. Further details on International logo contest/Vote instructions.

Picking a voting system for Stage 2


OK, let's vote on the voting system for the second stage (simple FPP voting -- option with the highest number of votes wins, only vote for one option):

General features of the vote that might be relevant to your decision on this issue:

The deadline for this vote is September 12, 20:00 UTC. This is to allow translators to translate the voting instructions before voting starts.

Average voting


Average voting (as used on article count reform, grade each logo 1(bad) to 5(good), winner has the highest score). This option currently has 21 votes and is in first place.

—Eloquence, DanKeshet, Maveric149, Tevi, Fantasy, Renke, Angela, MyRedDice, Pacific1982, Dhum Dhum, User:Anthere, Oliezekat :op, LDan, User:Stevertigo, TakuyaMurata, Gaucho, JDG, Bisho, CyberMaus, TwoOneTwo, Nanobug

  • (+) Allows voters to express different degrees of approval and disapproval. Example: if a voter really dislikes one logo, they can vote it "1" and all others "5".
  • (+) Easy to vote and count
  • (+)The final decision is easy to understand
  • (-) More complicated than approval voting and FPTP
  • (+/-) More strategic voting than Condorcet
  • (+/-) Strategically equivalent to approval voting. That is, if you want to maximize the impact of your vote, you would only give either one or five points to each candidate. If everyone does that, then the voting system is equivalent to approval voting anyway, which is simpler. (Article count reform vote showed that not everyone or not even the large majority of users act in this fashion.)
  • (-) Average voting only allows a limited number of gradations (five, in this case). You cannot express a preference between every pair of options.

Condorcet method


Condorcet method (rank the candidates, winner is the one preferred by a majority to every other candidate, if one exists). This option currently has 11 votes and is in second place.

Taw, { MB | マイカル }, Tillwe, AdamRaizen, Ryan Cable, James F., Toby Bartels, bdesham, Aoineko, RobLa, Gebu

  • (+) Allows voters to express their approval or disapproval of each candidate in relationship to each other candidate.
  • (+) Strong disapproval can be expressed as easily as in average voting by ranking a candidate low or last.
  • (+) least subject to strategic voting (the most effective vote is almost always the most sincere vote)
  • (+) no possibility of a majority preferring some candidate to the actual winner
  • (+) The basic system is at least as simple as average voting.
  • (+/-) You can rate two options as equal under Condorcet, e.g. A>B=C. This is strategically equivalent to tossing a coin between A>B>C and A>C>B.
  • (-) Some folks find this system confusing.
  • (-) It takes a long time to calculate the winner (unless it is automated) (see online tool available to do tabulation).
  • (+/-) There is the possibility of Condorcet paradox, where A > B, B > C, and C > A: if this system is chosen, then we will need to seperately decide which Condorcet resolution method we choose. This is discussed in #Condorcet cycle resolution. Some view this as an advantage, while others view it as a disadvantage.

Playoff voting


Playoff voting (All logos are paired against each another as in a cup. Each users votes which of the two logos that face each other that they think is best. The winner of each duel is paired against another winner until only one logo remains). Note: This option was added after several days of voting.


  • (+) quite simple
  • (-) takes very long time
  • (-) a lot of strategic voting
  • (-) easy to cheat: the organizers who decide on the pairings have a lot of influence on the outcome (unless the pairings are decided randomly).
  • (-) It is very unlikely that the number of candidates will be a power of 2.
  • (+/-) Element of chance.

Approval voting


Approval voting (vote for one or more candidates, candidate with the highest number of votes wins).

CGS, Dominus

  • (+) Simpler to vote and to count than everything except FPTP.
  • (+/-) Strategically equivalent to average voting (or is it? See above)
  • (+/-) Encourages voters to express tolerances over preferences
  • (+/-) Strategic voting necessary (and encouraged)
  • (-) Less expressive votes than everything except FPTP

Instant-runoff voting


Instant-runoff voting (rank the candidates and conduct a virtual run-off election, each time eliminating the candidate with the fewest first-place votes until one candidate has a majority of first-place votes).

Andre Engels

  • (+) No straightforward spoiler effect
  • (+) strategic voting often unintuitive.
  • (-) Highly prone to strategic voting by the clever (note: one strategic vote won't do much)
  • (-) In some cases lowering the ranking of a loser on some ballots may cause it to win. (ie, fails the monotonicity criterion). All voting systems are flawed... Well, this is one interpretation of en:Arrow's paradox... the real question is, which system is relatively best?

First past the post


First past the post voting (vote for one candidate, candidate with the highest number of votes wins).

  • (+) Very simple
  • (+) Most Americans are already familiar with it from political elections (elections in Europe are usually 2-stage not FPTP, what's a little less bad).
  • (-) Of all methods, this is the most prone to strategic voting.
  • (-) Prone to problems such as the spoiler effect.

Pro/con conversion


I converted the above stuff into pros and cons, bulleted. Here's the lucid prose I just excised, with commentary in italics:

Average voting: While commonly believed to result in similar strategies as approval voting, evidence from article count reform suggests this belief to be false at least in our case.

duplicates near-identical statement made in the "arguments against" section

Average voting: This guarantees (Wow!) a high level of transparency and makes it very easy for everyone to use their votes to the effect they find useful or necessary (summarise to "easy to vote"); that is, if they dislike a particular logo very strongly, they can exercise a high influence to prevent it from being successful, whereas they can also express moderate opinions or very strong approval. (repetition of point already made)

Average voting: If there are 4 different variants of a certain logo, you might have a clear preference among them, but you couldn't give them all a different score under average voting without seriously compromising one's ability to show a preference for others, and to make gradations among them. Condorcet does not have this problem.

statement of the blindingly obvious, rendered in a way that made it somehow very obscure - tried to improve.

Condorcet: Rank the candidates in order of preference, and the winner is the candidate preferred by a majority of voters to every other candidate. (how is a description of a system an argument for it?) Choosing one among 10+ ballots is vastly different from rating hundreds of comments, so there is no reason to think that a good rating system for comments would have any practical advantages for us. (orphaned: the parallel text in approval vanished)

Bulleted lists were invented for a reason... :) --MyRedDice



I should note that I will only organize the second stage of the vote if a reasonably simple method is used, i.e. approval voting, FPP voting or average voting. Otherwise someone else will have to do it.—Eloquence 00:41, 28 Aug 2003 (UTC)

I'm willing to organize it if Condorcet method or Instant-runoff voting is going to be used. Taw 01:20, 28 Aug 2003 (UTC)
I'm willing to organize it under whichever method. Andre Engels 10:46, 5 Sep 2003 (UTC)

Condorcet cycle resolution


For those who chose Condorcet voting - what are we going to use if Condorcet does not determine a winner? I might switch to it or not, depending on that. Andre Engels 10:46, 5 Sep 2003 (UTC)

In that case "candidate whose biggest pair wise defeat is the smallest of all candidates" wins. Taw 13:44, 5 Sep 2003 (UTC)
That's one resolution method (Minimax), but there are alternatives: In particular, Smith/Minimax is superior to straight Minimax. I'd suggest a seperate vote to determine which resolution method we use. --MyRedDice
It's extremely unlikely that Minimax and Smith/Minimax give different result. I don't have a problem with either. Taw 19:19, 5 Sep 2003 (UTC)
Fair point. Seems like we might be able to get consensus for Smith/MinMax, which would render a vote unnecessary. --MyRedDice
Isn't the obvious solution to have a last first past the post voting between those two alternatives IF that happens? BL 02:07, 6 Sep 2003 (UTC)
Obvious, perhaps, but sub-optimal. If we're going to choose a sophisticated voting system like Condorcet, we might as well choose a sophisticated Condorcet cycle resolution method.--MyRedDice

Condorcet "vote" on voting method


Oh no, let's vote Condorcet way on voting method:

Condorcet method > Instant-runoff voting > Approval voting > Average voting > First past the post -- Taw 01:17, 28 Aug 2003 (UTC)

Condorcet method > Instant-runoff voting > Approval voting > Average voting > First past the post -- Tillwe 11:43, 28 Aug 2003 (UTC)

Average > Approval > Condorcet > IRV > FPTP -- DanKeshet

Condorcet > Approval > Average > IRV > FPTP -- AdamRaizen

Condorcet method > Instant-runoff voting > Approval voting > Average voting > First past the post voting -- Ryan Cable 00:36, 3 Sep 2003 (UTC)

IRV > Condorcet > Average > Approval > FPTP -- Andre Engels

Approval > Average >>> Condorcet > IRV > FPTP -- MyRedDice (>>> marks the approval->disapproval point)

IRV > Condorcet > Approval > Average > FPTP -- Toby Bartels

Average > Condorcet > IRV > Approval > FPTP -- ant

Average > Approval > IRV > Condorcet > Playoff > FPTP -- LDan

Condorcet > Average > Approval > IRV > FPTP -- Aoineko

(no vote) (to highlight the problems of using something confusing) -- Angela

Condorcet > Average > Approval > IRV > FPTP > Playoff -- Gebu

Instantranoff> Condorcet > Approval -- Nerd

Discussion of complexity of Condorcet

  • All you have to do is rank order them. It's not that hard. - LDan
  • Whether it is hard or not is irrelevant. It is perceived as hard because it is unfamiliar to many people. I believe this will discourage people from voting.
The currently leading proposed voting method "Average vote" is that every user gives every logo a 1 to 5. Won't this also be an unfamliar and maybe confusing voting method? -- Tillwe 23:13, 7 Sep 2003 (UTC)
I don't buy this "voter confusion" argument. Real world elections, like for example to German parliament, are far more complicated than anything proposed here. Taw 23:24, 7 Sep 2003 (UTC)
That is true for the German parliament and probably for the US electoral college. In the German parliament vote the political parties actively capitalize on voter ignorance by using slogans like "Second vote for FDP" (where the second vote is actually the one that matters most). That's what you get when you use a process where the way the winner is decided is not understood by most voters. And don't expect most voters to understand the Condorcet tabulation and certainly not the determination of the Condorcet winner.
Why do you think that understanding how to determine the Condorcet winner is so hard? Have you tried to understand it, but couldn't?
Furthermore, is there actually any way to express equal feelings toward an option in Condorcet voting, or a stronger / weaker approval than simply "better than", "worse than"? That's what makes range voting so useful, in my opinion -- that you give all voters a direct way to express their full range of emotions, from "This sucks. This must never be the winning logo!" to "This is OK" to "This is definitely great" -- and it does not force you to pick one over the other. I can say "These two are great. This one is so-so. These three are crappy."
In Condorcet, there is no problem with ranking two candidates as equal in Condorcet, as in A>B=C>D, for example. In fact, Condorcet is better at this than average voting, because it doesn't arbitrarily limit you to 5 levels. You can rank as many logos as you want as equal and put the whole group above or below any other logo or logos.
There is a problem in voting A>B=C>D in Condorcet: it's bad strategy. On the other hand, "expressing your full range of emotions" in Average Voting is also bad strategy, though voting methods exist for scored votes that don't suffer from this collapse.
Disregarding strategy, ranked votes and scored votes are expressive in different ways. Ranked votes do not allow you to distinguish between Blair>Hague>>>Hitler, and Blair>>>Hitler>Stalin, though they are intuitively very different votes. Scored votes allow you to make that distinction, but don't allow you (with five levels) to express A>B>C>D>E>F. --MyRedDice
How is it bad strategy if you truly have no preference between B and C? -- AdamRaizen
It is not a bad strategy if you have absolutely no preference, though neither is it a good strategy: it's strategically equivalent to flipping a coin to choose between A>B>C>D and A>C>B>D.
However, it is a bad strategy in the "Clinton>Bush>>>Hitler" example: voting "Clinton=Bush>Hitler" is strategically inferior to voting "Clinton>Bush>Hitler". Although intuitively one might expect the former to be a stronger anti-Hitler vote, it is not. Note that these issues are parallelled by similar issues in average voting. --MyRedDice
It seems to me that people assume that just because people vote 5 or 1 on some options that critics of the methods automatically assume that this is "strategic" voting. Strategic in which respect? It reflects their opinion that this option should not win. You get real strategic voting in a public process where the current score is visible at any point of the vote -- but that may be the kind of strategic voting you actually want. "This is likely to win. If I vote it down to 1 I can reduce its chances to increase the ones of my favorite."
Strategic voting has a well-defined technical meaning, and that is voting in a manner that is not consistent with the voter's true preferences. For example, a voter's true preference may be as follows: A: 5, B: 4, C: 2, D: 1. However, the voter can increase the chance of B winning by rating B as 5 and decrease the chance of C winning by rating C as 1. Since the voter mostly approves of B and mostly disapproves of C, this increases the expected utility of the vote. The vote does reflect their opinion as to which candidates should or should not win, but the voting system allows a finer gradation of preferences. The voter is mispresenting his or her preferences slightly. This may have the effect of decreasing the likelihood of C winning and increasing the likelihood of A winning, but it also increases the likelihood of B beating A and D beating C. Strategic voting may be mostly harmless in some cases, and certainly strategic voting in average voting is much more benign than in FPP, but it complicates the vote, and it can lead to a decision which a majority of voters find less acceptible than some other option, so I think it should be avoided if possible.
In addition, if some people cast a well-informed strategic vote and some people cast a sincere vote, then the strategic voters' votes have a disproportionate impact on the outcome of the election, which isn't very democratic. Ideally, you shouldn't have to understand details of strategic voting methods in order to cast a well-informed vote, and Condorcet does better on this measure than any other system. -- AdamRaizen
en:Strategic voting may have a well-defined technical definition, but I think yours is wrong. If followed, it would render an FPTP vote for my favourite party as "strategic" because I am "misrepresenting" my opinions by ranking my second and third choices equally.
I prefer the definition I was given: strategic voting is when a voter changes hir vote on the basis of information or beliefs about how other voters are likely to vote. I like it because it's less focused on mechanics, and more on the underlying phenomenon. --MyRedDice

This is not compromise voting like in FPTP -- it is a way that voters express their strong or weak preferences for a particular logo. People don't vote for the "lesser of two evils", they vote against what they consider the greater evil. Because the method used is very transparent and very obvious, it is fair to everyone involved. As for it being confusing and different from what people know -- everyone knows how school grades work. Most people have been at or a similar site that uses grading systems. Range voting is very similar. On the other hand, I don't know a common real world process where Condorcet type ranking is used.—Eloquence 00:01, 8 Sep 2003 (UTC)
  • Rating is one thing - and voting for a single winner completely different thing. -- Taw?
    • I disagree. Voting for a single winner is simply picking the one of which people have the highest opinion -- the highest rated one. -- Erik?
  • I have no idea what are you talking about wrt school grades. -- Taw?
    • School grades are one of many real life forms of rating, also often averaged for various purposes. People are familiar with the concept. -- Erik?
      • Now what does taking mean of a few numbers have to do with voting ? -- Taw?
        • It has something to do with grading an idea or concept on a scale. That is the basic idea behind range voting. In response to the argument that range voting may be confusing, it is easy to respond that most people are familiar with the underlying concept. -- Erik?
  • In Condorcet, honest vote is almost always the best choice. In German system it usually isn't. In American system, it almost never is. -- Taw?
    • In range voting, the honest vote is the best choice as well. -- Erik?
      • You're deliberately lying now !!! Of course that's not true. In "range voting" there's huge space for tactical voting. -- Taw?
        • Any vote you make in range voting should reflect your honest opinion, whether that opinion is "This option is very good" or whether it is "Gee, compared to this other option that is likely to win, this one sucks, so let's make sure it doesn't win". Both are honest choices. -- Erik?
          • Now, now, there's no need to say that anybody is deliberately lying, even if you think that they are. The proper response is "I can't imagine what you could possibly mean by that. Of course that's not true. [...]". Then the other person will explain what they meant -- as Erik did.
          • Anyway, in response to Erik, I guess that you're saying that a strategic vote, at least when done deliberately and with understanding, is still an honest one. Very well. But this is exactly the problem with understanding! In a Condorcet system, the only possible strategic voting is to manoeuvre things into a Condorcet paradox, and that is rare; otherwise, ranking things according to your actual opinion is always the best thing to do. With range voting, it's harder to determine the best strategic vote. Hence potential confusion among the electorate.
          • To be sure, it's not as if confusion is impossible with the Condorcet vote! But when we say «Just list all of the options in the order that you prefer.», we can also justifiably add «It is highly unlikely that you can get a better result by listing the options in a different order through "strategic voting" to avoid the "spoiler effect". So there's no need to worry about it; just rank them.». -- Toby Bartels 18:29, 8 Sep 2003 (UTC)
  • You have bloody campaign to express your opinion. Voting is meant for voting. It's only because of stupid voting systems being used that some people in "real life" use voting to express their opinion. -- Taw?
    • Voting according to Webster's: "A wish, choice, or opinion, of a person or a body of persons, expressed in some received and authorized way". Voting according to Taw: "Voting" -- Erik?
      • And in Votes for deletion and similarly titled pages, "vote" means, even more generally, just an individual's opinion. Almost all of my own votes on Wikipedia have been of this fashion (although not the Round 1 logo vote). -- Toby Bartels 18:29, 8 Sep 2003 (UTC)

Discussion of Condorcet vote on voting method

I don't know who wrote the above, but please do not call the Round 1 system "approval voting". I know you probably know the difference, but the 10-vote limit makes it no longer "approval voting", so I don't want lots of people to become confused. Also, why on earth would we use FPTP voting to decide the voting system?! DanKeshet
Because it's simple, and for once I would like people to pick a clear favorite.—Eloquence
Because it biases the result in favor of the first choice of the one who called for a FPTP vote. -- AdamRaizen
There doesn't seem to be clear winner here (fortunately nobody seems to prefer FPTP), but few people expressed their opinion so far. What makes those of you who selected Average voting prefer it ? Is it only because it's simpler than alternatives or do you think it's more likely to select better logo ? Taw 16:04, 29 Aug 2003 (UTC)
Both LDan
FPTP = First Person To Post voting? :-P. But even if we use Condorcet voting for voting for the voting system, wont that also be unfair because you can already see the results so far? If you really like FPTP and your next alternative is Instant-runoff voting and you hate Condorcet voting for example, wouldn't you still place Instant-runoff before FPTP because the results so far from the Condorcet voting makes it seem that FPTP has no change to win and it is therefore better to vote for your second alternative? BL 03:31, 6 Sep 2003 (UTC)
In a Condorcet vote? No, you would not change your vote from FPTP to IRV, at least not without a better reason than that. Only if it seems reasonable that IRV might lose (pairwise) to FPTP and lead to Condorcet's paradox, then you might choose to reorder your votes. Even then, a minimax resolution of Condorcet's paradox would tend to favour your 2nd-place preference in just those situations where changing your vote would make a difference. -- Toby Bartels 16:56, 6 Sep 2003 (UTC)
That's the advantage of the Condorcet method; it's very rare that a strategic vote would actually help you. -- AdamRaizen

It's pointless to use two voting styles to define a voting style because we won't end up with a clear winner. As the organizer of this logo contest (and both a logo contributor and prize contributor), I think I should have some final say on these trivialities, unless you want to vote on which voting method we want to use to determine the voting method we want to use for the final vote. If you allow me the level of intervention necessary to prevent an infinite recursion, my decision is that this matter will be decided via FPTP voting, and the Condorcet votes above are invalid.—Eloquence 08:07, 5 Sep 2003 (UTC)

Okay, as long as you don't mind being seen to manipulate the vote. While I don't think you originally intentionally attempted to influence the outcome of the vote with your call for FPP, it does seem that way now. Why don't you at least change it to approval voting? Same format, just as easy to tabulate the winner. -- AdamRaizen
You can't change voting method mid-vote, if you want the result to have any credibility whatsoever. --MyRedDice
So maybe we should start over, since this is a bit messy. -- AdamRaizen

More on average voting


Another succesful example of this voting method is the comment rating system used by Kuro5hin, which is generally considered vastly superior to the ones used by other websites such as Slashdot.

This is extremely misleading. We're selecting single winner, on k5 it's only rating system, not elections of any kind. So K5 is not doing average voting elections. Taw 15:19, 5 Sep 2003 (UTC)

asking clarification

above is written

  • Approval voting (vote for one or more candidates, candidate with the highest number of votes wins): CGS
    • Arguments in favor: Strategically equivalent to average voting (that is, if you want to maximize the impact of your vote under average voting you would only give either the maximum or minimum number of points to each candidate). Simpler to vote and to count than everything except FPTP.

on en, it is written in approval voting, you may vote for one or several candidate, to approve only. Let say we have 3 candidates. What I understand is

  • you vote for A, it gets 1
  • you vote for B, it gets 1
  • you don't vote for C, it gets 0

You count the votes for A, B and C in the end, and the one with more points wins.

So, what is this reference to giving the maximum or minimum number of points ? How can this be strategically equivalent to average voting ?

It was a comparison of average voting to approval voting. If the voting system is average voting, and voters want to maximize their votes, they should give either 1 or 5 points (in our system) and never anything in between to each candidate. However, if everyone does this, then the vote ends up working just the same as approval voting anyway. Sorry for the confusion of putting this argument under approval voting instead of average voting. I've moved it. -- AdamRaizen
thank you very much for the explanation Adam :-) ant

Just wanted to say as long as it's not approval or FPTP, it's fine with me. (IRV's my favorite, but it's no big deal.) I find it amusing (mostly in the same way I find Darwin Awards amusing) that FPTP was used here. I'm of the opinion FPTP should never be used for anything of > 2 choices (at which point all systems effectively become the same anyway). It almost always gives the least clear results. -- Jake 03:21, 7 Sep 2003 (UTC)

  • My last edit appears to have screwed with formatting somehow. That's new. Think I fixed it now... will investigate (and not edit this page further). -- Jake 03:29, 7 Sep 2003 (UTC)
  • What's wrong with approval? The common misconception is that if you vote for more things, you have more power, but that only encourages people to vote honestly and gets rid of the spoiler effect. 20:03, 7 Sep 2003 (UTC)

Dont understand average system


Jai pô voté car jai rien compris au systeme "average vote" ...sinon, je suis contre les autres.

La methode Condorcet est decrit en francais sur notre wiki, mais pourrais tu expliquer la methode average, ou plutot comment elle va etre appliquée... car c'est ç`a que je comprend pô :o( Combien de vote par personne ? par logo et ses variantes ? etc.

Si seulement je m'y serais interressé plus tot, jaurais proposé une autre technique :

Chacun peut voter pour 2 ou 3 logos, et pour chaque vote on designe la variante préférée...ce qui me parait plus simple, non ? :op

Oliezekat 08:57, 7 Sep 2003 (UTC)

Its my message for Anthere...I said that I dont vote because i dont understand Average system ; how many vote per wikipedian ? by logo or variant ? etc
Oliezekat 15:07, 7 Sep 2003 (UTC)

You can vote on every variant that is among the finalists, and grade each variant on a scale from 1 (worst) to 5 (best). The logo with the highest average score wins. Simple, no?—Eloquence 15:10, 7 Sep 2003 (UTC)

No limit ? We can vote for each logos and variants ? :o. Oliezekat 16:49, 7 Sep 2003 (UTC)
Si on choisi cette methode on attribuera une note a toutes les variantes ou l'on desire s'exprimer. Chaque variante aura un compte separe. Aoineko 17:07, 7 Sep 2003 (UTC)

Maybe an update of that page later on would be interesting Electoral reform jargon list

Can we vote for 2 methods if we agree both ? Aoineko 15:31, 7 Sep 2003 (UTC)

No. Pick the one you prefer.—Eloquence
ok, thanks, already done ;o) Aoineko

Even if Average Voting can be manipulated, so can every voting system (with 3 or more candidates and no dictatorship) as proved by Arrow's Impossibility Theorem.

Condorcet's method passes only minimally weakened criteria, so this sentence is misleading. en:Arrow's_impossibility_theorem
In fact the impossibility theorem that's interesting w.r.t. "manipulation" of voting systems is not Arrow's. Instead, consider the fact that it is impossible for a voting system to be both deterministic and strategy-free. This is all deeply theoretical, though. --MyRedDice

Proposition for vote table


Assuming the "Average voting" will be the choosed method (not my favorite), what do you think to show result as this (see table bellow). I think is simple to vote and simple to count result. Please, before to disagree, look at the trick on the code ;o) Aoineko 09:03, 8 Sep 2003 (UTC)

Good idea, and very easy to manage :o) ...I add "titinette_who_give_only_3_votes" to show how its easy to support person who give only 3 (or few) votes :op Oliezekat 11:16, 8 Sep 2003 (UTC)
Wikipedian / Logos1 2 3 4 5 6 7 8 9 10 11
Test user 5 4 3 2 1 2 3 4 5 4 3
Toto 1 2 3 4 5 4 3 2 1 2 3
titinette_who_give_only_3_votes 5 5 5

I disagree Aoi. What is important here is not who voted what, but which logo was elected in the end. ant

« Ce qui est important ici n'est pas qui vote quoi, mais quel logo ete elu a la fin ». Heu... c'est bien ca ? Je vois pas trop le rapport ? Je proposait juste une facon, qui me semble simple, de voter et compter les votes. Aoineko 13:39, 8 Sep 2003 (UTC)
I'm not sure what you just said, Aoi, and I might be saying the same thing. But I think this table does group both the vote and the result together in a very efficient manner, when not looking at the HTML at least. It's much simpler than the previos layouts for average voting that we've used. 10:43, 10 Sep 2003 (UTC)