Is voting working properly?

[netactivist]

I was told I posted this in the wrong forum earlier
http://mmcxii.podcastalley.com/forum/showthread.php?t=135975
Others shared my concern there.

So I'm posting it again in this help section:


What caught my eye is, podcasts that usually get a handful of votes in an entire month, now have hundreds of votes in four days.

That could happen, of course, if a podcast with a large audience starts asking its audience to vote.

But those hundreds of votes would be accompanied by at least a couple of comments. But there are zero comments accompanying the hundreds of votes.

Just seems odd. There are other strange vote totals as well, involving oddly low comment numbers accompanying suddenly huge vote totals.

As well as a podcast jumping up 60 votes in 10 minutes just now, when its last episode was a week ago.

Are all these technical glitches left over from the last few days when Podcast Alley was down?

Or maybe the usual voting patterns don't hold any more?

Just wondering...
__________________

[WyethDigital]

Yep. The voting system seems to be busted. I was able to vote twice for my own show (well, I had to test it, ya' know ;) ).

Eric

[BrynM]

I'm taking a look at what's up now. I'll keep you guys posted.

[BrynM]

With theFerf's blessing I've turned off voting for a short bit. It would seem there are a few that would like to make this more trouble than it needed to be. They did, however, give me a nice list of emails to evauluate further action for.

[WyethDigital]

With theFerf's blessing I've turned off voting for a short bit. It would seem there are a few that would like to make this more trouble than it needed to be. They did, however, give me a nice list of emails to evauluate further action for.

Gotta luv it when Hansel and Gretel leave a trail of bread crumbs!

[BrynM]

Voting has been re-enabled (Yay!). I've got some scripts we're currently testing to clean up the bogus data, so the statistics and rankings should be fixed in the next couple of days. Sorry for the mishap everyone... and for those of you that took advantage... I know who you are.

[netactivist]

February's almost over and the fraudulent voting still stands.

Not only have some (most?) of the blatantly phony vote totals from the pre-voting-freeze period not been cleaned up two weeks later, but ALSO, I'm wondering if there is STILL ADDITIONAL phony voting going on as you read this.

Rapidly rising (within an hour, if not minutes) vote totals that I'm seeing with no additional comments, just seem to fit the prior vote fraud pattern. I can't be sure, however.

What I am sure of is:

Four out of the present Top Ten (two from the same host) have comment/vote ratios way out of line with everyone else, and way out of line with their prior month's ratios. Six podcasts out of numbers 11-20 have the same suspicious indicia.

Are overzealous listeners setting up one phony email account after another and voting?

Can Podcast Alley detect numerous votes from a single IP address?

What kind of safeguards are there to ensure the integrity of this voting, I ask.

Why did one outrageously phony vote total get its podcast thrown off the Top Ten, but another outrageously phony vote total has remained there for two weeks?

Will March's voting have adequate safeguards?

What's going on, please?

Please note: I assume it is overzealous listeners, not the podcasters themselves.

[netactivist]

Hmm... since I posted my last reply, seems like a remove bogus vote script was started up again. One podcast went from around 795 votes to 32.

Then after that, whom was the bogus vote script run on? One of the other podcasts with ridiculous comment/vote ratios?

Apparently not.

It was run on me! My ratio looked normal. I guess they just wanted to make sure the whistleblower was himself honest.

My vote total went from 434 to 417. A couple of overvotes, but essentially solid.

Now how about running the bogus vote script on the other 3 remaining podcasts in the Top Ten that are totally out of whack?

I know we don't pay to be listed here, so I hope my posts are always understood to be made with respect and thanks for the great site Chris has created here, and the hard work all his crew do to make it run smoothly (with a strong nudge needed now and then to attend to certain matters :-)

[yuzzy]

Made some tests myself. I wasn't able to vote twice... even with 2 different email accounts...

Maybe they fixed it. They better should. That's how it got out of hand on www.mp3.com back in the days.

The problem is.. the human. Where ther are humans, there will be humanry like this. Man is a wolf to man... Altough it might look like a community, everybody's here fort their own goods. Especially in a place where everybody wants to be a superstar... You'll see the worst from human being... be prepared...

[WyethDigital]

Hmm... since I posted my last reply, seems like a remove bogus vote script was started up again. One podcast went from around 795 votes to 32.

Maybe you need to start naming names (and providing links) if you think the cheating still stands?

Just a note: Not everyone comments when they vote, so that's not always an indication of cheating. PCA does block multi-votes from IP addresses, but if someone uses a screening service or switches IPs and email, then they can vote more than once. It's honestly a crap shoot. People cheat these things, so unless it's incredibly blatant, like this last month, then I tend to ignore it.

Eric

[netactivist]

I was asked to post some specifics.

Here goes.

I assume this is overzealous listeners, not the podcasters themselves.

What follows is not, as far as I know, how the script used by Podcast Alley works. It is how I determine that there are bogus votes.

I use this process:

First, I subtract the February votes and comments from the totals, which yields the inception-thru-January vote/comment ratio.

Second, I compare that to the February vote/comment ratio.

I can eyeball it and tell in a second. But since I'm posting this publicly, I actually did the precise math for what follows on the Friday, Feb 23 numbers, when I wrote this.

Slight variations are common. For example, Blast The Right changed from 2.6 to 3 votes per comment from pre-February to February.

But FTL went from about 5 votes per comment before this month, to 18 votes per comment in February. That's statistically impossible.

Common Sense and Hard Core History, both had similar statistically impossible results.

Common Sense went from 9 votes per comment to 21 votes per comment.

Hardcore History went from 4 votes per comment to 21 votes per comment.

No high-ranked podcast, before this month where there was all the cheating, had a ratio anywhere near 18 or 21 votes per comment, to the best of my recollection.

This statistically impossible jump in the number of votes per comment in FTL, Common Sense and Hard Core History, is the exact same pattern that was exhibited by Acorns, Urban Coffee and all the others that the script did correct by knocking out a huge number of their votes.

So I'm not sure why FTL, Common Sense and Hard Core History haven't been corrected. But they should be.

To repeat, I assume this is overzealous listeners, not the podcasters themselves.

[BrynM]

Ok. Here's some official word now that I have a few spare moments. The stats for February are correct. I have run the cleanup script and looked at the data within the database myself.

Before anyone starts thumping their chests, I'd like to point out a few things. First, commenting is not tied to voting. People have an option to comment when they vote, but a vast majority of voters actually do not. Trying to gauge the number of votes from the number of comments is like projecting apple sales from the number of oranges eaten. The two do not and will never actually correlate to each other.

Secondly, the cleanup tool (as I have pointed out to Netactivist a few times so far) does not target any particular feed. The idea that we would be "targetting" feeds is plain unmanageable with the number of feeds on the site. When we run the script, we run it against the feeds in the order that they come.

Finally, accusing a feed of having an inflated vote count by only looking at the PCA vote to comment ratio is not seeing the whole picture. Free Talk Live is a syndicated radio show carried by 20 local affiliates ( http://freetalklive.com/affiliates.php ). FTL not only has a "PCA Vote drive" going on their show this month, but they've also got a few million potential listeners from the syndication alone. Their numbers rising the way they had is not unreasonable given this (and again I have looked at the DB to verify). They have earned their rank rightfuly.

I hope this clears up some of the confusion. There are parts of the process that I will simply not go into, but voting is now more solid than it had ever been. Everyone had bogus votes cast when the hole was announced in the public forums (rather than being reported to PCA admin). The hole was fixed right away even though the cleanup took a bit, it is done and the statistics are correct.

I was asked to post some specifics.

Here goes.

I assume this is overzealous listeners, not the podcasters themselves.

What follows is not, as far as I know, how the script used by Podcast Alley works. It is how I determine that there are bogus votes.

I use this process:

First, I subtract the February votes and comments from the totals, which yields the inception-thru-January vote/comment ratio.

Second, I compare that to the February vote/comment ratio.

I can eyeball it and tell in a second. But since I'm posting this publicly, I actually did the precise math for what follows on the Friday, Feb 23 numbers, when I wrote this.

Slight variations are common. For example, Blast The Right changed from 2.6 to 3 votes per comment from pre-February to February.

But FTL went from about 5 votes per comment before this month, to 18 votes per comment in February. That's statistically impossible.

Common Sense and Hard Core History, both had similar statistically impossible results.

Common Sense went from 9 votes per comment to 21 votes per comment.

Hardcore History went from 4 votes per comment to 21 votes per comment.

No high-ranked podcast, before this month where there was all the cheating, had a ratio anywhere near 18 or 21 votes per comment, to the best of my recollection.

This statistically impossible jump in the number of votes per comment in FTL, Common Sense and Hard Core History, is the exact same pattern that was exhibited by Acorns, Urban Coffee and all the others that the script did correct by knocking out a huge number of their votes.

So I'm not sure why FTL, Common Sense and Hard Core History haven't been corrected. But they should be.

To repeat, I assume this is overzealous listeners, not the podcasters themselves.

[WyethDigital]

But FTL went from about 5 votes per comment before this month, to 18 votes per comment in February. That's statistically impossible.
It's not statistically impossible. You can't just say that when you don't know all the variables. It's irresponsible. You're putting way too much emphasis on comments. For example, I have a voting box (http://www.wyethdigital.com/vote_at_podcast_alley.htm) on my site that allows folks to vote at PCA without entering a comment. In fact, a comment field isn't even an option.

What I find ironic about all of this is that when FreeTalkLive first joined PCA they were complaining about the same thing you are. Only then, it was accusations of Dawn and Drew cheating the vote. If things keep up like this, next year you'll be the one accused of cheating. ;)


Eric

[netactivist]

I appreciate ByrnM's taking the time to respond. Frankly, I find his assertions off-point to the issues I raised.

February's about over, but I don't like loose threads when my careful and solid logic is attacked, so I will respond :-)

I put my responses in blue in the quote.

Ok. Here's some official word now that I have a few spare moments. The stats for February are correct. I have run the cleanup script and looked at the data within the database myself.

Before anyone starts thumping their chests, I'd like to point out a few things. First, commenting is not tied to voting. People have an option to comment when they vote, but a vast majority of voters actually do not. Trying to gauge the number of votes from the number of comments is like projecting apple sales from the number of oranges eaten. The two do not and will never actually correlate to each other.

This is statistically wrong. If you have thousands of events occurring, and a certain pattern emerges, you wouldn't all of a sudden get a massive change in that pattern, unless a variable changed massively. The ratio of votes/comments is such a pattern. When that changes massively, some other variable must have changed. This month there was widespread cheating, which would account for the change we saw in vote/comment ratios on all the sites that were bounced off the Top Ten.

No other hypothesis has been advanced that stands up for FTL, Common Sense and Hardcore History.

Saying that FTL has 23 stations goes to the quantity of votes it can get, not the ratio of comments to votes.

WyethDigital says he has comments turned off as an option. Totally irrelevant, since FTL's link offers the comment option.

Secondly, the cleanup tool (as I have pointed out to Netactivist a few times so far) does not target any particular feed. The idea that we would be "targetting" feeds is plain unmanageable with the number of feeds on the site. When we run the script, we run it against the feeds in the order that they come.

Bryn, you never bothered to read my last pm to you, when I addressed this. If your script doesn't target individuals, then some other process of yours obviously does, since it was days apart when Acorns, and then Urban Coffee, disappeared off the Top Ten with massively diminished vote totals.

If you ran the script twice, why didn't it catch UC the first time?

There was obviously some other process used.

I know this, because the second time, along with UC, my podcast had its vote total diminished by 14, and that didn't happen days before when Acorns vanished.

Finally, accusing a feed of having an inflated vote count by only looking at the PCA vote to comment ratio is not seeing the whole picture. Free Talk Live is a syndicated radio show carried by 20 local affiliates ( http://freetalklive.com/affiliates.php ). FTL not only has a "PCA Vote drive" going on their show this month, but they've also got a few million potential listeners from the syndication alone. Their numbers rising the way they had is not unreasonable given this (and again I have looked at the DB to verify). They have earned their rank rightfuly.

Again, this has nothing to do with the massive change in the vote comment ratio, but has to do with the quantity of votes they might get.

And has nothing at all to do with Common Sense and Hardcore History.

I hope this clears up some of the confusion. There are parts of the process that I will simply not go into, but voting is now more solid than it had ever been. Everyone had bogus votes cast when the hole was announced in the public forums (rather than being reported to PCA admin). The hole was fixed right away even though the cleanup took a bit, it is done and the statistics are correct.

If the voting is working now, then why did Distorted View post a week ago on the other thread about this, that he noticed -- as did I -- "as of two days ago I was at number 10, but in the last day and a half I shot up to number 6. Obviously someone is jacking up the votes."

Hopefully March will go smoothly.

[netactivist]

You seemed happy I raised the issue initially, now you're singing a different tune, when the issue hasn't yet been fully resolved.

Even though it's the end of the month, I don't like being accused of being irresponsible, among other things, so I will take a few moments to respond below, in blue.


It's not statistically impossible. You can't just say that when you don't know all the variables.

I obviously meant it's statistically impossible without the variables changing. The variable I said changed, was the sudden inclusion of bogus votes. Just like the other podcasts that got booted out of the Top Ten.


It's irresponsible.

Please don't call me irresponsible, when...

You're putting way too much emphasis on comments. For example, I have a voting box (http://www.wyethdigital.com/vote_at_podcast_alley.htm) on my site that allows folks to vote at PCA without entering a comment. In fact, a comment field isn't even an option.

...you've just suggested another explanation that had you spent 10 seconds going to the FTL site, you would have seen is irrelevant, since the FTL vote link does have a comment option.

FTL's having 23 channels is also irrelevant - -that goes to the quantity of votes, not the vote/comment ratio.

What I find ironic about all of this is that when FreeTalkLive first joined PCA they were complaining about the same thing you are. Only then, it was accusations of Dawn and Drew cheating the vote. If things keep up like this, next year you'll be the one accused of cheating. ;)

I hope I'm number one, and then would be happy to be accused of cheating http://mmcxii.podcastalley.com/forum/images/icons/icon7.gif

Also, as I said in both my first sentence and last sentence of the prior post, I'm not accusing the podcasts, but overzealous listeners.

I've corresponded with the FTL guys via email, and even advertised on their site. I know they are men of integrity, and aren't sitting there clicking away. It would have been their overzealous listeners.

Same with Dan Carlin. I don't know him, but I can tell from his podcast he's a person of intergrity as well.
Eric

Richard Pryor in one of his famous routines, said to his wife after being caught in bed with another woman, Who you gonna believe, me or your lyin' eyes?

I believe my eyes.

But we'll have to agree to disagree.

[WyethDigital]

At the time you initially brought this up, things were broken. The reason I'm criticizing you now is that the issue has been fixed as well as it's going to be. Period. Continuing to denounce the admins and other podcasters or even their listeners will eventually lead to unecessary ill will. That is why it is irresponsible.

Also, instead of one thread going on about it, you've got numerous threads, as well as PMs to Bryn. It's way overboard. IF the vote totals are still wrong, and IF PCA has fixed the leak, then March will be a month of self-correction.

Further, one of the points I was making when showing you the commentless voting script is that you don't need to comment to vote. That's just ONE reason why you can't judge vote accuracy based on comments. That's as true now as when I first joined, and is hardly irrelevant to the discussion. I have seen FTL's site and do know that their voting links do include space for comments, but as I said, they are optional, and many people don't leave them.

You seriously need to get over this...

Eric

[netactivist]

I was planning to respond to the criticisms of me -- some of an unnecessarily personal, and inaccurate to boot nature -- made by Wyeth Digital and hittman, but the March voting situation means that that type of back and forth needs to take the back seat.

I'll only say for now, that I do not believe I was demeaning towards anyone, as WD charged. If I was, I apologize. My intent is always to afford everyone the respect and dignity I would like afforded to me.

Getting on to the subject of this post:

Towards the end of February I complained that the voting results still contained lots of bogus votes, and, that bogus voting was still continuing.

The latter was confirmed by Distorted View, who posted on the 21st I think it was -- after the voting had supposed been fixed -- that someone had jacked up his votes, rapidly jumping him up from #10 to #6. I had noticed the same thing and myself concluded that it was bogus voting. I commend DV for his honesty.

Well, now we're in March, and right off the bat, a huge bogus vote for Coffee & Tea: 118 votes with 1 comment, as of 12:15 pm March 1. I was thinking about posting about that, when lo and behold, another honest podcaster, the host of Coffee & Tea, posted himself, asking whoever was jacking up his votes to stop. He said he had never even asked his listeners to vote for him!

So we now have proof that bogus voting is still possible in March.

Maybe it's a bug in the voting system itself that inserts large blocks of votes without comments into the totals of a podcast, I don't know.

But something clearly is still wrong.

Since we know something is still wrong, I will point out additional voting abnormalities, which point in the same direction.

This relates to two podcasts I had flagged in February for suspicious voting patterns, Common Sense and Hard Core History.

By 8 pm on March 1, HH had 60 votes and no comments. It's now, as I type this, at 136 votes and 3 comments. Its home page contains a normal voting link, with comments allowed.

Does this podcast have unusually taciturn listeners? Did the host demand that no one comment? Those would be two possible explanations. Someone jacking up the votes would be another.

As always, I assume the podcaster himself is honest. Maybe whoever did Coffee & Tea is manipulating the HH totals as well.

Interestingly, the other podcast I want to bring to the attention of the administrator, is from the same host.

On Friday night, I was starting to wrap up my work, and checked the Top Ten at about 10:45 pm CA time. I was number 7, CS was several votes behind at #8. It had 130 votes with just seven comments. Already suspicious, given the voting patterns of every other podcast.

The CS home page contains a normal voting link, with comments allowed.

Then just before I turned off the computer, I checked again, and what do you know, CS had jumped up to #6, having gotten 20 votes in 45 minutes, with no additional comments.

A Friday night at 11:30 pm west coast time, the most recent CS podcast having not come out that day but a couple of days before, and 20 votes without comments suddenly flow in.

Are Friday nights a "listen to CS and cast votes all at once and without comments" tradition?


Even worse: At the present time, CS has moved up to #4, 208 votes with -- guess how many comments? Just 8!!

It got 58 additional votes with only 1 comment.

CS & HH together now have together 344 votes with only 11 comments. A 33 to 1 ratio.

In February I said a 20 to one was unbelievably abnormal compared to other podcasts.

33 to 1 is flat earth territory.

Something is radically different about these two podcasts, than every other podcast getting substantial numbers of votes.

I suggest it may be the phantom bogus voter, who afflicted Coffee & Tea. Or maybe it's an overzealous listener who's figured out how to cast multiple votes. Maybe it's a bug in the voting software.

If there's another explanation that has evidence to support it, please let us all know.

But I maintain, something's still going on with the voting, producing inaccurate results, and I ask the administrators to please investigate and report back to the Podcast Alley community.

[WyethDigital]

I was planning to respond to the criticisms of me -- some of an unnecessarily personal, and inaccurate to boot nature -- made by Wyeth Digital and hittman, but the March voting situation means that that type of back and forth needs to take the back seat.

Care to tell me where I was getting "personal?" I told you that I thought you were overreacting to something that already has the attention of the admins, and that it was irresponsible to continue to do so (and I stated my reasons as to why I think that -- none of them personal). If you think that's a personal attack then I would suggest that you're a little overly-sensitive.

Eric

[netactivist]

Another day, another phony voting report is needed. The March 2007 Bogus Voting Derby continues.

I check earlier today, about 1:45 pm CA time, and I'm not in the Top Ten anymore. Got kicked out overnight.

Who's now in the #10 spot? Urban Coffee.

Let's take a look at this overnight rise of UC.

UC suddenly has 201 votes, with a grand total of 2 comments.

It hasn't had a podcast since Feb 21, although one seems to be scheduled for later tonight. Its website link to vote allows comments.

UC is one of the podcasts that when the correction script was run by the administrators on it in February, had its vote total reduced by 90%.

Seems like another run of the ol' script is called for.

Could the Podcast Alley adminstrators respond, please?

I'm back in the Top Ten as I write this, but it's really unfair for me, or anyone now losing a spot, to have our hard word made for naught by voting irregularities, whether they're from deliberate cheating, a bug in the voting software, or some other cause.

[netactivist]

Care to tell me where I was getting "personal?" I told you that I thought you were overreacting to something that already has the attention of the admins, and that it was irresponsible to continue to do so (and I stated my reasons as to why I think that -- none of them personal). If you think that's a personal attack then I would suggest that you're a little overly-sensitive.

Eric
Eric, you've applied the following adjectives and phrases to me:
sour grapes
get over it
overreacting
irresponsible
overly-sensitive

These are not factual descriptions of what's happening with the voting, but evaluations of my character or motivation.

Your attacks are obviously not the most vicious in the world, but I felt it wise to try to nip it in the bud before your mischaracterizations of me escalate.

And by the way, if I was irresponsible to claim the cheating was still going on, how do you account for this post directly below taken from the other thread, where Coffee And Tea complains that his podcast, which he never even asked his listeners to vote for, had gotten over 100 votes the first day in March?

My show Coffee and Tea Show just got 120 votes on the first day of March. I had NEVER asked people to vote for that show because I am just not big on self-promotion and yet somebody, out of a misguided sense of generosity or else as a joke, made my poor little weak show #1 for a few hours today. I don't even personally know these people but I am asking for them to STOP.

Now I am getting mean comments in my blog that I'm a cheater and I feel like Carrie at the prom, waiting to be hit by a bucket of blood. I am having to go around doing damage control. I am embarrassed and mortified as I've always wanted a more popular show but wanted to go the "more listeners" route.

[/quote]

I just can't understand how people like you can take the position, aw, let the cheating go on. Maybe if you have nothing to lose, that seems a good position.

But since I've just lost a spot in the Top Ten (currently 11) this cheating affects me personally.

As far as the administrators, where are they?

I haven't seen a reply from them about the continuing March problems, have you?

[Metaphore]

Dude, seriously. You can sit here and rage against the machine all you want to but nothing is going to happen.

Voting has been a gimped cheated mess since DAY 1 on podcastalley. Do you honestly believe that you are the first and only person that hasn't beaten this dead horse? Much bigger casts than you have tried and failed to get the system fixed. Sure someone might respond and slap a wrist once every 6 months, but then two days later they are right back in the top 10 or spamming the forums or whatever someone is whining about.

It all boils down to what Eric said (either in this thread or the OTHER one you started about this) This site isn't worth the effort. It drives almost no traffic to go to this much stress over. Everyone knows the voting system is gamed here, no one cares anymore.

[WyethDigital]

Eric, you've applied the following adjectives and phrases to me:
sour grapes
get over it
overreacting
irresponsible
overly-sensitive

These are not factual descriptions of what's happening with the voting, but evaluations of my character or motivation.
...
And by the way, if I was irresponsible to claim the cheating was still going on, how do you account for this post directly below taken from the other thread, where Coffee And Tea complains that his podcast, which he never even asked his listeners to vote for, had gotten over 100 votes the first day in March?
First of all, "applying those phrases" is not character assassination. It's opinion. You either agree with it or you don't. You deal with it, or you... complain about it. Alot.

Secondly, Coffee and Tea's problems are not your problems. They are irrelevant to your reactions. They may be similar in nature, but they are still theirs. Why muddy the waters of your issue by tossing around theirs?

I just can't understand how people like you can take the position, aw, let the cheating go on. Maybe if you have nothing to lose, that seems a good position.
"People like me?" I'm not sure what that's supposed to mean... :-?

"Maybe if I have nothing to lose?" Nothing to lose? At the risk of sounding like I'm escalating things: Getting just a bit pretentious, aren't we?

The thing is, you're assuming that I think the voting counts for much, short of a small bump in downloads each month or an ego trip. News for you: It doesn't. We don't even promote it on the show. We apply a tiny amount of our overall effort to it each month, and put the rest of our energy into more important ratings builders. Things that will help long-term growth.
But since I've just lost a spot in the Top Ten (currently 11) this cheating affects me personally.
You mean it affects your ego personally.

As far as the administrators, where are they?

I haven't seen a reply from them about the continuing March problems, have you?
They're probably working on something more important to the functioning of PCA and PodShow.

Eric

[keifer]

I think the only vote that matters is the one someone gives when they choose to subscribe to your show ... outside of that :rolleyes:

[netactivist]

WyethDigital, you'll have the last word here. You're too intelligent and solid a guy for me to be having this back and forth with here. We'll just agree to disagree on this issue, ok?

Metaphore, you wrote:Voting has been a gimped cheated mess since DAY 1 on podcastalley. Do you honestly believe that you are the first and only person that hasn't beaten this dead horse? Much bigger casts than you have tried and failed to get the system fixed. Sure someone might respond and slap a wrist once every 6 months, but then two days later they are right back in the top 10 or spamming the forums or whatever someone is whining about.
I appreciate your historical perspective, and knowing that bigger casts failed to move Chris to do the right thing. I see the forum spammers with 2000 phony replies. I'd be curious as to whom you've seen getting "right back in the Top Ten" over and over.

[Metaphore]

Just type in "Cheating vote" to the forum search and read the repetative threads in the way back machine.