We should probably hammer this out before we start getting in ballots.

 

This poll will be weighted with favour to the top 3, maybe even top 10. The question is how?

 

Let's get some options out there and set up a vote.

 

Dylan mentioned Laparde method.

 

How was Smark's choice done?

I'm no expert on this, but why does it need to be weighted? When I think about my own ballot, I don't think of the top three or ten as some roped off area.

I'm no expert on this, but why does it need to be weighted? When I think about my own ballot, I don't think of the top three or ten as some roped off area.

People are going to be a LOT more careful in the top 3 or so. The different between 1 and 5 is bigger than the difference between 96 and 100.

That's interesting. I don't necessarily see a huge gap between my No. 1 and my No. 13 or whatever, and I'm not sure I'd want my top couple of votes to carry exponential weight. But if that's out of step with the way most people see it, that's fine. I guess a weighted system makes it harder for wrestlers to dominate just by appearing on every ballot.

I strongly prefer a weighted system.

 

The Laprade method minus the subtraction of points if someone doesn't appear on every ballot is the one I prefer.

 

I'm not violently opposed to doing it with points allotted via inverse order of placement, but the weighted system puts value in who people actually see as the greatest of the great, and also serves as something of a balancing act regarding those who are writing off entire styles from their personal lists for one reason or another.

I strongly prefer a weighted system.

 

The Laprade method minus the subtraction of points if someone doesn't appear on every ballot is the one I prefer.

 

I'm not violently opposed to doing it with points allotted via inverse order of placement, but the weighted system puts value in who people actually see as the greatest of the great, and also serves as something of a balancing act regarding those who are writing off entire styles from their personal lists for one reason or another.

I can't find a description of that system online, could you lay it out?

From a thread at Classics:

 

About your point system, I wouldn't do from 100 to 1. I've been doing these kind of rankings for Quebec wrestling every year for 10 years, and I based myself on something the NHL did 15 years ago.

There should be more of a difference between #1 and #2, and between #2 and #3 and so on for the top 10. Because without that, there's nothing special about being voted #1 and even if 10 person would vote for the same guy as #1, if one voter place him #10, he's losing all of his edge.

Ex: 10 x 100 + 1 x 90 = 1090
Ex: 8 x 99 + 1 x 98 + 2 x 100 = 1090

So someone who would be voted 10 times as the best wrestler would share first place with someone who has been voted only twice, only because one voter put him 10th once.

You should start at 125 points.

1- 125 pts
2- 115 pts
3- 110
4- 105
5- 102
6- 99
7- 96
8- 94
9- 92
10- 90
11- 89
12- 88
13- 87
All the way to
98- 2
99- 1
100- 0

Plus, every one not voted in by someone should receive -1 (minus 1) point.

So if we take the same example:

10 x 125 + 1 x 90 = 1340
8 x 115 + 1 x 110 + 2 x 125 = 1280

Now the same guy has a 60 points lead and it would take more than one voter to affect his ranking. He would need a few more 10th position to lose his edge and if that's the case, well maybe he deserved to lose his first place.

I like everything about that.

Does that system maybe undervalue simply appearing on the ballot? Like if 5 people place someone at 100 that's not as good as 1 person placing them 94th. This probably isn't that important to anyone as it won't affect the upper tier of results but just thought I'd throw it out there.

If we used the weighted scale, would we adjust the scoring slightly so that 100th place earned 1pt and there wouldn't be minus points for not featuring on a ballot?

My preference is to use exactly what they did in Smarkschoice, so the comparison is exact.

Smarkschoice simply did 1 point for a #100 spot, 2 points for a #99 spot ... ... ... 100 points for a #1 spot. Then they ranked EVERYONE who got even one vote, with all the people who got one #100 vote tied for last place.

 

I do like the method Dylan proposed. However, this isn't a statement about the method as much as it is a hope that if we go that route, the ranking difference between #2 and #3 is about the merits and not an attempt to strategically place wrestlers in certain spots to ensure a good final overall ranking. But I guess there's no way to control that, and it would probably happen under any method. It just really stood out to me when I read that more than it had when the topic of strategic voting had come up before.

My preference is to use exactly what they did in Smarkschoice, so the comparison is exact.

I don't see that relevant, since Smarkschoice had no nomination process and you could vote for anyone. Smarkschoice had no 18 month build up with threads for each person.

 

The process has changed in every aspect from 2006, why stop at the way tabulate votes?

On the other hand, why does it absolutely have to be different?

On the other hand, why does it absolutely have to be different?

It doesn't have to be different for the sake of different. However, I'm really on board that the top few votes are more important than the rest and should be treated differently. There is a lot of arguments about who is number one. I don't see anybody arguing about who should be number 76. The majority of people will be way more effort at the top of their list than the bottom and the voting should reflect that. At least that is my thoughts and Dylan's.

That's a fair point, I didn't know Smarkschoice was more of a snap poll than this much more rigorous process.

 

My impression is that people deliberate a lot on their top 5s and to a slightly lesser degree on their top 10s and 20s and then get more crap shooty beyond that, so I agree with the general principle that something should reflect that.

What was the basis in that method of the dropoff in the amount of variance between spots, for instance 3-4 is 5 point variance but 4-5 is 3 point variance?

There were "discussion" threads just like we have here for the previous GWE, but they weren't nomination threads, and you could vote for anyone with or without a discussion thread.

What was the basis in that method of the dropoff in the amount of variance between spots, for instance 3-4 is 5 point variance but 4-5 is 3 point variance?

Yeah it does seem weird to not drop off as gradually as possible. 5 point variance to 4 points to 3 and so on.

 

 

Edit: actually no it's not

That weighted method would really hurt great luchadores just because dummies don't "Get" lucha

That weighted method would really hurt great luchadores just because dummies don't "Get" lucha

 

What? Doesn't weighting it mean that you can do better while appearing on fewer ballots? Shouldn't luchadores be hurt for failing to appeal to everyone? Can you suggest a system where 'great luchadores' aren't hurt because some people leave them off other than just disregarding ballots that dont' include Negro Casas?

My suggestion was to produce different lists, not just one overall list. For example, "of those that voted for luchadores, these were the results" etc., but that was shot down pretty quickly.

My suggestion was to produce different lists, not just one overall list. For example, "of those that voted for luchadores, these were the results" etc., but that was shot down pretty quickly.

 

I'm sure Mookie can include stuff like that in a deeper dive stats report, if he's willing. But there needs to be one overall list. We all understand style biases are baked into it--comes with the territory.

That weighted method would really hurt great luchadores just because dummies don't "Get" lucha

I think it would help great luchadores, because nobody would rank them lower to hurt them. Somebody like Shawn Michaels will be hurt because he may get some top votes, but he will get votes all over. While Negro Casas will either get high votes or no votes, so the hurt won't be as much.