Question Here is your Kommunity Top 10. Do you agree?

[Frostbarrier]

Why do you people keep posting ranks for characters, A Good any effin character can beat a top or whatever character ... and again, Screw Tiers n shit ! ERMAC is awesome !

 

[TakeAChance]

Why do you people keep posting ranks for characters, A Good any effin character can beat a top or whatever character ... and again, Screw Tiers n shit ! ERMAC is awesome !

The point of this thread was to grab everyone's collective top 10, and see if we collected all data, what characters would land in the top 10 given the communities responses. This was not an individualized top 10.

 

[Frostbarrier]

The point of this thread was to grab everyone's collective top 10, and see if we collected all data, what characters would land in the top 10 given the communities responses. This was not an individualized top 10.

I know, don't get me wrong I respect people's opinions but I will always stick with Ermac IMO

but hey, good job doing all the math and calculating all that lol ... good effort

 

[TakeAChance]

That's not the right way to look at it though. Sonya didn't have 4 extra chances to land at #1, she and Cyrax had the same amount of chances because everyone was able to pick both characters for their Top10, it just happened that in the opinion of 4 people Sonya was a Top10 material while Cyrax wasn't. This should help her place above Cyrax on the list, but instead it ended up hurting her.

The point is, calculating averages is not the best method. I could have been the only person to even have Kano in Top10 and place him at #1. His avg would then end up 1.00 and he'd be the best in the game. See the problem?

The best way to make a list, imo, would be to add pts to characters for placing in Top10s. For example, on each user's list character placed #1 should get 10 pts, #2 should get 9 pts, #3 should get 8 pts, ... #10 should get 1 pt. Then you sum up character pts from all the submitted lists and get a total for each character. This way characters would get penalized for failing to make Top10s 'cause they wouldn't get any pts (and placing in Top10, even at #10, would be rewarded in contrast to not placing in Top10 at all). It'd be interesting to see Cyrax's and Sonya's final rank using this method as well as the rest of the Top10 since there are more cases like this in the current Top10 (like Skarlet vs Smoke, he got into a lot more Top10s, yet she ended up ahead of him just because her avg was better even though many people didn't even have her in their Top10).

Btw, thanks for answering. Also, I like the project idea and appreciate your effort, that's why I'm debating about the method as I think it's not the best. I don't want you to think I'm being mean or something. I'd just like this to be as perfect as possible, even if it's just for fun.

What you are saying is correct. There is a system in place to make a bare minimum baseline ammount of votes required to be considered. Jax has an overall score of 9, but could not be counted simply because he only had 9 votes.

The average number of votes needed to be considered was 35.26 adjusted to 35. This number was gathered by taking the total votes (670) and diving them by the total number of individual characters named (19). If any character had under 35 votes, they did not meet the statistical baseline for votes per character and thus their results were not counted.

That was how I determined a baseline for votes.

 

[TakeAChance]

I know, don't get me wrong I respect people's opinions but I will always stick with Ermac IMO

but hey, good job doing all the math and calculating all that lol ... good effort

Thank you, it was actually a lot of fun!

This is not to persuade anyone from playing certain characters. I appreciate player skill more so than any particular chart, but I found some of the results of the stats interesting and I will be using them for further discussion points.

 

[PaletteSwap]

I think it should be Kung Lao, Kitana, Skarlet.

I haven't played much with Skarlet so I can't say for sure.

 

[GrandMasterson]

Tier lists are made from match up numbers, not "well this character is really really good so I put them here lol"

The only way to do this correctly in this format is to have everyone vote on every character's match up numbers and let the math speak for itself. It sounds tedious, but it would leave the least amount of room for people to whine. That is, if you think all of that is worth the effort, and if it isn't, then why even bother doing it this way in the first place?

Personally I'm really not that interested, but if you all think this is something worth doing, then it's worth doing right. I applaud you patience, Chance.

 

[TakeAChance]

Tier lists are made from match up numbers, not "well this character is really really good so I put them here lol"

The only way to do this correctly in this format is to have everyone vote on every character's match up numbers and let the math speak for itself. It sounds tedious, but it would leave the least amount of room for people to whine. That is, if you think all of that is worth the effort, and if it isn't, then why even bother doing it this way in the first place?

Personally I'm really not that interested, but if you all think this is something worth doing, then it's worth doing right. I applaud you patience, Chance.

This is the common misconception in this post. This isn't supposed to prove anything other than the communities perception. Currently, according to the data collected, this list is what the community thinks is the top 10.

The problem with match up numbers is they are just as subjective as me saying that Jade-Kung Lao is 4-6 and not 3-7. We would still have people debating and that is why it is not worth the effort at this time

 

[GLoRToR]

 

[Deftonesrc]

I think people are just afraid of the maths...

 

[Ninj]

This isn't supposed to prove anything other than the communities perception.

Perfectly said. How is there an argument in this thread from anyone?

The only thing you can say is "I didn't get a chance to vote.". If you did vote, these are the results.

Sent from my Galaxy Nexus using Tapatalk 2

 

[fr stack]

Perfectly said. How is there an argument in this thread from anyone?

The only thing you can say is "I didn't get a chance to vote.". If you did vote, these are the results.

Sent from my Galaxy Nexus using Tapatalk 2

quoted so more people might read

 

[Dramatica]

Thanks to STB TakeAChance for doing this, it is always interesting to look at poll results.

With Chance's permission, I have taken his data and used it to recreate the list for my own entertainment. However instead of using averages I have given weight to each and every vote based on its position 1-10.

Using linear weighting (10 pts for #1, 9 pts for #2, 8 pts for #3, etc) Cyrax drops below Sonya and Skarlet drops two places:

1. Kabal 668
2. Kenshi 519
3. Kung Lao 481
4. Sonya 439
5. Cyrax 431
6. Freddy 359
7. Smoke 190
8. Cage 168
9. Skarlet 160
10. Kitana 123

(Reptile 67, Liu 22, Jax 18)

I tried out some non-linear weighting schemes but they don't provide any interesting results.

The main (obvious) conclusions:
- Top 3 are very consistent (MK is clearly ruled by the "KKK", if you will)
- Sonya and Cyrax are a total toss-up for 4th and 5th
- General consensus is good in the Top 6, after which point totals drop dramatically
- Skarlet is now below Smoke and Cage which makes sense to me

I hope there are more community-driven experiments done!

These are my own opinions, not those of STB TakeAChance. Big thanks to him for the data, without which this post would not exist..

 

[Deftonesrc]

Thanks to STB TakeAChance for doing this, it is always interesting to look at poll results.

With Chance's permission, I have taken his data and used it to recreate the list for my own entertainment. However instead of using averages I have given weight to each and every vote based on its position 1-10.

Using linear weighting (10 pts for #1, 9 pts for #2, 8 pts for #3, etc) Cyrax drops below Sonya and Skarlet drops two places:

1. Kabal 668
2. Kenshi 519
3. Kung Lao 481
4. Sonya 439
5. Cyrax 431
6. Freddy 359
7. Smoke 190
8. Cage 168
9. Skarlet 160
10. Kitana 123

(Reptile 67, Liu 22, Jax 18)

I tried out some non-linear weighting schemes but they don't provide any interesting results.

The main (obvious) conclusions:
- Top 3 are very consistent (MK is clearly ruled by the "KKK", if you will)
- Sonya and Cyrax are a total toss-up for 4th and 5th
- General consensus is good in the Top 6, after which point totals drop dramatically
- Skarlet is now below Smoke and Cage which makes sense to me

I hope there are more community-driven experiments done!

These are my own opinions, not those of STB TakeAChance. Big thanks to him for the data, without which this post would not exist..

This is great! I think we need some more unbaised, mathematical approaches to our list. I wonder what would happen if we took the frame data and somehow assigned "value" to each move based on its advantage/disadvantage. I'm a big fan of math and science, so this kind of thing is awesome to me. Out of curiosity, what non-linear models did you try? Exponential/log?

 

[Nephrite]

Cool stuff Dramatica! I thought Sonya & Cyrax might switch that way and that Skarlet might drop.

Fun thread, thanks Chance!

 

[TakeAChance]

So, what we can assess from both sets of data is that as Deftonesrc has said, the top 3 is pretty much set in stone within our community. Sonya and Cyrax could go either way but are a clear 4 and 5. Freddy is pretty much set in stone at number six as well. Smoke seems to be the general consensus for number 7, but 8,9, and 10 are all over the board in terms of votes.

Standard deviation in the original statistics reflected this discrepancy and explains these new statistics quite well.

Excellent work Deftonesrc

 

[Dramatica]

This is great! I think we need some more unbaised, mathematical approaches to our list. I wonder what would happen if we took the frame data and somehow assigned "value" to each move based on its advantage/disadvantage. I'm a big fan of math and science, so this kind of thing is awesome to me. Out of curiosity, what non-linear models did you try? Exponential/log?

I have considered this before (doing a mass analysis of frame data, not necessarily to rate players, but to automatically reveal ALL the situations against ALL of the cast where you are at a certain level of advantage/disadvantage), but I would need to think a lot harder about the pitfalls and frivolous aspects of it before I go further..

 

[Rampage254]

It's been almost two years, and we still can't get an accurate tier list, let alone an accurate top 10 because people either cry that their characters are too low, or too high up.

 

[Lulzlou]

I can't believe some people actually think Kenshi isn't top 2

 

[Dramatica]

My thoughts are:
- To be making the most ideal community Top 10 list, we should be ranking the whole cast, or at the very least the top 15. Less attention could be paid to very low tier rankings because they won't affect the holy Top 10. Doing this would allow us to understand, for example, how Reptile compares, since in the last poll so many people could not even include him.
- If ranking most of the cast, we could allow for one or two "ties", thus preventing people from having to make seemingly arbitrary choices regarding who's better in a tight pair. This would result in the two chars carrying the same mathematical weight in the analysis. No problem.
- If a Top 10 list gets made again using only ten spots, these spots should NOT include the KKK, i.e. Kabal Kenshi Kung Lao, or at least Kabal and Kenshi. This #4-13 list would be way more informative.

Let it be overwhelmingly clear to the haters: WE KNOW that lists are just lists and do not inherently have meaning. We do shit like this because we are analysts and it is interesting and fun. If you hate on a statistical analysis here just cause the output happens to be a "list", then you have announced your measurable lack of intelligence to the rest of the forum, and I will immediately think nothing of what you have to say on the subject. Flawless victory

 

[Chongo]

Whatevs bruh my lagtile is still top 1 online #yolo #swag #BeastCity #JumpOuttaHitd4

 

[NKZero]

My thoughts are:
- To be making the most ideal community Top 10 list, we should be ranking the whole cast, or at the very least the top 15. Less attention could be paid to very low tier rankings because they won't affect the holy Top 10. Doing this would allow us to understand, for example, how Reptile compares, since in the last poll so many people could not even include him.
- If ranking most of the cast, we could allow for one or two "ties", thus preventing people from having to make seemingly arbitrary choices regarding who's better in a tight pair. This would result in the two chars carrying the same mathematical weight in the analysis. No problem.
- If a Top 10 list gets made again using only ten spots, these spots should NOT include the KKK, i.e. Kabal Kenshi Kung Lao, or at least Kabal and Kenshi. This #4-13 list would be way more informative.

Let it be overwhelmingly clear to the haters: WE KNOW that lists are just lists and do not inherently have meaning. We do shit like this because we are analysts and it is interesting and fun. If you hate on a statistical analysis here just cause the output happens to be a "list", then you have announced your measurable lack of intelligence to the rest of the forum, and I will immediately think nothing of what you have to say on the subject. Flawless victory

MATHEMATICALITY!

 

[Groove Heaven]

I don't think Cyrax belongs that high. His damage output is crazy when he has meter and you don't, but his footsie game and zoning game are too weak for him to be #4. I think Sonya and Freddy are better than him.

 

[SLy]

I don't think Cyrax belongs that high. His damage output is crazy when he has meter and you don't, but his footsie game and zoning game are too weak for him to be #4. I think Sonya and Freddy are better than him.

Are you high?

 

[NKZero]

I don't think Cyrax belongs that high. His damage output is crazy when he has meter and you don't, but his footsie game and zoning game are too weak for him to be #4. I think Sonya and Freddy are better than him.

His footsie game is incredible man what are you talking about? Maybe the best D3 in the game, one of the best D4's in the game, stupid ass D4-chestsaw, net is always scary, bombs are very good for forcing opponent positioning, F2 despite being slow grants 11 plus frames on block and it has great range, AA is good for mobility or for actually AA'ing lol and he has a ton of other shit.

Not to add the dumbest mix-up in the game....command grab death loop vs reset death. His damage is the best and on top of that he gets free mix-ups after taking 70+% of your life. He gets out zoned by good zoners but well placed/well timed bombs are a good way to force a zoner out of position and once Cyrax is in he's very dangerous. His rushdown is highly underrated.