Return of the PM Character Survey

[Choice Scarf]

@Saproling I think I got it for you. I also found out how to recall the URLs to edit responses individually so if you want to double check yourself I can send your link in a private message.

 

[Choice Scarf]

Okay, the spreadsheet seems good enough to go live, so if reddit pulls through we'll hopefully get more data soon.

EDIT: for the sake of not continuously posting I'm just going to use this post an observation log thing will reddit blows my numbers up:

• Some people ran into issues where they still had multiple submissions, not sure I can do anything from my end but manually delete duplicates
• Thanks to the memesters to test my discounting thing works. For anyone curious, the red texted ones are what I do manually so that even if they toggle back to YES I'm taking it seriously I still remember who they are
• There are more professionals compared to no experts at the time of writing this. Maybe "international threat" isn't established enough in our meta quite yet. Perhaps professional should also include if you actually play this game for a living, but since it's PM that probably won't work
• Cells containing open-ended answers doubled in width for a bit more visibility, not a perfect solution until this data can be sent to a doc or something - SLOWLY TAKING TOP PRIORITY
  
• First to columns are frozen to keep track of which player you are looking at. Hopefully you should remember their character while reading the data.
• Not sure how to stress not writing so much to the point it makes your row ridiculously huge. Limit how many times you press ENTER?
• Why are there blank rows all of a sudden? Maybe ones that started but never submitted?
• Sheilda - distinct enough for separate discussion?
• Gap between Intermediate and advanced is too big. Along with expert and professional difference, definitions should be reviewed.
• Make clear that if people don't know MUs or ratings that skipping the answer is an option; Don't Know option added just in case.
• Mobile maybe looked at, but since volume of data is so large it may not be feasible (laggy even by desktop)
• MUs will stick to +/- 4 range as apparently some of them can be that bad (according to Yata!). Keeping track of chart though since its best and worst MUs are currently at +/- 3.

 

[shairn]

According to the spreadsheet atm Bowser has a better matchup spread than Fox or Falco

Hm...

 

[Farquaad]

According to the spreadsheet atm Bowser has a better matchup spread than Fox or Falco

Hm...

Not remotely. According to the spreadsheet (assuming it had anywhere near the amount of participants then as it does now) Bowser has negative matchups pretty much across the board.

 

[shairn]

Not remotely. According to the spreadsheet (assuming it had anywhere near the amount of participants then as it does now) Bowser has negative matchups pretty much across the board.

My bad I was looking at the standard deviation instead of the average
Carry on

 

[4tlas]

I mentioned this before in the Reddit thread, but I'll reiterate it here since nobody else has commented:

There is Sheilda content that does not fall under Zelda or Sheik, and thus there is nowhere to place this info. I don't think a full Sheilda category is necessary by itself as it will be difficult to account for who is posting to it or how they are rating data, but I also don't think posting separately in the Zelda and Sheik categories is enough. There is strategy and tech involved in Transforming and in matchups and stage selection where you plan to use both characters as opposed to the "better" one. I also cannot completely fill out Zelda or Sheik because I do not use either character in 100% of scenarios, so there are many where my response would be "Don't know, I use the other half here" and cannot possibly explain thoroughly enough without delving into things irrelevant to mains of the individual characters.

 

[Choice Scarf]

Progress update

I working on a proper method to get weighting implemented without it screwing up the visibility of the results, so it can be reinserted into the sheets. The idea is that the skill level you choose determines the weight of your ratings, which should help make some of the numbers look a little more accurate. I currently have the weight location and failsafe control in and am working on getting the formula to ignore blank cells, so it's almost there. But before I add it in for real, I need to decide how much weight each level of player skill should have.

A lot of this will probably be readjusted with each iteration of the survey, but with what we have now:

• The difference between Casual/Beginner and Expert/Professional seem very similar to each other in practice because of the way it was worded in the survey, so I feel they would have similar weights
• Experienced/Intermediate/Advanced falls under an ambiguous mid tier of players (dipping into low and high) so there would be small but distinct differences between those 3 and major differences between the lowest and highest rungs.

I'm pulling these numbers out of thin air but if no one has any objections I'm thinking of going with the following:

• Casual: 1
• Beginner: 2
• Experienced: 5
• Intermediate: 8
  
• Advanced: 14
• Expert: 20
  
• Professional: 25

EDIT (12/26/15): Weight added to ratings