Raw Voting Data Now Available (ie da Feedback)

Founder
Posts: 1,138
Joined: 2002.04
Post: #31
>to distinguish votes from the same voter?
Indeed I do, but I didn't pull out such info from the voting system. I will at some point if it would be helpful.

Cheers,

Carlos A. Camacho,
Founder
iDevGames
Quote this message in a reply
Member
Posts: 131
Joined: 2004.10
Post: #32
Something I noticed is that the voter details page mentioned that it only needed to be filled out once, when I took a gander at the results I noticed a lot of my voted didn't include the voter details because I only did them for a few times and then never again.
Quote this message in a reply
Moderator
Posts: 335
Joined: 2002.04
Post: #33
Remember that the suggested price will be biased slightly by there being a larger balance of shareware developers in the voting mix, so it would probably tend towards the non too surprising result of $10-$15.
Quote this message in a reply
Moderator
Posts: 771
Joined: 2003.04
Post: #34
Average Scores & Number of Votes (vs Time)

Code:
Date    Gameplay    Graphics    Audio    Votes
Nov-09    3.654    3.808    3.551    234
Nov-10    3.424    3.664    3.327    217
Nov-11    3.457    3.674    3.295    129
Nov-12    3.455    3.597    3.221    77
Nov-13    3.404    3.640    3.167    114
Nov-14    3.380    3.589    3.287    129
Nov-15    3.794    3.912    3.706    34
Nov-16    3.245    3.628    3.234    94
Nov-17    3.808    3.923    3.385    26
Nov-18    4.044    4.133    3.689    45
Nov-19    3.371    3.671    3.286    70
Nov-20    3.000    3.096    2.890    73
Nov-21    3.029    3.360    3.131    175
Nov-22    3.370    3.467    3.250    184
Nov-23    3.597    3.702    3.435    124

Graph: here
  • The sections with big variations seem to be the result of the low number of votes at that time.
  • More votes at the beginning and end of the voting period than in the middle.
  • Average daily score for all categories seemed to go _down_, not much and not necessarily statistically significant, but surprising since many entries fixed bugs/added stuff during the voting period.
  • The average daily score for Graphics was always higher than Gameplay's, and Gameplay's daily average was higher than Audio's (except on Nov. 21)

Edit: Direct linking to the graph didn't work, so I made a page with all the graphs - link updated.
Quote this message in a reply
Moderator
Posts: 771
Joined: 2003.04
Post: #35
Average Scores by Voter Type
Code:
Are you a gamer or developer?    Gameplay    Graphics    Audio
Gamer    4.31    4.36    4.12
Gamer; Developer    3.92    3.95    3.73
Developer    3.69    3.70    3.46
Not specified    3.15    3.41    3.07

Graph here (same as before)

People who considered themselves as "Gamers" (only) gave the highest scores, and those who considered themselves "Developers" gave the lowest (from those who answered the question). Those who didn't answer the question gave the lowest scores of all.

Interesting... Sneaky
Quote this message in a reply
Moderator
Posts: 771
Joined: 2003.04
Post: #36
Average Scores & Number of Votes by "Gamer Type"

Code:
1=Casual 4=Hardcore    Gameplay    Graphics    Audio    Votes
Not specified    3.137286759    3.39642567    3.054427295    1231
Casual (1)    4.201680672    4.168067227    4.016806723    119
Semi-Casual (2)    4.222222222    4.281481481    4.074074074    135
Semi-Hardcore (3)    3.945454545    4.066666667    3.733333333    165
Hardcore (4)    4.226666667    4.253333333    4.053333333    75

Graph here

Those who didn't answered the question about what kind of player they are gave the lowest scores, followed by those who classified themselves as "Semi-Hardcore"

Correlation Analysis
Code:
    Gameplay    Graphics    Audio    Votes
Gameplay    1            
Graphics    0.993976237    1        
Audio    0.997226678    0.990538771    1    
Votes    -0.978115058    -0.981032998    -0.960675511    1

When grouped by "Gamer Type", all categories' scores have a strong inverse relationship to the number of voters. Also, Gameplay, Graphics & Audio have a strong direct relationship between them.
Quote this message in a reply
Moderator
Posts: 771
Joined: 2003.04
Post: #37
Correlation Analysis (based on each entry average scores)
Code:
    Gameplay    Graphics    Audio    Originality    Story    Overall    Price
Gameplay    1                        
Graphics    0.77503823    1                    
Audio    0.80941343    0.94937233    1                
Originality    0.57114400    0.58839934    0.63655862    1            
Story    0.35554912    0.49295929    0.55910539    0.49428052    1        
Overall    0.85332665    0.92678016    0.95789815    0.76790004    0.66771110    1    
Price    0.78441473    0.83907031    0.79797292    0.58630223    0.58644433    0.86428460    1
  • Overall scores have a strong relationship with Graphics & Audio, a relationship to "Estimated Price", and practically no relationship to Story & Originality (note: for a relationship to be considered relevant, it should have a correlation value of +0.8 or higher (or -0.8/lower).
  • Gameplay scores have a relationship with Audio (and, to some extent, with "Estimated Price" and with Graphics).
  • Graphics scores have a strong relationship with Audio scores, and a relationship with "Estimated Price".
  • Audio scores seem to have a very weak relationship to "Estimated Price".
  • Story and Originality have practically no meaningful relationship between them.


That's all the stats I have for now. Anything interesting? Smile
Quote this message in a reply
Moderator
Posts: 916
Joined: 2002.10
Post: #38
shweet.

if you have anymore, keep posting them, I LOVE reading this stuff
Quote this message in a reply
Member
Posts: 30
Joined: 2002.10
Post: #39
I may seem stupid, but what is CSV-format? Also I can't unstuff that file. Can anyone send me a "readable" file by email? I'd also like to read those funny comments ...

Mac games free download: http://www.zhanggames.com
Quote this message in a reply
Moderator
Posts: 916
Joined: 2002.10
Post: #40
have you tried opening it in text edit
Quote this message in a reply
Member
Posts: 328
Joined: 2002.04
Post: #41
Story and originality were peer categories, so that may help explain some of why the didn't correlate: they were very different samples. And you showed that "developers" tend to vote differently than "gamers", at least when averaged together :-)
Quote this message in a reply
Moderator
Posts: 437
Joined: 2002.09
Post: #42
Still looking forward to the judge's scores/comments in the polish category... (No hurry though.)

Measure twice, cut once, curse three or four times.
Quote this message in a reply
Founder
Posts: 1,138
Joined: 2002.04
Post: #43
CSV=Comma Separated Variable
It is a regular text file, with the fields/variables divided by a comma. You can use Excel, or a Spreadsheet to open it in a nice to read display. Probably other ways to view CSV in a nice formatted way as well.

>Still looking forward to the judge's scores/comments in the polish category
OK, keep reminding me. Also, keep in mind that their comments were not for EVERY game. For the most part, some mentioned a game or two, or commented on the contest as a whole. Next year, I will add a comment field after each game, for the judges who have the time to comment on Polish.

Well, there is a lot of data in this thread. It would be VERY good for someone to summarize it in a report, as was done in uDG 2002. Anyone?

Carlos A. Camacho,
Founder
iDevGames
Quote this message in a reply
Moderator
Posts: 771
Joined: 2003.04
Post: #44
Some new tables/charts: http://www.reversecode.com/powermacx/udg04stats.html

Please note that some of the charts & decimal places in the table's values are slightly different from what I previously posted (not much, ie: my previous comments about them are still valid). Why? Well... there is this game called l1neum, which is called different in it's .csv file so that messed up some data when I sorted it, and in the end I forgot to add the votes of one game to my basic "all votes" spreadsheet. BTW, I think that the "number of votes" stat presented earlier (not by me) had incorrect values for the same reason (l1neum/OneNeum).

Note: For the 2D vs 3D charts, I assumed that BumperCars was 2D, but I never played it, so if it actually was 3D I'll have to update that chart. I think Carlos said that the game will be disqualified, but I made all my stats before I read that post, so the votes for BumperCars are included. I really don't want to calculate everything again but in any case, there are about 10 votes for this game so it wouldn't change the end results much.

New charts: Gamer Type vs. Age, Number of Votes by Age and 2D Games vs. 3D Games.
Quote this message in a reply
Moderator
Posts: 771
Joined: 2003.04
Post: #45
DaFalcon Wrote:Story and originality were peer categories, so that may help explain some of why the didn't correlate: they were very different samples. And you showed that "developers" tend to vote differently than "gamers", at least when averaged together :-)

Yes, that's why the charts only compare the public categories. Perhaps I should add an additional correlation table with only developer's votes taken into account?
The problem with this is that, for instance, many of my votes in the public categories (those I can recognize by my comments) didn't have any info! I entered it once the first time I voted, but in many of my votes it didn't show up. Wacko

Also, I would like to complete the correlation table with Polish scores, as soon as that info is available.
Edit: I just found out the official results Excel files, so I'll add Polish to the stats tomorrow.
Quote this message in a reply
Post Reply 

Possibly Related Threads...
Thread: Author Replies: Views: Last Post
  uDG08 Voting Results and Methods AnotherJake 111 49,048 Mar 20, 2009 05:15 PM
Last Post: pinguoren
  uDG voting DoG 11 6,020 Mar 7, 2009 09:57 AM
Last Post: FlamingHairball
  Questions regarding voting? FlamingHairball 20 9,752 Mar 6, 2009 10:00 AM
Last Post: FlamingHairball
  How will public voting work? pinguoren 7 5,297 Mar 2, 2009 11:22 AM
Last Post: Carlos Camacho
  Networking Minicontest "Voting" aarku 26 11,426 May 9, 2008 08:36 PM
Last Post: PowerMacX