The Wii U Scorecard Rankings List: Day 3: Call of Duty: Black Ops II

#1PaulKagebeinPosted 1/18/2013 10:14:14 AM(edited)
Rate Call of Duty: Black Ops II on a scale of 1 to 10! - Results (72 votes)
10/10
6.94% (5 votes)
5
9/10
27.78% (20 votes)
20
8/10
31.94% (23 votes)
23
7/10
6.94% (5 votes)
5
6/10
8.33% (6 votes)
6
5/10
0% (0 votes)
0
4/10
4.17% (3 votes)
3
3/10
1.39% (1 votes)
1
2/10
0% (0 votes)
0
1/10
12.5% (9 votes)
9
This poll is now closed.
ZombiU received 75 votes with an average rating of 7.27, 7.52 without outliers. The mode was 8. The median was 8.

Today's game is Call of Duty: Black Ops II.

This is a daily "Rate the Game" series for Wii U games! I add all of the scores and divide by the number of votes for each game, figuring out the median within standard deviation, and create a rankings list.

I would like to give a big thanks to user SychinLegacy for assisting me in this project yesterday and helping me figure out how to calculate standard deviation.

The Rule: I would prefer that people actually experience or play the game before voting. I have no easy way of proving this, so everyone is welcome to participate, but please don't make your score lower or higher than it otherwise would be just to change the averages. Give your honest rating on the day's title.


GAME RANKINGS LIST:
RANK/GAME/MEAN WITHOUT OUTLIERS/MEAN/#VOTES/MODE/MEDIAN/PUBLISHER
1. Nintendo Land (8.15/7.57)(70 votes)(Mode of 8)(Median of 8)(publisher: Nintendo)
2. ZombiU (7.52/7.27)(75 votes)(Mode of 8)(Median of 8)(publisher: Ubisoft)

Also, nominate your choice for tomorrow's game! The top nominated game will be the game we rate tomorrow. We can also rate eShop games. I've given a few games (the ones with the top communities on Miiverse) honorary nominations to encourage participation.

Nominations:
Assassin's Creed III x5
Batman: Arkham City - Armored Edition x5
New Super Mario Bros. U x5
Ninja Gaiden 3: Razor's Edge x5
Scribblenauts Unlimited x5
Sonic & All-Stars Racing Transformed x5
Darksiders II x1
Little Inferno x1
Mass Effect 3: Special Edition x1
Mighty Switch Force! Hyper Drive Edition x1
NBA 2K13 x1
Skylanders Giants x1
Tank! Tank! Tank! x1
Tekken Tag Tournament 2: Wii U Edition x1
Trine 2: Director's Cut x1
---
Pokemon White FC - 3225 1582 0350
#2PaulKagebein(Topic Creator)Posted 1/18/2013 11:17:00 AM
Boost!
---
Pokemon White FC - 3225 1582 0350
#3coreyerbPosted 1/18/2013 11:54:36 AM
Kinda hard to determine outliers in a 1-10 poll, isn't it? Did you throw out all the 1s, but not the 10s? How do you decide what's an outlier?
---
NNID: LoeliaPonsonby / [56] iS*LoPo - http://ironsightge.webs.com
#4DaLaggaPosted 1/18/2013 12:00:45 PM
coreyerb posted...
Kinda hard to determine outliers in a 1-10 poll, isn't it? Did you throw out all the 1s, but not the 10s? How do you decide what's an outlier?



My question exactly. In a 1 to 10 poll, nothing should be considered an outlier. Outliers are more of an issue when you're dealing with data covering a very broad range (for example, average height, average amount of soft drink consumed in a day, etc.). But in a poll like this, especially with such a small sample size, you can't just go throwing out data because you don't like the results.
#5PaulKagebein(Topic Creator)Posted 1/18/2013 12:17:10 PM(edited)
DaLagga posted...
coreyerb posted...
Kinda hard to determine outliers in a 1-10 poll, isn't it? Did you throw out all the 1s, but not the 10s? How do you decide what's an outlier?



My question exactly. In a 1 to 10 poll, nothing should be considered an outlier. Outliers are more of an issue when you're dealing with data covering a very broad range (for example, average height, average amount of soft drink consumed in a day, etc.). But in a poll like this, especially with such a small sample size, you can't just go throwing out data because you don't like the results.


There's still a standard deviation, but I will gladly rank them by regular mean instead of mean with standard deviation, which wouldn't change the rankings at all at this point, it simply lowers the primary score we go with.

I didn't throw out any data whatsoever, and refuse to do so rather I like the results or not. There's a reason I included both totals, but I agree that overall mean should take precedence, as it will.

"ZombiU received 75 votes with an average rating of 7.27, 7.52 without outliers."

There was 75 votes in that poll. As stated, I counted all 75. I figured out outliers by figuring out the standard deviation. It's an extra figure thrown in to the statistics for those who care, but everyone's vote counts in the primary ranking method. No worries :)

If you want to check my math on both polls, by all means, I welcome that. A second calculation to make sure I didn't make a mistake is appreciated, but if a mistake is made, that's all it is. I solemnly swear to not let any bias manipulate my calculations.

---
Pokemon White FC - 3225 1582 0350
#6coreyerbPosted 1/18/2013 2:43:38 PM
I was asking since the means without outliers are significantly higher than the raw means. Outliers are going to be biased in the direction of whether most people like or hate the game - if it's an average 7.5 game, 10s aren't going to be outliers, but 2s or 3s might be. You throw out the low votes, which might be valid, while keeping the 10s, some of which might not be valid. Same with a 3.5-rated game. A legit 9 might be tossed but nonsense 1s might be counted.

Tossing outliers makes bad games seem worse and good games seem better. Some people might legitimately rate a "good" game as poor, since game enjoyment is subjective. Those votes are useful and combat group-think. Obviously there is room for abuse, but in such a small sample, the best you can do is what you did - ask people to only vote if they play the game, and vote realistically.

Like you said, to me, ranking by pure mean is more useful in a 1-10 scale. Keep listing the median as it tells us what most people ranked the game, but a weighted mean doesn't offer any different information. Just a thought.
---
NNID: LoeliaPonsonby / [56] iS*LoPo - http://ironsightge.webs.com
#7NerdstickPosted 1/18/2013 2:48:52 PM
One of my first planned purchases. As much slack as CoD gets, you have to give them props for not only supporting off-tv play but two player multiplayer via gamepad and TV is fantastic.
#8PaulKagebein(Topic Creator)Posted 1/18/2013 4:01:16 PM
coreyerb posted...
I was asking since the means without outliers are significantly higher than the raw means. Outliers are going to be biased in the direction of whether most people like or hate the game - if it's an average 7.5 game, 10s aren't going to be outliers, but 2s or 3s might be. You throw out the low votes, which might be valid, while keeping the 10s, some of which might not be valid. Same with a 3.5-rated game. A legit 9 might be tossed but nonsense 1s might be counted.

Tossing outliers makes bad games seem worse and good games seem better. Some people might legitimately rate a "good" game as poor, since game enjoyment is subjective. Those votes are useful and combat group-think. Obviously there is room for abuse, but in such a small sample, the best you can do is what you did - ask people to only vote if they play the game, and vote realistically.

Like you said, to me, ranking by pure mean is more useful in a 1-10 scale. Keep listing the median as it tells us what most people ranked the game, but a weighted mean doesn't offer any different information. Just a thought.


You have a very valid point. At any rate, I will make the true mean the primary ranking method as opposed to the weighted mean, and consider tossing the mean with standard deviation all together. It's just an extra statistic, but with a such a subjective process it is meaningless compared to the main mean, median and mode. Thanks for the feedback!
---
Pokemon White FC - 3225 1582 0350
#9coreyerbPosted 1/18/2013 4:23:29 PM
Thanks for doing it. It's an interesting experiment.

I'd be interested to see how 007 Legends scores for Wii U since a lot of the feedback has been better than expected, but it obviously should wait till after the bigger titles.
---
NNID: LoeliaPonsonby / [56] iS*LoPo - http://ironsightge.webs.com
#10DTY3Posted 1/18/2013 4:26:05 PM
9/10

Yeah CoD is stale, but it does what it does brilliantly.
---
Metroid Zero Mission and Super Metroid are some of the finest 2D games ever made and are the best in the series.
Metroid Fusion is pretty cool I guess.