|
Update 3/24/20
Sorry I didn't get around to this sooner. I made the report into a pdf 2 weeks ago but I've been too lazy to figure out a way to share it. This post skips some stuff in the report that is too much of a pain to re-format.
In case you missed the survey, here were the questions and available responses:
How much do you enjoy watching top-level Terran vs Zerg Starcraft II games?
- I love watching top-level TvZ
- I like watching top-level TvZ
- I feel indifferent about watching top-level TvZ
- I dislike watching top-level TvZ
As a Starcraft II player, what is your main race?
- Terran
- Zerg
- Protoss
- I play Starcraft II but do not have a main race
- I do not play Starcraft II at all
These were the results:
Terran Zerg Protoss No Main Don’t Play Total Love 59 49 11 11 4 134 Like 17 6 13 3 3 42 Indifferent 6 0 7 2 1 16 Dislike 5 1 0 2 2 10 Total 87 56 31 18 10 202
Unfortunately this isn't enough data to do the test for independence as-is, so I combined the categories "No Main" and "Don't Play", as well as "Indifferent" and "Dislike", to have high enough expected values (all expected values at least 1 and 80% or more at least 5).
<insert modified data and expected value tables here>
The chi-squared variable comes out to 29.238 with 6 degrees of freedom, and a p-value of 0.00005484
What does this mean?
It means that if race played and how much the viewer enjoys watching top-level TvZ really are independent, there is a 0.005484% that a survey of a random sample of viewers would give results with this much variation from the expected values for each category. (note: You can challenge how much of a "random sample" of viewers my data collection methods were)
So it is pretty safe to conclude that that viewer enjoyment of top-level TvZ is NOT independent of what race they play.
***
Update 3/9/20
Thank you everyone who participated. I am closing the survey now and will post the results in a day or two.
***
Greetings, fellow Starcraft II viewers.
***
TLDR
Please participate in my survey of how much you like watching top-level Terran VS Zerg. It will be open for about a week. THIS IS NOT ABOUT BALANCE OR TRYING TO GET BLIZZARD TO CHANGE THE GAME, but it is fine if your perception of the matchup balance affects how much you enjoy watching it. THIS IS NOT ABOUT SPECIFIC PLAYERS WINNING OR LOSING, so please consider the matchup in general and disregard specific tournament results or which player is your favorite. PLEASE DO NOT DO ANYTHING THAT WILL HARM THE INTEGRITY OF THE DATA, such as voting multiple times or telling your 10000 Zerg stream viewers to vote; I am trying to reach a conclusion about all Starcraft II viewers, which means I need data that represents all Starcraft viewers in as unbiased a way as possible.
You may discuss why you do or don't like watching TvZ in this thread, but only responses on the survey count for the study. I will post my personal thoughts in a second post.
***
I have to do a statistical research project for a class, so I decided to pick a topic of interest to me (Starcraft II).
I'd love to do a study on balance, but chances of getting decent data for that are pretty much zero, so I decided to ask a question that I might be able to get some reasonably unbiased data for:
Does how much Starcraft II viewers enjoy watching top-level Terran VS Zerg depend on what race they play?
Some people consider Terran VS Zerg to be the "classic" sci-fi matchup, of Humans trying to survive against the countless Swarm. If this matchup has universal appeal, then the amount viewers enjoy watching TvZ will not depend upon which race they play.
How will I test this theory? I will use something called a contingency table and a test for independence. The basic idea is to assume the amount people enjoy TvZ is the independent of which race they play. After collecting all the data, I calculate what amount of people should be in each category if this is true. For example, if 25% of people are Terran, then 25% of the people who love watching TvZ should be Terran, 25% of the people who dislike watching TvZ should be Terran, etc. I do a calculation that compares observed values to these "expected" values and measures how different they are.
If they are different enough to be statistically significant, we can say that viewers' enjoyment of TvZ depeneds upon which race they play. If they are not different enough, we say there is not enough evidence to conclude that viewers' enjoyment of TvZ depends upon which race they play (this does not prove they are independent). Because the second conclusion is the "default" (if there is less data, it is harder to reach the first conclusion), it is not very interesting unless there is a substantial amount of data.
***
Here is where to find the survey. It is on Google Forms and may require a Google sign-in to prevent multiple "votes". Sorry for the inconvenience.
<Survey Closed>
Once again, PLEASE DO NOT DO ANYTHING THAT WILL HARM THE INTEGRITY OF THE DATA.
Thank you for your time.
|
Just 2 questions, takes 5 seconds.
I hope your project is a success, gl!
|
Submitted, good luck with your project .
|
Done Suggestion for your next statistic project: Order all 6 matchups how enjoyable to watch and compare to all 9 matchups and how enjoyable it is to play
|
Where's my "I like it but it used to be way better" option
|
On March 01 2020 05:32 Elentos wrote: Where's my "I like it but it used to be way better" option
Well I'm not sure if you're joking, but please make your decision based on the current state of the matchup.
I wanted to keep the choices relatively simple, as more answer choices means more reading means less likely someone will be willing to take the time to answer the survey.
|
On March 01 2020 05:32 Elentos wrote: Where's my "I like it but it used to be way better" option
I know what you mean
|
Turns out, if the matchup is centered around low/mid tier fast units that are vulnerable to splash, combined with good defender advantage and reinforcements that have to come from home, it creates exciting game where its harder to tell who is winning and with more action and come back potential. Also the tech units are vulerable too mass low tier units so they arent massed and instead are used only as support.
ZvT bio historically in Sc2 has been the best designed matchup. They should take examples of what make this matchup better and try to make protoss more like it.
Biggest protoss offenders are offensive warpins+ warpprism micro. turning a favorable protoss engagement into a game ending scenario where opponent cant stabilize because of warprism.
Also recall having 0 counter play is frustrating. U could kill the mama core before to prevent. Now all you need is a nexus with 25 energy O_o. No counter play.
|
On March 01 2020 09:15 Snakestyle11 wrote: Turns out, if the matchup is centered around low/mid tier fast units that are vulnerable to splash, combined with good defender advantage and reinforcements that have to come from home, it creates exciting game where its harder to tell who is winning and with more action and come back potential. Also the tech units are vulerable too mass low tier units so they arent massed and instead are used only as support.
ZvT bio historically in Sc2 has been the best designed matchup. They should take examples of what make this matchup better and try to make protoss more like it.
Biggest protoss offenders are offensive warpins+ warpprism micro. turning a favorable protoss engagement into a game ending scenario where opponent cant stabilize because of warprism.
Also recall having 0 counter play is frustrating. U could kill the mama core before to prevent. Now all you need is a nexus with 25 energy O_o. No counter play.
A lack of counterplay is always bad. Blinding Cloud and Abduct are just as bad for the game and need to go.
|
Should have done this poll before Serral ran over every top terran besides Maru, which might happen tomorrow.
|
On March 01 2020 11:46 sc2isawesome wrote: Should have done this poll before Serral ran over every top terran besides Maru, which might happen tomorrow.
As I mentioned in the post, I don't want people making their choice based on specific tournament results. However, I do think it has been a good opportunity to refresh viewers' opinions on the state of the matchup.
Personally I'm always looking forward to what the top Terrans can do in the matchup, even if the result is disappointing.
For better or worse, the timing of this project is pretty much out of my control. The report is due in a week and a half, and the topic was greenlit only a few days ago.
|
I like bio tvz but watching mech makes me wanna turn off the stream
|
I am a little saddened by the fact that Random is not treated as a race here. Good thing there was a similar option with not having a main race. I've been playing from launch. February 2020 was the first time I played non-random 1v1. My main is Random.
|
done.
personally i love watching TvZ, as a person that considers myself a Terran player cause that's the only race i've ever played, although i haven't played sc2 at all in a few years. i especially love watching Mech but bio games can be amazing too.
out of curiosity: what kind of class is this a project for?
|
Terran without Bunker
Beta : indifferent, WoL : Like, HoTS ; Love, LoTV : indifferent (start : like)
|
Would have said I love it in the past, but I only like now. The current meta isn't as exciting as it used to be. Not a single game went to proper late game in Maru vs Rogue or Inno vs Serral. The Terrans always tried to end the game with timings (I assume they feel top Zergs are favoured in late game scenarios).
|
|
On March 01 2020 17:57 Drfilip wrote: I am a little saddened by the fact that Random is not treated as a race here. Good thing there was a similar option with not having a main race. I've been playing from launch. February 2020 was the first time I played non-random 1v1. My main is Random.
If there's not enough people in a category, I will have to combine it with another category to be able to use this statistical method. I did this preemptively so that it would cover Random and players who consider themselves "dual main". If I don't get enough people who don't play StarCraft at all, I will have to merge with that category as well. And if that's still not enough I would probably have to merge it with Protoss.
|
On March 01 2020 19:00 Schelim wrote: out of curiosity: what kind of class is this a project for?
Probability and Statistics II (upper division college math)
|
Done. It's hard to answer actually, I used to love it but as so many others am not to keen on it now.
|
|
|
|