The Results of the 1999 Survey

Modifications by Times Played

I separated the subjects in 4 groups: People who had not finished the game yet (7, 3.1 %); people who had played the game one time (100, 44.4 %); people who had played the game 2-3 times (95, 42.3 %); and finally people who had finished the game not inclusively more than 3 times ( 4+ times played) (22, 10.2 %). This distinction was made to group the people, well, in absolute newbies, newbies, standard gamers, and hardcore fans. One person who was in the data analyses above included had a missing value in the times played.

Notice the important fact that the groups have very diffent sizes. This makes the standard procedures to test the significance of the differences a little tricky, and this lead to the fact that here no every tiny diffference will be highly significantly. In fact, almost nothing went significantly. I had planned to show the results of hte comparison for every map, but I learned, when I had arrived at Bafford (after 4 hours of annoying work), that this is not a work for me, but a work that probably only masochists would enjoy. So, I will reduce the presentation of the results. I will only describe the few relevant and significant results. I tested the significance of the differences with a Mann-Whitney-Test, just as comment for the really statistic freaky guys who might read this.

The results

The grouping of the subjects after the number of times played had some interesting, but not very significantly results. So was it very funny, i.e. to see that the few people who did not finish the game one single time were only male. Funny to see, as I said, but the question is of course: So what? When you have only abot 10 women in your survey, then you can't expect that in this small number is a woman. This is not very probably, you know.

But I had two major interesting and significantly results:

1.) The few evil people who did not finish the game one single time had indeed given the last few maps a worser score than the other groups. The difference is mainly significant in 'Bedfellows', but easy to watch in the other maps, too. As an example the distribution of the ratings of this evil group of the subjects about 'Bedfellows', compared to the ratings of the other groups:

"A"	"B"
"C"	"D"

Note the fact that I have choosen % as the way to visualise the differences. This overexaggerates the number of subjects who are in the two extreme groups (0 and more than 4 times). While the differences between the B and C is not significantly, the distribution of A differs from the expected values. The score 12 was given many more times than it was to expect. 12 is the score given if you simply score the last few missions with the number of their place in the sequence of the missions.

The D (the 'hardcore fans')-distribution is odd, too, but I decide to ignore that oddity for this time.

2.) The other result was more interesting for me, because it matched absolutely my expectations about the preferences of the 'hardcore fans' about RTC. Sadly, the hardcore fans had not prefered the Zombie missions per se, but it matched perfectly to my hypothesis about: The more times someone has played the game, the higher he will score RTC. Well... if my enthusiasm for RTC is too much for you, just ignore the following pics, even if they are really cool:

"A" (N = 7; mean = 8.29; Std = 2.36)	"B" (N = 100; mean = 7.53; Std = 3.66)
"C" (N = 95; mean = 7.36; Std = 3.82)	"D" (N = 22; mean = 5.22 (!); Std = 4.25)

Hm, maybe should I mention at this time that I had definitely not filled out the survey more than one time, because these results match so absolutely good to my enthusiasm about RTC. :-)

Let's ignore the ratings of the group A (they were never in RTC, I presume, or stuck there). The distribution of B seems to be, err, not so very good, but it has an amazingly huge Std. Same with group C - the Std gets even larger. This distribution is almost equal distributed. And then to group D, the hardcore fans who played the game over and over again: It is odd, but this distribution is definitely not equalised or indicating that the people did not like the mission. More: These odd freaks loved the mission, at least gave 41 % of them this mission the score 1.

Well, this leads to some interesting questions such as: Did the hardcore fans vote differently from the others? Are they freaks? Or are they simply completely nuts? Err... Or did they give RTC a high score because they are members of the gameforum and knew about my preferences? ... :-)

I can't answer all of these questions. Or, better to say, I don't want to do this now, because this would mean some intensively reading of the feedback, some analyses if this subject might be a certain person... No. But I give a short summary for the group of hardcore fans:

Their votes differed not significantly from the other groups. The sequence of the mean scores differed slightly, but not significantly - so, i.e. did this group give 'Sword' (mean = 5.2273) a higher score than the other groups. But these differences were in the range of the normal Stds. I mentioned the mean score of 'Sword' so exactly because of a funny coincidence: This is excactly the 'RTC' mean score (mean score, more exactly = 5.2273), so that both missions share place 2. :-)

The group prefered 'Assassins' (mean = 3.68) most of all missions. In this rating, and more or less in the other ratings, the group of hardcore fans behaved as the other groups.

The differences in the other variables are not significantly, neither. The group pretends to be a very normal group. One point is highly interesting: Which gaming style do they prefer?

I have to admit that I choose that picture also because it is really looking funny. 'Mis-sion—mat-' means: Just matching the mission. These preferences about the gaming style indicate either that the hardcore fans enjoy automatically after a while the sneaker-stlye, or that I know them indeed from the gameforum. Another point: This group preferred also the Expert difficulty, and that definitely more than the other groups did.

So, I think it is time now to stop my ranting about this minority of hardcore fans of Thief... only because I am a member of that group and because the scores of that group matched my expectations about the results of the survey more than the entire survey did, it is not in the place here to argument more about this group.

It differed from the other groups in some certain ways, but was more or less normally in other ways.

Summary and Comments:

The examination of the subjects about differences that correlated with the number of times played that two significantly results:

- There had been an effect of the sequence of the missions in the game: The people who had not finished the game one time, had given the later missions worser scores.

- The group of hardcore fans differed slightly from the other subjects: They gave RTC a higher score, so that RTC and Sword shared their place two. Their number one was, nevertheless, Assassins. Also in the other ratings did the small group of hardcore fans not differ significantly from the others.