It looks like you're new here. If you want to get involved, click one of these buttons!
^ Back to Top
The MI6 Community is unofficial and in no way associated or linked with EON Productions, MGM, Sony Pictures, Activision or Ian Fleming Publications. Any views expressed on this website are of the individual members and do not necessarily reflect those of the Community owners. Any video or images displayed in topics on MI6 Community are embedded by users from third party sites and as such MI6 Community and its owners take no responsibility for this material.
James Bond News • James Bond Articles • James Bond Magazine
Comments
Fair enough. My other list is likely far more dissident, so that's fine with me. :)
1. On Her Majesty's Secret Service
2. Licence to Kill
3. From Russia with Love
4. GoldenEye
5. The Man with the Golden Gun
6. The Living Daylights
7. Casino Royale
8. Thunderball
9. Tomorrow Never Dies
10. Goldfinger
11. The Spy Who Loved Me
12. Dr. No
13. Spectre
14. Live and Let Die
15. You Only Live Twice
16. Octopussy
17. Skyfall
18. Die Another Day
19. For Your Eyes Only
20. Quantum of Solace
21. A View to a Kill
22. Diamonds Are Forever
23. The World Is Not Enough
24. Moonraker
I did a correlation test on Excel for every users rankings relative to the average rankings. User Tanaka123 currently has the highest correlation at .952173913. Interestingly, while several other users are fairly close (below .2) only one user has a negative correlation, that being BMW_With_Missiles who scores at -.015990207.
Correlations:
Birdleson: .726086957
GoldenGun: .693913043
Stun_Harvesting: .207826087
Barryt007: .48956522
Thunderfinger: .677391304
suavejmf: .86608696
bondjames: .784347826
thelivingroyale: .533913043
w2bond: .61242564
1. Casino Royale
2. From Russia With Love
3. Octopussy
4. Skyfall
5. Thunderball
6. On Her Majesty's Secret Service
7. GoldenEye
8. For Your Eyes Only
9. The Spy Who Loved Me
10. Goldfinger
11. The Living Daylights
12. Dr No
13. Licence to Kill
14. You Only Live Twice
15. A View To A Kill
16. Live And Let Die
17. Tomorrow Never Dies
18. Moonraker
19. The World Is Not Enough
20. Quantum Of Solace
21. Spectre
22. The Man With The Golden Gun
23. Die Another Day
24. Diamonds Are Forever
Yes. It is a measurement of the level of correlation between two datasets. In this case, each individual ranking vs the average rank of each film. It is measured relative to 1. In other words, 1 would be a perfect correlation. If the number is negative then that means that there is no correlation at all, i.e. it is a negative correlation.
I'm quite sure my ranking of some of the Moore films is higher than most. I too have TLD (& GF for that matter) much lower than most, and conversely have TB higher than most. That probably accounts for most of it.
I more think that it is something interesting to look at than anything else. I am not passing judgement as to whether one person's ranking is superior to someone else's on the basis that it is more closely correlated with the average. Either way, your number of .78 is not especially low.
@w2bond
We still don't have your ranking with YOLT included.
It's interesting to see the info anyway. I wouldn't mind a lower correlation actually, because there's something good in all the Bond films, even the ones that aren't all that popular, and having a score contrary to the average suggests one appreciates that fact.
Yeah, I just find the results interesting. To me, the correlation is just about telling us how close ones rankings are to the average. That is still a very interesting exercise though.
@w2bond
I am not entirely sure why that is the case. I would have to take a closer look at both of their rankings to get a better idea of why we are seeing that.
Probably a stupid question but numbers were never my strongest point ;)
Your ranking is fairly mid-level as far as the correlation is concerned. The positions of TMWTGG and TND jump right out, along with your comparatively low rankings of GF and SF as reasons why you are not more closely correlated.
@w2bond
Yes, I computed it on Excel. With stun_harvesting, it is important to remember that his "objective" list is not what I was using. I have been using his "subjective" list for his personal ranking.
The cumulative rankings currently look like this:
1.) Casino Royale (4.8512)
2.) On Her Majesty's Secret Service (5.1250)
3.) From Russia With Love (5.5620)
4.) Goldfinger (7.3058)
5.) The Living Daylights (8.3140)
6.) Skyfall (9.1157)
7.) Dr. No (9.4298)
8.) Goldeneye (9.5041)
9.) The Spy Who Loved Me (9.7355)
10.) Thunderball (10.8264)
11.) Licence to Kill (11.1653)
12.) Spectre (12.2314)
13.) For Your Eyes Only (12.5620)
14.) Live and Let Die (13.6364)
15.) Octopussy (14.1074)
16.) You Only Live Twice (14.5667)
17.) Quantum Of Solace (15.2893)
18.) Tomorrow Never Dies (16.0248)
19.) The Man with the Golden Gun (17.3719)
20.) Moonraker (17.6033)
21.) Diamonds Are Forever (17.6364)
22.) The World Is Not Enough (17.7851)
23.) A View To A Kill (18.9669)
24.) Die Another Day (21.1653)
As you were saying, flipping the positions of TLD and QOS (#10 and #19 respectively) would raise your correlation to .815652174 from .72173913, a substantial jump.
No problem. It is interesting. We are still waiting for w2bond to complete his ranking though. He didn't place YOLT.
I agree with that. I was thinking that I could run a regression analysis on the rankings to test that exact question as I entered all of the rankings in order going backwards by taking everyone's most recent ranking. It would not catch subsequent changes since I first entered the numbers though and I have updated a few users rankings since the rankings were initially entered.
As an example, I realize CR is #1 around these parts, but is that on account of a lot of #1 placements and a lot of #10 placements or is it because a lot of people ranked it #4 or #5 (which is closer to its 4.85 average rank)?
Such a measure could isolate anomalies given our small sample size.
I'm pretty certain FYEO & LALD will sit between 10 and 14 on a lot of people's lists (I've seen these two show up close together a lot of times).
Yes, I have calculated both the Standard Deviation and Coefficient of Variation for every film. The first number is the Standard Deviation and the second is the Coefficient of Variation.
Dr. No 5.6610 0.6003
From Russia with Love 4.7309 0.8506
Goldfinger 5.0610 0.6927
Thunderball 6.2911 0.5811
You Only Live Twice 4.6159 0.3169
On Her Majesty's Secret Service 4.4449 0.8673
Diamonds Are Forever 5.1023 0.2893
Live and Let Die 4.9041 0.3596
The Man with the Goldun Gun 4.3668 0.2514
The Spy Who Loved Me 5.0905 0.5229
Moonraker 4.9941 0.2837
For Your Eyes Only 5.1086 0.4067
Octopussy 5.1199 0.3629
A View to a Kill 4.1492 0.2188
The Living Daylights 5.3759 0.6466
Licence to Kill 6.1378 0.5497
Goldeneye 5.6937 0.5991
Tomorrow Never Dies 5.2017 0.3246
The World Is Not Enough 5.4179 0.3046
Die Another Day 4.1299 0.1951
Casino Royale 4.5197 0.9317
Quantum of Solace 6.4076 0.4191
Skyfall 5.8938 0.6466
Spectre 6.1991 0.5068
You also like YOLT better than the average user does and of course your ranking of Spectre is last.
The median for Casino Royale is currently at 3, which tells us that negative outliers are actually dragging it down slightly. However, OHMSS median is also 3 leaving them both tied in first in terms of median. Being that CR has a better average by almost .4, which means that this is a bit more pronounced in the case of OHMSS.
So if I'm interpreting this correctly, the controversial ones with the greatest variability in rank (as measured by standard deviation) among users are TB, SP, SF, & LTK.
DAD, AVTAK & TMWTGG aren't too controversial based on standard deviation, and given their low overall rank it appears that most users uniformly feel poorly towards them.
Interestingly, I note that CR, OHMSS & FRWL (the highest ranked films here) are also the ones with the highest coefficient of variation. What does that mean?
My update.
1. Dr No
2. On Her Majesty's Secret Service
3. Goldfinger
4. GoldenEye
5. The Living Daylights
6. From Russia With Love
7. Live And Let Die
8. Thunderball
9. Octopussy
10. Diamonds Are Forever
11. The Man With The Golden Gun
12. The Spy Who Loved Me
13. Casino Royale
14. Tomorrow Never Dies
15. For Your Eyes Only
16. You Only Live Twice
17. The World Is Not Enough
18. A View To A Kill
19. SPECTRE
20. Moonraker
21. Die Another Day
22. Skyfall
23. License To Kill
24. Quantum of Solace
(It's so much easier updating your list when you have your old one to refer back to.)
I am seeing what you are seeing in terms of the standard deviation. The four films that have a deviation over six are TB, LTK, QOS and SP. In terms of variance in numerical terms, those four films seem to have the largest variance.
The reason why the COV is showing what it is showing is because the COV defines the standard deviation relative to the mean. In other words, it simply looks at what the size of the standard deviation is relative to the average. This is very important in terms of understanding the standard deviation in context. In other words, if the figure is 11,555,678, a standard deviation of 1,312 is actually extremely small. This may not register though if viewing the number 1,312 in isolation.
Being that COV is giving us the standard deviation relative to the mean and that the numbers that we are analyzing can only possibly range between 0-24, it isn't a particularly useful number in this case because it will rapidly inflate as the averages get closer to zero. It is very useful in some applications, but in this case I am not sure that it tells us much of value. Standard deviation is a good measurement here though.
@Mendes4Lyfe
Great. I updated your ranking.
Among the top 10 loved films around these parts, CR & OHMSS are pretty rock solid based on standard deviation (under 4.6) and TB & SF are the ones more likely to get knocked out of the hallowed 10 in the future based on standard deviation (above 5.8).
Yes, no problem. I agree entirely regarding COV.
I can't help but be amused by the results for Die Another Die. The average and stdev paint a picture that is just staggeringly bad.
TWINE would likely have joined it if not for a few users who think highly of the film.
DAD is ranked dead last 37℅ of the time and in the bottom five 79℅ of the time. Too funny.
TWINE is not particularly good, but I don't feel as though it is nearly as bad as DAD. It sits at 22/24 though currently, so it is still doing very poorly on here.