GameFAQsContests.com

About the SC2K4 SFF-Adjusted Standings

These standings are an attempt to adjust for the influence of Same Fanbase Factor in various Summer Contest 2004 matches.

There were two parts to figuring out the amounts by which to adjust the characters: comparing the affected characters' 2K4 strengths to their 2K3 strengths, and additional constant. This constant was 7.24%—which was the factor implied by keeping Cloud's strength constant across 2K3 and 2K4, and is essentially estimated to be the amount by which Link's strength increased. The method by which I arrived at this constant is rather complicated, but I will try to explain it.

After the chosen characters were adjusted for SFF, I took the resulting strength values for 2K4 and calculated the sum of squared differences for the difference between 2K3 and 2K4 strengths for the characters who were in both contests—with certain exceptions (Auron, Tails, Ness and Sora, because it is thought they may have been affected by SFF in SC2K3; Lara Croft and Sam Fisher, because of extreme strength fluctuations; and Gordon Freeman, even though he didn't change much, because he lost to Sam Fisher in 2K4). I then calculated the scalar factor by which 2K4 strengths could be multiplied to reduce the sum of squared differences. The resuling value was about 6.65%. (here is a least squared differences example in Excel, using made up data)

This number was actually rather close to the value for keeping Cloud constant. So, rather than using the calculated number, I simply assumed Cloud's strength had remained unchanged and adjusting the SFF adjustments accordingly.

To use Ganon as an example: his strength value in 2K3 was 34.72% against Link. His strength value in 2K4 was 12.10% against Link (who he faced directly). This means his performance would need to be adjusted upwards by 186.95%. However, since the evidence is clear that Link increased, even without SFF Ganon would only have been expected to get about 32.38% against Link anyway. Thus, Ganon was adjusted upwards by about 167.58%. This then also increased Alucard's strength, as Ganon defeated Alucard.

The characters who were adjusted for SFF, along with the level of their adjustments and who they lost to:

• Ganon: +167.58% (Link)
• Magus: +44.96% (Crono)
• Bowser: +41.52% (Mario)
• Yoshi: +38.13% (Link)
• Squall: +29.04% (Cloud)
• Mega Man: +11.11% (Link)

The thought of SFF in Link vs. Mega Man is somewhat controversial, but was included as without it, everything else gets really off from what you would expect. It has been hypothesized that perhaps some factor other than, or in addition to, SFF caused Mega Man to underperform in that match.

Other possibilities for SFF include Zero vs. Protoman and Magus vs. Luca Blight, but no adjustments were attempted for these matches.

In addition to the standard rankings, these standings also include columns that show estimates of what each character would have been expected to receive against the past champions. These comparisons across the various contests is done by assuming that Mario's strength was constant across years, and adjusting everyone else accordingly by a scalar factor. The assumption of keeping Mario constant was one that resulted in a low sum of squared deviations in character strengths across years—i.e., low year-over-year volatility relative to the assumptions of keeping most other characters constant. Also, this assumption seems to be pretty sensible as Mario is someone whose strength you would expect to be consistent.