In addition to each player's performance in the six key performance measurements creating a game score, every player also has a player score for each of the performance measurements. The six performance measures are as follows:
Out Wit
Out Play
Out Last
Modified Merge Baseline:
The difference between between the calculation of game scores and player scores for vote clustering, accuracy, avoidance, and challenge performance is a modified baseline. This is to account for the fact that 697 players have played 839 games (or 1.2037 games per player). So the baseline value is multiplied by 1.2037 to create a new baseline. For example, Francesca Hogi played twice (season 22 and 26) and didn't meet the minimum amount of votes in either season to use her marginal accuracy vote rate. The rate in both seasons was adjusted to the merge baseline (3.7174 votes cast).
Season 22 vote accuracy game score:
1 vote cast and it was not accurate (cast for a person that was not voted out)
0.0000 vote accuracy - .1111 probable vote accuracy = -.1111 marginal rate
-.1111 marginal rate * 1 vote cast = -.1111 actual marginal rate
3.7174 baseline votes - 1 vote cast = 2.7174 projected votes
2.7174 projected votes * .4843 marginal accuracy rate for entire population = 1.3160 projected marginal rate
-.1111 actual marginal rate + 1.3160 projected marginal rate = 1.2049 total marginal rate
1.2049 total marginal rate / 3.7174 baseline votes = .3241 accuracy game score
Season 26 vote accuracy game score:
1 vote cast and it was not accurate (cast for a person that was not voted out)
0.0000 vote accuracy - .1000 probable vote accuracy = -.1000 marginal rate
-.1000 marginal rate * 1 vote cast = -.1000 actual marginal rate
3.7174 baseline votes - 1 vote cast = 2.7174 projected votes
2.7174 projected votes * .4843 marginal accuracy rate for entire population = 1.3160 projected marginal rate
-.1000 actual marginal rate + 1.3160 projected marginal rate = 1.2160 total marginal rate
1.2160 total marginal rate / 3.7174 baseline votes = .3271 accuracy game score
To calculate her vote accuracy player score (or career vote accuracy score), an adjustment is made to the baseline. Vote accuracy baseline for player score = 3.7174 merge baseline * 1.2037 games per player = 4.4746.
Francesca Hogi's vote accuracy player score:
2 votes cast and neither was accurate (cast for a person that was not voted out)
0.0000 vote accuracy - .1053 probable vote accuracy = -.1053 marginal rate
-.1053 marginal rate * 2 votes cast = -.2106 actual marginal rate
4.4746 baseline votes - 2 vote cast = 2.4746 projected votes
2.4746 projected votes * .4843 marginal accuracy rate for entire population = 1.1984 projected marginal rate
-.2106 actual marginal rate + 1.1984 projected marginal rate = .9878 total marginal rate
.9878 total marginal rate / 4.4746 baseline votes = .2208 accuracy player score
For a player who met the minimum amount of votes in each of the seasons they played, no further adjustment to their vote accuracy game score was needed. For example, Amanda Kimmel played three seasons and had a sufficient number of votes cast in each season that there was no adjustment to her score (marginal accuracy rate = accuracy score). Therefore, her vote accuracy player score is calculated the same as her vote accuracy game score, but with the accumulated inputs from all three seasons.
Amanda Kimmel's vote accuracy player score:
28 votes cast and 24 were accurate (cast for a person that was voted out)
0.8571 vote accuracy - .1458 probable vote accuracy = .7113 marginal rate
.7113 marginal rate * 28 votes cast = 19.9164 actual marginal rate
19.9164 marginal rate / 28 votes = .7113 accuracy player score
Accumulated Statistics:
Since the measurement for safety and idols is a linear weight based model, their individual game scores can be added together to calculate player scores. For example, Tyson Apostol played four seasons and his player safety score is an accumulation of his four game safety scores. Tyson Apostol's career safety score of 28.57 ranks him 76th overall out of 697 total players, which fell into the 89.2 percentile.
Total Player Scores:
A total player score can be calculated by taking their aggregate percentile rankings in the six key performance measurements and weighting them, to create a composite score that reflects their overall performance in ALL of the seasons in which they played Survivor. For example, Russell Hantz had a weighted player score of 85.7.
Vote Cluster percentile rank = 88.3% x 100 = 88.3 x 11.64% = 10.2781
Vote Accuracy percentile rank = 97.2% x 100 = 97.2 x 12.99% = 12.6263
Challenge Performance percentile rank = 41.6% x 100 = 41.6 x 0.61% = .2538
Immunity Idols percentile rank = 97.2% x 100 = 97.2 x 25.66% = 24.9415
Vote Avoidance percentile rank = 70.1% x 100 = 70.1 x 22.55% = 15.8076
Safety Earned percentile rank = 82.1% x 100 = 82.1 x 26.54% = 21.7893
Weighted player score = 10.2781 + 12.6263 + .2538 + 24.9415 + 15.8076 + 21.7893 = 85.6966
All individual player scores were calculated and ranked against the entire population of player game scores. A percentile rank was calculated for the ordinal rank. Russell Hantz's player score of 85.7 ranks him 29th overall out of 697 total player scores, which fell into the 95.9 percentile.
To view a full list of player scores, please click button below:
*** updated through season 46 ***
Comments