[Complex] Survivor: Barely Survivor COMPLETE!

[Crazy]

So quite a few years back, I developed a scoring system for Survivor games, which I have applied both to the real show and to games played on MS. Unfortunately, I have faded out of the MS community recently so I haven't been keeping up with most recent games. However, I am planning on correcting this and eventually getting all Survivor games scored that I am able to. I also owe a big credit to Supervisor Silverclaw for displaying enough of an interest in this to get me interested again; and as a bonus, he also scored the previous game run on the site. (Flash Mob)

You can see the master spreadsheet here, which includes scoring for all the games that have been done, and a few overall player rankings and other stats that some might find interesting. Keep in mind, of course, that this is a work-in-progress and several recent games are not yet included.

If you're a fan of the Survivor TV show, you might enjoy this spreadsheet where I have scored all 40 seasons of US Survivor as well as the 5 modern Australian seasons.

If you're enough of a nerd that you actually want to to understand how this works, then see below:

Spoiler: Scoring Rules

Overview

In this scoring system, each Survivor player is scored based on an average of their scores for each individual Tribal Council they attended. This score is then multiplied by a certain percentage (from 10-100%) based on how long they survived in the game.

For each Tribal Council, all surviving players will be scored with a 2, 3, 4, or 5, depending on factors such as whether they voted in the majority or not, whether they received votes, and even depending on whether the vote could be considered "dynamic" or not. (More on this later.) The eliminated player will always receive a score between 0 and 1, depending on how many votes they did NOT receive. For example, a player who was voted off with a 5-2 vote would receive a score of 2/7 (0.29), because 2 of the 7 votes were for another player. A score of "0" is typically reserved for quits, and a score of "1" for those special circumstances when a player is eliminated without receiving any votes that round.

For each player, after taking the average score of all Tribal Councils they attended, it is then multiplied by 20, so that the theoretical maximum score is 100. Just to note, this maximum score is impossible to achieve from a practical standpoint, and in fact, nobody on MS has ever scored above 90 - though four winners on the real show did.

Finally, we take a percentage of the raw score based on how long the player survived. For players who made the Final 3, this percentage is 100%, so nothing is deducted from their raw score. For everyone else, 5% is deducted for every placement worse than 3rd place. For post-mergers, this caps out at 55%, like this:

• 100% - Final 3
• 95% - 4th place
• 90% - 5th place
• 85% - 6th place
• 80% - 7th place
• 75% - 8th place
• 70% - 9th place
• 65% - 10th place
• 60% - 11th place
• 55% - 12th (or worse) place

For pre-mergers, instead of counting down, we count up in increments of 10%, depending on how many Tribal Councils the player attended. This maxes out at 50%:

• 10% - 1 Tribal Council
• 20% - 2 Tribal Councils
• 30% - 3 Tribal Councils
• 40% - 4 Tribal Councils
• 50% - 5 (or more) Tribal Councils

After taking the raw score and applying the necessary multiplier, we arrive at the final score, which every player is ranked by. Now that we have the overview, we'll go into how each specific Tribal Council is scored.

Specifics

Earlier in this post, I mentioned that each surviving player in a Tribal Council will be scored with a 2, 3, 4, or a 5. To determine what score to give them, we first need to determine whether the vote should be considered "stable" or "dynamic." (I know this part sounds bizarre, but bear with me.)

In practical terms, an "dynamic" vote is one that could possibly be considered a #BigMove, a power shift, a blindside, or something along those lines. To be considered dynamic, a vote must pass both of the following tests:

"Pagonging" Test - The vote must not eliminate someone who was already in the minority.

To pass this test, the eliminated player must not have been on the outside of an already established majority. This is based on who voted together on the most recent dynamic Tribal Council. For example, if the Final 9 featured a 5-4 vote, and at the Final 8 Tribal Council, the players eliminated someone else apart from the 5 that voted together, then that vote would fail this test, and would be considered stable.

A post-swap or a post-merge tribe that hasn't gone to Tribal Council yet can still fail this test if the boot was a member of the minority based on previous tribe divisions. For example, if a tribe swap resulted in one tribe having 5 "Brawn" and 2 "Beauties," then the vote would fail this test if one of the Beauties were voted off. If there has been more than one swap/merge, then the eliminated player must be part of EVERY possible minority (e.g. looking at both pre-swap and post-swap tribes) in order for the vote to fail this test.

"Landslide" Test - The vote must not be a landslide majority vote.

A landslide vote is when far more players than necessary vote with the majority. If this happens, the vote is considered stable. What counts as a "landslide" vote first depends on how many players are at the Tribal Council:

• Votes with 5 or fewer players cannot fail the Landslide Test. They are only considered stable if they fail the Pagonging Test.
• Votes with 6-8 players simply must not be unanimous in order to pass this test. 5-1, 6-1, and 7-1 votes fail, but anything else is okay.
• Votes with 9 or more players must have at least 3 players voting in the minority in order to pass this test. This includes split votes. For example, an 8-2 vote fails, but even if the majority splits the vote so it ends up 4-4-2, it still fails.

If a vote fails either of the two above tests, then it is considered stable. Stable votes are very easy to score, because each player can only score a "3" or a "4."

Stable scoring

• 4 - A player will score 4 if they had a perfect round, meaning they received no votes and voted in the majority.
• 3 - A player will score 3 if they received any votes OR voted in the minority.

Stable Tribal Councils have a lower range of scores, because the game is not typically built on them. The reason why it is not possible to score a "5" for an stable Tribal Council is so that players who float along while remaining at the bottom of the majority alliance don't get a higher score than they deserve.

Dynamic scoring

If a vote passes both the Pagonging Test and the Landslide Test, then it is scored as follows:

• 5 - A player will score 5 if they had a perfect round, meaning they received no votes and voted in the majority.
• 4 - A player will score 4 if they received EXACTLY one vote.
• 3 - A player will score 3 if they received more than one vote.
• 2 - A player will score 2 if they voted in the minority.

For pre-merge Tribal Councils, the criteria for "3" and "2" are switched, because typically receiving votes pre-merge is worse than receiving votes post-merge, while voting in the minority post-merge is more critical.

If more than one criteria applies - for example, a player BOTH received votes AND voted in the minority - then they will receive the lowest applicable score.

Final Tribal Councils

For the Final Tribal Council, winners will receive a score depending on the percentage of jury votes they receive. The formula is:

• 2.5 + [2.5 * (fraction of jury votes received)]

This results in a winner receiving a score between 3.75 and 5 for the Final Tribal Council.

Losing finalists, on the other hand, will receive a negative score, depending on how many jury votes they receive:

• -10 * [0.5-(fraction of jury votes received)]

A finalist who receives no jury votes will score a -5 for the Final Tribal Council, while a finalist who receives almost half the votes will receive a negative score much closer to 0.

The Final Tribal Council scoring is weighted to account for different game sizes/lengths. For finalists, 10% of their final score will be based on their Final Tribal Council, and the remaining 90% from the other Tribal Councils they attended.

There are more rules for special situations such as ties, idols, and advantages, that I don't feel are necessary to include in this post, since I don't expect most people's interest levels will run that deep. If you are so curious, additional rules are listed on the "Rules" tab on the spreadsheet, and I am also open to any questions or comments.

[Crazy]

Hi Crazy, nice to see you. I have a question about how you handle "strategic minority votes" i.e. votes that for various reasons are knowingly cast in the minority. A lot of times people are knowingly voting in the minority to send a message or to create paranoia or to pander to the Jury or for any number of other reasons. I know it gets very messy to take these into account, especially if players are voting for the people they legitimately want to see gone, knowing they won't go, but I just wanted to know how you've handled this in the past and how much it should be taken into consideration. I know that split votes in the past have not counted against players, but I wanted to know if this logic extended to other "strategic minority votes".

To add to what Silverclaw said, the criteria I use is:

1. The player had knowledge of who was actually getting voted off.
2. The player fully approved of the elimination that actual occurred.
3. The vote against the majority was made for purely strategic reasons. (In other words, not out of loyalty or sentimentality.)

This is very much a case-by-case basis as Silverclaw said and is definitely the most difficult part of implementing the scoring system. To add to this, for most recent games I've scored, I haven't accounted for split votes or other strategic minority voting. Really, anything post-One World I'm not confident on, apart from Flash Mob and Barely Survivor. So I may need some insider knowledge on these games at some point.

Fyi you mixed the identities on Gummy Bear and Snow Leopard in the chart there, tho the boot timing and thus the stats should be good still I assume.

Thanks, the image is too much effort to fix but I have corrected the mistake in the spreadsheet.

is the first TC only taking into account the landslide rule or something?

The first TC was obviously very unusual, so it was a challenge to score. The landslide rule does not apply, though, in any Tribal Council with less than 6 players. And since there was no established minority since it was the first TC of the game, all of the vote-offs there were considered dynamic by default.

Yo Crazy, weird question. Are Lake Gaston and Willamette Valley going to end up on the sheet eventually? Neither are on the game list.

I can certainly score them if anyone has the voting history available. I have lately been considering, though, which games to include as "canon" for the scoring system. I would like to include all long-form forum Survivor games with voting records available, but I would prefer not to include any marathon games. Obviously, the games you mentioned are pretty much in their own league, so I'm willing to follow public demand regarding whether those games should be included in the overall ranking. I do have scores available for Blowing Rock, which I have removed from the overall ranking, but I can easily add them back. Similarly, I also removed scores I had done for Summer of Spookz and my own game Expect the Unexpected, since both of those were essentially Big Brother games.

[Crazy]

The first TC was obviously very unusual, so it was a challenge to score. The landslide rule does not apply, though, in any Tribal Council with less than 6 players. And since there was no established minority since it was the first TC of the game, all of the vote-offs there were considered dynamic by default.

interesting, aren't first tc's often stuff like "all aboard the inactive guy bandwagon"?

idk, haven't read much survivor, but if that's the case i'd argue it's pretty non-strategic

Generally, yeah, and in many of those cases the votes would be scored as stable. The first TC this game was an exception due to the small tribes. But any vote that ends up being 5-1, 6-1, 7-1, etc. would be scored as stable due to failing the Landslide Test. Even in this game, most of the early pre-merge votes after the first TC were unanimous and thus scored as stable.