Predicting the “Dateability” of Survivor Players

With all the great pre-season Survivor coverage happening on RHAP, I was inspired to do my own pre-season project.

We have all witnessed the awesome power of “Nicoletradamus” when it comes to winner picks, but what about predicting the “dateability” (i.e. whether a contestant makes the merge) of Survivor contestants?

To address this age old question, I went back through the RHAP archives and re-listened to the last six Survivor preview shows (Blood vs Water to Kaoh Rong) from Nicole and Rob Cesternino to collect their merge predictions about each cast member.

I also took a look at Angie Caunce’s boot predictions from these seasons to determine who she thought ahead of the game was likely to make the merge and those that would go out prior to the merge and hence be undateable.

So who’s psychic powers reigns supreme?

Is it the great Nicoletradamus? “The System” developed by Angie Caunce? Or the Rob that Sucks?

Let’s find out.

Dateability Predictions, Season by Season

The table below shows the percentage of correct predictions made by Rob and Nicole for each of the 6 seasons. Unfortunately, I only have Angie’s results for Worlds Apart, Cambodia and Kaoh Rong.

	Rob	Nicole	Angie
Kaoh Rong	44.4% 8 out of 18 correct	50% 9 out of 18 correct	61.1% 11 out of 18 correct
Cambodia	50% 10 out of 20 correct	55% 10 out of 20 correct	60% 12 out of 20 correct
Worlds Apart	61.1% 11 out of 18 correct	61.1% 11 out of 18 correct	55.6% 10 out of 18 correct
San Juan Del Sur	55.6% 10 out of 18 correct	66.1% 11 out of 18 correct	N/A
Cagayan	66.7% 12 out of 18 correct	66.7% 12 out of 18 correct	N/A
Blood vs. Water	80% 16 out of 20 correct	75% 15 out of 20 correct	N/A

Analyzing the Season by Season Results

In Survivor Blood vs Water, Rob correctly predicted merge versus pre-merge for 80% of the cast, while Nicole correctly predicted 75%.

The only mistakes Rob made were with John Cody, Laura Morett, Katie Collins and Rachel Foulger.

To put this into context, if we simply picked “merge” for every cast member, then we’d get 55% of these predictions correct since 11 out of 20 cast members made the merge.

That is pretty amazing.

When I first saw this result, I thought perhaps Rob did so well because the cast is partly returning players so he has a lot more information to go on than with a newbie season.

However, this trend did not continue for Cambodia where he only successfully predicted 50% of the cast’s fate.

Cambodia was a particularly tricky season to predict.

Besides all the swaps and a pre-merge med-evac, Rob and Nicole had a lot of personal connections with the cast that perhaps muddied their psychic prowess. Further, the merge happened earlier than any season in history. No one outside of perhaps Jeff Probst could have seen this coming.

It does appear that Rob and Nicole’s predictions have got worse over time. Perhaps all the podcasting has diminished their psychic abilities or perhaps Blood versus Water was an anomaly.

Only time will tell :-).

Overall Results

Here we have how Rob and Nicole have done overall with predicting the fate of all 112 contestants across 6 seasons and Angie with the 3 seasons I had data for.

 

Rob	59.8% 67 out of 112 contestants
Nicole	60.7% 68 out of 112 contestants
Angie	58.9% 33 out of 56 contestants

 

Analyzing the Overall Results

Through the 6 seasons I collected data for, there was a total of 112 castaways with 70 (62.5%) making the merge and 42 (37.5%) going out pre-merge. So, if prior to a season, we simply predicted everyone was going to make the merge, across these 6 seasons we’d be correct 62.5% of the time.

This is actually higher than both Rob, Nicole, and Angie’s overall scores, but it’s not much fun to always “predict” merge.

Because the number of people that make the merge each season is larger than the number that do not, every time you guess that someone is pre-merge, you are taking substantial risk with that prediction. The weighted probability of being correct is much lower for guessing pre-merge versus merge.

It’s basically equivalent to flipping an unfair coin where the heads side weighs more and as a result appears 62.5% of the time rather than an expected 50% with a fair coin. For a coin like this, it’s always better to call heads, just as with Survivor dateability, it’s always better to guess merge.

Making pre-merge and merge picks is further complicated by the fact that Rob and Nicole do not know prior to the season what the split will be in terms of merge and pre-merge. Angie does not have this information either, but she predicts boot order, so she’s not effected by the variability in jury size.

Based on this knowledge, we can take a look at who is the bigger risk taker between Rob and Nicole. That is, who is willing to put it all on the line and guess pre-merge more often?

Not too surprisingly, there’s a strong positive correlation between Rob and Nicole’s cast assessments and a smaller positive correlation between Angie and Rob.

This is because Rob and Nicole are not making their predictions in independent vacuums, they tend to influence each other’s opinions during the cast preview.

Nicole slightly edged out Rob’s overall score by predicting the dateability of one more contestant correctly than Rob.

Although this is not a statistically significant result, I am willing to give Nicoletradamus the win :-).

Final Thoughts

Hopefully you enjoyed this little pre-season fun.

Predicting anything about Survivor is difficult, especially when all you have is a CBS bio and a short video.

Rob, Nicole and Angie stick their neck out each season to make bold predictions and craft narratives based on very limited information, all for our entertainment.

I greatly appreciate the effort :-).

I think a fun follow-up project, that would likely require a lot more work, would be to go back through these podcasts and detail all the bold predictions made by Rob, Nicole and Angie and see if any of them panned out.

For example, Nicole correctly predicted that Peih-Gee and Abi would not get along in Cambodia.

If you have any thoughts, let me know in the comments.

Cheers!