iiinova — where your digital self keeps evolving

Dan Oehm @danoehm.bsky.social · 02/10/2026

Random #Survivor stat - There hasn't been a season where every player has attended tribal council within the first two votes since S40. It almost happened in #Survivor51 but since Lewis was on Exile Island he missed the first vote and so only 20/21 players have been to tribal council.

161

Dan Oehm @danoehm.bsky.social · 02/10/2026

#Survivor51 confessionals per hr I don't have a variable in the model to account for severing part of your thumb, so Brady now has the second highest edit behind Lewis.

041

Dan Oehm @danoehm.bsky.social · 02/10/2026

#Survivor51 e2 data is up and ready to download. 💾 survivorstatsdb.com/download 😺 github.com/doehm/survivoR And the confessional timing app is ready to go for e3 on desktop and mobile. 💬 survivorstatsdb.com/confessional...

survivorstatsdb.com

SURVIVOR STATS DB - Confessional App

000

Dan Oehm @danoehm.bsky.social · 01/10/2026

They give people options and opens up the game play. It's a nice change!

020

Dan Oehm @danoehm.bsky.social · 30/09/2026

In the New Era, only 20% of the first tribal councils were NOT unanimous votes, meaning there was at least one player not voting with the majority and not voted out. The classic era was about 65%. Larger starting tribes are better #Survivor

182

Dan Oehm @danoehm.bsky.social · 26/09/2026

#Survivor51 confessional time and CPH Lewis, Kristin, and Devin are the standouts wrt CPH in episode 1. This accounts for Lewis's exile but not the journey-like challenge which would give him a few more cfnls. Similar I didn't account for Kristin's negotiation. I'll adjust the model next week.

021

Dan Oehm @danoehm.bsky.social · 26/09/2026

#Survivor51 data is now up! Apologies, for the delay. In personal news I have just relocated to the UK 💂 and arrived about 11 hrs ago! I've caught up and back to regular programming. I appreciate all those that reached out to see where it's at. 💾 github.com/doehm/survivoR

github.com

GitHub - doehm/survivoR: Data sets from Survivor in R

Data sets from Survivor in R. Contribute to doehm/survivoR development by creating an account on GitHub.

130

Dan Oehm @danoehm.bsky.social · 30/05/2026

That's awesome. Nice work, Josh!

130

Dan Oehm @danoehm.bsky.social · 27/05/2026

Git: + github.com/doehm/survivoR Websites using the data that are worth checking out: + survivor-reference.com + survivorelo.com + survivordle.com I commit to a 24hr window to correct issues. My record is 12 minutes.

github.com

GitHub - doehm/survivoR: Data sets from Survivor in R

Data sets from Survivor in R. Contribute to doehm/survivoR development by creating an account on GitHub.

020

Dan Oehm @danoehm.bsky.social · 27/05/2026

survivoR v2.3.12 is released and includes complete US50 and AU12 data. Any issues please log an issue on git. The R package is approaching 30k downloads which is wild! More considering the untracked git installs and xlsx downloads. If you use it I'd love to hear about it. #Survivor #Rstats

150

Dan Oehm @danoehm.bsky.social · 27/05/2026

42 is wild with Drea holding 4 trinkets at one time.

010

Dan Oehm @danoehm.bsky.social · 26/05/2026

I've added a new data set 𝚊𝚍𝚟𝚊𝚗𝚝𝚊𝚐𝚎_𝚝𝚒𝚖𝚎𝚕𝚒𝚗𝚎 to the survivoR package. No new info but makes analysis way easier e.g. now you can choose a point in time and see which advantages were in play. Want to guess which season/episode had the most advantages/idols in play at TC? #Survivor

140

Dan Oehm @danoehm.bsky.social · 22/05/2026

Rizo has held an idol for 18 tribal councils. Second only to Ozzy on 19 and tied with Tai. Of those three, only Tai has played an idol successfully (three in fact) #Survivor

060

Dan Oehm @danoehm.bsky.social · 22/05/2026

That's a wrap on #Survivor50. Preliminary data is up and ready to go. There are still a few things to add, which I'll complete in the next few days. The next release will also include a new advantages table. Download the xlsx from Git. Any corrections, let me know. 🔗 github.com/doehm/survivoR

020

Dan Oehm @danoehm.bsky.social · 18/05/2026

At this point starting with 2 tribes of 10 would be a bigger twist than anything in 50.

020

Dan Oehm @danoehm.bsky.social · 18/05/2026

Mildly interesting - Cirie has voted for the winner of the season 4 out of 4 times she has been on the jury. Will she make it 5 for 5? Only 3 people in either the F5 or the jury have never voted for the winner. #Survivor #Survivor50

051

Dan Oehm @danoehm.bsky.social · 16/05/2026

I'm pretty sure that Rizo has never attended a tribal council up to the F5 without holding an idol - 15/15. Rizo is 3rd on the list of players who have held at least 1 idol at tribal council. Ozzy is 1st with 19 tribals, and Tai is 2nd on 18. #Survivor #Survivor50

0101

Dan Oehm @danoehm.bsky.social · 15/05/2026

While Ozzy didn't beat Boston Rob's IIC record, he still holds the highest overall career score #Survivor #Survivor

010

Dan Oehm @danoehm.bsky.social · 15/05/2026

Jon, Joe, Aubry in the FTC is my pick too. Rizo loses fire again!

020

Dan Oehm @danoehm.bsky.social · 15/05/2026

Just to give the model some cred - About 7/10 times, one of the top 2 players wins - The most likely player to win has won in the last 6 seasons.

020

Dan Oehm @danoehm.bsky.social · 15/05/2026

Look at this. One of the strongest indicators of winning is confessionals, particularly the difference between the observed and expected values, and the trend leading into the finale. While the other 4 have trended down, Jonathan's trended up and above the line.

110

Dan Oehm @danoehm.bsky.social · 15/05/2026

What I like about this chart/model is that it picks up on patterns that you may have missed. What it misses is all the social elements that aren't captured in the data. To adjust for that, I listen to who others think is the favourite is and apply a prior. Even with the prior Jonathan is on top.

100

Dan Oehm @danoehm.bsky.social · 15/05/2026

How to read the chart: - It's an estimated probability, not a prediction - There's still a 60% someone other than Jonathan will win - The data is saying that Jonathan is in a more similar position to other winners in the past.

100

Dan Oehm @danoehm.bsky.social · 15/05/2026

That's quite a swing from the last episode! For the record, I would still put my money on Aubry winning, but this is why I like looking at these charts, it shows things you may be missing. Jonathan is actually in a good position. But it always comes down to the pitch. #Survivor #Survivor50

371

Dan Oehm @danoehm.bsky.social · 10/05/2026

If you want to look into it yourself, here's the curated data: docs.google.com/spreadsheets...

docs.google.com

Final 5 Vote

010

Dan Oehm @danoehm.bsky.social · 10/05/2026

The only rationale I can see is that by not voting with the majority the player doesn't burn any bridges with who was voted out and stacks the jury in their favour. Maybe. Honestly, I'd err on the side of this is mostly down to just normal variation. 9/

120

Dan Oehm @danoehm.bsky.social · 10/05/2026

There's a chance that this is just normal variation. If you asked me to guess, I would have said the chances would be lower, not higher. But this is saying that either the final 5 vote isn't any more important than other votes, or NOT voting with the majority can increase your chances 8/

110

Dan Oehm @danoehm.bsky.social · 10/05/2026

- The final 5 is arguably more important since S35 (forced fire). But the way I see it, at the final 4, the only way a player can not be in the majority and survive is with an idol play. If it's a tie, it goes to fire anyway. So the final 5 vote is still that key vote. 7/

110

Dan Oehm @danoehm.bsky.social · 10/05/2026

But I dunno, that's pretty wild, so here are the caveats: - The 10 include the infamous Chris Underwood case, an outlier in itself 6/

110

Dan Oehm @danoehm.bsky.social · 10/05/2026

Simulating the possible scenarios, I get the probability distribution below. The probability of observing 10 or more cases is ~2.8%, which is pretty low. To summarise, there is some evidence to suggest that NOT voting with the majority may INCREASE the players' chances of winning 🤨 5/

110

Dan Oehm @danoehm.bsky.social · 10/05/2026

If we assume there's no effect on the chances of winning, we would expect ~6 players to still win the season. We have seen 10 players win after NOT being in the majority at the final 5 vote. Counterintuitive to what I would expect. The question is, is this significant? 4/

110

Dan Oehm @danoehm.bsky.social · 10/05/2026

I'll focus on the majority. The hypothesis is, if a player doesn't vote with the majority at final 5 but survives the tribal council they have a lower chance of winning the season. 3/

110

Dan Oehm @danoehm.bsky.social · 10/05/2026

There are two main ways of looking at this: - Who voted for the F5 boot and - Who voted with the majority The difference is that a player could have voted with the majority but the votes were nullified and someone else went home, therefore, unsuccessful boot but still in the majority. 2/

110

Dan Oehm @danoehm.bsky.social · 10/05/2026

Ok, this is kind of interesting. Consensus is that the final 5 vote is one of the most important in the game. I wanted to see how the final 5 vote affects the players' chances of winning. The results are not what you'd expect. 🧵 1/ #Survivor

131

Dan Oehm @danoehm.bsky.social · 08/05/2026

Random stat: There have been 7 individual immunities up for grabs, and they have been won by 7 different people. First time in Survivor history. How about that. #Survivor

0101

Dan Oehm @danoehm.bsky.social · 08/05/2026

Yeah, anything can happen from here. I could see Tiff winning. Jonathan has had a good game so far, honestly wouldn't be surprised if he takes out the season either.

010

Dan Oehm @danoehm.bsky.social · 08/05/2026

There's been a shift in the leaderboard with Ozzy voted out. Aubry now has the best odds, followed by Jonathan. I'm still hopeful this is Cirie's season though. #Survivor50

Probabilities of winning estimated from confessionals, challenge success, voting patterns, advantage plays and jury favorability.

110

Dan Oehm @danoehm.bsky.social · 08/05/2026

The top 3 players on the CPH chart are now Devens, Cirie, and Jonathan, the only 3 still in the game with a +ve edit. Jonathan has the largest difference between obs and exp. He's had a couple of really good episodes tbh. #Survivor50

010

Dan Oehm @danoehm.bsky.social · 05/05/2026

5 year old: Dad, can we watch Ghostbusters every day after school like you did when you were a little boy? But you have to say what they say before they say it. Me: Yes 🥹

030

Dan Oehm @danoehm.bsky.social · 02/05/2026

Yep those challenges are included

000

Dan Oehm @danoehm.bsky.social · 02/05/2026

I don't count journeys or auctions as challenges either, and exclude the ones where players aren't eligible. The only other thing that stands out to me is the Q&A type challenges. There are 3 on Cirie's list. Do you count those? I'll have a closer look when I get a chance.

000

Dan Oehm @danoehm.bsky.social · 02/05/2026

I *think* (assuming I haven't missed any edge cases or a data descrepency - please reach out if so) the next challenge will be Cirie's 100th! This includes challenges where she sat on the bench. Imagine if she won individual immunity. Huge. #Survivor50

271

Dan Oehm @danoehm.bsky.social · 02/05/2026

Standings after episode 10. Ozzy is still in the lead. What this means is that all the data points are saying that Ozzy is in a more similar position to players who have won in the past than the other players. #Survivor50 #Survivor

011

Dan Oehm @danoehm.bsky.social · 02/05/2026

#Survivor50 confessionals/hr At this stage Devens has the largest over-edit wrt to CPH. It's also interesting that Rizo is low given his position. The mean CPH for a winner at episode 10 in the new era is 4.2. All are under, but Ozzy is closest to that benchmark. #Survivor

141

Reposted by Dan Oehm

coolbutuseless @coolbutuseless.bsky.social · 29/04/2026

#RStats Playable Doom! Figured out some key handling stuff in my custom graphics device, so I can now actually control the game! Content Warning: I'm not good at doom :)

1123

Dan Oehm @danoehm.bsky.social · 27/04/2026

That's how I've been thinking about it as well. And yeah, honestly, just give me a straightforward episode!

020

Dan Oehm @danoehm.bsky.social · 27/04/2026

Two for the price of one! Does two names on one parchment equate to one or two votes? What if they had to write the names on separate parchments? If we assume one of those votes was a forced vote by production, does that mean Christians' vote for himself doesn't count? Important questions!

110

Dan Oehm @danoehm.bsky.social · 27/04/2026

How did you handle the tied destinies episode? Did you count it as 1 or 2 VFB? I think there's an argument for both options, but annoying to have to make an exception.

330