@nbashaw Ha, good eye. I added that because I had to do that search for our own internal unicorn league so many times and never found any good results. "Build the search results you want to get", right? :)
You think Product Hunt's valuation will be only $30M one year from now, @buster!? Challenge accepted. :)
This is an awesome idea, although I wish we could place shorter term bets, where the "reward" of winning comes much less than 1 year out. Of course the challenge with private companies is that most data isn't public. Any thoughts on this, @buster?
@rrhoover I do hope you beat my prediction. At 12 guesses so far, looks like my guess is the median (that's what the numbers in the gray bar are, which isn't super clear yet)... so it's not just me.
And yeah, I agree that shorter term bets would make for a better engagement loop. I've gone back and forth on ways to make this game move faster... I could make the 1-year time frame something you could change. Would you feel comfortable making predictions in 3 month chunks? 1 month chunks?
I'm open to ideas!
@rrhoover Also... about the problem with the data not being public. That's part of why I built this. There is information out there in many cases (# employees, amount raised, whispers on the streets) that could be reverse engineered into a valuation. Hopefully, by picking the median of the guesses (I haven't started weighing high confidence guesses more than low confidence ones, but will try when there's more data), I think we may actually approach a realistic "best guess" for this data. Wisdom of crowds, etc.
Part of why I'm releasing this super early is so I can see what the data looks like and start playing around with it to find out what it all actually means.
@buster unless you measure something more granular than overall valuation, it may be hard to simple reduce the prediction time from 1 year to 1 or 3 months.
Other tangential metrics to valuation could be number of employees, Alexa ranking, or Mattermark score (cc @daniellemorrill, @SparksZilla).
What do you think about allowing anonymous guesses, @buster? I imagine many VC's would enjoy playing with this but not if their name is publicly tied to their prediction.
@rrhoover Ooh, I like the idea of predicting other more verifiable pieces of data! That's interesting... I'll think on it.
As for anonymous guesses... I think you're probably right, but my own selfish interests in building the game make me want to keep it all public. Putting your name on the line is actually the part that makes this exciting to me (since there's no money involved).
@buster agreed. The same is true for Product Hunt. We could anonymize all upvotes but it would feel less like you're a part of a "tribe" and the community side is of course important.
This is a very simple prototype of a game to see how good we are at predicting the current and future valuations of existing companies. I'd love any and all feedback and am open to questions.
@buster I love the idea. It seems like it could be a lot of fun. My only issue is that the Twitter auth allows this to post for us? I don't mind sharing things, but I hate writing a blank check to things I may end up not using / liking. I prefer to make the decision a la carte. Any reason why we have to give that explicitly up front?
I usually just don't signup for things that allow themselves to post on either Twitter or FB. Given your position with Twitter, I thought it might be interesting to discuss this with you.
@EAWharton I totally understand. When I set up the token I didn't know if I would want to allow people to tweet directly from the app... but now that I'm not doing that I'll go and change the permissions to read-only.
Question for Product Hunters: how should the scoring system be designed? My first thought is that you should get reputation points of some sort for being able to be near the median on current valuations, and the valuation a year from now (based on the median current valuation in a year). It could get more creative though... thoughts? The goal being to make the game feel more exciting right now, even though each guess takes a year to verify.
Report
@buster well, it looks like Current numbers are more provable today than the 1 Year numbers (fact vs prediction), so points for that make sense. But say once you reach a 100 or so guesses for a company, the median should theoretically lock in on the current truth (even though in this industry the 'truth' changes quite frequently), which makes it quite easy to get points for basically restating the prevailing fact. Also, how does the timelock work? if I guess in Sept, and you guess in Dec, our numbers will be off because information has changed - which makes distributing points tricky because we'll have different 1 year marks and predictions.
P.S. Have you considered hiding the results until a user submits a guess? This way it's a game you play rather than watch....just a thought. Love the idea!
This is very cool - I've thought about different small products in the area but none as simple. Really good stuff.
@buster I would definitely add at least a 5y prediction. I can predict a number of companies having higher val next year but then dropping down or the other way around.
@buster or 3, but 1 year seems like tomorrow to me.
Mostly probably because valuation data points are very discreet and randomly placed in the life of a company.
@stefanobernardi I definitely see your point. I'm curious to see how much crowdsourced valuation moves relative to actual valuation events. Once I have a few months data it will be possible to learn a lot more about how this stuff moves.
@buster this is a great start! I would love if this was similar to fantasy sports. So we get to pick a team initially . (Maybe require leagues so i can only play with friends or play for a specific prize). I get to draft companies based on public valuations and funding raised to date. So if i want a company thats close to IPO i have to draft them higher. Or put more of my cap (the over all amount of money available to me ) in the deal. if i pick poorly or if i go long on all my bets i don't have money for new deals that come up .( similar to waivers where i have to drop a player to add more players that can hep me win now). I can get friends to let me in on deals at smaller money. ( so we have deal flow) When new valuations are released if my portfolio has appreciated in value i now have more money to play with on new companies.
(Possibly use external signals to ascertain valuations such as getting on the front page of Product Hunt or Hacker News. Getting into a top accelerator or Mattermark score. )
Eventually maybe create funds where if I've had some good hits i can team up with other players to create funds that can buy more and give returns .
@solfrombrooklyn I like a lot of these ideas. The tough part is that this is not going to have the trusted external source of data to provide *source of truth* for each period of time. I still think there may be something here. Maybe we keep the core version of the site as-is, and let the "league" live off of the consensus data that people can add for free.
Lots to explore. I think after a month or so we'll have enough data to look at to see how far we can go with it.
Do you have a favorite fantasy league? Which one is implemented in the cleanest, least-gameable, way? Or are they all about the same?
@buster Most fantasy leagues are the same. the differences arise when you look at daily leagues vs seasonal leagues. In the daily leagues you have to play more like a GM where you have a budget and have to pick players based on the budget. In seasonal if a player is available at my draft slot i get him. So seasonal leagues are more like the draft where daily leagues are like putting the roster together.
An example of a daily league site is draftkings.com
While Yahoo, espn and the NFL sites have seasonal leagues.
Also there is obviously a difference when you play for something. It keeps people honest. You really try to win when theres something on the line. Even if its just bragging rights.
Report
Great idea. Even though I never played any Fantasy League and I'm probably horrible at predicting, this seems like a fun challenge. I noticed a little error on your site. After clicking on a user, it shows me I'm logged in as that person even though I'm not logged in.
Lex
Why Are We Yelling?
Product Hunt
Why Are We Yelling?
Why Are We Yelling?
Product Hunt
Why Are We Yelling?
Product Hunt
Why Are We Yelling?
Strikingly
Why Are We Yelling?
Why Are We Yelling?
Unreplied
Why Are We Yelling?
Unreplied
Why Are We Yelling?
Blab
Why Are We Yelling?
Blab
Why Are We Yelling?