Live data from Hacker News

Show HN: March Madness for Hackers

codersbracket.com

51–60 of 63 posts

Re: Show HN: March Madness for Hackers

#51
post #11

This is awesome! I went to UConn, so I'll obviously hard-code a UConn win there. But I heard on NPR today that, historically, a 12 seed has a 50% chance of upsetting a 5 seed in the first round. If I'm biasing towards the higher seeds everywhere else, that means my code looks like this: function (game, team1, team2) { var winner_team = "Connecticut"; if (team1.name == winner_team) { team1.winsGame(); } else if (team2…

WOuldn't this not take into account 12 seeds winning? Since you do team1.seed > team2.seed before checking if team2.seed is 12. Also, this would mean 12 seeds would go all the way anyway, you're not limiting it to the first round only.

Yeah, trg2 would need to put all the edge case rules towards the beginning. That, or just put the basic rules at the beginning and have separate conditionals at the end to handle the edge cases.

Re: Show HN: March Madness for Hackers

#53
post #25
post #24

Could you add the last couple years' brackets and stats as an algorithm tester? Would be cool to run your algorithm against previous years to see how it would have done so you can refine it before submitting. I'm imagining coloring the final bracket green/red for correctly/incorrectly predicted games in the simulation.

Hey yeldarb, Nimit here from Fullstack. Good idea! We did have the 2013 data up last week before the 2014 bracket came out yesterday but will work on adding more years and the comparison to how the previous brackets turned out!

Any idea if you'll have this functionality up before the bracket deadline ends? I'd love to be able to build out some functions and test them on previous years before the contest officially starts.

Re: Show HN: March Madness for Hackers

#54
post #43

Would also be cool to be able to include more detailed data. There's some useful data points exposed at http://madness.io including Tempo-free stats and (not advertised) json endpoints, http://madness.io/teams/uaa.json Full disclosure: I released this site a few weeks ago, mostly as a side project. I submitted to Show HN but it didn't get much traction. Might be of use though.

Nice site. I agree, I'd love to be able to use some of your data points in my algorithm.

Re: Show HN: March Madness for Hackers

#55
Hey Guys!

I just saw this today and got pretty interested. I was interested in implementing a 'weighted' scoring function, but needed a way to save state somehow. I ended up finding a pretty fun solution using the console. It's really fun to play with the weights and see how the bracket changes!

https://github.com/hstove/Coderbracket-Weighted-Scorer

Re: Show HN: March Madness for Hackers

#56
post #54
post #43

Would also be cool to be able to include more detailed data. There's some useful data points exposed at http://madness.io including Tempo-free stats and (not advertised) json endpoints, http://madness.io/teams/uaa.json Full disclosure: I released this site a few weeks ago, mostly as a side project. I submitted to Show HN but it didn't get much traction. Might be of use though.

Nice site. I agree, I'd love to be able to use some of your data points in my algorithm.

There's a few exposed json endpoints for data, the matchup page, team page, and player page.

eg team: http://madness.io/teams/fak.json eg player: http://madness.io/players/1772.json eg matchup: http://madness.io/matchup/dae/vs/oad/1395332100.json

Re: Show HN: March Madness for Hackers

#57

Earlier quoted context omitted.

WOuldn't this not take into account 12 seeds winning? Since you do team1.seed > team2.seed before checking if team2.seed is 12. Also, this would mean 12 seeds would go all the way anyway, you're not limiting it to the first round only.

Yeah, trg2 would need to put all the edge case rules towards the beginning. That, or just put the basic rules at the beginning and have separate conditionals at the end to handle the edge cases.

Good call - thanks for the CR! :)

Re: Show HN: March Madness for Hackers

#58
post #55

Hey Guys! I just saw this today and got pretty interested. I was interested in implementing a 'weighted' scoring function, but needed a way to save state somehow. I ended up finding a pretty fun solution using the console. It's really fun to play with the weights and see how the bracket changes! https://github.com/hstove/Coderbracket-Weighted-Scorer

hold on.. there's a developer console somewhere in the webapp? I don't see it and would love to use it. Am I missing something?

Re: Show HN: March Madness for Hackers

#59
post #58
post #55

Hey Guys! I just saw this today and got pretty interested. I was interested in implementing a 'weighted' scoring function, but needed a way to save state somehow. I ended up finding a pretty fun solution using the console. It's really fun to play with the weights and see how the bracket changes! https://github.com/hstove/Coderbracket-Weighted-Scorer

hold on.. there's a developer console somewhere in the webapp? I don't see it and would love to use it. Am I missing something?

The browser console. (in chrome) View -> Developer -> Javascript Console. You can enter commands there, run functions, access variables (in the page's global scope), etc..

Re: Show HN: March Madness for Hackers

#60
post #56
post #54

Earlier quoted context omitted.

Nice site. I agree, I'd love to be able to use some of your data points in my algorithm.

There's a few exposed json endpoints for data, the matchup page, team page, and player page. eg team: http://madness.io/teams/fak.json eg player: http://madness.io/players/1772.json eg matchup: http://madness.io/matchup/dae/vs/oad/1395332100.json

Is there any endpoint for teams.json or would I need to request each individually? Even if it was just the season stats without the game stats it would be useful.
Post reply on HN