I'm actually going to be a little more serious in a future post, since I'll be thinking about this all month... bu for this first post, I thought I'd just lay out in an easily interpretable Table just what a poor guesser I am... apparently you're gonna have to click on it... sorry... In any case, the deadline for guesses is April 5, which is also the deadline for my summer Adjunct Faculty position application...hence a very important monetary date for me...
OK, more Ice-Classic blogwash to come...
P.S. I'm not feeling clever today. Don't bother looking for a "joke" in this post other than the obvious ones...
Posts
2 comments:
Why would you consistently guess previous years' dates/times? If anything, that would be the least likely guess (in my humble, non-scientific opinion).
Yea, this comment raises interesting issues that are mostly way over my head... essentially "probability theory" stuff... Different types of events have different probability functions... the functions that most people (me certainly included) intuitively grasp are the "binomial" and the "normal"... Binomial example: flip a coin... it's gonna be one of two outcomes. Flip it another time, same equal chance. And so on. In this case, even if you get 4 heads in a row, the next flip still has a (1/2) chance of being heads and a (1/2) chance of being tails. Thus: "Chance has no memory." However, the "event" of getting 5 heads in a row is (1/2)*(1/2)*(1/2)*(1/2)*(1/2)= (1/32). Pretty unlikely. Don't bet on it... still, if you're just betting on that 5th toss alone, you've still got 50/50 odds, don't think that just cause it was heads 4 times in a row it's gotta be tails this time...
Of course the Nenana Ice Classic dates/times are not a binomial distribution. I think the date is normally distributed, with the "hump" of the curve (function) being April 30, and the time *seems* to be normally distributed around the hour of 1-2pm. There's lots of famous normal curves... human height (few really short people, few really tall people, lots of 5'10" guys like me). Grades another... IQ another (though that's probably a bogus construct to begin with)... etc. So, yes, it was admittedly DUMB of me to guess dates/times just because they happened before, but not quite for the reason the above comment suggests... Take 9:33AM May 11, one of my dumber guesses last year... It's was just as likely in 2007, regardless of it occurring before... but it was unlikely the first time it happened, and independently it was pretty unlikely to happen in 2007... Indeed, I was 2 freaking weeks off. My worst quess yet. I was definitely throwing away $2.50. Not quite sure why I do dumb things knowingly... I guess if I had actually won, by some freak of weather, I would probably not have to share the pot with anybody! Yea, it turns out that almost all of my closest guesses (in yellow on my Table), have been those for April 30, which is the most common day...kinda a no-brainer really... So the next post I'm going to do address whether that most common day has been shifting over the years... is the ice breaking up on average earlier??? I.e. has April 30 Always been the most common day? Spoiler: it is getting earlier, but not in the way some people think... And then in another post I'm gonna ask whether there's any other way other than just "guessing" based on probability functions. Spoiler, there is: thicker ice on the river tends to mean a later breakup... Duh! So ice-thickness is a really great "predictor variable." Alright... off my damn lectern, sorry...
Post a Comment