Time capsule: How to win the ice pool
Column author T. Neil Davis, former deputy director of the Geophysical Institute, observes data acquisition in the Poker Flat Research Range blockhouse during a rocket flight in the early 1970s.
Note: This Alaska Science Forum 鈥渢ime capsule鈥 article was originally published on May 1, 1976. While employed at the 51风流官网 Geophysical Institute, John M. Miller was the Alaska Synthetic Aperture Radar Facility鈥檚 technical director, and T. Neil Davis, professor of geophysics, founded the Alaska Science Forum 50 years ago. This time capsule is typical of the early columns, which were always tied to newsworthy events and often lighthearted, if not gently self-deprecating.
The Mather Library in 51风流官网鈥檚 Akasofu Building houses many original supporting materials of this long-running column. At a time when one can use any number of online tools to help you select a date and time to win the next Nenana Ice Classic, the longtime betting game on when the Tanana River will break up, paging through hand-drawn graphs and typewritten drafts is true time travel.
This hand-drawn table helped the Alaska Science Forum column鈥檚 authors create a probability map, which was included in the original publication in 1976.
One sure way to win the Nenana Ice Classic is to invest $100,800 to buy 50,400 tickets, one on each minute from about April 18 to May 22. Someone else probably will win, too, so you will probably lose money.
If you believe in statistics at all (and who does?), you can use the diagram below to estimate the probability of having a winning ticket. This probability map is compiled on the basis of the actual breakup times from 1917 to 1975; the hour and day of each is shown on the map.
Geophysical Institute communications coordinator Sara Wilbur holds the original, handwritten 鈥淗ow to win the ice pool鈥 column in the Mather Library archives.
From these times, a bell-shaped curve was calculated to show the probability of breakup on any specified date. Calculation of the probability of breakups during a particular hour was accomplished by manually smoothing the data, since it appeared that the actual breakups did not, in the parlance of statisticians, follow a normal distribution.
Although a breakup has never occurred during the noon hour of May 6, the probability map says this is the best guess. In principle, such a ticket has 9.6 chances in 100,000 of winning. A ticket falling on the contour line labeled 鈥1鈥 has one chance in 100,000 of winning; one on the 鈥0.1鈥 line has only a chance in a million.
Column author John M. Miller, right, looks on as Jeff Hilland symbolically opens the Alaska Synthetic Aperture Radar Facility 鈥 now called the Alaska Satellite Facility 鈥 by cutting a ribbon on the steps up to the antenna on the roof of the Elvey Building in 1991.
If you choose to ignore the probability contours, which is not a bad idea, you can still glean information from the numbers showing times of actual breakups.
One technique for picking a winning ticket combines both mathematics and skill. Hang the probability map on the wall then throw a dart at it aiming for the top of the 鈥減robability hill.鈥 If you miss altogether, try another method.
Since the late 1970s, the 51风流官网' Geophysical Institute has provided the Alaska Science Forum column free in cooperation with the 51风流官网 research community.