This week I launched Rank the Things, a project in which people can vote to form a crowdsourced ranking of anything from pizza toppings to Pixar movies. I wanted to make it as simple to use as possible, so one is given a choice between just two options and has to pick which they prefer (or skip if they really can’t make a decision). To get a decent sample size, they’re given a bunch of opportunities to vote, but not too many to prevent one person from having too much influence on the final results.
Of course, there is the challenge of how to take all these votes and turn them into a ranking. The simplest way is to divide the number of votes for something over its total matchups, but that runs into the problem that there isn’t necessarily going to be a balanced strength of schedule. For instance, as of this writing, about three times as many people have weighed in on pepperoni vs. banana peppers as they have banana peppers vs. black olives. Therefore, I opted to go for a maximum likelihood estimation instead, which adds some complexity but isn’t excessively fancy.
Continue reading Rank the Things: How the ratings work (feat. math)
The Presidential election has come and gone, and all 50 states and DC have finally finished counting every single one of the 158,394,605 votes. Now that we know the will of the people and not what we can estimate based on polls, I wanted to update the the electoral ladder that I had made about three weeks before the election.
There’s been a ton of reporting that “Biden won Georgia” or “Trump won Ohio by a lot more than expected”, but I thought it would be more useful if those statements could be put into context within how all the other states did and how things would look compared with a neutral environment. For instance, Biden won Pennsylvania by over a percentage point, but the state had actually drifted to the right compared with 2016. Meanwhile, the formerly competitive state of Colorado is starting to enter the company of the Democratic stalwart of Illinois.
If you were to compare the previous chart with this one, you’d see that there were several good predictions (Nevada, North Carolina), while others missed the mark quite a bit (Florida, Georgia). This doesn’t account for the difference in national popular vote, though the previous edition was done around Biden’s high point in the polls. To be fair, this was a difficult year for polling, thanks to insane turnout, many people’s ways of life being disrupted, and the sharp increase in early voting.
This chart is aligned such that the 0 point is the hypothetical scenario where Trump and Biden tied in the popular vote. In effect, that would be a 4.5-point shift towards Trump compared with the actual results. This was done to make it easier to compare how partisan each state is with previous elections. The higher the state is on the chart, the more Democratic it is compared with the nation as a whole.
Continue reading The Electoral Ladder: Now with actual results
Another four-year cycle is coming to a close, and that means it’s time for me to get way too obsessed about election polls and demographic projections and the like. The idea of trying to use data and statistical samples to predict what will happen while having few prior cases for such an important event is extremely fascinating to me. Does that poll herald a sign of long-term change, or was it an outlier? How much should we trust our gut versus the data we see today versus what actually happened several years ago? There’s an incredible amount of data one can dive into, though it’s surprisingly easy to get lost in the woods and miss the big picture.
Continue reading The Electoral Ladder: The political leaning of every state, to scale
Apologies for the long delay, but the results of the Least Popular Answers Contest is here! As implied in the title, this was basically the opposite of the previous entries in this series, as this time people had to pick valid answers that was shared by as few other people as possible. Of course, everyone else was doing the exact same thing, so it wasn’t as easy it looked at first glance. There were a total of 93 entrants, which was amazing to see! I’m glad people are continuing to enjoy this, and I have a few other twists up my sleeve in future iterations.
Continue reading The Least Popular Answers Contest: The Results
Another month has passed, and so it’s time for yet another contest! Previously I had you all guess what you thought the most popular answer to a series of prompts would be, but this time you have to do the exact opposite. Fill out this form as per usual, but now you need to come up with valid answers that you think the fewest number of people would pick. The more people that choose the same answer as you, the more points you get, so this time the person with the lowest number of points will win.
Continue reading The Least Popular Answers Contest
The time has come to reveal the results of the third iteration of the Most Popular Answers Contest! As in previous editions, everyone had a list of 10 prompts and they had to give what they thought would be chosen the most. This time around I tried asking a few more subjective questions, and it definitely led to some interesting results, including what is easily the least agreed-upon prompt in the short history of this contest. I’d like to thank all 58 of you that took part, and I’d love to see everyone back next month for the Least Popular Answers Contest.
Continue reading The Most Popular Answers Contest III: The Results
Hey guys, I’m back with another round of the Most Popular Answers Contest! As with previous iterations, all you need to do is fill out this form with what you think the most popular answer to each of the 10 prompts will be. The more people that choose the same answer as you, the more points you get. As per usual, the person with the most points will win.
Continue reading The Most Popular Answers Contest III
A couple weeks ago I published my second quiz in which I asked people to give what they think would be the most popular answer to each of 10 prompts. This time around I tried making the best answers slightly less obvious, and I think I may have succeeded too well — fully eight out of ten had at least fifteen unique responses given! Still, it was a lot of fun and 54 people took part this time around, so I might end up making another one of these in the future. I have a couple of ideas for twists to use.
Continue reading The Most Popular Answers Contest II: The Results
Last month I ran The Most Popular Answers Contest, and it was popular enough I decided I wanted to do another one! Once again it’s very simple: Just go to this form and give what you believe will be the most common answer to each of the 10 prompts. For each question, you get one point per person who picks your answer (including yourself), and your scores for each question will be combined to get a final tally. The winner will once again receive untold fame and fortune, or at least until I run another one of these contests.
Rules are mostly the same as last time, but with a couple clarifications:
- Answers need not be correct. If I asked people to name a color on the American flag and 3 people said green, they would each get 3 points.
- Misspellings and synonyms where it’s obvious what the person is going for count as part of the original answer. So for instance, if I asked for a type of electronic and 10 people said television, 4 said TV, and 1 said tellevision, they would each get 15 points.
- Answers that are specific cases of other answers are not grouped. Continuing from the previous answer, if another 3 people said “flat-screen television”, they would only get 3 points (and the other 15 would still only earn 15 points).
- Not attempting to give an answer (e.g. “I don’t know”, “No clue”, etc.) will only get one point, unless someone else happens to answer the same way verbatim.
- To get a final score for each person, I will add up the logarithm of the scores of each of their questions. (AKA, I will be using the “sum of log” score from last time.)
Good luck and have fun! You have until 12 PM Eastern Time on Sunday, May 24th to give me your answers, and results will be posted the following day.
Last week, I asked people to enter what they think the most popular answer would be to each of 10 prompts. Sixty-seven of you responded, and here are the results! I had a lot of fun putting this together, I learned a bunch of lessons, and you guys seemed to like it a lot, so I’m probably going to run a second version of this in a few weeks. But without further ado, here are the results:
Continue reading The Most Popular Answers Contest: The Results