CLOSE
Original image
Jeopardy! / Photographed By Chris Higgins on His Couch

6 Elements of Arthur Chu's Jeopardy! Strategy

Original image
Jeopardy! / Photographed By Chris Higgins on His Couch

Arthur Chu has made waves the last three nights by employing excellent strategy on Jeopardy! He won first on Tuesday, tied to win on Wednesday, and won decisively on Thursday. Here's a rundown of what makes his play so special.

UPDATE: Chu won again on Friday. Check out our in-depth interview, conducted right before his fourth win aired.

1. Playing to Tie

Chu's most interesting move so far was on Wednesday's show, when he entered Final Jeopardy in the lead...and intentionally wagered so that he might tie with another player, Carolyn Collins.

Chu faced the same problem confronting every dollar leader heading into Final Jeopardy—he wanted to stay in the game, returning the next day to win more dough—but unlike most, he did not bet to win. Instead, he assumed that Collins would bet everything. He proceeded to calibrate his wager so that if Collins bet everything, and answered correctly, and Chu answered correctly too, they'd end up in a rare tie, with both Chu and Collins moving on to compete again the next day.

But this was not the first time Chu had wagered to tie! The night before, on Tuesday, Chu had bet $17,200 in a bid to tie with the returning champion on that show (he had only a slight lead heading into Final Jeopardy). The reason there weren't dual winners on Tuesday is that only Chu responded correctly to the Final Jeopardy clue.

Now, the big question here is why would you play to tie, instead of playing to win? It turns out, in short, "Because game theory suggests that this is smart." A longer analysis comes to us from Keith Williams at The Final Wager, a blog analyzing Jeopardy! wagers. Chu is a fan of the blog, and used the techniques outlined by Williams when planning his tied game. I won't steal his thunder; here's Williams analyzing the tied Chu game (warning—there's some light math in here):

Neat, huh? For much, much more on game theory as it applies to Jeopardy!, check out The Final Wager. (Note that Williams also broke down the wagers in the third game, which wasn't as exciting because Chu went into Final Jeopardy with nearly three times the dough of his nearest competitor.)

2. Knowing His Weakness

In Chu's second game, he scoured the board for Daily Doubles. When he hit one in the $1000 slot of the category IN THE SPORT'S HALL OF FAME, he immediately admitted that he knew very little about sports, and wagered only $5. When the clue was read ("Eddie Giacomin,
Herb Brooks, Conn Smythe"), Chu promptly said, "I don't know," and lost the five bucks as he smiled and the audience chuckled.

This performance is important for several reasons. First, Chu lost hardly any money on the Daily Double; a more aggressive player might have bet a few hundred bucks just because it felt like an opportunity (especially because heading into this Daily Double, Chu was $400 behind the other two players, who were tied for first!). Chu knew the potential for loss in the arena of Sports Trivia outweighed the possibility of gain, and he chose to keep the game moving. The other benefit to this move (attributable to his Daily Double-searching strategy in general) was that he prevented the other players from accessing that Daily Double—they may well have known the answer. (For the record, it was: "What is hockey?")

While some would argue that Chu should have thrown out some random sport's name just for kicks (in the hope he'd win $5), I think his move here was gutsy. Hey, if I got a sports Daily Double, I'd bet almost nothing too, and keep the game moving so there was time to clear the board—and find the remaining Daily Doubles.

(An aside on math: in the third game, there was a whole category devoted to math word problems, and it was poorly received by everyone, Chu included. Chu tweeted, "As this game demonstrated doing actual math is too hard when you're playing." Fortunately, when he had a moment to do his ciphering, namely for his wagers, Chu was just fine.)

3. Mastering Daily Doubles...Mostly

In his first game, Chu wasn't afraid to go for a "True Daily Double," meaning a wager that would double his current winnings if he got it right—which he did. This smart move catapulted him into the lead halfway through Double Jeopardy, which helped set him up for the win.

His performance with Daily Doubles in the second game was mixed; he made a blunder with a clue about paint types and lost his second attempt at a True Daily Double (!). At that point he wagered $4,195, the amusingly specific number left over after his aforementioned loss of five bucks on hockey. But by scouring the bottom of the Double Jeopardy board, Chu quickly racked up another $10,000 and wagered half of that (successfully) on the final Daily Double of his second game—again, very gutsy and in the end, it worked out.

4. Taking Advantage of Others' Mistakes

In his first game, Chu spotted a mistake made by fellow contestant Cesar Perez-Gonzalez. Trying to identify an antelope shown in a video, Perez-Gonzalez said "prog-horn," causing Trebek to take a tiny pause and say, "Oh gosh, no." Chu buzzed in immediately with the correct response, "What is a prong-horn?" That response pushed Chu to $20,000, again in the lead. Whether Chu simply knew the correct answer, or recognized Trebek's implication that "prog-horn" was close to correct is for Chu to explain. But he was fast on the buzzer when Perez-Gonzalez barely missed that one.

5. Talking About His Wife with Trebek

During the interview segments with Alex Trebek, Chu repeatedly mentions his wife Eliza Blair, relating a story about buying her a meteorite, and working to support fibromyalgia research. After Chu's appearances on the show, Blair began retweeting troll comments about her husband, gleefully fanning the flames as Chu lovers and haters sparred online. The most notable exchange is when Blair tweeted a photo of pizza boxes in bed, referring to Kevin Clancy's comment. (If you ready the Clancy/Chu/Blair exchanges, they're actually delightful, although often profane.)

6. Knowing His mental_floss Trivia

Okay, we're biased here, but this was a big moment for the mental_floss crew. The $1000 clue in "THE 'RENT' IS DUE" on Wednesday was: "The Mental Floss History of the World" is this kind of "romp through civilization's best bits." Chu rang in quickly, correctly answering, "What is irreverent?" We salute you, Mr. Chu!

(For the record, that clue refers to this lovely book we put out in 2008.)

Note: We interviewed Arthur on Friday. Look for that over the weekend.

Original image
iStock // Ekaterina Minaeva
technology
arrow
Man Buys Two Metric Tons of LEGO Bricks; Sorts Them Via Machine Learning
May 21, 2017
Original image
iStock // Ekaterina Minaeva

Jacques Mattheij made a small, but awesome, mistake. He went on eBay one evening and bid on a bunch of bulk LEGO brick auctions, then went to sleep. Upon waking, he discovered that he was the high bidder on many, and was now the proud owner of two tons of LEGO bricks. (This is about 4400 pounds.) He wrote, "[L]esson 1: if you win almost all bids you are bidding too high."

Mattheij had noticed that bulk, unsorted bricks sell for something like €10/kilogram, whereas sets are roughly €40/kg and rare parts go for up to €100/kg. Much of the value of the bricks is in their sorting. If he could reduce the entropy of these bins of unsorted bricks, he could make a tidy profit. While many people do this work by hand, the problem is enormous—just the kind of challenge for a computer. Mattheij writes:

There are 38000+ shapes and there are 100+ possible shades of color (you can roughly tell how old someone is by asking them what lego colors they remember from their youth).

In the following months, Mattheij built a proof-of-concept sorting system using, of course, LEGO. He broke the problem down into a series of sub-problems (including "feeding LEGO reliably from a hopper is surprisingly hard," one of those facts of nature that will stymie even the best system design). After tinkering with the prototype at length, he expanded the system to a surprisingly complex system of conveyer belts (powered by a home treadmill), various pieces of cabinetry, and "copious quantities of crazy glue."

Here's a video showing the current system running at low speed:

The key part of the system was running the bricks past a camera paired with a computer running a neural net-based image classifier. That allows the computer (when sufficiently trained on brick images) to recognize bricks and thus categorize them by color, shape, or other parameters. Remember that as bricks pass by, they can be in any orientation, can be dirty, can even be stuck to other pieces. So having a flexible software system is key to recognizing—in a fraction of a second—what a given brick is, in order to sort it out. When a match is found, a jet of compressed air pops the piece off the conveyer belt and into a waiting bin.

After much experimentation, Mattheij rewrote the software (several times in fact) to accomplish a variety of basic tasks. At its core, the system takes images from a webcam and feeds them to a neural network to do the classification. Of course, the neural net needs to be "trained" by showing it lots of images, and telling it what those images represent. Mattheij's breakthrough was allowing the machine to effectively train itself, with guidance: Running pieces through allows the system to take its own photos, make a guess, and build on that guess. As long as Mattheij corrects the incorrect guesses, he ends up with a decent (and self-reinforcing) corpus of training data. As the machine continues running, it can rack up more training, allowing it to recognize a broad variety of pieces on the fly.

Here's another video, focusing on how the pieces move on conveyer belts (running at slow speed so puny humans can follow). You can also see the air jets in action:

In an email interview, Mattheij told Mental Floss that the system currently sorts LEGO bricks into more than 50 categories. It can also be run in a color-sorting mode to bin the parts across 12 color groups. (Thus at present you'd likely do a two-pass sort on the bricks: once for shape, then a separate pass for color.) He continues to refine the system, with a focus on making its recognition abilities faster. At some point down the line, he plans to make the software portion open source. You're on your own as far as building conveyer belts, bins, and so forth.

Check out Mattheij's writeup in two parts for more information. It starts with an overview of the story, followed up with a deep dive on the software. He's also tweeting about the project (among other things). And if you look around a bit, you'll find bulk LEGO brick auctions online—it's definitely a thing!

Original image
quiz
arrow
Name the Author Based on the Character
May 23, 2017
Original image
SECTIONS
BIG QUESTIONS
BIG QUESTIONS
WEATHER WATCH
BE THE CHANGE
JOB SECRETS
QUIZZES
WORLD WAR 1
SMART SHOPPING
STONES, BONES, & WRECKS
#TBT
THE PRESIDENTS
WORDS
RETROBITUARIES