Phil vs. the Computer

[Philip Coates]

The thing I like about Jeopardy is it has 'general knowledge' questions of the kind which are not too trivial, not memorize the encyclopedia stuff so that once you hear the answer to the ones you miss, many people will often say "oh yeah". I'm really good at it, getting a strong majority of the questions, so I decided to score myself against the IBM computer tonight.

I won!!

Kent Jennings - 1000. Brad (I don't know his last name) - 5000. IBM Watson - 5000. Phil Coates between 7200 and 8000. (I couldn't write the numeric scores fast enough and two or three I'm not sure if I was fastest or a shade slower on correct answer.)

The competition between the two former champions and Watson goes on for three nights so, if I'm home, I intend to continue to see if I'm competitive the next two nights:

Where I often fall down/most often do not know the answer is the final -written by the contestant- question where you can bet any amount of your money. It's usually not 'general' knowledge that a well-read but non-encyclopedic person might know, but something more obscure. And 'trivial'.

(When I lived in L.A. where it's filmed, I was often tempted to see if I could get on the show, but someone told me it has as much to do with pressing the button at the right instant as having a broad range of knowledge in various fields and other 'screening' factors, so that discouraged me from doing it.)

[Philip Coates]

ADVICE: If you watch the show (7:30 Eastern on ABC)and want to score yourself, the tricky and concentration-damaging part for me is that you have to listen not only to the question but to its value -- 200, 400, 600, 800, 1000 - and then write down your numerical score *quickly* while that smugly coiffed Canuckistani Trebek is already asking the damn next question!

That can cost you a bit. (Don't even attempt to score the three TVcontestants, just write down their scores at the end or at the commercial. Mute the damn commercial. It will break your concentration. Have a piece of chocolate. It helps. At least it does for me.)

No cheating on the scores, please.

[Michael Stuart Kelly]

Phil,

Dayaamm!

When I saw the title, "Phil vs. the Computer," I thought you were talking about the quote function.

Michael

[caroljane]

Phil,

Dayaamm!

When I saw the title, "Phil vs. the Computer," I thought you were talking about the quote function.

Michael

[9thdoctor]

I won!!

Kent Jennings - 1000. Brad (I don't know his last name) - 5000. IBM Watson - 5000. Phil Coates between 7200 and 8000.

Brad Rutter. Are you sure you rang in ahead of the others? They've been advertising the online test lately, time to prove yourself, Pudd'nhead Coates.

When I saw the title, "Phil vs. the Computer," I thought you were talking about the quote function.

Dream on.

What was really neat was seeing the computer’s responses, they showed three for each question (ahem, answer), with percentages next to each one, so there’s an algorithm at work guessing at the likelihood that different replies are the right one. Also, the computer isn’t hearing the other replies, at one point Ken Jennings got one wrong, and the computer then chimed in making the same mistake.

[caroljane]

Way to go Phil! I'm getting the T shirt made.

The buzzer pushing I always thought was the killer. My kids and my mother (a major Trebek groupie) always told me I should go on that show. (My husband like me knew I would fall totally apart and disgrace myself in public) , but nobody offered to pay any airfare anyway. so I rest on my laurels defeating the Americans at my intellectual peak age 15.

You are an example to us all, Sir.

[merjet]

I won!!

Kent Jennings - 1000. Brad (I don't know his last name) - 5000. IBM Watson - 5000. Phil Coates between 7200 and 8000.

Phil "won" in a very qualified sense. He did not have to hit the response button before Ken Jennings, Brad Rutter, or Watson, then answer correctly. Two contestants on the show cannot win $ on the same question. Only one can. That restriction did not apply to Phil.

Here is a show about the preparation of Watson for Jeopardy and Watson's programming.

Edited February 15, 2011 by Merlin Jetton

[Philip Coates]

> Phil "won" in a very qualified sense. He did not have to hit the response button before Ken Jennings, Brad Rutter, or Watson, then answer correctly. Two contestants on the show cannot win $ on the same question. Only one can. That restriction did not apply to Phil. [Merlin]

Good point. I just counted my answer as winning if I blurted it out before anyone on the program did.

[kiaer.ts]

> Phil "won" in a very qualified sense. He did not have to hit the response button before Ken Jennings, Brad Rutter, or Watson, then answer correctly. Two contestants on the show cannot win $ on the same question. Only one can. That restriction did not apply to Phil. [Merlin]

Good point. I just counted my answer as winning if I blurted it out before anyone on the program did.

On that basis I have never lost.

[william.scherk]

Are you sure you rang in ahead of the others? They've been advertising the online test lately, time to prove yourself, Pudd'nhead Coates.

Online test . . . here.

[merjet]

Are you sure you rang in ahead of the others? They've been advertising the online test lately, time to prove yourself, Pudd'nhead Coates.

Online test . . . here.

This year's online test for adults took place last week, February 8-10. The next opportunity will be about the same time next year.

Taking the test will likely prove nothing. I have taken the test two or three times. You are not told how well you did. You will know you passed only if you are later called for an audition, and names are drawn randomly (from those who passed) for that. I haven't found any source indicating what fraction of people pass, what fraction of those who pass are later called for an audition, nor what fraction of those who audition eventually appear on Jeopardy. There is a little more info here.

Incidentally, Pudd'nhead Wilson was a pretty bright guy. The novel by the same name is terrific.

Edited February 15, 2011 by Merlin Jetton

[9thdoctor]

Incidentally, Pudd'nhead Wilson was a pretty bright guy. The novel by the same name is terrific.

And so is Phil! I was just making a pun (or something like that) on the phrase “the proof’s in the pudding” and thought of Twain.

[merjet]

Incidentally, Pudd'nhead Wilson was a pretty bright guy. The novel by the same name is terrific.

And so is Phil! I was just making a pun (or something like that) on the phrase “the proof’s in the pudding” and thought of Twain.

Fine. "Pudd'nhead" is rarely used nowadays, but it meant "nitwit" in Mark Twain's time, as the Wikipedia article says.

The clues and answers for yesterday's Jeopardy round are here. The Double Jeopardy round is not there, which is unusual.

[Philip Coates]

> The clues and answers for yesterday's Jeopardy round are here. The Double Jeopardy round is not there, which is unusual.

The Double Jeopardy round didn't take place last night. They're stretching two complete games into three nights so they could include more background information on Watson and how it operates and how it's been programmed by IBM.

[merjet]

The Double Jeopardy round didn't take place last night. They're stretching two complete games into three nights so they could include more background information on Watson and how it operates and how it's been programmed by IBM.

Aha. I wasn't home yesterday when the show aired on tv. Today's show will be at the usual time (3:30 pm Central).

[Philip Coates]

Subject: Building up a Wide Knowledge Base

We lived in Japan for a year when I was a little boy and it sparked my hunger for knowledge about new things, in this case a distinctive and very unusual culture and people. My father was a career army officer who was always very critical: "Philip, you don't have the common sense God gave a Little Pig." So I decided I'd show him. If he was Colonel Coates, I would pull rank and be General Knowledge. :-) [Ok, that's a bit exaggerated for humor.]

The good news is a range of general knowledge is very useful and makes your mind more productive. You have a fund of information that you can always use to interconnect things, find examples and contrasts.

The bad news: You can't acquire a broad base quickly. The analogy is to when I was teaching in California to heavily Asian immigrant students. They were in high school wanting to get into an Ivy League college, likely to get close to 800 in math SAT, but perhaps 400 in verbal. They wanted me to tutor them on the side, help them quickly acquire an "sat vocabulary". Like teaching a little pig to have common sense, can't be done in a hurry if at all. Takes a lifetime for native English speakers to have high level fluency with sophisticated concepts.

Also, in the case of general knowledge in the sciences and humanities and arts and literature and current events and so on, (by contrast to cramming the dictionary for the National Spelling Bee) you can't just read a reference work cover to cover and try to remember all that unintegratable stuff.

And if you don't -love- it, your mind will tend to discard or be bored. In my case, I've spent the last thirty years eagerly reading all kinds of fascinating stuff and my mind just naturally tends to soak up a lot of facts, even though I'm not trying to.

I like to read several things simultaneously over a space of six months: Right now, a book called "Modern Drama", a textbook on soil science, Bertolt Brecht's Galileo, some math stuff, Kenneth Clark's Civilisation, a history of the Hellenistic to early Roman period.

Edited February 15, 2011 by Philip Coates

[caroljane]

Subject: Building up a Wide Knowledge Base

We lived in Japan for a year when I was a little boy and it sparked my hunger for knowledge about new things, in this case a distinctive and very unusual culture and people. My father was a career army officer who was always very critical: "Philip, you don't have the common sense God gave a Little Pig"

Snap! I do not have any common sense and never have had any. People were always telling me to use it and asking me where it was. In the 90s Ontario elected a government on the slogan"the Common Sense Revolution". Their revolution consisted in finding the parts of the system that worked well and wrecking them. Since then I have been proud of my disability, not ashamed like formerly. The Common Sensers are coming back now, since everything in politics is cyclical, and this time I'll be ready for them.

[merjet]

Wow! Watson dominated today. He now leads with $35,734. Brad has $10,400. Ken Jennings has $4,800.

I thought the Final Jeopardy clue was pretty easy (but I live in the area). This city's largest airport is named after a WWII hero and its 2nd largest after a WWII battle.

The answer is 'What is Chicago?' (The airports are O'Hare and Midway.) Ken and Brad answered correctly. Watson guessed Toronto, but bet less than $1000 dollars.

[Philip Coates]

I also got the Final Jeopardy question and I had bet 10,000. As soon as I thought of Midway and knew that was a WWII battle as well as the other major airport in Chicago, I had it even though I had never heard of a WWII hero named O'Hare. Chicago is a patriotic city (think Soldier Field), so if I were a guessing man I might have gone that way but didn't need to.

End of second day, the computer is ahead of me: 35,700 to my score of between 24,000 and 31,000. The reason for not being sure where I fell in that range is (i) one of the daily doubles for 6435 [ i rounded off] was a -very- big score and I'm not sure if I got it faster [addition] or a tie [no subtraction]. (ii) The other problem is I can't always follow the question and write down the point value so on a couple questions, I'm not sure of my actual score. iii) I tend not to guess, so I seldom had to -subtract- points for a wrong answer.

Whoops, I forgot to subtract a couple from yesterday where I did have a wrong answer, so adjust my above score down by 1000: So my score was: 23000-30000.

Still well behind that damn Watson.

I think Brad Rutter? is third with 10,400 and Ken Jennings is last with 4800.

(The problem I had today is yesterday the computer was getting less than half the questions, today it seemed like it was getting well over 80% inlcuding each of the 'big bucks' double jeopardies. I always ...and this goes back years... seem to get pushing 2/3. I tend to miss the Final J. question, though... but I got it today and I placed the biggest bet of all, 10, 000.)

I have to take a trip tomorrow, but now I'm hooked so I'm going to try to be home in time to watch the third and final installment. I have the sense I'm not going to beat that goddamn computer, but I'll be very pleased if I beat the two humans.

(I'm curious if button pressing at "computer speed"/speed of light vs. human reflexes is an issue: A number of the questions Watson got tonight were extremely easy and I'd be amazed if neither K or B knew any of those answers. Didn't seem to be a problem the first night though.]

Edited February 16, 2011 by Philip Coates

[Brant Gaede]

The thing I like about Jeopardy is it has 'general knowledge' questions of the kind which are not too trivial, not memorize the encyclopedia stuff so that once you hear the answer to the ones you miss, many people will often say "oh yeah". I'm really good at it, getting a strong majority of the questions, so I decided to score myself against the IBM computer tonight.

I won!!

Kent Jennings - 1000. Brad (I don't know his last name) - 5000. IBM Watson - 5000. Phil Coates between 7200 and 8000. (I couldn't write the numeric scores fast enough and two or three I'm not sure if I was fastest or a shade slower on correct answer.)

The competition between the two former champions and Watson goes on for three nights so, if I'm home, I intend to continue to see if I'm competitive the next two nights:

Where I often fall down/most often do not know the answer is the final -written by the contestant- question where you can bet any amount of your money. It's usually not 'general' knowledge that a well-read but non-encyclopedic person might know, but something more obscure. And 'trivial'.

(When I lived in L.A. where it's filmed, I was often tempted to see if I could get on the show, but someone told me it has as much to do with pressing the button at the right instant as having a broad range of knowledge in various fields and other 'screening' factors, so that discouraged me from doing it.)

You might try still. I'm sure there are practice games for pushing the button. Unlike you, I couldn't do this show. Watching Get Ben Stein's money years ago, I figured I could tie him on a good, day but beat him no more than one time out of 10. Jeopardy is two or three levels harder.

--Brant

it's too easy to be discouraged, and remember that first round is just to settle you down and get use to the button pushing

[Mike Hansen]

(I'm curious if button pressing at "computer speed"/speed of light vs. human reflexes is an issue: A number of the questions Watson got tonight were extremely easy and I'd be amazed if neither K or B knew any of those answers. Didn't seem to be a problem the first night though.]

Phil,

That had me wondering tonight as well. A lot of the questions seemingly could have been solved by a search engine. Especially the Final Jeopardy question. Let's hope that changes for tomorrow. I want to see Watson sweat.

I would think that the IBM guys, with all the other 'human' features they gave Watson, would have measured the time it takes for a typical player to ring in and factored that into Watson's answers/questions. Otherwise it's nowhere near a fair contest. I remember reading something about how there's a "time penalty" for ringing in too quickly. Given that the questions/answers were pretty easy, and that Ken and Brad have probably lost the rhythm they once had, I can see the timing being a big issue.

Did anybody else see David Ferrucci's face after Watson completely blew the Final Jeopardy answer/question? But since the category was U.S. cities, I guess "Toronto" probably wouldn't elicit the happiest response from him.

Mike

[merjet]

How quickly a contestant presses the button is clearly a factor. It obviously is for humans. Suppose two people both think of the answer, but one who thinks of it before the clue is completely given and the second thinks of it a few nanoseconds after the clue is completely given. The first has the speed advantage.

A lot of the questions seemingly could have been solved by a search engine.

In post #7 I gave a link to a tv show about Watson. During that show it was said that Watson had Wikipedia and previous Jeopardy shows in its memory. A search engine is a component of Watson, but there is more to it.

Did anybody else see David Ferrucci's face after Watson completely blew the Final Jeopardy answer/question? But since the category was U.S. cities, I guess "Toronto" probably wouldn't elicit the happiest response from him.

Watson's answer to the Final Jeopardy clue was indeed a surprise. There is at least one city named Toronto in the U.S. -- in Ohio with a population < 6,000. In addition to Toronto, Ontario, Canada not being a U.S. city, I learned the following using a search engine.

Pearson International is Toronto's largest airport. The airport was named in honor of Lester B. Pearson, the 14th Prime Minister of Canada and recipient of the Nobel Peace Prize. Pearson did military service during WWI, but didn't attain hero status. I'm not sure which airport is 2nd largest, but two candidates follow. John C. Munro Hamilton International Airport is in Hamilton, Ontario, Canada. It is named for John C. Munro, a Hamilton Member of Parliament and cabinet minister. Billy Bishop Toronto City Airport was named after a famous WWI fighter pilot. Hence, none of the names of the three airports fit the Jeopardy clue.

Edited February 16, 2011 by Merlin Jetton

[9thdoctor]

How quickly a contestant presses the button is clearly a factor. It obviously is for humans. Suppose two people both think of the answer, but one who thinks of it before the clue is completely given and the second thinks of it a few nanoseconds after the the clue is completely given. The first has the speed advantage.

Here’s one extra little detail: you can’t ring in until the question (answer, whatever) has been read in full, and there’s a signal light that goes off, which viewers at home can’t see, that lets contestants know they can try to ring in. If they try to ring in too soon, they’re locked out momentarily. So, what’s the deal with Watson? Is the computer getting an electrical signal at the same nanosecond that Jennings and Rutter are getting a visual cue? My point being that Jennings and Rutter are probably at a disadvantage because of reaction time.

[merjet]

How quickly a contestant presses the button is clearly a factor. It obviously is for humans. Suppose two people both think of the answer, but one who thinks of it before the clue is completely given and the second thinks of it a few nanoseconds after the the clue is completely given. The first has the speed advantage.

Here’s one extra little detail: you can’t ring in until the question (answer, whatever) has been read in full, and there’s a signal light that goes off, which viewers at home can’t see, that lets contestants know they can try to ring in. If they try to ring in too soon, they’re locked out momentarily.

That is why I phrased the hypothetical the way I did.

So, what’s the deal with Watson? Is the computer getting an electrical signal at the same nanosecond that Jennings and Rutter are getting a visual cue? My point being that Jennings and Rutter are probably at a disadvantage because of reaction time.

I suspect the different ways were carefully tested to avoid giving either Watson or humans an advantage.

[Philip Coates]

A better test and quiz show based on general knowledge of people (and computers) would not be about button pressing or reaction times at all. I think there was a quiz show from the 50's in which each contestant was in a separate glass booth so he couldn't hear anything said or be affected by anyone else's answer. And they had more than an instant (five? ten? seconds) to rack their brains while a bit of background music played.

That's the proper way to do it. More about accuracy and knowledge than speed. That way each of the three contestants gets to give his (or in this case its) answer. Points then should be awarded to each right answer, not the fastest, and not only the fastest gets to answer. Fraction of a second is really, really dumb. As is having to worry about pressing too fast. Give me a @#$%^&*## break.

This is just one symptom of a cognitive disorder, of superficiality. We live in an age where speed is prized over careful reflection, over knowledge, over mulling over: The lightning quiz. Spritzing out bits of attention across the internet. Multitasking. Short attention spans.

DGMS (don't get me started).

Edited February 16, 2011 by Philip Coates