A Computer Scientist in a Business School

It's quite worthwhile to consider the OTHER notabl...

2009-11-08T21:27:02.431-05:00

It's quite worthwhile to consider the OTHER notable way of analyzing the results of a bet, which is to consider the probability of making money on the bet.

And THAT offers conclusions of quite a different shape.

Thus, consider, for a wager, N, what is the probability P(N) that you break even given that wager?

We actually only need to consider wagers of the form $2^n$, where $n$ is a positive integer.

- If you wager 2 dollars, then it is certain that you will get a return of at least 2 dollars, the worst case being where the tail appears in the first throw.

- If you wager 4 dollars, then your chances of breaking even are 1/2

- If you wager 8 dollars, the probability of breaking even is 1/4

And this relationship continues. The larger the wager, the lower the probability of breaking even.

And this is so, irrespective of the fact that the expected payoff is infinite.

It would certainly be worth my while to pay 2 dollars to participate in this "lottery." But the value of engagement at higher prices is not nearly so obvious.

Expected value is by no means sufficient to support correct conclusions.

Stazybo: Of course "not betting on negative expect...

2009-11-07T09:53:27.078-05:00

Stazybo: Of course "not betting on negative expected profit bets" does not imply that someone will bet on bets with positive expected profit. But often people that claim not to bet due to the negative expected return of the bet, they will base their argument solely on the expectation of the return. (Which is easy to show that is not sufficient.)

It is a good example for introducing the concept of "risk aversion", i.e., the need to take into consideration the variance of the bet and not only the expectation.

rather, "positive/negative expected profit", I sho...

2009-11-07T06:05:40.808-05:00

rather, "positive/negative expected profit", I should've written. Expectancy is 0 to 1.

"I never enter a game with a negative profit expec...

2009-11-07T05:33:39.173-05:00

"I never enter a game with a negative profit expectancy" doesn't imply "I am willing to enter one with a positive expectancy".

p=>q does not imply ¬p=>¬q

For me it is very easy: I do not bet. Period. I've...

2009-11-07T04:12:08.264-05:00

For me it is very easy: I do not bet. Period. I've worked in horse racing and have learned "ο παίζων χάνει και ο πίνων μεθά" (he who drinks gets drunk; he who gambles loses).

But then again I am not the general case of non-gamblers and I admin trying to calculate prior reading the rest of the post :)

Never drink and derive :-) yes, concave...

2009-11-07T00:46:36.884-05:00

Never drink and derive :-)

yes, concave...

maybe you mean concave ?

2009-11-07T00:27:41.706-05:00

maybe you mean concave ?

Anonymous, I find the tone of the comment borderli...

2009-11-04T17:03:51.108-05:00

Anonymous, I find the tone of the comment borderline inappropriate, especially when hidden behind an anonymous comment.

Simple answer: processing videos in order to be posted online takes time.

Notice that VideoLectures.net still has not posted the videos online from KDD, that took time in June 2009 (see http://videolectures.net/kdd09_paris/) and they were actually a paid service, charging more than $20K to videotape the conference.

Are the videos gonna be uploaded anytime soon?

2009-11-04T16:42:24.613-05:00

Are the videos gonna be uploaded anytime soon?

When I was in grad school and ran information retr...

2009-11-03T12:51:27.865-05:00

When I was in grad school and ran information retrieval experiments with people, I had them fill out an informed consent form that promised not to cause them any (long term? noticeable?) injury. I don't recall ever having to submit these study proposals to an IRB. It seems that something similar as a pre-test for an MTurk HIT would be sufficient.

What was the outcome of your IRB proposal you mentioned in January?

My sense is that publishers justify their costs by...

2009-11-02T18:12:00.892-05:00

My sense is that publishers justify their costs by producing over-produced content. The value of the article is not in the quality of typesetting, but in the quality of the ideas, as long as the typesetting meets a minimum level of quality. This level is more than achieved for conferences, and thus should be achievable for journals as well. It should be possible for the ACM to offer its digital library at no additional cost if it stops competing with print publishers. I've written more about this in my post Academic papers want to be free.

This may be of interest. Actual costs of publishi...

2009-11-01T01:38:13.548-05:00

This may be of interest. Actual costs of publishing are virtually impossible to pin down, because most publishers are for-profit (even nominally non-profits, such as scholarly societies, often use publishing profits to subsidise other activities). Thus no one will open up their books and tell you what things actually cost, lest you find out that they are price gouging.

I've tried to estimate toll-access revenue per article, though the estimate doesn't include such things as charging for reprints. Given that the major publishing houses regularly report operating profits of between 30 and 40%, even in these straitened times, one could make simplistic estimates of operating costs from total number of papers.

That would still need considerable adjustment when talking about society publishers though, for reasons described upthread (volunteer labor, etc).

[ Please remove my previous comment, due to format...

2009-10-29T08:38:21.864-04:00

[ Please remove my previous comment, due to formatting garbage ]

Some random notes / thoughts:

- IEEE charges $9 per paper its members for papers that they normally do not have access to.

- I am in a similar situation like bob_d. This has transformed me into a digital beggar. I find myself always trying to find a friend (or a friend of a friend) to help out whenever I hit a closed door. And I know that you do not talk about for-profit publishers, but come on Springer, you charge $34 for http://www.springerlink.com/content/mn2u24251n05n45j/ written in 1957?

- Where I work we publish a magazine that reaches more than 110K people weekly. People divide the cost in: creating the issue, printing it and sending it. Guess what cost is fixed, what varies and where we saved from after giving the option of email distribution.

- Tim O'Reilly says that printing is less than 20% of the cost of a book (ref). But then again he speaks about books and not magazines so TMMV.

Anonymous: I am talking about publication from ass...

2009-10-27T14:06:01.982-04:00

Anonymous: I am talking about publication from associations, not from for-profit publishers.

Copyediting is actually a value-subtracting process at this point. When do they use the "well structured XML document" and I have never seen it?

Software, there is plenty of open source software (see DSpace).

1Tb of data? $100 from Best Buy. Can store the full digital library. Replicate 10 times for backup and protection. We are talking about numbers well below $10K for storage and backup.

Email, furniture, etc are being paid by the volunteers that actually serve as editors and are paid by universities.

There is no tax for non-profits.

Digitization is an one-time process. Sunk cost. Done once, serves for ever.

Digitizing printed material from several decades a...

2009-10-27T12:35:27.895-04:00

Digitizing printed material from several decades and putting them into databases and making them available online?

Company car to visit the Professors that serve as ...

2009-10-27T12:34:34.061-04:00

Company car to visit the Professors that serve as Editor-in-Chiefs? Marketing material? Payign for travels to meet scholars at conferences and exibitions?

Mathematicians and Economists (who are believed to...

2009-10-27T12:32:54.707-04:00

Mathematicians and Economists (who are believed to do well with numbers) are the ones with most difficulties of understanding business numbers. Yes, pubishing has a cost as all other business as well. Ever thought of turning a crappy LaTeX (well, still far better that crappy MS Word) into a well-structured XML document? Takes 2 hours in average. Oh yes, the guy that does this needs a salary for those two hours. Ever thought of asking a company or hiring an IT staff to develop software for you (rates start at 100 USD/hour for an experienced developer)? Count several dozens of hours per month. Ever thought of the real cost of a professional server environment (I am not talking about that crappy server setup under your office table)? Backup? If your data amounts 1 TB you will quickly have to switch to harddisk arrays and network attached storage solution. Oh yes, that costs several thousands bucks and needs to be replaced after four years. Telephone lines? E-mail? Computers for the staff members? Some furniture for the office? (Yes, the editors and publishers also want to have a desk and a chair and a decent computer with a screen)... Tax? Further education and training of your staff? Time and money lost on hiring lazy guys and girls that needs to be kicked out of the company again?...

Thanks for the data Bob. The $2K/article is comp...

2009-10-26T13:29:38.333-04:00

Thanks for the data Bob.

The $2K/article is comparable to the cost of a conference publication, but I am willing to accept that there is a cultural element there, preventing the switch to pay-to-publish.

BUT, given that ACM is also supported by membership subscriptions, and in addition gets the subscription fees from all the research libraries, I am still wondering why ACM is still not Open Access, at least partially.

The journal Computational Linguistics just went fu...

2009-10-26T13:21:01.047-04:00

The journal Computational Linguistics just went fully open access this year. Submissions and editorial are handled by the editor and production/distribution by MIT Press. It's paid for out of membership dues and conference fees paid to the Association for Computational Linguistics.

My recollection is that it's in the US$20K-30K per year range all in. It's published quarterly. That includes partial salary for an editorial assistant, professional copy editing, and production/distribution.

This works out to less than the US$2K/article or so that the pay-to-publish journals charge. Indeed, NIH will pay publishing costs for these journals. NIH requires all NIH-supported work to be published free at PubMed Central. Many proprietary journals open their articles after a year or so.

Like other journals, the scientific side is handled by a volunteer editor, a volunteer editorial board, and volunteer reviewrers.

I've stopped reviewing for or submitting to anything that's not open access.

Bob, I will not disagree with you about Open Acces...

2009-10-26T09:47:51.444-04:00

Bob, I will not disagree with you about Open Access. Actually, I doubt that anyone would disagree. However, most publishers cite the cost of publishing as the main reason they cannot offer the articles for free to the public.

So, I am looking to find out the real cost of publishing. I somehow suspect that we are feeding a huge bureaucracy, and not any value-adding service. But I am open to see the data and be convinced otherwise.

As someone with a very broad, and often not very d...

2009-10-26T09:26:26.569-04:00

As someone with a very broad, and often not very deep knowledge of many areas of science, I still want to be able to read (sometimes) very detailed and complete articles.

I'm not an academic, but I still have interests ranging from CS, Engineering, Biology, Genetics,Human relations, geopolitics, the law etc. etc. No university access. There is no way that I would subscribe to journals in all of these areas in the hopes that there is an occasional really important article that is not important enough to be well covered (and commented upon) by the millions of blogs, thousand news sources but isn't available because it is to important to not be secret? That wouldn't make sense. I coud pay per use, but in most of these sites you can't tell enough if you are going to scan the paper and find it is useless...then too late to ask for a refund.

Since much of these articles are supported by govt funds or are used to further the economic interests of the writer (tenure, speaking engagements, importances in the field) then making the reader pay seems backwards ... the writer gets the benefit from every reader.

shouldn't it be considered unethical for supported research to be published under restricted terms?

also, i hate registering for a site, buying an article (maybe once a couple years) knowing that there may be only a 1 in 2 or 3 chance on a specific site that the article will be really as important or interesting as it looks. i'm out the money and my time anyway.

Oh, you could be right, I don't know specifics.

2009-10-26T02:44:08.709-04:00

Oh, you could be right, I don't know specifics.

Conferences are a profit center for ACM, as far as...

2009-10-26T00:44:39.700-04:00

Conferences are a profit center for ACM, as far as I know.

Although the CACM does disseminate information, it...

2009-10-26T00:42:09.930-04:00

Although the CACM does disseminate information, it's also there to encourage people to pay membership dues that let the ACM as a whole hold conferences, publish proceedings, and carry on other activities that lose or break even on money!

>If you pertubate a little bit the input before fe...

2009-10-14T17:47:03.039-04:00

>If you pertubate a little bit the input before feeding it to the simplex algorithm, then it is almost impossible for the pertubed input to generate an exponential running time.

This starts to look an awful lot like using randomized algorithms. E.g., choosing the pivot in quicksort randomly makes it virtually impossible to get the worst-case running time of O(n^2). Basically, you're using random bits as noise.