Book Release! and more
Oh my god, people, today’s the day! I’m practically bursting with excitement and anxiety. I feel like throwing up all the time, but in a good way. I want to go into every bookstore I walk by, find my book, and throw up all over it. That would be so nice, right?
Also, I wanted to mention that Carrie Fisher, who is a SUPREME ROLE MODEL TO ME, has just started an advice column at the Guardian. How exciting is that?! So please, anyone who still mourns the loss of Aunt Pythia, go ahead and take a look, she’s just the best.
Also! I’m into this new report, and accompanying Medium piece, by Team Upturn on the subject of predictive policing. It explains the field in a comprehensive way, and offers a convincing critique as well.
Also! It turns out I’ll be in Berkeley next Friday, here’s the flier thanks to Professor Marion Fourcade:
I hope I see you there!
mathbabe 
Amazon’s notification email showed up in my inbox at 9 PM sharp Pacific time last night so I promptly downloaded the book to my iPad and now I’m reading it. It’s easy to see early on that this is going to be a great read!
LikeLike
Congratulations!
Being in the UK I find myself wondering about the vagaries of the international publishing industry with which you are now entangled. Your “buy now” link on the WMD website took me to Amazon.com (not really a surprise as most US website designers tend to forget the rest of the world – even when writing is the author’s main source of income and they need to maximise the chance of a sale). Highlighting the title and doing a google search came up with Amazon’s UK site as the top result and I was thus soon comparing the USA and the UK.
It’s a bit weird. Not only do the publishers vary by country and format (which is not unusual though seems a bit wasteful as they’re all Random house subsidiaries) but the USA has no kindle version and no “look inside” option; the UK has a kindle version (with “look inside” of course which must help sales) and the new paperback but no hardbacks, save for used copies from the marketplace.
In this case being in the UK is a definite advantage even though the kindle price was “set by the publisher” and thus at a level at which it would normally just result in the book going on my “price too high” wish list to wait for the inevitable price drop or the appearance of the “£0.01 plus p&p” used versions. However, in this case I bit the bullet, hit the “buy now” button for immediate gratification and started reading. I’m well into the book and so far it is living up to expectations.
LikeLike
Of course, if Mathematrucker is in the US there must be a US kindle version which is just not showing up when I look at amazon.com. This would be weird though as amazon normally shows me the kindle versions – it just stops me buying them from the US.
LikeLike
Congratulations!!
LikeLike
Congrats! Keep the movement and the book going by all means.
LikeLike
Same experience … The Book trickled onto my Kindle last night, noticing it just before midnight, and couldn’t resist starting a read.
Congratulations!
LikeLike
Waiting excitedly for the printed book – and for your Berkeley talk! Congratulations!
LikeLike
Am just glad I got my copy before you came upon it and hurled on it! And you even have an endorsement blurb from Ralph Nader in the beginning — I am SO impressed.
LikeLike
Congrats, Cathy. Now the really hard work starts! But a great experience for a data journalism teacher. Hope you do for Stats what Piketty did for Econ. And more.
LikeLike
congratulations on your book. i read a brief summary on another site and was curious. if one’s zip code results in a greater number of police officers, one still needs to violate the law to be arrested, right? just because one lives in a “higher crime rate zip code” doesn’t mean they will get arrested, right? therefore, that person wouldn’t need to worry about sentencing algorithms. additionally, if that person made less money, their credit would only be poor if they spend too much and couldn’t pay their bills. however, if that person managed their money appropriately, it wouldn’t matter where they live and their credit score would be good, therefore it would bode well for them when getting a job, no? are the numbers (or the math) skewed against people because of where they live or their friends/family or instead because of the decisions they make?
LikeLiked by 1 person
I’m sure Dr O’Neil can respond for herself. On the other hand, she’s probably busier than heck right now. So, as someone who is reading her book and is a little familiar with her points at her blog (which I would also recommend), I’m putting in my response.
First, from the sound of it, you haven’t read the book. I think it’s only fair to suggest that you do so, hearing her more complete explanation, before posing a question which, according to your Comment, is only based upon a summary read elsewhere. For example, one thing Dr O’Neil covers in her book is the increasing use of credit scores for judging suitability of being hired. It should be obvious to anyone with quantitative sense that there’s a potential for a negative feedback loop there, but it apparently doesn’t occur to those using this technique. Someone, for instance, could take a hit on a credit score because they went deeply into debt to pay unexpected medical bills, lose their job because they were a caretaker, and end up taking lower and lower paying jobs, just to make ends meet, and be tarnished by a low credit score and therefore unhireable. This is yielding responsibility for decisions to mindless and heartless machinery.
Second, setting abuses of the data and algorithmic kind Dr O’Neil describes aside, the inappropriate application of statistical results to criminal justice is infamously large. I would recommend a careful look at Joseph Gastwirth’s (ed.) Statistical Science in the Courtroom, 2000, Springer, and some of the publications of Stephen Fienberg.
Third, even setting aside the notion that arrest does not mean conviction — which really oughtn’t be — the idea of conditioning sentencing upon Zip Code is wallowing in something which statisticians call the ecologic fallacy. This is the idea that the nature of individuals, in this case their risk to society, is predictable from the group to which they belong, in this case, the quite accidental one of Zip Code. The notion of doing this comes from online marketing, where a product will be pushed because a group has a penchant to like the product and the visitor to a site has been identified as a member of the group. For the merchant, they want to get a sale and the worst thing that can happen if the classification is done wrong is that they don’t. From their perspective they are improving their odds. But in the case of sentencing and comparably serious question, the idea of ignoring Loss when making a mistake is outright criminal itself. If it were military apparatus being used, there is an ethical argument to be made that it could be a war crime.
Read the book, in its entirety.
LikeLiked by 1 person
Hi,
Love hearing you on Slate Money. You should look into this algorithm about you tube.
LikeLike
Congratulations, Cathy! The books looks great. Can’t wait to read it. I found it this morning at Barnes and Noble on the ‘New non-fiction” table, which is very close to the entrance to the store – it is one of the measures by which some people judge popularity (i.e. success) of a book). May your book always be placed on a display as close as possible to the entrance to every bookstore. I started reading it right there at the store and a few store ‘browsers’ asked about it.
LikeLike
I have started listening to your book and am floored, in a very good way. I am a senior data analyst. A few years ago I lost confidence in the data algorithms used to manage the business aspects of my organization. I could not put it into the elegant terms (as you have in the book) but I knew the algorithms were considered too correct, the logic was very elusive, and did not allow the users to impact change to either get the results the algorithm was meant to produce or provide the information to know what part of the process to change. I have come full circle since I was a new analyst. Now, my business intelligence tools focus on distinct parts of processes (instead of nebulous algorithms), always includes the raw data and the logic, and always emphasizes human (on the ground) interpretation. I am learning every day. Leadership always pushes for simple “RED, YELLOW, GREEN” Algorithmic BI Tools of complicated multi-system processes but I purposefully put most of my focus on the process parts and less focus on the leadership Algorithmic BI Tools. My focus (which has been much more fair and impactful) is making information available and useful to the people doing the job and their supervisors. This availability and usefulness almost always impacts the overall mission and the Algorithmic BI Tool the leadership had wanted. Your book put into words what my gut is/has been saying. I look forward to finishing your book and have told many of my friends and colleagues about it too.
LikeLike
Cathy,
I’ve just read a brief article about your book. I’ve not read it. However I am excited that there’s someone out there researching this topic. It’s tangible and thus it can be evaluated and reformed unlike some of the ambiguous arguments that are being championed that are more difficult to prove, argue, articulate or that are outright falsehoods. These metrics DO disenfranchise people and they’ve created barriers that need to be softened. They’ve effected me personally and I’ve often wondered how hard it would be for someone with less access to information/education. Please keep it ethnicity neutral – we have a lot of people that could be bigger contributors in our economy that are being sidelined and they come from all backgrounds. This should be a HUGE topic today as we can make changes in this area!! I congratulate you and thank you for taking up the torch! I’ll try to grab the book Cathy and thank you again. Thumbs up Cathy!!
LikeLike
Congratulations on book day! Can’t wait to read it 😀
LikeLike
Interesting…. Such algorithms will never cease to exist as they are much too imbedded in the business, engineering, socio-economic, and political environments that our world revolves around. The “ones” that control things, you know, the “ones” with the majority of wealth in the world, will never let such dark corridors into the light. Algorithms manage and control complex business empires, medical devices, weapon systems, transportation systems, and so on. Sure one can write books, articles, etc. demanding accountability for such diabolical math but it to I’m afraid is highly likely to go the way of tax reform, health reform, education reform, and so on. Not pessimism, unfortunately, realism…
LikeLike
Maybe! I’m not fatalistic.
LikeLike
In my opinion, the responsibility for “ending” use of improper and inappropriate algorithms lies with the engineers, data scientists, and statisticians charged with implementing them. Sure, you can get the situation as happened with CDOs (subprime crisis, collateralized debt obligations) that the quants who complained about the inappropriateness of the calculations got replaced by those who did not, but, still, having a group say quite loudly that something is incorrect and improper can help give those who remain courage to stand up. This is nothing new.
Engineers in industry have stood up and complained about inappropriate practices and designs. Sure, there were consequences, but with a professional position comes a responsibility for ethics. Nearly every professional association has such as code. For example, here is the code from the American Statistical Association.
LikeLike
I have not read the book, but based on some of the points I’ve read, have to say the mortgage zip code part is dated. As of 2008, Dodd-Frank Act and the CFPB will not let a company discriminate based on a zip code. This would result in a major lawsuit and is not worth it for legality purposes. The mortgage industry is now OVER regulated, so let’s clear the air here and settle into reality! Stop will the dated Occuppy Wall Street crap
LikeLike
I’m not sure what you’re referring to, but most of what I talk about in the book refers to new-fangled types of big data algorithms, where there is very little regulation indeed.
LikeLike
Well done. Years ago I sued a homeowner insurer for redlining discrimination — for using various underwriting tools, although facially neutral, to exclude or price based on race. For example, refusing to issue insurance when the replacement value is higher than the market value, which cuts out many urban (black) homes and favors suburban (white) homes. This is called “disparate impact” and it is illegal under the Fair Housing Act. I remember deposing a PhD actuary for many hours — a very smart and kind man. He held to the idea that all he was doing was math — just running data on how various underwriting would, or would not, tend to affect urban area zip codes. He said he had no opinions, this was all just math and his charts were all just data. But the person to whom he sent his work then used the data to adopt criteria that would, in fact, hurt people based on race — all thus perpetuating the cycle of under-insurance/under-development in black urban neighborhoods (and the opposite for white suburban ones). We finally hit a point in the deposition where he understood, and conceded, that his superiors might use the data in precisely this way. And that they had in fact used the data in this way, and that they used his skills for this very purpose. I will never forget the look of sadness of his face — he was so smart, and so mathematical, but he had been used, and he understood it. I have taken 250 depositions in my career. That’s the one I will probably always remember best.
LikeLike
“Carrie Fisher, who is a SUPREME ROLE MODEL TO ME”.
Well, some context would be helpful on why a Hollywood actress is a role model. In what ways is she different from the usual crop? What are her roles? A slave princess in a bikini/harem dress (an apt description for most big heroines in entertainment) in the thrall of Jabba the Hutt (a moneylender and general big time shadow operator, again an apt description for who truly controls Hollywood). Is that the SUPREME ROLE MODEL?
LikeLike
Cannot pretend to know Dr O’Neil’s admiration of Carrie Fisher, but there is Fisher’s novel, SURRENDER THE PINK ( https://en.m.wikipedia.org/wiki/Surrender_the_Pink) which was quite a journey of self-discovery.
I actually find your comment rather boorish.
LikeLike
You are the boor. When challenged, you show a romance novel by her. There are hundreds of thousands of them. It is not a priory obvious why this particular one is special. You are not able to explain why the novel is meritorious or cite acclaim from literary circles, or at the very least, high sales like J K Rowling got. You simply assert, it is “quite a journey of self-discovery”. I just had a brief look in some book websites and there are negative comments as well. Therefore I would have to take your unexplained opinion with a pinch of salt.
LikeLike
Guys, I’m obviously not talking about the characters she’s played, but the woman herself. She’s incredibly hilarious and feisty, look into it.
LikeLike
Well done! Thank you for the inciteful work.
LikeLike
I just read the article about your and read some details about your work. It is amazing. Well done.
LikeLike
Thanks for quantifying something I see every day as I work in an industry that depends on a math model for almost every decision the company makes.
LikeLike
Congratulations on carrying this through to release and beyond. You’ve done a great service for us all.
LikeLike
Cathy, Congrats! PLEASE give us a list of where you’ll be talking!!! The people demand to know! Cheers, Joel
LikeLike