The "What I Expected; What I Got" Thread

[P A]

I have yet to finish my bachelors in statistics, but I can't help but wonder what I'll be thinking whilst sitting in a 5x5 cubicle assessing risk for some corporate beehive 10+ years from now. I mean, I'm fully aware and appreciative of what my chosen major might entail, but is anyone else currently attending university similarly curious about what will happen on the other side? Who regrets their chosen major and the career(s) that resulted therefrom? Who expected something from their education that they didn't get? Or how about vice versa? Has anyone been pleasantly surprised by an opportunity irrespective of their education that landed them in an unexpectedly awesome career?

 

[double ewe]

i actually do assess risk for a corporate beehive (quant at a bank), and i would say it's better than i expected. i work with interesting people, solve interesting problems, and rarely stay past 5. however, i definitely interviewed at other risk assessment/corporate beehive type operations that would have made me miserable.

 

[P A]

In your experience, is the job market for financial mathematicians anything like it used to be, what with the whole '07 credit crunch thing?

And what exactly is a 'quant?' Some kind of financial analyst or actuary?

 

[double ewe]

quant is short for quantitative analyst - basically anybody who builds/tests models. i mostly do pricing/valuation of loans and securities. i actually had intended to be an actuary when i started my masters (I've passed the first two SOA exams), but chose this position instead.

there's definitely still plenty of hiring in the actuarial and math finance fields. banking regulation and healthcare reform, among other things, have created a lot of opportunities. i would say that programming experience and communication skills are two of the best ways to get yourself noticed.

 

[P A]

banking regulation and healthcare reform, among other things, have created a lot of opportunities.

Would you say that these new opportunities cater equally well to the skill set of a statistician, or are most stat jobs still routinely government- and/or research-oriented?

 

[double ewe]

not quite sure what you mean - there are a lot of statisticians in both the public and private sector. the examples i mentioned were a reference to the banks and insurance companies increasing hiring to keep up with the demands of the new government regulations (although the govt hires plenty of stats/math/actuaries as well).

that said, statisticians are becoming more and more necessary in nearly every industry as the huge boom in data availability creates untold opportunities to improve their business. i don't think there's a company out there with enough people to analyze all the data that's available.

is there a specific field your aiming/leaning towards?

 

[P A]

not quite sure what you mean

the examples i mentioned were a reference to the banks and insurance companies increasing hiring to keep up with the demands of the new government regulations (although the govt hires plenty of stats/math/actuaries as well)

Well, this is basically what I was asking about, particular re. insurance companies, healthcare stuff, etc., which, I presumed, constitute a large share of the stat/actuary/analyst job market today.

statisticians are becoming more and more necessary in nearly every industry as the huge boom in data availability creates untold opportunities to improve their business. i don't think there's a company out there with enough people to analyze all the data that's available.

This jives with what I've been hearing, and is certainly good to know; thanks, W.

is there a specific field your aiming/leaning towards?

This brings me to another question I wanted to ask you: To what extent are the skills that a stat/applied math major attains at the advanced undergrad level (probability theory, advanced calc, other proof-intensive shit) actually utilized in a typical occupational setting? I know that you're not a job recruiter or whatever, so I don't expect you to know all the ins and outs of the 'math sector,' but I'd really like to hear from someone who actually works in the field nevertheless. I mean, I doubt seriously that the typical daily workload of a 'quant' at a bank includes deductive proof and complex vector analysis - or am I just showcasing my profound ignorance at this point? I guess the question is, succinctly, what kind of actual math usage can I reasonably expect as someone who cannot and will not major in pure maths and would prefer to work in some industry or another, i.e. someone who doesn't particularly care for lengthy proofs and complexity for its own sake? While I'm sure that the answer varies by the particular industry in question, there must be some kind of 'mean' that exists throughout the job market as whole...right?

To answer your question, I'm currently leaning toward biomedical research, although that may be too intellectually demanding to sustain in the long run. I could easily see myself becoming quite weary of analyzing data and drafting reports if the job didn't compensate my efforts with $$$$.

 

[Transcendence]

 

[double ewe]

This brings me to another question I wanted to ask you: To what extent are the skills that a stat/applied math major attains at the advanced undergrad level (probability theory, advanced calc, other proof-intensive shit) actually utilized in a typical occupational setting? I know that you're not a job recruiter or whatever, so I don't expect you to know all the ins and outs of the 'math sector,' but I'd really like to hear from someone who actually works in the field nevertheless. I mean, I doubt seriously that the typical daily workload of a 'quant' at a bank includes deductive proof and complex vector analysis - or am I just showcasing my profound ignorance at this point? I guess the question is, succinctly, what kind of actual math usage can I reasonably expect as someone who cannot and will not major in pure maths and would prefer to work in some industry or another, i.e. someone who doesn't particularly care for lengthy proofs and complexity for its own sake? While I'm sure that the answer varies by the particular industry in question, there must be some kind of 'mean' that exists throughout the job market as whole...right?

i would say almost all jobs would have at least two things in common:

-you will do a lot of programming/number crunching with excel, sas, matlab, r, spss, etc.

-you will have five minutes to explain something you spent two months learning to someone who's got three other things on their mind and is really just ready to get the hell out of this meeting already.

if you want to make yourself more marketable, those are great things to work on. and if either of them sound completely miserable to you, you're probably going to have a bad time. in terms of the math you'll actually use, it really depends. i work with a lot of different kinds of models, so i sometimes have to go pretty deep into the math to really figure something out. i'm certainly not proving anything, or quoting statistical properties (WolframAlpha ftw) - but i think learning those things is a really good way to get the deeper understanding of math/statistics that you need to succeed in/enjoy this career path. and there are also jobs with a lot less theory and a lot more programming/data-manipulation - depends on the position and the company. at the entry level, you're going to have to learn the job as you go no matter what you studied, so versatility is usually going to trump specific expertise.

likewise, math definitely isn't the only way to get into math finance/risk analysis. there are a lot of quants/actuaries with degrees in other fields (engineering, physics, etc), but that's also because their backgrounds demand so much math that it's almost second nature to them.

 

[P A]

explain something you spent two months learning to someone who's got three other things on their mind and is really just ready to get the hell out of this meeting already

I have never experienced any difficulty in constructing declarative sentences, whether in verbal or written form. I foresee no difficulty in communicating complex, abstract ideas to harried, pragmatic people.

programming/number crunching

While not excruciating in and of itself, this sort of thing can become draining, even for those in the IT professions, I'd bet. I'm tempted to say that perhaps even manual labor would be preferable to a job that requires extensive IT training/expertise. Again, this may be my sheer naivety betraying me as a fool, but if I wanted to be a console jockey, I would be majoring in IT, not stat. I understand that computer models are where it's at these days, but are these jobs really so reliant upon bottom-up manipulation of software that I must double-major in CompSci just to give myself a competitive edge? If so, I want out. Before switching my major a second time, though, I'd like to know what you mean by 'programming.' As in, C++, Java, et. al? Or are you referring more to user-friendly applications whose settings are in constant need of fine-tweaking? I'm perfectly fine with crunching numbers for a living (it's what I thought I had signed up for, after all), but I literally cannot fathom the thought of hunching in front of a computer screen performing tasks for which I was not even formally well-trained trained - day in, day out.

at the entry level, you're going to have to learn the job as you go no matter what you studied, so versatility is usually going to trump specific expertise.

Versatility of what sort? Mathematical, or interdisciplinary?

 

[muvolution]

I studied:
Architecture

I expected:

I got:

My job is actually alot better now, but I thought I was going to off myself for a while there.

 

[double ewe]

I'd like to know what you mean by 'programming.'

here's a good example of a programming task you might run into:

you have a lot of data in some format (SQL, Excel, comma-delimited, etc) that you want analyzed. so you write a small script in SAS or Matlab to import/clean/organize the data, run it through your analysis (whatever that may be), and then output results/summary statistics to excel for some final manipulation/graphing/reporting. so it's kind of a mix of the technical programming to get the data in and out of your analytical engine, the statistical programming to conceptualize and design the analytics, and the general statistical/mathematical knowledge to know what tests to run and how to interpret their results. it's not the kind of thing you need a comp sci degree to do, but you also need to be experienced enough with programming in general to figure it out quickly.

I'm perfectly fine with crunching numbers for a living (it's what I thought I had signed up for, after all), but I literally cannot fathom the thought of hunching in front of a computer screen performing tasks for which I was not even formally well-trained - day in, day out.

i think you're making it out to be more repetitive and mindless than it is (at least in my experience).

you're going to be using a computer, and you're going to be doing things that are only tangentially-related to what you learned in school (particularly at the undergrad level). this is true with nearly every white collar profession you can come up with. but personally, i see it a lot more optimistically. i come up with creative solutions to interesting, challenging problems using sophisticated software and logical mathematical reasoning. i really like that i'm not doing all the same things i learned in school, and i'm very grateful to be in one of the rare careers where i will always be learning ("I feel like I'm getting dumber" is a common complaint of a lot of my working friends). and the only related arena where you won't see as many spreadsheets is in academia, but that just means you'll be using more obscure programs to solve more abstract problems.

the reaction i get from pretty much everyone when i tell them what i do is "that sounds absolutely miserable." so there's no shame if it's something that just plain doesn't suit you. but there's a reason "mathematician" and "actuary" are always at the top of job satisfaction surveys - for the people who like this stuff, it's a very satisfying, well-compensated job in a relatively comfortable environment.

Versatility of what sort? Mathematical, or interdisciplinary?

a little of both. some days you may be working in SAS, some days you may have to figure out what a couple lines of VBA code mean. and you'll likely also need to learn a bit about whatever industry you're in as well (i.e. credit markets, health insurance, biochem, etc).

 

[P A]

Thank you so much for the insight, W; this little chat has been a really big help for me.

"mathematician" and "actuary" are always at the top of job satisfaction surveys - for the people who like this stuff, it's a very satisfying, well-compensated job in a relatively comfortable environment.

This fact, if truth be told, constituted 90% of my rationale re. my choice of major. It seems that people with abstract, 'math-y' job descriptions are consistently rated as being better off financially and psychologically than those whose tasks are comparatively concrete or non-numerical in nature. I guess that math-intensive jobs are among the few professions that truly allow you to 'do what you love' (if you love math) for a living, with fewer hurdles and ancillary responsibilities to bog you down. Or that's my impression, at least.

you're going to be doing things that are only tangentially-related to what you learned in school (particularly at the undergrad level)

Okay. This is exactly what I needed to know. I'm going for a masters in applied maths or stat, and I'm not under the illusion that I'm going to be doing anything resembling undergrad classwork for a living once I graduate and enter the job market. I was just mildly concerned about the conceptually difficult, abstract nature of higher-level coursework and whether any of this would be extensively applied 'in the field,' since I'm not a 'natural' of any sort. I've gleaned from your answer that it isn't something that I should be particularly concerned about at this point. Thanks again.

it's not the kind of thing you need a comp sci degree to do, but you also need to be experienced enough with programming in general to figure it out quickly.

That's a relief. I expected to acquire many/most of these rudimentary computing skills in a graduate program, job training, etc. Is this reasonable, or do you think that I should begin familiarizing myself with the software right now, just to be sure?

and you'll likely also need to learn a bit about whatever industry you're in as well (i.e. credit markets, health insurance, biochem, etc).

Of course. Are you aware of the minimum expectations for an 'average' biochem/biomed applicant? Would I need to have taken advanced courses in some area of specialty (i.e., practically double-majored in biology or something), or is the work that is ultimately given to mathematicians generally confined to the number-crunching/programming/analysis aspects of those fields? I'm working under the assumption that it's mostly the latter, but I could easily be very wrong.

 

[Titus]

Apologies for butting into your convo. I know nothing about the fields your talking about but found this thread to be an interesting read, the back and forth exchange between you two.

I have a question. I used to play poker (live and on the internet) part-time and earned OK money for this being a hobby of mine. For a while I did it after work on a computer, while lying in bed, watching TV. The online games have since gotten harder as the competition stiffened, so I stopped as my earnings were declining. The college kids (no insult intended, I'm 45 is all) who were the smartest in math were dominating the online games. I don't know if either of you have played poker, or understand there is math involved, both basic and more complex like game theory. There were (I've been away from it for a few years or so) college kids making $100,000 and up playing internet poker part time while in college. Some making substantially more. And now the game of poker is dominated (both online and live) by the younger generation, in particular by those who have a sophistocated understanding of complex math.

I'm curious to know if either of you have played internet poker for profit or know of any fellow students who have done so?

 

[P A]

I'm curious to know if either of you have played internet poker for profit or know of any fellow students who have done so?

I've actually read (what seemed to be) firsthand accounts of math majors being diverted into the world of on- and off-line professional gambling, especially poker and blackjack. Knowing probability theory would certainly give you an edge over someone who doesn't - you're playing a game of chance, after all - but the ability to perform mental calculations at the rate that I imagine would be necessary to win the average hand of poker would be quite a feat in itself, a real talent that can only be honed, not learned nor taught. This is, at least, my limited understanding of the phenomenon. For what it's worth, no math or science major that I personally know has gained any notable infamy as a card-shark or counter. I also can't attest to the sweeping allegation that online poker has been taken by storm by college kids (presumably the stars of 21). This sounds more like a parody or urban myth than a plausible Darwinian uprising of supergenius whipper-snappers. The fact of the matter - as I understand it anyway - is that card counting is a tedious, difficult, and painstaking task, and is relatively moderate-risk, low-reward in the short term; and any other applications of PT would likely be similarly haphazard and strenuous in such fast-moving environments as poker or blackjack games.

However, if, as you say, more and more plucky mathematicians are signing on to play professionally, the ante may well have been upped for those who are used to playing solely by fair odds. The more card counters there are at any given table, the more difficult it is likely to be for (statistically) untrained players to make 'fair,' successful bets. One doesn't need a degree in math to figure that one out.

 

[Titus]

Well card counting is different. The math involved in that, in the game of blackjack, is actually pretty basic. I've counted cards and did so for a profit, but it was a relatively low pay grind.

With poker there is math involved at a whole other level. There is the basic math around understanding the probabilty of you hitting your hand when on the draw, factoring in 'pot odds', implied odds in getting paid off.. and then there is calculating 'hand ranges' of other players. Poker is a game of luck, with a skill factor involved (actually a variety of skills). Luck is the same for everyone. The greater your skill, the greater your chances of winning over the long run. But the calcuation of hand ranges of other players and how that factored into the game was above my skill level.

I don't know how many professional card counters there are nowadays. It is not illegal to count cards, and the casino's have many built in counter measures. You only have a slight edge even when you do it perfectly (the hard part is actually maintaining the count with all the distractions in a casino while at the same time trying to vary your bet to make it as much as possible when the count is in your favor without this being picked up upon by casino employees). I made like $500 a weekend, but that was working 14 hour days, and it was in the largest casino in the world so a little easier to fly under the radar.

When done as part of a team, like the movie 21, which was a true to life movie as there actually was a harvard blackjack team- one former member, Andy Block, is a pro poker player now and has been for some time- you can make lots of money because the abilty to bet large when the count is in your favor is done by only 1-2 members of the team and they are always betting large so it doesn't look unusual (or didn't until casino's discovered the existance of blackjack card counting teams).

When 'caught' counting cards casino's will tell you your welcome to play any other game in the casino but blackjack and have a tresspass violation put on you if you are caught playing blackjack there again. Doing it as I did my bets would vary from $10 (the minimum) most of the time, to a max of $250-$300 per hand when the count was in my favor if I could get away with that size of a bet variation. Big jumps in bet size (or big drops) instantly will get the dealer to yell out "Checks!" (what they call betting chips in a casino) and that will bring the pit boss right over to the table, the dealer will point out the player-if it isn't obvious to the pit boss- and he will stand there counting cards as the game continues while watching you play, paying particular attention to how you vary your bet size, if at all, depending on the count.

With poker, your playing against the other players, not the house, so the house could care less how skilled or unskilled you are (aside from being willing to notify pro's who frequent the casino they work at when a big money fish shows up, or getting you into the table with the most fish, for a tip). When I played on the internet I eventually played 8 tables simultaneously. Some people played 12, I heard of one guy who played 16. Eventually I found myself sitting at the 8 tables with 6-7 of the same people at most all of the tables I was sitting at. Obvously they are skilled and playing purely for profit, NOT what you want at your table, unless you have a distinct edge over them.

It's not at all an urban myth that there are a lot (by a lot maybe 10 thousand) college kids, college grads, college drop outs, who are successful pro poker players making 100K/yr and up. In the early days of online poker the lack of skill was ridiculous, meaning those who got in early cleaned up easy and big. What it essentially comes down to is your looking for opportunities to 'flip a coin' where you have a 70% chance to their 30% chance of winning, while betting even money on the coin toss. So playing online, at 8 or 12 the opportunity to engage in such coin tosses come up 8 or 12 times more often.

The reason I asked is b/c you guys are in or recent college grads in the math field and what was happening 6 or 8 years ago is that after some unskilled young guy won the world series of poker, and it was televised, a lot of people started playing. A lot of them were college kids and some of them started reading the basic books that explain the basic math around the diferent poker games. But when some of them started making money, being young they were bragging to their friends about how they were making X ammt of money they were making relative to their friends working at the local pizza joint, everyone and their brother read a few books and started doing it. Most were probably at my level, ok not great. I had to step back when it got to the point that 1/2 or more of my table was filled with players who had the basic skills to be profitable.

But the really smart math guys could dominate all of us, build higher bankrolls, and play at higher stakes games. The one's that dominated there, some where making several hundred K/yr and decided to go pro. Both online in cash games and tourniments. Some moved to vegas where it's the ultimate place to play because people go there from all over the world to party, get drunk and wild, and many go with X thousand that they basically decide is gambling money they really have no intentions of coming home with.

The tourniments that happen all over the world have poker pro's who travel year round to play in them. The average winning pro doing this was in their 40's 10 years ago. Now the guys making the final tables are mostly all in their early 20's. In addition to the math knowledge they have other advantages that come with youth- better memories, better stamina...

So it's definitely not a myth. I was a member of a very large online poker forum where people discussed stragegy, etc and info flowed (still does I'm sure) about who was who and there was/is softwhere out there that most people used that would overlap the poker tables online and keep running stat's on all the players. It's called poker tracker. People built databases on one another and could see important stat's on the way they played as well as their earnings, etc. The negative side to youth is that a lot with a lot of potential took too much risk and played above their bankroll and short term risk took them out. The wise one's didn't do this and remain successful to this day.

 

[double ewe]

very interesting stuff, especially about the data-oriented approach. i imagine there are a lot of overlaps between that and "big data" marketing in terms of quantifying and predicting behaviors.

several of my peers (twenty-something math guys) have been very successful in poker. one of them paid his way through yale undergrad w/ online winnings, and another dropped out of a graduate econometrics program at columbia to play full-time in vegas. the third, in nanotech, quit playing underground cash games when he had kids. personally, i've played a good bit of poker, but it never really stuck with me. my math background was definitely an advantage (although i probably owe my bs in psych just as much credit), but it's simply not my thing. really just more stress/pressure than i want from a hobby.

and i would definitely agree that it's more about that deep, instinctive knowledge of probability than actually calculating particular odds (although those definitely help too).