Debating with Aashir on the NBA

Aashir is quite the NBA fan.

Instead of preparing for his upcoming MCAT, he typically spends his late nights watching Youtube videos of obscure NBA events. He’s also extremely generous, so he occasionally shares these videos with me — the most recent being this eleven minute compilation of NBA fights.

In middle school, Aashir never made the school basketball team because of his weight issues. Despite these failures, he was still a self-proclaimed top 5 basketball player in his grade. I remember us discussing his prospects of making the NBA, and whenever I would say only one in a million make it, he would always confidently respond with “what if I’m that one?”

In the midst of MCAT preparation, he still made time to watch most Playoff games and even attend two of them. His loyalty has sometimes been questionable, like when he forced us to leave a Rockets game early and miss T-Mac’s historic 13 in 35. Nevertheless, the impossible trade scenarios he routinely creates for the Rockets prove his commitment to the organization.

Unfortunately, our discussions on the NBA are often… inadequate? I can’t think of the right word to describe our conversations on the subject, so maybe an example will help.

Yesterday we were comparing LeBron and Jordan. Aashir said Jordan was without a doubt the better player. I said it was impossible to conclusively pick one player over the other, so I decided to play devil’s advocate and back LeBron.

Aashir’s first argument was that Jordan’s team would suffer more than LeBron’s team would without their respective superstars. Fortunately, we were both at computers, so I checked his theory. The Chicago Bulls went 57-25 in the 1992-93 season with Jordan and then 55-27 in the following season without Jordan. The Cleveland Cavaliers went 61-21 in the 2009-10 season with LeBron and then 19-63 in the following season without LeBron. Aashir said my evidence was invalid; there are too many varying variables (I think he meant confounding variables).

Aashir’s second argument was that Jordan made his teammates better, even more so than LeBron does. I pointed out that Jordan had only one season averaging above 6.5 assists per game, and LeBron has already had seven. Aashir said my numbers were meaningless; assists are only one way of measuring the impact on teammates.

He reaffirmed the “helping teammates” argument with a more specific example — Scottie Pippen. Aashir claimed that Jordan turned Pippen into a Hall-of-Fame player when Pippen won four championships with Jordan. I was taken aback by this, since I had always thought they won all six championships together. A quick search on Google confirmed I was right, and I pointed this out to Aashir.

He acknowledged the mistake, but he claimed his original point was still true; Pippen was a horrible basketball player before he was on the Bulls. Not knowing much about Pippen, I checked his stats online. The first NBA team Pippen played for was the Bulls.

Again, Aashir acknowledged his mistake… followed by yet another clarification. Pippen may have started off playing for the Bulls, but apparently Chicago never thought he was going to be good when they drafted him. Of course Chicago didn’t. Pippen was drafted by the Seattle SuperSonics with the fifth overall pick.

Aashir concluded the discussion saying majority of NBA Analysts agree with him, so he must be right. He then told me I know nothing about basketball, and I will never know because I have never played at a high level. I said playing basketball shouldn’t be a requisite for analyzing the sport; we don’t only go to doctors who have had cancer before when we look for cancer treatment. He didn’t appreciate my response, so he ended the conversation with the last words — “you’re done boii.”

I can assure you, none of this is made up. None of this was even exaggerated at all. If you ever want to experience a debate without any facts, Facebook message Aashir any night this summer. Perhaps the most frustrating thing about Aashir is that he beat me in two Fantasy Basketball leagues this past year. To him, this serves as perfect evidence of his superior analytical capabilities.

Law School and More Discrimination

As a business major in McCombs, I’ve learned three main things: investment banking is $$$, consulting is badass, and everything else is useless. But limiting myself to 2 career paths seems a bit premature as someone with no real work experience, so I try to keep an open mind. One option I’ve always thought about is stripper law. I don’t know any lawyers, so I instead turned to google for more information about the law school atmosphere and lawyer profession. Here are my findings:

• Don’t Go To Law School, a book by a law school professor
• Some quotes suggesting you shouldn’t go to law school, by another law school professor
• Inside the Law School Scam, a video by two law school professors
• Mamas Don’t Let You Babies Grow Up To Be…Lawyers, an academic paper by (sigh) another law school professor

Maybe law school isn’t such a good idea. I decided to check the opinions on Reddit:

• “As an existing lawyer, stop minting new ones please.”
• “All I have to say is do not go to law school unless you really, really, really, really, really, really, really, really, really, really, really, really, really, really, really, really, really, really, really, really, really, really, really, really, really, really, really, really, really, really, really, really, really, really, really, really, really, really, really, really, really, really, really, really, really, really, really, really, really, really, really, really, really, really, really, really, really, really, really, really, really, really, really, really, really, really, really, really, really, really, really, really, really, really, really, really want to push paper the rest of your life. It’s not all Law and Order, people.”

Looks like the law profession really does suck. But remember, open mind! So again, I went to Reddit to get more information. Fortunately, I found two AMAs by law school admissions officers. Here are some brutally honest answers that really surprised me:

• “Performance is valued over majors.”
• “There is no pretending that LSAT and GPA aren’t the two most important factors.”
• “I would say it [LSAT/GPA/Other weights for admission] is more like 50/40/10.”

So of course I went to Google again. First result for “law school numbers”, lawschoolnumbers.com. They have (self-reported) data on admissions’ results for each school. Just for fun, I looked at Harvard’s:

Above is a graph of Harvard Law School’s 11-12 and 12-13 admissions’ results based on LSAT and GPA. Looks like the “50/40/10″ ratio is pretty accurate. I drew the diagonal lines myself — above the top line means acceptance, between the lines means maybe, and below the bottom line means rejection. But there are a few acceptances scattered below the bottom line; how’d they manage to get in?

Two words, under-represented minorities:

Above are the admissions results of under-represented minorities from the same pool with the same blue lines drawn. As an (unsuccessful) applicant to several top undergraduate universities, I’ve always known that affirmative action is used almost everywhere, but I never knew its impact was this substantial.

Apparently, I’ve taken it for granted that all semi-intelligent people acknowledge the undesirable consequence of affirmative action, reverse discrimination. Just a few weeks ago, I read an article by a black girl for affirmative action, and it was unanimously praised by my friends on Facebook. It’s funny how under-represented minorities often argue for affirmative action on the basis of racism in America. In our post-9/11 world, brown people aren’t exactly treated the best either. Can racism help me get into law school too?

On a side note, I’m glad topics like gay marriage are being questioned on a federal level. Marriage equality is a perfectly understandable request that should be granted immediately; let’s make sure to leave it at ending discrimination and not starting any reverse discrimination like we do with race. Oh wait…

Greatest Comeback in NBA History?

Just four days after giving up on my team, the Rockets are halfway to becoming the first to ever come back from an 0-3 deficit. We owe much of game 5 to Omer Asik (the beautiful man pictured above) for hitting 72% of his free throws during Scott Brook’s pathetic attempt at slowing us down. And of course, I can’t forget the new KD stopper (and I’m not talking about Russell Westbrook). After a miserable game 1, this has turned into quite the series.

Regardless, it’s still too early to celebrate. Historically, when the team with homecourt advantage is up 3-2 in a 7 game series, they win 91.9% of the time (148-13). In game 6, however, that team wins only 50.9% of the time (82-79). With the momentum in our favor, I expect to win game 6. It’s the last game in OKC that I’m worried about, especially with our most recent game 7 performances.

I’ve thoroughly enjoyed the playoffs thus far. Unfortunately, some comments I’ve seen / heard recently have made me lose all hope in the average NBA fan:

• “Carmelo Anthony is better offensively than Kevin Durant.” Except Durant shoots better than Melo from literally every location on the floor.
  
• “James Harden sucks ass.” Yeah, 10 turnovers in a playoff game is pretty bad, but he isn’t the only good player to have a bad playoff performance. If we’re going to rate players on a 1 game sample size, should we make a move for Kwame Brown this offseason?
• “Take Asik out!” Omer Asik shoots 56% from the free throw line. The “Hack-Asik” gives us two free throws each time. Simple arithmetic says 1.12 points per possession. The league average is 1.06 points per possession. Not to mention, getting a team with a short bench in foul trouble is always good.

I’ve come to realize that the average NBA fan is as knowledgeable about basketball as the average political junkie is about politics. The passion is admirable, but the misinformation is widespread. While the political information gap really kills all of us, hopefully Dork Elvis can take advantage of the basketball analysis gap this offseason.

Four in a Row?

Coming into this series against the historically amazing Oklahoma City Thunder, the Houston Rockets pretty much guaranteed themselves a first-round exit. The only real debate was whether or not we’d make it to game 5. Nevertheless, as any sports fan knows, extraordinarily improbable upsets can happen. Throw in Westbrook’s unfortunate injury — after never missing a game since middle school — and it looks like the Rockets have a legitimate chance.

And then we lost game three…

So I wanted to reevaluate our position; do the Rockets have any chance at winning this series? Well, teams that have gone down 0-3 in a best-of-seven series are 0-93 in NBA history. But 93 isn’t that big of a sample, right? So I decided to actually calculate the probability of this situation happening. Below is a graph showing how likely it is for a team to lose three in a row and then win four in a row given their probability for winning 1 game. Keep in mind, I made this quickly and did not account for any confounding factors (rest, homecourt, etc.):

The most probable scenario is when the team has a 57% (4/7) chance of winning 1 game. And even there, the probability of going 0-3 and finishing 4-3 is only 0.84%. Just in case you thought the 0-93 record was insane, 0.84% translates to a record of 1-119. And that’s assuming the max probability.

But the 0-3 has already happened for the Rockets; what we’re more interested in is our probability of just winning 4 games in a row. Below is a graph showing how likely it is for a team to win four in a row given their probability for winning 1 game. Again, I did not account for any confounding factors:

Suppose the Rockets have a 20% chance of winning each game against the Thunder. That puts us at a 0.2% chance of winning 4 in a row. But the last two games were (heartbreakingly) close, so lets say we have a 40% chance of winning each game! That gives us a 2.6% chance of winning 4 in a row. Even a really good team, one that has an 80% chance of winning each game, only has a 41% chance of winning 4 in a row.

So to put it simply, let’s start discussing who to sign this offseason with all our cap space.

Let’s Talk About Discrimination

Apparently, the Supreme Court is doing something regarding gay marriage this week. I don’t know exactly what’s being discussed, but I honestly couldn’t care less. From what I’ve heard, the Supreme Court doesn’t intend on making a nationwide ruling, and leaving the decision to the states won’t lead to much progress anytime in the near future considering over 40 states currently forbid gay marriage. If the Supreme Court or Congress really wanted to get serious about this issue, they’d invoke the 14th amendment:

Section 1. All persons born or naturalized in the United States, and subject to the jurisdiction thereof, are citizens of the United States and of the State wherein they reside. No State shall make or enforce any law which shall abridge the privileges or immunities of citizens of the United States; nor shall any State deprive any person of life, liberty, or property, without due process of law; nor deny to any person within its jurisdiction the equal protection of the laws.

Section 5. The Congress shall have power to enforce, by appropriate legislation, the provisions of this article.

That makes it pretty clear that gay marriage is a constitutionally protected right (…along with incest, lol). I’m glad my friends on Facebook feel the same way, but really, how much courage does it take to support something that 81% of 18-29 year olds are already in favor of? I admire the passion, but the effect is very marginal considering very few teenagers publicly oppose gay marriage (at least not here in the hipster capital of Texas). Regardless, at least I agree with their stance, and I’m sure the next generation of Congressman will pass a law promoting marital equality considering the growing support among young people and even among Republicans.

How about we switch our focus to help another group that faces discrimination — single people. A big reason gays want the right to get married is for the 1,138 provisions in which marital status is used to give out benefits. Marriage may have been a social good years ago, but is that a real reason to subsidize it? Asian kids are statistically among the smartest, should the government pay them to keep having kids?

Giving married couples a joint tax deduction seems pretty unfair to the single people, especially when marriages typically yield children (aka non-tax paying citizens for 15+ years). Single people still contribute just as much as married people to public schools, playgrounds, infant nutritional programs, and a variety of other programs they will never need or use. Soon enough, single people in America will face even more discrimination than gay people. If there ever comes a wave on Facebook in support of single people, perhaps then I’ll join the bandwagon and change my profile picture too.

MIT vs UT: Not Even A Competition

One of my great friends is on his way to MIT next year, but I’ve been thinking about convincing him to come to UT instead. Forget the academic reputation. The only reason to maybe pick MIT over UT is in hopes of doing better off after graduating, but even that’s not entirely true.

Nine months’ tuition and fees at MIT is $42,050. Tuition at UT varies by major, so let’s just pick the school with the most expensive tuition – McCombs, of course. Tuition for each semester at McCombs is $5,369 (for in-state residents), so yearly tuition at UT is $10,738. Assuming tuition stays the same (which it won’t), after four years, MIT costs $168,200, and UT costs $42,952. Surely the MIT degree immediately pays itself off, right?

Apparently not. Here’s a table of average starting salaries for graduating students from several majors. Not all the majors were the exact same (“Business” at UT, “Management” at MIT), but I did my best to pair them up. I only chose majors that I had data for and are somewhat relevant at UT. Sorry, liberal arts kids.

Major	MIT	UT
Mathematics	$76,869	~$48,000
Physics	$56,225	~$38,000
Management / Business	$70,256	$56,190
Electrical Engineering	$80,193	$69,014
Mechanical Engineering	$74,403	$69,044
Civil Engineering	$56,410	$55,131
Computer Science	$80,193	~$79,000
Biology	$38,778	~$38,000
Aeronautics / Aerospace Engineering	$59,643	$62,499
Chemistry	$42,267	~$48,000
Chemical Engineering	$66,842	$72,738
Biological / Biomedical Engineering	$49,400	$61,859

Looks like picking MIT over UT pays off a bit more than half the time. But really, unless you major in Math, Physics, Business, or Electrical Engineering (three of which I’ve been studying at UT…) you won’t be that much better off coming from MIT. And that’s not even considering the massive tuition difference. So really, why would you pick MIT over UT?

…Okay that wasn’t completely serious. Here’s my real question: how the hell are employers getting away with paying MIT students barely more than they pay UT students when kids like Lawrence go to MIT and kids like this go to UT?!

50 Days of Sleep

If you’ve known me well since I’ve been in college, you probably know how much I enjoy sleeping. If you talk to Solomon, Anuj, Nakul, or Akash, you’ve probably heard exaggerated stories of me sleeping for 30 hours straight or from noon to 8:00 PM multiple days a week. People legitimately believe I sleep for about 12 hours a day. Honestly, half of these rumors are at least partially true: I slept for more than 24 hours straight once in college, I did often sleep for extended periods in the evenings, and I did sleep 12 hours a day half the time. But I always balanced the 12 hour sleep days with 4 hours sleep days. Of course nobody remembers those though…

I was determined to prove everyone wrong and silence the false rumors, so I begin tracking my sleep. Here is my sleep log for the first 10 days of tracking:

Blue is college, and orange is Thanksgiving break, but it honestly doesn’t matter. My sleep schedule was equally screwed up during both periods. And that wasn’t just an abnormal 10 days; if I had begun tracking my sleep earlier, my entire semester would have looked the same as the blue period. Regardless, I averaged 8.5 hours of sleep a day over this duration, so the rumors about me sleeping too long were clearly inaccurate. However, my timings were definitely irregular. The main issue was I had little reason to wake up. The only classes I regularly attended last semester were BA 151H (Honors Lyceum) from 4:00 – 5:30 PM on Wednesdays, and EE 302 (Circuits Lab) from 9:00 AM – 11:00 AM on Fridays. My only early morning requirement was work in the BHP Office at 8:00 AM on Wednesdays.

Although I wasn’t averaging too much sleep, my cycle made Wednesdays especially difficult. I had to be awake at about 7:00 AM, and I worked the entire day till class was over at 5:30 PM. Unfortunately, I often couldn’t fall asleep til 2:00 or 3:00 AM, and this didn’t give me enough sleep to last throughout the entire day. So I made it a goal to sleep and wake up at the same times everyday. Here is my progress thus far:

Green is college, and red is winter break. My sleep certainly isn’t perfect now, but it’s a huge improvement over before, and it’s lasted quite some time now… even throughout break. I averaged 8.14 hours of sleep per day over this duration, so not too different from the earlier period. The real difference comes from my variance in sleep durations:

Period	Average	SD
College (Pre-Fix)	7.92	4.25
Thanksgiving Break	9.20	4.40
College (Post-Fix)	7.72	2.47
College (Post-Fix, w/o All-Nighters for Studying)	8.31	1.57
Winter Break	8.80	1.45

So the key for me was not letting myself have days where I sleep too little or too much. If I had no desire to wake up and do anything, I could easily let myself sleep for 11 or 12 hours a day. But sleeping this much gives me energy to be awake longer, which means I’ll likely fall asleep later the next day. On the other hand, sleeping too little doesn’t give me enough energy to stay awake long, which means I’ll likely fall asleep earlier the next day. The amount of sleep should correlate to the time til I fall asleep again. As an example, if I sleep at 2:00 AM, wake up at 12:00 PM, and stay awake til 4:00 AM, 10 hours of sleep gave me 26 hours between sleep times. Here’s a scatterplot of the lengths of my sleep vs the hours between my sleep, along with a line of best fit and it’s 95% confidence interval:

I notice several details immediately:

• There are many blue and orange points far from the center; this explains the high standard deviation in the table above when I had my broken sleep cycle.
• The best regression fit expresses diminishing marginal returns, which makes sense. Sleeping an additional hour when I’ve only had 2 hours of sleep should give me more “awake time” then sleeping an additional hour when I’ve already had 12 hours of sleep.
• There is a y-intercept, which doesn’t make sense. Sleeping 0 hours should not give any “awake time.” Hopefully this will be gone as I accumulate more data.
• The line of best fit intersects y = 24 at x = 7.61. This suggests that sleeping for 7.61 hours means falling asleep about 24 hours later. So if I want to fall asleep and wake up at the same time everyday, I should aim to sleep for 7.61 hours a day.

Once I get the final equation mapping length of sleep to hours til next sleep, I can plan out my days accordingly. For example, I didn’t get sleepy til about 6:00 AM last night. However, I have to get up at 7:00 AM tomorrow for work. I want to comfortably get 8 hours of sleep tonight, which means I can assume I’ll fall asleep at 11:00 PM. Sleeping at 6:00 AM one night and at 11:00 PM the next means the hours til next sleep should be 17. Substituting this into the equation above gives me x = 3.25 hours of sleep. Unfortunately, I didn’t hear my alarm at 9:30 AM, 10:30 AM, or even 11:00 AM…

I guess this serves as a decent New Years Resolution. I’ll continue tracking my sleep, and I should completely fix my sleep cycle by the end of the year. This means sleeping about 8 hours a day, and sleeping and waking up at the same times each day. If anyone else is interested in tracking their sleep, I can share you my Google Doc so you can start doing the same.