Diaperpalooza

(I'll say up front that maybe this is a stupid post at the moment. Very high jinx potential. Maybe I should wait for the wave to crest. But kind the whole point is I don't believe in voodoo or superstition, so here we are.)

A few months ago, in order to try to erase that sad scrunchy face that Sara would get whenever I talked about my sports bets, I attempted to reframe the conversation. I spoke of the winnings in terms of diapers. A diaper costs roughly 25 cents. So if I won a $20 bet, then I'd say "We won 80 diapers today." Notice also the use of the word "we." I wanted to make her feel less icky about me gambling online while expecting a baby.(1) Also--genuinely--I wanted to convey to her that any winnings I might produce would go to the baby and not back into my personal pot.
This strategy has been about as successfully as I could have hoped. No, of course she doesn't ever support my degenerate efforts, but at least to a small degree she shares in my excitement with winning. Or at least she is happy when I have good gambling news, as opposed to rejoining with something like this: "That's great, maybe doing something legal would set a better example for our unborn."

That out of the way, I'm on a hot streak right now. "Hot streak" there being slightly misleading, because it connotes that it's been mostly about luck or guessing. I've got a system now. It's come from a couple months of relative immersion in the world of smart (analytical) sports gambling. Maybe I'm in the minority, but I find it all incredibly fascinating. I'm much more addicted to the theory behind exploiting an edge than I am to actually exploiting it. This might not be 100% healthy, but it's many times moreso than simply rolling the dice and hoping to win.

Anyhow, to the system. It pertains to college basketball and contains two components. First is a strong rating system. Taking a poll amongst your friends doesn't count. Neither does reading up on the subject and trying to rank teams yourself. From my extensive research on this matter, the best college basketball rating system is put out by Ken Pomeroy. The second, and equally important component, is a betting calculator set by the Kelly Criterion. The rating system will tell what team is most likely to win and the calculator will tell you how much to wager. Case closed.
The trick to the system is to trust it and to not think about it. If kenpom says that Ohio State is 78% likely to beat Illinois, and your betting website is offering OSU at only a 70% rate, then you take OSU, no questions asked, no subjective judgement allowed.(2) Elimination of the subjective element is of paramount importance. It prevents you from getting emotional and reacting foolishly to a hot streak or to a single unlucky loss. It prevents you from being just another gambler, basically. The computers are smarter than you are, so trust them.
--Important aside here. You ask, if the computers are smarter than me, then why shouldn't I also assume that the bookmaker is smarter than me? Shouldn't those cancel out? Yes and no. Bookmakers are usually even smarter than computers, this is true. A lay person will never beat a bookmaker on his own over a large sample size, and in theory a single computer should not either. Thankfully for me, it's not that simple for the books. They have to take into consideration public opinion because if they set a "true" line that disagrees at all strongly with public opinion, then they'd really be opening themselves up to risk, which they of course don't want to do. So, the books will shade their lines toward public opinion, meaning there is still room for value to be exploited. To a smaller degree, books will also sometimes put up initial lines just to take the temperature of the public; they will offer a line they know is off just to see what happens. Or they might not devote enough attention to a line by the time they put it up, so in the time it takes them to correct it, there is a window of opportunity.(3)
Let's talk for a minute about edge. Edge is really the key to the whole thing. Most point spread bets are sold at -110. This means you pay $110 to win $100. It also means that you need to win more than just half your bets, you need to win 52.4%. This is not insignificant. If you bet all 16 NFL games some weekend, and go 8-8, you did not break even. You lost. If you bet $10 on every game, you'd have won $80 but also lost $88 for a net of -$8. If you only go about 50% over a long time, the loss total adds up. That most people don't consider 50% a loss is probably what amounts to much of the money Vegas makes.
So a -110 bet requires a 52.4% likelihood of success to be a winner over time. If I have a system that tells me a team being sold at that price has a 60% likelihood of winning, then--voila--I've got a 7.6% edge. I should bet. This works across all types of wagers. My system only spits out a win% expectation for straight-up bets. It doesn't work for point spreads, so I don't bet on point spreads.(4) In every game where I have an edge, I run the numbers into the Kelly calculator, which requires inputs for the bet price, the win expectation, and your total bankroll. This last factor is basically what determines how much it tells you to bet. The Kelly criterion is meant to be "maximally aggressive," but it's not meant to bankrupt you overnight. If your edge is small, then it will tell you to bet small; if it's big, then it'll tell you to go big. It does the heavy mathematical lifting so I never have to worry about being bold or being timid.
Which brings us back to the point: trust the system. It's a very liberating thing, actually. I used to make bets based on personal preferences, maybe including stuff I've read or stuff I attempted to discern using some statistics-oriented sites. I would make almost all my wagers the same amount. I knew nothing about exactly what my edge was or when to hit the gas and when to hit the brakes. Of course I was going to lose over the long run.
Yes, I keep fairly detailed records of my bets. During the whole NFL season, I went exactly 42-42 and finished up about $30 (this thanks to often taking underdog moneylines, which pay out more than even of course. It also is thanks to getting lucky and winning $160 just on the Super Bowl). For college football, I went 39-46 and lost $113. Factor in my initial $300 deposit plus the free $60 bonus the website gave me, and after the Super Bowl, my balance was $291, but it fluctuated a lot in-season. Twice my total balance dropped below $60, and at one point it was over $500. I scuffled with college hoops too, running about -$80 before fully implementing the system.
As of right now, taking into account pending wagers, my balance is $415. Since starting the fully objective and analytical system one week ago, I am up $189.77. This is slightly higher than would be expected, because I've also benefited from some luck, but a majority portion of it is legit. Over the 58 bets placed during this time (yeah I know, it's a lot), the average price given me corresponded to a 68.8% break-even line, meaning my bets would have to win at least that often to break even.(5) The expected win% of these games based on the kenpom preditions was 75.2%. So you can clearly see that my average edge over this sample is 6.4%. That's good. My actual win% was 79.3%. I've been more than 10% better than the break-even line, which pretty easily explains why over 58 bets I'm up $189. Some small regression might be in order, but I'm extremely confident that over the next 58 bets I place, I will do better than 10% worse than break-even, meaning that in the end I will finish up.
The mutually happy part of this whole story is that next Monday I will be making my first-ever online gambling withdrawal. Depending on how the system fares between now and then, I expect to withdraw close to $200. That's 800 diapers. Over the first year, a baby will go through on average about 8 diapers per day. Those 800 diapers would last 100 days, or at least as long as we will be living in NYC after the baby comes. So, thanks to daddy's adoption of the system, baby girl folger will be shitting for free in NYC.


1. Granted, this is something I did think about. But then the money I deposited was from my personal stash and not our family's, and also I did so well before she became pregnant. Gambling online requires an upfront deposit, so in a sense the money was lost right from the get-go. Finally, by the time we were safely into the pregnancy, my initial $300 investment was down to like $100, so at that point, what's the use cutting and running? Might as well stick around and try to bump it back up to even.
2. Just yesterday I broke this rule of doubting the system. UConn was playing St John's in the first round of the Big East tournament. The system said that St John's was a 50/50 shot to win the game, and my suggested bet amount was $35 at a price of +170, meaning I would stand to win about $60. But it didn't look right to me. I couldn't believe St John's was as good as UConn. I let subjective analysis (UConn was playing without their coach for a few games, St John's was really playing at home in MSG, UConn is playing for an NCAA tourney bid, etc, etc) talk me out of trusting the system. So I didn't bet. This is the only time since starting the system that I overruled an output so deliberately (I have nudged my bet amounts slightly up or down based on my perceived risk avoidance, though). I'm sure I don't have to tell you that St John's won the game by 22. Of course they did. This "loss" made me angrier than all the others I've suffered combined, and I didn't even technically lose any money. I lost the opportunity to collect $60. Because I didn't trust the system.
3. This paragraph is actually why it's often better to bet on obscure games. Books don't pay as much attention to them and so you're more likely to get a line that is off. Obscure lines are also more susceptible to shifts in movement based on small samples of bettors, so not as many idiots can move a line farther in your favor than if the whole public is watching it. Just last Friday, I made plays on all of the following: Yale, Hofstra, Citadel, Wisconsin-Milwaukee, Harvard, Drexel, Bradley and Long Beach St. My first wager Saturday was on Stony Brook. Trust the system.
4. I've placed at least 75 straight moneyline bets since my last point-spread bet. No, I'm not missing them, thank you.
5. This is roughly true, because a loss in a -700 line is not the same as one on a -130 line. Over a large sample though, this is a good shorthand.