Former actuary now gambles full time on sports

[DW Simpson]

http://www.casinocitytimes.com/news/article.cfm?contentID=140664

...Fezzik, meanwhile, was making a good living in the corporate world before he took up gambling full time. After graduating from Northwestern University, he worked as an actuary and as a vice president for a big insurance company in California. A couple of years ago, he gave up the insurance game for sports betting -- where total liability means a losing bet on the over/under, and a rider is a basketball team favored against Siena. "I had been flying into Nevada for every football weekend, and I was making the same amount of money gambling as I was in corporate America -- although with violent fluctuations in my bankroll," said Fezzik, 40. Like many high-level gamblers, Fezzik -- a one-word name, like Nico -- treasures anonymity. Hence the pseudonym and his use of disguises during public appearances. (At last year's seminar, he sported sunglasses and a frizzy Afro wig.) The low profile allows him to make large wagers without drawing unwanted scrutiny, Fezzik said. And his scheme works -- usually. Fezzik said one Las Vegas sports book manager once barred him from the book for a year, telling him, "It's not my job to buy you a new car every year."

[Bama Gambler]

I didn't think it was possible to beat the 10% juice, much less make a living.

[DW Simpson]

http://www.lvasports.com/fezzik.cfm

[Bama Gambler]

# College Football ATS Record (last two seasons, ignores ties): 169-142. 54.3%

# NFL ATS Record (last two seasons, ignores ties): 236-178. 57% Wow did better in the anyone can win on any given week league (aka NFL).

[2pac Shakur]

I didn't think it was possible to beat the 10% juice, much less make a living.

Just bet the opposite of me every week.
Works for investing, too.

[Wigmeister General]

Fezziwig didn't earn his letters. Man, that guy is a genius.

[duodenum]

Is fezzik the giant in princess bride too?

[Sotally Tober]

Is fezzik the giant in princess bride too?

Yup.

[jared]

Doesn't surprise me. The juice is not 10%. If you pick 52.4% winners, you make money. When you add in the value of line shopping, the takeout is reduced further. The problem is having discipline to weather out losing streaks.

[Bama Gambler]

The juice is not 10%.Really? When I used to bet sports it was. In other words, if you bet 50 and won you got 50, but if you lost you paid 55. That's 50 x 10% = 5.

Your 52.38% is accurate to break even. I'm still not convinced that 52.38% can be beat.

[Sotally Tober]

The juice is not 10%.

I've never seen "the juice" not be 10%.

[jared]

Example: If I bet both sides on the same game at 11:10, I will lose $1. The casino has taken in $22, and has profited $1. The takeout is
1/22 = 4.5%. For baseball bettors the takeout is approximately half that.

If I win 53 out of 100 games, I win 53 and lose 47 + 10% (47)= 51.7.
So I would come out ahead picking 53%

[DW Simpson]

Your 52.38% is accurate to break even. I'm still not convinced that 52.38% can be beat.

It's been awhile since I did this, so I might screw this up, but...

The 57% is artificially high because ties are ignored. Pushes happen. Assume they happen approximately ten percent of the time. Add 40 games. His 57% would now be 236 / 454 = 52.0%.

But ignore that. Assume his 236 wins out of 414 games are real.

Grab a normal distribution table (http://people.hofstra.edu/faculty/Stefan_Waner/RealWorld/normaltable.html).

What's the probability of winning 236 or more games out of 414, assuming that it's a coin flip? So assume that there isn't line shopping, and that he's just guessing.

Z = [ 236 - (414/2) ] / [ (414/2) ^ 0.5] = 2.02

P (Getting 236 or more games correct out of 414) = 0.500 - 0.478 = 2.2% (you only want the right side of the distribution)

OR

Assume 40 pushes (10% of the time):

P(Getting 236 or more games correct out of 454) = 0.500 - 0.225 = 27.5%

My guess is that a record like his happens around 30% of the time, at least. Depends on how many ties there are.

[3rookie]

The point spread is not set where there is a 50% chance of one side winning and a 50% chance on the other side. It is set where the money will be bet on both sides equally. 2 very different things.

Instead of point spreads, money lines are also heavily wagered upon in Vegas on most sports, not just baseball. No juice, per se, but a bid-ask spread.

For the people who are successful in sports gambling long term, it is a JOB to them. When it is your living, you research everything and do whatever it takes to be successful. Any claims above 60% success are very likely false.

[DW Simpson]

The point spread is not set where there is a 50% chance of one side winning and a 50% chance on the other side. It is set where the money will be bet on both sides equally. 2 very different things.

I am well aware of that. I stand by my calculations for the purposes of getting a rough idea of what's going on.

[Gandalf]

Your 52.38% is accurate to break even. I'm still not convinced that 52.38% can be beat.

It's been awhile since I did this, so I might screw this up, but...

The 57% is artificially high because ties are ignored. Pushes happen. Assume they happen approximately ten percent of the time. Add 40 games. His 57% would now be 236 / 454 = 52.0%.

But ignore that. Assume his 236 wins out of 414 games are real.

Grab a normal distribution table (http://people.hofstra.edu/faculty/Stefan_Waner/RealWorld/normaltable.html).

What's the probability of winning 236 or more games out of 414, assuming that it's a coin flip? So assume that there isn't line shopping, and that he's just guessing.

Z = [ 236 - (414/2) ] / [ (414/2) ^ 0.5] = 2.02

P (Getting 236 or more games correct out of 414) = 0.500 - 0.478 = 2.2% (you only want the right side of the distribution)

OR

Assume 40 pushes (10% of the time):

P(Getting 236 or more games correct out of 454) = 0.500 - 0.225 = 27.5%

My guess is that a record like his happens around 30% of the time, at least. Depends on how many ties there are.
1. Where did you get that denominator of (414/2)^.5? Shouldn't it be (414/4)^.5? Z = 2.85 P = .22% not 2.2%

2. Your second calculation seems to be the probability of randomly picking at least 236 right out of 454 when random p(right) = .5. If you recognize ties as an event with nontrivial probability, then p(right) for random picking < .5.

[Bama Gambler]

jared, we agree on what the casino takes in. But juice is defined at the extra amount you pay when you lose a bet. Therefore I stand by my orginal statement that the juice is 10%. I'm not saying that the casino makes 10% and I'm not saying that you have to pick 60% right in order to win. I'm simply saying that the juice is 10% and that is really hard to overcome. If it were easy none of us would work.

[DW Simpson]

That's where It's been awhile since I did this, so I might screw this up, but...

comes in handy. If someone would fix it instead of tearing it apart, then we can get an idea of how 57% success falls in the realm of likely scenarios.

[3rookie]

What's the probability of winning 236 or more games out of 414, assuming that it's a coin flip? So assume that there isn't line shopping, and that he's just guessing.

Z = [ 236 - (414/2) ] / [ (414/2) ^ 0.5] = 2.02

P (Getting 236 or more games correct out of 414) = 0.500 - 0.478 = 2.2% (you only want the right side of the distribution)

OR

Assume 40 pushes (10% of the time):

P(Getting 236 or more games correct out of 454) = 0.500 - 0.225 = 27.5%

My guess is that a record like his happens around 30% of the time, at least. Depends on how many ties there are.
This is faulty. This would be the answer if those 40 pushes were treated as losses. Your 2% figure is more valid. Pushes are non-events and should not be included in the analysis (btw, I see an NFL push rate closer to 5%).

[3rookie]

The point spread is not set where there is a 50% chance of one side winning and a 50% chance on the other side. It is set where the money will be bet on both sides equally. 2 very different things.

I am well aware of that. I stand by my calculations for the purposes of getting a rough idea of what's going on.
This was not directed at you, DWSW.

[DW Simpson]

What's the probability of winning 236 or more games out of 414, assuming that it's a coin flip? So assume that there isn't line shopping, and that he's just guessing.

Z = [ 236 - (414/2) ] / [ (414/2) ^ 0.5] = 2.02

P (Getting 236 or more games correct out of 414) = 0.500 - 0.478 = 2.2% (you only want the right side of the distribution)

OR

Assume 40 pushes (10% of the time):

P(Getting 236 or more games correct out of 454) = 0.500 - 0.225 = 27.5%

My guess is that a record like his happens around 30% of the time, at least. Depends on how many ties there are.
This is faulty. This would be the answer if those 40 pushes were treated as losses. Your 2% figure is more valid. Pushes are non-events and should not be included in the analysis (btw, I see an NFL push rate closer to 5%).

One might argue that those pushes should be treated as losses, because you lose your vigorish.

But, yeah, the probabilities are off even if pushes are considered losses. Just trying to get a rough idea without spending more than two minutes on this.

[3rookie]

You don't lose vig on a push in Vegas.

[Gandalf]

That's where It's been awhile since I did this, so I might screw this up, but...

comes in handy. If someone would fix it instead of tearing it apart, then we can get an idea of how 57% success falls in the realm of likely scenarios.
I did fix part 1: .22% instead of 2.2%.

The comment on part 2 was an observation that the part 1 calculation is the relevant calculation. Pushes should be irrelevant, or nearly so. Records of 236 W - 178 L - 40 T, 236 W - 178 L - 20T, 236 W - 178 L - and 236 W - 178 L - 60 T all have the same gambling payoff, and all involve about the same amount of skill. A calculation that counts a record of 240 W, 214 L, 0 ties in 454 games as better than 236 W, 178 L, 40 T is silly, yet that is what a probability of >= 236 W does.

I'll be amazed if anyone can come up with a valid calculation that says guessing at random will have a net win greater than 236*100 - 178*110 out of 454 games much different than .22% of the time, as long as they assume p(W) = p(L) for random guessing, regardless of what they use for p(T).

[DW Simpson]

You don't lose vig on a push in Vegas.

Oh. Cool.

[Gandalf]

Some exact numbers for achieving a result better than or equal to 236 * 100 - 178 * 110 = +4020:

414 games, p(w) = p(l) = .5: 0.25%
454 games, p(w) = p(l) = .5: 0.28%
454 games, p(w) = p(l) = .45, p(tie) = .1: 0.22%
454 games, p(w) = p(l) = .40, p(tie) = .2: 0.18%

That success percentage is extremely impressive.

The results for a simple scoring of W - L >= 236 - 178 would be similar.

[Klaymen]

Yeah, if you believe all that stats rubbish.

[jared]

If anyone read the link, Fezzik's advice was awful. He said that New England would not make a 4th down conversion (they did) and that the game had a less than 10% chance of going into OT (it didn't because of a late rally and field goal). Also, Fezzik is not his real name, so we don't know if he got his letters.

[working girl]

Also, Fezzik is not his real name, so we don't know if he got his letters.

He got his FSA. Here's the latest story.


NFL: Fezzik Wins Las Vegas Hilton Supercontest Again
written January 4, 2010 by John Kelly
NFL: Fezzik Wins Las Vegas Hilton Supercontest Again


BACK-TO-BACK JACK.....A bronze statue of Man o' War sits outside the Las Vegas Hilton SuperBook. Man o' War won 20 of 21 lifetime starts with his only defeat coming to a longshot appropriately named Upset. The Blood-Horse magazine recently ranked Man o' War the champion horse of the 20th Century, beating out the likes of Secretariat, Citation and Kelso.

If the Hilton wants to honor another champion, a bronze statue of professional sports gambler Fezzik should be commissioned. Without question, "Fezz" is the sharpest sports handicapping contestant in Las Vegas.

Yesterday, the 45-year-old professional gambler pulled off the greatest accomplishment in the 20-year history of the prestigious Las Vegas Hilton SuperContest when he successfully defended his title. No contestant has ever won the Hilton SuperContest twice, let alone winning in back-to-back years.

Fezzik defeated a field of 328 contestants this season and beat a group of 349 players last season. His pointspread record of 53-29-3 (64.6%) fell just short of a $10,000 bonus awarded to a contestant hitting over 65%. Fezzik earned the bonus last season by hitting 67.5% of his plays with a salty pointspread record of 54-26-5.

Fezzik's victory this season came at the expense of Week 17 leader "Big E." Fezzik trailed "Big E" by a half-point (wins are worth one point, ties are worth a half-point) heading into the final Sunday of the NFL season. Among their five final-week selections, Fezzik and "Big E" had a pair of common plays, but more importantly, one contrasting selection: Fezzik liked Tennessee -4, while "Big E" landed on Seattle +4.

Jay Kornegay, executive director of the Hilton SuperBook, predicted before the day started that the Tennessee-Seattle game would determine which contestant would take down the first-place prize of $196,800.

Kornegay was right.

Fezzik needed either a win or a tie with Tennessee -4 to provide the historic tournament victory. He got his desired result when Tennessee rallied to defeat Seattle, 17-13.

The Tennessee-Seattle game highlighted Fezzik's general edge in the NFL and his specific advantage in handicapping contests. "Fezz" beat Tennessee's closing line by two points (contest line -4, closing line -6).

The advantage gambler preaches line value. Better yet, he applies his successful approach in the real world of high-stakes sports gambling.

Ten years ago, Fezzik began his public handicapping career with only a few followers and countless detractors. Today, "Fezz" enjoys friendships and partnerships with a legion of loyal supporters, while his enemies have nearly vanished.

Emphasis on the word "nearly."

Not every member of the sports gambling fraternity is impressed with Fezzik's incredible tournament run. When I told college basketball expert Alan Boston about Fezzik's multiple tournament titles, Boston replied, "F*** Fezzik. And you can tell him I said that."

Computer Bob, a trusted observer of the sports gambling scene, congratulated Fezzik for his tournament success, but cited his poor 2009 record in pro and college football. "Fezz" has lost over 30 units this season for his clients and followers at his website, www.lvasports.com.

To be fair, Fezzik's subpar results on his website are the exception, not the rule. If there were any questions about Fezzik's sports betting credentials, he erased all doubts yesterday with his back-to-back Hilton championships. "Fezz" deserves his position as this city's most informed, most intelligent and most accomplished sports handicapping contestant. His legend continues to grow.

[obominable snowman]

Just bet the opposite of me every week.
Works for investing, too.

Works for political wisdom too.

[jared]

In the first post, it says a casino banned him from sports betting. The casino would be better off letting him bet and then over-adjust the line.