Why people fail..

[phillygodfather]

Why do people fail at gambling???



Since the invention of the internet there has been an explosion of information available to the average Joe.. With everyone knowing everything and every possible stat available to everyone from injuries to situational factors and weather conditions why do so few actually profit from Gambling? Does everyone know everything but can't put the pieces together to come to a valid conclusion? Here comes the Philly out of me that makes things very simple.. Have you ever gone into a ice cream store and they have 999 flavors and you can't decide what you want.. In the Old days Ice cream stores had 5 flavors and you were in and out the door faster than a speeding bullet with a winning flavor..



I will tell you right now 95% of statistics handicappers talk about have no fucking value and will not help you determine the outcome of a game or the final score..Odds and statistics are based on infinity and none of us is going to live that long.. As the Market constantly changes so are the spreads in diff situational factors changing also.. Anything you are thinking about the Sharpest in the world have beat you to it.. I have never claimed to be a genius but I am smart enough to know their are some really fucking sharp people out there and they are who I have pursued over the years.



Human nature, tendencys and conditioning levels of diff atheletes and their ages play big role...


For the sake of the average Joe who has been brainwashed with the word UNIT I will Use the word UNIT.. LMFAO.. You will always hear Handicappers talking about money management and then tell you to only bet between 1-3 Units a game. The only way you can properly use Money managment is to know the true edge you have in each and every game you wager. If My guys have determined we have a superior edge on a certain wager why would we cut our own throat and not bury the game.. Betting too much on a side or betting too little on a side is bad money management. The only time betting one to three units a game will be profitable is if the game only warrants a 1-3 unit bet, but there will be times when the wager warrants a 5 unit bet and even a 10 unit bet and once in a blue moon like the Cotto/ Margarito fight I bet 30 times what I usually bet. Of course always keep your bankroll in mind and never risk more then 10% of it on a single wager.



Margarito is blind in one eye and the New York state Boxing commision was not going to approve ...


Follow me on Twitter @Phillygodfather

[Beyond the Bets]

(06-04-2012 09:19 PM)phillygodfather Wrote:  I will tell you right now 95% of statistics handicappers talk about have no fucking value and will not help you determine the outcome of a game or the final score

Agree. And stats are certainly more relevant in some leagues than others. For instance, the point spreads in Cincinnati Bengals games last year were ALL between one and seven points. The majority of games had spreads of four or less.

In other words: The Bengals played a lot of toss-up games against similar teams. It can be argued that the majority of the stats extracted from their games (assuming they are correctly applied) can be useful.

Compare that with Cincinnati Bearcats games, where their point spreads ranged from 1 point to 7 points to 16 points to 41 points. How many folks try to take Bearcats stats (i.e., YPP or TO differential or whatever) and turn them into something that doesn't have meaning or predictive power?

You're much better off using gut instinct and observation than incorrectly applying stats that will send you down a path toward something that isn't. All stats aren't earned equally. A 500-yard passing game against Austin Peay doesn't provide you with much insight when the Bearcats have to take on Tennessee a week later.

[lindetrain]

(06-04-2012 09:19 PM)phillygodfather Wrote:  For the sake of the average Joe who has been brainwashed with the word UNIT I will Use the word UNIT.. LMFAO.. You will always hear Handicappers talking about money management and then tell you to only bet between 1-3 Units a game. The only way you can properly use Money managment is to know the true edge you have in each and every game you wager. If My guys have determined we have a superior edge on a certain wager why would we cut our own throat and not bury the game.. Betting too much on a side or betting too little on a side is bad money management. The only time betting one to three units a game will be profitable is if the game only warrants a 1-3 unit bet, but there will be times when the wager warrants a 5 unit bet and even a 10 unit bet and once in a blue moon like the Cotto/ Margarito fight I bet 30 times what I usually bet. Of course always keep your bankroll in mind and never risk more then 10% of it on a single wager.

Just don't see how this is good advice to a novice bettor who likely can't determine their true edge on a game...

Yes, for YOU it makes sense to vary bet sizes as you've been doing this for quite awhile and are more than capable of determining your edge...but can you really expect a majority of amateur bettors to be able to do the same? You tell them to bury a side they love, then they drop 10x an average bet on the game and get their ass handed to them because they've miscalculated their edge and read on the game. What works for you likely won't work for the Average Joe - and I think making mistakes like this one are the reason 99% of bettors lose.

And while I'm aware stats may not always hold the same capping value as subjective factors such as situations, human tendencies, etc., are we really gonna say they can't be utilized to some extent? You can draw some strong conclusions as to how a game may play out and potentially find value if looking at appropriate stats - it's when people bark up the wrong trees with misleading/useless stats that they find themselves in trouble.

[phillygodfather]

(06-05-2012 03:36 AM)lindetrain Wrote:  

(06-04-2012 09:19 PM)phillygodfather Wrote:  For the sake of the average Joe who has been brainwashed with the word UNIT I will Use the word UNIT.. LMFAO.. You will always hear Handicappers talking about money management and then tell you to only bet between 1-3 Units a game. The only way you can properly use Money managment is to know the true edge you have in each and every game you wager. If My guys have determined we have a superior edge on a certain wager why would we cut our own throat and not bury the game.. Betting too much on a side or betting too little on a side is bad money management. The only time betting one to three units a game will be profitable is if the game only warrants a 1-3 unit bet, but there will be times when the wager warrants a 5 unit bet and even a 10 unit bet and once in a blue moon like the Cotto/ Margarito fight I bet 30 times what I usually bet. Of course always keep your bankroll in mind and never risk more then 10% of it on a single wager.

Just don't see how this is good advice to a novice bettor who likely can't determine their true edge on a game...

Yes, for YOU it makes sense to vary bet sizes as you've been doing this for quite awhile and are more than capable of determining your edge...but can you really expect a majority of amateur bettors to be able to do the same? You tell them to bury a side they love, then they drop 10x an average bet on the game and get their ass handed to them because they've miscalculated their edge and read on the game. What works for you likely won't work for the Average Joe - and I think making mistakes like this one are the reason 99% of bettors lose.

And while I'm aware stats may not always hold the same capping value as subjective factors such as situations, human tendencies, etc., are we really gonna say they can't be utilized to some extent? You can draw some strong conclusions as to how a game may play out and potentially find value if looking at appropriate stats - it's when people bark up the wrong trees with misleading/useless stats that they find themselves in trouble.

My advice to the Novice better is to do what me and some friends did in the late 80s when we decided to learn how to count cards playing Blackjack.. We practiced 5-6 hours a day at home for 6 months with imaginary money before we decided we were ready to gamble our hard earned cash, practice makes perfect..

[BetATL]

(06-04-2012 11:48 PM)Beyond the Bets Wrote:  

(06-04-2012 09:19 PM)phillygodfather Wrote:  I will tell you right now 95% of statistics handicappers talk about have no fucking value and will not help you determine the outcome of a game or the final score

Agree. And stats are certainly more relevant in some leagues than others. For instance, the point spreads in Cincinnati Bengals games last year were ALL between one and seven points. The majority of games had spreads of four or less.

In other words: The Bengals played a lot of toss-up games against similar teams. It can be argued that the majority of the stats extracted from their games (assuming they are correctly applied) can be useful.

Compare that with Cincinnati Bearcats games, where their point spreads ranged from 1 point to 7 points to 16 points to 41 points. How many folks try to take Bearcats stats (i.e., YPP or TO differential or whatever) and turn them into something that doesn't have meaning or predictive power?

You're much better off using gut instinct and observation than incorrectly applying stats that will send you down a path toward something that isn't. All stats aren't earned equally. A 500-yard passing game against Austin Peay doesn't provide you with much insight when the Bearcats have to take on Tennessee a week later.

Good posts BTB/PGF.

Stats can be your friend but also make things more complicated then they should be. I found myself getting too caught up in stats early on in baseball season and I was struggling because of that.

I was noticing that a lot of my "gut feeling" plays were winning but I was leaving them off the card most nights, letting stats keep me away.

Loosened up a little bet and let my knowledge of the game be the main factor in my handicapping. Have turned it around quite nicely since then.

[ChiefZilla]

I think it's really a "to each his own" thing. I personally don't use a ton of stats in sports betting, but I know there are people out there who have had success doing so. (Look at the MLB price machine guys.)

But if you want to be more of a "feel" bettor, have at it.

Bottom line, though, I agree with the original thoughts. It's information overload out here on the internet, so you have to decide what stats/websites are useful to you and then go from there.

[mount187]

(06-05-2012 09:57 AM)ChiefZilla Wrote:  I think it's really a "to each his own" thing. I personally don't use a ton of stats in sports betting, but I know there are people out there who have had success doing so. (Look at the MLB price machine guys.)

We've always maintained that our project is based solely on statistics, and that other factors should be taken into account. I bet every game we identify, but only because I know other people are blindly following the plays and I want to stand by my work. If I was looking at it from an outside perspective, I'd factor in travel schedules, trends, umpires, weather, etc. There could be information available that we could never account for with statistics. A pitcher's arm angle in SF during day games may match up really well with the batter's eye, and increase K's against a team with a high K-rate. It would be nearly impossible to automate that. It could be that the grounds crew at Wrigley let's the grass grow higher in the middle of the summer to keep it from burning up, which disadvantages teams that hit a lot of ground balls, and increases that viability of a sinker-ball pitcher. To be able to factor something in like that would be mind boggling.

While I still contend that 2,400 games is a large enough sample size to have our system be profitable, the point PGF makes about not playing "infinity" games does hold weight. We may say that Felix Hernandez, at home, should be a -155 against Santana (LAA), but he's only a -135 (!), it's still a one-time match up. Technically, for number theory to work in its purest form, you'd want to be able to bet Hernandez, at home, vs Santana 100 times and you'd expect to make a profit paying 57.45% and winning at a 60.78% clip. If you only get that match up twice in a season, you may take two pieces of the 39.22% losing side, and never have it balance back out long term. However, you're doing that with any bet you make. Even if you think GB covers -8 at home against the Jags 57% of the time on -110, there's still a shot you end up on the 43% side.

It's an interesting discussion to have, and people have made some great points here. Good sports betters are the ones that are able to take a seemingly infinite amount of available information, pare it down to what's relevant, and then use instincts to determine if what the numbers say make sense outside of a vacuum of a pure "on paper" scenario (things like grudges, letdowns, etc).

[Edit] One last thing to keep in mind about systems like ours, is it's still software, and software is designed by humans. It's impossible for a human to create a perfect model that doesn't have some sort of built in bias. That's how you get conflicting models like one saying global warming is going to kill us all in a decade, and the other saying we're headed for an ice age. In the end, your model is only as good as the human that's feeding the machine the data. The person modeling their football model may have a bias that "defense wins championships" and may make slight adjustments to their theories based on that. Someone making a baseball model may give more weight to projected stats than actual stats. Even the most complex "learning" software that tries to reverse engineer results from stats has someone programming the learning functionality, which is going to make the software "learn" in the same way that human would. If the programmer sucks at transferring how they learn into software, you're going to have software that learns in a stupid way.

[jimmysingh9]

(06-05-2012 06:17 AM)phillygodfather Wrote:  My advice to the Novice better is to do what me and some friends did in the late 80s when we decided to learn how to count cards playing Blackjack.. We practiced 5-6 hours a day at home for 6 months with imaginary money before we decided we were ready to gamble our hard earned cash, practice makes perfect..

THIS!

I think people want to use stats however because its easier to look up data on the internet and say this team gets more yards passing than this team and ranks #2 in the country while the other teams ranks #312 against the stat.

It is alot easier to do that than actually learning a sport and watching a game and believing in the skill that you and you're handicapping abilities.

I've always said you can find a stat to back any team. The Miami Heat are undefeated every June 2nd during a Leap Year when the Series is tied 2-2 and they won the first 2 games of the Series but the Boston Celtics are undefeated during every even year in the calendar when the groundhog has seen its shadow.

There is always an argument.


As for the Novice Bettor, they are a NOVICE BETTOR. Does someone go into the stock market and throw 20 grand on commodity futures right away? Would it be smart for someone to go to Vegas with their mortgage and play Blackjack if they didn't know how to double down?



The one thing I never understood was why the frig stats in college sports mattered 10 years ago if the same players aren't playing, perhaps the same coaches aren't coaching????

Why in the hell would a stat involving Joe Montana at QB against Michigan St matter when Brady Quinn is now QB for Notre Dame? How can passing yards from a totally different variable allow you to predict a score for today?

[gynecologist]

Its quite simple, its because they have to make a decision and dont bet the right amounts..

[GreyMatterStats]

Hi phillygodfather,

I respect your opinion but disagree with some of what you are saying here.

I feel that statistics can be a valuable tool in the handicapping process.

For team sports, the opening lines are derived largely from statistics.

Line movement may come from from the "feel" or inside information of sharps, but the genesis of the lines are mathematical, for the most part.

In my opinion, it is important to note that stats by themselves hold little value. What is important is how you use them and incorporate them into models. As BTB stated, (para-phrasing) if you are going to use stats you must apply them correctly.

We use current, historic, and situational data to project outcomes without regard to released lines, essentially creating our own number for individual games.

If we see a significant variance from our "projected line" versus the actual line we begin to take a much closer look at that game, and look to make a case either for or against a side.

I suspect you are aware of the fascinating book called "The Smart Money", written by Michael Konik, that goes into great detail how this process was done several years ago, back in the days of data entry and punch cards.

We don't pretend to have anywhere near the expertise or the resources detailed in the book, but it does go to show how stats have been successfully utilized by some very sharp gamblers.

I want to be clear that in our process, the use of data and projections is just one part of the equation. Once we identify games of interest via data, our handicapping process becomes a much more traditional one.

I will also add that at the end of the day, "real information" is always more valuable than historic data.

If a QB is listed as questionable and you have the connections to know whether that means he is 95% or he is really hurt and going to try to play anyway, it goes without saying that is an enormous advantage. But most handicappers, ourselves included, are not privy to these types of things.

That said, there is an overwhelming amount of information, statistical and otherwise, that is available and it is our hope that sites such as ours help the handicapping public to make better use of it.

PGF, I enjoy your contributions on here and Twitter. Much of what you have written I have found fascinating, and you always seem to stir up interest.

A few questions for you if that is OK?

1. You mentioned 95% of the stats have no value, what 5% do?

2. I am making these numbers up, but say under Reid the Eagles are 25% ATS on the road in non-conference games - Do you care about that at all when handicapping say Eagles at Houston?

3. I share you Philadelphia roots, so this one really perplexed me,

in your analogy,

"Here comes the Philly out of me that makes things very simple.. Have you ever gone into a ice cream store and they have 999 flavors and you can't decide what you want.. In the Old days Ice cream stores had 5 flavors and you were in and out the door faster than a speeding bullet with a winning flavor..

Why did you go ice cream instead of water ice?????

(06-04-2012 11:48 PM)Beyond the Bets Wrote:  

(06-04-2012 09:19 PM)phillygodfather Wrote:  I will tell you right now 95% of statistics handicappers talk about have no fucking value and will not help you determine the outcome of a game or the final score

Agree. And stats are certainly more relevant in some leagues than others. For instance, the point spreads in Cincinnati Bengals games last year were ALL between one and seven points. The majority of games had spreads of four or less.

In other words: The Bengals played a lot of toss-up games against similar teams. It can be argued that the majority of the stats extracted from their games (assuming they are correctly applied) can be useful.

Compare that with Cincinnati Bearcats games, where their point spreads ranged from 1 point to 7 points to 16 points to 41 points. How many folks try to take Bearcats stats (i.e., YPP or TO differential or whatever) and turn them into something that doesn't have meaning or predictive power?

You're much better off using gut instinct and observation than incorrectly applying stats that will send you down a path toward something that isn't. All stats aren't earned equally. A 500-yard passing game against Austin Peay doesn't provide you with much insight when the Bearcats have to take on Tennessee a week later.

[illadelphia]

(06-05-2012 04:19 PM)GreyMatterStats Wrote:  3. I share you Philadelphia roots, so this one really perplexed me, in your analogy,

"Here comes the Philly out of me that makes things very simple.. Have you ever gone into a ice cream store and they have 999 flavors and you can't decide what you want.. In the Old days Ice cream stores had 5 flavors and you were in and out the door faster than a speeding bullet with a winning flavor..

Why did you go ice cream instead of water ice?????

Gotta go ice cream because people who haven't been to Philly have no idea what wudder ice is. I remember living on Cape Cod one summer trying to find water ice, only to have people/massholes look at me like an alien everywhere I asked for it. Apparently everyone else calls it italian ice or flavor ice.

Damn I wish I was home for a quart from Rita's or Philly Flavors and some soft pretzels.

[Weebay]

Here is my question, what stats are the most important when handicapping a game? I personally use very little stats when it comes to betting. I could have all the statistics available to me but for example on nfl Sundays my book has team A +7 and current line is 4.5 why would I not hammer the +7 even if every meaningful stat indicates I shouldn't?

[Beyond the Bets]

In college football? Almost nothing.

At least for me.

[mattrud]

This is the best advice ever, and what I (attempt) to model my gambling style off of. Extreme selectivity, a large disregard for most stats (and virtually all trends) and the confidence to bury some bets (usually on my few favorite teams that I know exceedingly well).

[mount187]

I think it's also important to take into account what sport you're using stats for. In football, basketball, and hockey, the number of possessions / time of possession can vary greatly game to game, where in baseball, the road team gets at least 27 outs and the home team gets at least 24 outs. There are also a lot more moving parts in other sports that can affect how one team defends the other, where baseball is pretty much defended the same from game to game (outside of the occasional shift). If the A's are a bad defensive team, they're not all of the sudden going to play lock-down defense because they have a favorable match up against the Mariners. The defenders are going to play how they play in baseball, so you can more easily boil the match ups down to, "How do these batters stack up to the available pitchers." Obviously stats don't perfectly translate to how the game is going to turn out, but I feel like they hold a lot more value in baseball versus other sports.

[louiemdj]

Great thread.

The obvious problem with stats is you can pretty much find some random stat to back up any opinion and some sports are much more oriented around numbers rather than making a play with your "gut".

[SkyBlues42]

(06-06-2012 10:04 AM)louiemdj Wrote:  Great thread.

The obvious problem with stats is you can pretty much find some random stat to back up any opinion and some sports are much more oriented around numbers rather than making a play with your "gut".

I know it was mentioned before.. but this ^. There's a stat to justify a play in almost any situation. The key is filtering out the irrelevant stuff to develop a solid opinion. I don't want to say some stats aren't useful, because in certain situations, they can be.

The thing that bothers me the most, is when people refer to ATS data from 10+ years ago. Why is that relevant? Especially when there are new coaches, players, opponents, refs, etc. It grinds my gears to see people like Lawrence use that data in magazines.

If its the same coach, then fine. It can be somewhat relevant. Kirk Ferentz is a good example. He's been at Iowa for a long time and his spread data is actually relevant. Not because of his overall ATS record in 2000. But because it shows that he prepares his team well off a loss. Covering the spread in a game following a loss shows that the coach knows how to prepare his players.

Some of the older stats can be relevant, others are just there to fill space. Again, the key is filtering out the garbage.

[@TracktheBet]

In College Football, I find stats to be much less meaningful than finding solid situations

[louiemdj]

(06-06-2012 01:09 PM)SkyBlues42 Wrote:  

(06-06-2012 10:04 AM)louiemdj Wrote:  Great thread.

The obvious problem with stats is you can pretty much find some random stat to back up any opinion and some sports are much more oriented around numbers rather than making a play with your "gut".

It grinds my gears to see people like Lawrence use that data in magazines.

[BattaBing]

The odds are against you, so it should be more about finding the games with the best odds and not the best stats. Most handi-cappers are lifetime losers. It reminds me of an old joke a bookie told me about a handicapper who loses every week even though he spends all day studying stats.

The bookie tells him to give hockey a try.

"Hockey? I don't know anything about hockey!" the handicapper says.

"Exactly." says the bookie.


Most gamblers have disproportionate ego's to their actual skills and abilities such as critical thinking and rational decision making. Everyone thinks if they just find the right angle they could be like the guys they see in movies or read about.

The other suckers are looking for an easy way, like buying picks from a self proclaimed expert.

I've thrown handicapping out the window and just play lines now. Have developed a few models that are very profitable based solely off line patterns.

Handicapping is for suckers. Unfortunately, with more stats and information, it will only further solidify the illusion that the gambler can figure out a new angle the linesmaker hasn't thought of.