What method do you use to cap games?

[BuddyBear]

A number of diifferent methods are used by cappers to pick winners...just curious as to what others use. I"ll list a few but I am not sure if I've included them all so if I missed one just post it.

1. Breakdown Method: Very time consuming and cumbersome and usually what you'll find is that the linesmaker has taken everything into account to create a pretty tight line. Can be rewarding if you find something that isn't taken into account in the line

2. Bet against the Public: My personal favorite and the one I use the most. Here you identify who the public is on and do the opposite of them. Usually you'll be bettting underdogs. Great b/c you don't have to do too much other than montior the public and bet against them.

3. Trend Analysis: Relying on trends to give you plays. Some trends can be very recent others very inclusive with data as far back as 30 years. Some people swear by trends that they work while others insist trends are factored inot the lines.

4. Smart Money: A variation of bet against the public theory. Here you wait till the line moves in a direction that it's not suppose to and then bet it. For example, let's say the spread in the Detroit/Green Bay game is -3 Det and 80% of bets are on Detroit and 20% on GB. You would expect the line to move to DET -3.5 but instead moves to DET -2.5. Basically what's happened is a large sum of "smart money" has been placed on GB moving the line the other way. Very time consuming but is a favorite of the sharps.

5. Value betting: Another offshoot of bet against the public accept this time you wait till you get the possible value. I use this one but it's hard to conceptually distinguish from bet against the public since most bets are on the underdog. Great for dog line players.

6. Line Service/tout/friend: Here you rely on a line service or friend to provide you picks. For whatever reason you can't do the research or analysis on the game. A poor method in my opinion since most touts and services are not reputable. But then again, some people insist that they win using this method.

7. Blind betting: Here you just pick whoever just to have some action on the game. Often times, bets with this method are placed based on emotion and very little, if any, research on the actual game. Not very good, but you have a "50/50" chance of winning so sometimes this might be appealing especially if you lack time.

8. Ask your girlfriend/wife method: In this method, you've completely given up and rely on your significant other to make the pick for you. Somedays, I feel like my girlfriend could do a better job than me

[clonecat]

I have combined #2 with my version of #6 -borrowing (I hope it's not stealing) picks from those I respect on the SBR. I do ask a friend what he thinks and bet the opposite way, which has been successful.

[Senator7]

Always #1 and occasionally I'll factor in some of #3.

Senator 7

[freebie]

BuddyBear, I would like to know where do you gather the info on public bets? excluding wagerline, sportsbook and sportsinsights, because I don't think they're accurate.

I like to ride a hot winning team, once I picked out the the team. Then I double check to see if 80%-90% of public are on them. If not, then hit the enter button for wager.

This method can be real hot and sometimes it can get real cold. If it's cold, I'll try to follow other cappers whom have good percentages winnign record.

[BuddyBear]

BuddyBear, I would like to know where do you gather the info on public bets? excluding wagerline, sportsbook and sportsinsights, because I don't think they're accurate.

Good question Freebie. I don't think Wagerline and sportsbook are accurate at all at least not scientifically that is...but if you look at the methodology of sportsinsights.com you'll find that is is really really good and about as good as you are going to get to discovering who the public is on. So the short answer I use sportsinisights.com...I've found it very reliable and recommend it but there are a few critiques I have of it.

But there are other ways to find out who the public is on but I'll caution you they are not scientific at all. you could listen to "experts" like Ron Jaworski and Sean Salisbury and find out who they are on..."experts" almost always pick favorites. You could listen to sports talk radio....you could go to an Internet sports forum and see who everyone is playing. You could call you friends who don't bet but keep up with sports as often times their opinions should reflect the majority opinion.

Basically what I am saying is that the public is an abstract concept...it doesn't exist in the same sense that the computer you are typing on exists. You need real world observations to create a concpetulization of who the public is and who they are playing. To me, Sportsinsights is as scientifically close as you are going to get it in my opinion however if they sampled in more books I would feel more comfortable as they only sample 6 books which is a bit troublesome...I would think 10-14 books should suffice and a lot of their books they use for sampling purposes are more less top tier books where professionals play.

[freebie]

Thanks, BuddyBear

I do use sportsinsights from time to time to double check before I pull the trigger.

[freebie]

For example today Cleveland with C.Lee pitching. I really want to bet 3X my normal wager on Cliff Lee. But seeing 78% or more was on Cleveland. I backed out and missed my chance to make good money.

[BuddyBear]

That's the other thing I should mention as each of these methods above have advantages and disadvantages...what's important is to recognize what they are and how they come into play.

It's important to recognize that the public is right on certain games too. However, as long as Vegas is a multi billion dollar industry and the online sportsbook industry is booming I'll keep betting against the public

[Brick Tamland]

Some of #4 but before the line moves

[bigboydan]

good info sources

[onlooker]

#8

:105084828

[Illusion]

1-5 are good. People who use options 6-8 should seek help.

[pags11]

I'd consider myself a value player...

[jjgold]

I just see how a team has played the last few games and keep it simple

[Ganchrow]

I'd say that there's another method of handicapping, but one which can be rather dangerous to put into practice. Specifically, if one can identify systematic misestimations of win/cover probabilities by the bookmakers, then one can find value in early lines even without having any estimator of "true" fair value.

This is what you often see utilized by the neural net bettors out there. Without even using the current price as an input, these programs can often be quite effective at picking wnners based on earlyish lines simply by identifying situations where the books tend to systematically misvalue (oer even ignore) input factors. What's more, these input factors are not even specifically identified by the neural net. (Think principal component analysis.) The problem of course becomes when the books eventually catch on (and they always do,) the neural net must will then need to train intself in the opposite directions which can result in many false positives (and generally noisy selections).

[Mudcat]

At this point, I bet strictly based on long-term mathematics-based trends. I'm not sure "capping" is even the right word for it. I don't even think about who is pitching, streaks, injuries - any of that. I just look for a situation with the numbers and place the bet.

I used to have a bit of success with #1, straight breakdown capping, but I don't have time for it now.

[THESCOUT]

5.

[EBone]

I'd say that there's another method of handicapping, but one which can be rather dangerous to put into practice. Specifically, if one can identify systematic misestimations of win/cover probabilities by the bookmakers, then one can find value in early lines even without having any estimator of "true" fair value.

This is what you often see utilized by the neural net bettors out there. Without even using the current price as an input, these programs can often be quite effective at picking wnners based on earlyish lines simply by identifying situations where the books tend to systematically misvalue (oer even ignore) input factors. What's more, these input factors are not even specifically identified by the neural net. (Think principal component analysis.) The problem of course becomes when the books eventually catch on (and they always do,) the neural net must will then need to train intself in the opposite directions which can result in many false positives (and generally noisy selections).

Ganchrow, you seem to be light-years ahead of me with statistical breakdown if you can understand the neural net bettors and what their concepts are. Just so I'm clear about this post: are you saying there are programs out there that predict games just on specific input values, say how a team performs in a certain role (Road Underdog, Home Underdog, Road Favorite, Home Favorite), # of returning starters for a college football team, avg. power rating from a number of sources, avg offensive yardage accumulated vs. defensive yardage given up in football, etc........? These programs would spit out a prediction and you, as the bettor, decide if that is good data or not.

Or is it much more complex than that? My guess is that the more data gathered for a specific season (college basketball?), the better a model would be. I have toyed with this idea for years. I've never quite gotten a good grasp on it because of outside factors (no time to manually input data for a proposed model, etc...). I know how some models behave strictly from experience but have not continued to monitor their progress because of the time involved in monitoring.

I'd just like to know more about the specific input factors these "neural net" bettors are using. As has been discussed in this thread, most people, I think, are betting on "perceived" value and people they trust as "cappers". "Perceived" is the optimum word.

E

[Ganchrow]

Are you saying there are programs out there that predict games just on specific input values, say how a team performs in a certain role (Road Underdog, Home Underdog, Road Favorite, Home Favorite), # of returning starters for a college football team, avg. power rating from a number of sources, avg offensive yardage accumulated vs. defensive yardage given up in football, etc........?

I;m not talking here about programs which estimate cover probabilities, but rather those which identify situations where there are likely to be mispricings in quoted lines.

You need to distinguish between a model which purports to first estimate the "true" line on a game and then compare that line with the market to come up with a betting decision (and I think this is to what you are referring) and one which makes its determination largely irrespective of the current market price. Note, that I'm not saying that I particularly endorse this latter method, just that it's a different flavor of handicapping from those mentioned by the OP, and one which seems to be at least marginally effective.

As far as neural nets in general go, they are particularly well-suited to pattern recognizition problems and are specifcally constructed so as to adapt to changing conditions. Because bookies (and to an extent also the betting public) tend at times to systematicaly over or undervalue certain characteristics (which can be simple or complex), these patterns may then be sussed out by the neural network software and taken advantage by the bettor.

[EBone]

Ganchrow, you are correct. I am not differentiating between the 2. The concept of "identifying situations where there are likely to be mispricings in quoted lines" is new to me. It certainly interests me........... I think I'm gonna look into it and see if I can get up to speed. Thanks.

E

[Antonio]

1. Value

2. Probabilites (index Rating)

3. Overlays

4. Service Fades

5. Square Fades


Imo Trends Sucks And Are A Nice Fade Tool. Smart Money Has Been Hitting Around 50/50 Which Is A Losing Prop. Steam Is Nice If Your Getting On Their Number , But After 30 Seconds Value Is Lost.

[pags11]

yeah, I use jjgold's method as well...