sports betting system

"Fading the Tout" can be yours for 99 dollars a season. Just kidding. I'm sure this idea has been tossed around here before...track the sports touts that you can actually monitor and if they reach 60-70-80% winners over a decent sampling period play against them for X number of units or remainder of season. We know that the best/luckiest touts (one in the same) hit 55-58% over a season if they "release" enough games. Yes/No? Ideas?

Generally speaking, fading doesn't work because you must overcome 11/10.

Be a bookie and put 11/10 on your side.

Quote: stoneynv

"Fading the Tout" can be yours for 99 dollars a season. Just kidding. I'm sure this idea has been tossed around here before...track the sports touts that you can actually monitor and if they reach 60-70-80% winners over a decent sampling period play against them for X number of units or remainder of season. We know that the best/luckiest touts (one in the same) hit 55-58% over a season if they "release" enough games. Yes/No? Ideas?

This is not how reversion to the mean works.

Quote: MakingBook

Generally speaking, fading doesn't work because you must overcome 11/10.

Be a bookie and put 11/10 on your side.

No doubt making book gets you on the right side of the bet but as they say "if you can do the time don't do the crime".....I knew a bookie out in Phoenix that made a nice living with 40 or so 100 dollar players. He ended up getting busted. Did a little time, less than a month. Tax evasion issue popped up and property seizure under Rico statutes.

Quote: sodawater

This is not how reversion to the mean works.

Somehow I knew I would stand corrected ....sorry sodawater.....were did I go wrong?

Quote: stoneynv

Somehow I knew I would stand corrected ....sorry sodawater.....were did I go wrong?

you can't expect to something to go under the mean, just because it has a history of going over the mean. You can expect it at the mean, and over time, the ratio approaches the mean.

To sum it up, you wouldn't expect the tout to do WORSE than 50%, and then it would just be unprofitable for you because of the vig.

Quote: sodawater

you can't expect to something to go under the mean, just because it has a history of going over the mean. You can expect it at the mean, and over time, the ratio approaches the mean.

To sum it up, you wouldn't expect the tout to do WORSE than 50%, and then it would just be unprofitable for you because of the vig.

Example....tout releases 300 games/season NBA....half way through season said tout has 100 wins 50 losses. I'm sure you would agree that's good/lucky....actually lucky....too many point spread influences during game to predict/handicap, particularly in football......(ball not round). So it just seems to me that the second half of season tout will not realize same success since 55% is a lot to ask for even for the luckiest of handicappers (using term loosely) over the course of the season. The wheel always wants to balance out over time.

Quote: stoneynv

the wheel always wants to balance out over time.

The way it balances is not how you think. Over time, it will approach the mean, because the expected value of each spin or game or flip is the mean. You cannot expect it to go over or under the mean, just because it previously went under or over the mean. The way it evens out is that the mean is the expected outcome.

Simple example. If you see a coin hit heads 7 out of the last 10 flips, the expected number of heads the next 10 flips is 5, not 3. The "balancing out" is that the ratio of results will approach the mean.

Quote: sodawater

The way it balances is not how you think. Over time, it will approach the mean, because the expected value of each spin or game or flip is the mean. You cannot expect it to go over or under the mean, just because it previously went under or over the mean. The way it evens out is that the mean is the expected outcome.

Simple example. If you see a coin hit heads 7 out of the last 10 flips, the expected number of heads the next 10 flips is 5, not 3. The "balancing out" is that the ratio of results will approach the mean.

I'm sure I'm not falling for the "gambler fallacy" here......sports touts under the best of circumstances might win 58% over a large sampling (a season) provided that they are "releasing" more than just a handful of games. The touts that hit 58% fall in a group that probably represents 2% of all handicappers for any given season.

Here is what you are missing: You cannot expect a tout to do WORSE than 50% picking games. You would need him to do WORSE than that to make money laying 11 to 10 to fade his picks.

Quote: sodawater

Here is what you are missing: You cannot expect a tout to do WORSE than 50% picking games. You would need him to do WORSE than that to make money laying 11 to 10 to fade his picks.

I'm just saying that regression towards the mean is very real in sports handicapping. Point spread outcomes should be treated differently than the roulette ball randomly bouncing around the wheel. Fading handicappers that are hot is different than believing red has a better chance of coming up because 10 previous trials showed black. Yes I understand the 11-10 lay. This is just a real world observation I've made through the years. Maybe I'm wrong.

I feel like I am repeating myself over and over here but I will say it one more time.

Yes, mean reversion is real. But the method of reversion is NOT to "overcompensate" towards the other side of the mean. So there is no angle for you to play. The very best you could expect is the mean, 50 percent.

O.K......I was hoping that just once we could apply real world observations and put the math off to the side.....NOPE.......

what is math if not a way to describe the real world?

Quote: sodawater

what is math if not a way to describe the real world?

O.K. sodawater you wore me out....you win

I concur with SodaWater.

Let me try to explain this a different way, StoneyNV:

When we talk about something approaching the mean, or expectation, we're not necessarily saying that happens by virtue of an event (in this case, the picks of a tout) also falling worse than the mean over a period of time.

Your most specific example had a tout picking 100-50 for a record of .667. The reason SodaWater stated that you can't expect someone to do worse than the mean of .500 is because you can't expect ANYONE to do worse than .500. A random picker should pick at .500. My 83 year-old Great-Grandmother should be able to pick .500 NFL point spreads, and she's probably lived more years than she's watched games.

Now, you're wanting a tout to not only pick worse than he's been picking,(expected) for the rest of these 150 games, but you're also wanting this tout to pick worse than someone picking completely at random, (not expected) and significantly worse, at that.

MakingBook points out that have you have to have him get 11 games wrong for every ten he gets right...and then you'll break even. Over 150 games, you'd essentially need him to get 71 right and whiff on 79 games, discluding possibility of a push.

The thing that you have to keep in mind is that he will still finish the season significantly better than the mean, if you essentially break even, the tout would be 171-129 for .570 for the season.

This regression to the mean is not meant to be a short-term affair, and you're counting on it to be a short-term affair. The way it works is that you eventually approach the mean, not by doing worse, but by running at or near the mean for an extended period of time. For example, let's take your .667 100-50 tout and bring him down to the expectation of a random picker for the rest of his career. He finishes that season, and he picks 300 games per season for 10 more years, and ends having picked an additional 3150 games at 1575-1575.

If you include that great start to the season, he is now 1675-1625 over the selected period of time, or .508.

That's basically how it works, there are ups and downs, and simulations of a coin-toss would help give you an idea of how swings can go with Variance and Standard Deviations. Simulations are long run affairs, though. Just because he went 100-50 once doesn't mean he'll do it on the second half of the season...and he shouldn't...but nor should he do worse than 75-75.

Basically, what you're talking about IS the gambler's fallacy. Your negative expectation is the commissions on the bets. If you wish to do things as you describe, then you might as well just watch a Craps Table and wait for 100 Points Made/Missed vs. 50 of the opposite, or whatever. With a lower House Edge on those bets that the Commission on a sports bet, you're better off.

I also agree with sodawater. Most of us do.

Quote: Mission146

I concur with SodaWater.

Let me try to explain this a different way, StoneyNV:

....

Great explanation, Mission. I didn't have the stamina to type all that up but you did a great job.

Thanks for the compliment, AcesAndEights!

Quote: Mission146

I concur with SodaWater.

Let me try to explain this a different way, StoneyNV:

When we talk about something approaching the mean, or expectation, we're not necessarily saying that happens by virtue of an event (in this case, the picks of a tout) also falling worse than the mean over a period of time.

Your most specific example had a tout picking 100-50 for a record of .667. The reason SodaWater stated that you can't expect someone to do worse than the mean of .500 is because you can't expect ANYONE to do worse than .500. A random picker should pick at .500. My 83 year-old Great-Grandmother should be able to pick .500 NFL point spreads, and she's probably lived more years than she's watched games.

Now, you're wanting a tout to not only pick worse than he's been picking,(expected) for the rest of these 150 games, but you're also wanting this tout to pick worse than someone picking completely at random, (not expected) and significantly worse, at that.

MakingBook points out that have you have to have him get 11 games wrong for every ten he gets right...and then you'll break even. Over 150 games, you'd essentially need him to get 71 right and whiff on 79 games, discluding possibility of a push.

The thing that you have to keep in mind is that he will still finish the season significantly better than the mean, if you essentially break even, the tout would be 171-129 for .570 for the season.

This regression to the mean is not meant to be a short-term affair, and you're counting on it to be a short-term affair. The way it works is that you eventually approach the mean, not by doing worse, but by running at or near the mean for an extended period of time. For example, let's take your .667 100-50 tout and bring him down to the expectation of a random picker for the rest of his career. He finishes that season, and he picks 300 games per season for 10 more years, and ends having picked an additional 3150 games at 1575-1575.
If you include that great start to the season, he is now 1675-1625 over the selected period of time, or .508.

That's basically how it works, there are ups and downs, and simulations of a coin-toss would help give you an idea of how swings can go with Variance and Standard Deviations. Simulations are long run affairs, though. Just because he went 100-50 once doesn't mean he'll do it on the second half of the season...and he shouldn't...but nor should he do worse than 75-75.

Basically, what you're talking about IS the gambler's fallacy. Your negative expectation is the commissions on the bets. If you wish to do things as you describe, then you might as well just watch a Craps Table and wait for 100 Points Made/Missed vs. 50 of the opposite, or whatever. With a lower House Edge on those bets that the Commission on a sports bet, you're better off.

Thanks for taking the time for a hopeless dreamer. I totally understand what you are saying. I understand the math. I understand the gamblers fallacy. I'm simply saying that for whatever reason, sports touts have an uncanny way of working their way back or up to 50/50. It's just been my observation though the years. They run hot/cold but in the end they approach 50/50. If tout is winning 70% fade him....if tout is losing 70% play him. 300 plus "releases" I believe is a good sampling. I'm not sure that the gambler fallacy completely applies here. Again, just a real world observation. Thank for the response.

Quote: AcesAndEights

Great explanation, Mission. I didn't have the stamina to type all that up but you did a great job.

Just read it myself, and I agree. That's a damn good explanation Mission! Well done.

Quote: MakingBook

Just read it myself, and I agree. That's a damn good explanation Mission! Well done.

Yes we all agree that Mission did a great job explaining why my observations are useless...

I think you understand the concept. And I think everyone might agree with you if there wasn't the thing called "the vig". Your person running "hot" right now will most likely finish the remainder of the season around 50%. Maybe a little more, maybe a little less the rest of the way, but even if he finishes a little less than 50% and you go against him on every wager, the vig will negate the worse than average run and eat up your profits.

ZCore13

Quote: Zcore13

I think you understand the concept. And I think everyone might agree with you if there wasn't the thing called "the vig". Your person running "hot" right now will most likely finish the remainder of the season around 50%. Maybe a little more, maybe a little less the rest of the way, but even if he finishes a little less than 50% and you go against him on every wager, the vig will negate the worse than average run and eat up your profits

ZCore13

Yes, the vig......always the take out.....just a stupid idea that I tossed around.....no way to win at sports betting either....thanks everybody for setting me straight. The collective has saved me thousands upon thousands...... I'm just a dreamin' chump. Out

Quote: MakingBook

Just read it myself, and I agree. That's a damn good explanation Mission! Well done.

Thanks for the compliment, MakingBook!

Quote: Zcore13

I think you understand the concept. And I think everyone might agree with you if there wasn't the thing called "the vig". Your person running "hot" right now will most likely finish the remainder of the season around 50%. Maybe a little more, maybe a little less the rest of the way, but even if he finishes a little less than 50% and you go against him on every wager, the vig will negate the worse than average run and eat up your profits.

ZCore13

It seems as though your "system" would be very testable. Check past results where someone has gone 70% halfway through the season. How did they finish the season? My guess is those were the few times that the average was moderately better than a 50/50 split. Let's see some data.

Quote: stoneynv

Yes, the vig......always the take out.....just a stupid idea that I tossed around.....no way to win at sports betting either....thanks everybody for setting me straight. The collective has saved me thousands upon thousands...... I'm just a dreamin' chump. Out

It's cool, there's no reason to be upset.

It's clear that you're a smart guy, so there's no reason to feel stupid. If you were stupid, then you would have made a post that equalled or exceeded mine in length in an attempt to dispute an infallible point. Besides, it's just a paraphrase of stuff you learn from studying math, the knowledge is not innate and it goes against pattern-recognition tendencies that our brain creates for us.

There's nothing wrong with having ideas. It's just that some will be viable and some will not. I'm sure that people thought Thorp was crazy, people make a living card counting.

Quote: Mission146

It's cool, there's no reason to be upset.

It's clear that you're a smart guy, so there's no reason to feel stupid. If you were stupid, then you would have made a post that equalled or exceeded mine in length in an attempt to dispute an infallible point. Besides, it's just a paraphrase of stuff you learn from studying math, the knowledge is not innate and it goes against pattern-recognition tendencies that our brain creates for us.

There's nothing wrong with having ideas. It's just that some will be viable and some will not. I'm sure that people thought Thorp was crazy, people make a living card counting.

Thanks for kind words....I'll move on to the Wheel of Forune....I'm sure that game can be cracked.

Quote: stoneynv

Thanks for kind words....I'll move on to the Wheel of Forune....I'm sure that game can be cracked.

O.K......nobody liked the "fade the tout" idea.....so now I releasing (kidding) yet another system (yes I know nobody cares)...why would you? Let's use the NBA. As season progresses the gambler assigns a plus or minus value to teams as it relates to their performance against the spread. Example team A covers 50 out of 80 games....+ 20 against the spread minus vig. Or conversely -20 against spread with team B. Straight bet AGAINST team A (+20). WITH team B that's (-20). Again, I don't believe I'm falling for the gamblers fallacy here.....the wheel doesn't change but the sports bettors perception changes all well as a team performances Sports betting is not as static as the wheel. NBA can be streaky and teams that outperform tend to go through periods where they underperform. Conversely teams that underperform will find mojo at some point and over perform. Idea can be used for totals. Betting large and trying to win a unit or two. O.K. let me have it....

Quote: stoneynv

O.K......nobody liked the "fade the tout" idea.....so now I releasing (kidding) yet another system (yes I know nobody cares)...why would you? Let's use the NBA. As season progresses the gambler assigns a plus or minus value to teams as it relates to their performance against the spread. Example team A covers 50 out of 80 games....+ 20 against the spread minus vig. Or conversely -20 against spread with team B. Straight bet AGAINST team A (+20). WITH team B that's (-20). Again, I don't believe I'm falling for the gamblers fallacy here.....the wheel doesn't change but the sports bettors perception changes all well as a team performances Sports betting is not as static as the wheel. NBA can be streaky and teams that outperform tend to go through periods where they underperform. Conversely teams that underperform will find mojo at some point and over perform. Idea can be used for totals. Betting large and trying to win a unit or two. O.K. let me have it....

Just taking the air out of the variance

Spam removed, Member Banned.

-Mission146

Quote: stoneynv

"Fading the Tout" can be yours for 99 dollars a season. Just kidding. I'm sure this idea has been tossed around here before...track the sports touts that you can actually monitor and if they reach 60-70-80% winners over a decent sampling period play against them for X number of units or remainder of season. We know that the best/luckiest touts (one in the same) hit 55-58% over a season if they "release" enough games. Yes/No? Ideas?

Great thread, thanks for thinking out of the box StoneyNV. Let's take Stoney's idea deeper into the bell curve. Assume you find a legitimately monitored tout riding along at an insane +90% clip; say 30 of 32. (93.75%) He's on the extreme tip of that bell curve. Banking NOT ON Regression to the MEAN, but on Regression, I'd like to fade this tout till his overall percentage goes sub -90%, then stop the fade on him. Plausible?

Quote: DHass22

Great thread, thanks for thinking out of the box StoneyNV. Let's take Stoney's idea deeper into the bell curve. Assume you find a legitimately monitored tout riding along at an insane +90% clip; say 30 of 32. (93.75%) He's on the extreme tip of that bell curve. Banking NOT ON Regression to the MEAN, but on Regression, I'd like to fade this tout till his overall percentage goes sub -90%, then stop the fade on him. Plausible?

I'm guessing it would be difficult to find a monitored sports tout that could hit 30 out of 32, but possible. Your idea makes sense to me but I'm very vulnerable to drinking out the "gambler fallacy" punchbowl. The forum will tell you that the tout can be expected to pick 50% winners for remainder of season/segment. Touts run hot and cold so I would be inclined to "fade the tout" in your scenario. Let me know if you find a "tout" that's running 90% winners over a decent sampling.

or a tout that is picking 90% losers.....and play with him