[mehoolio]

If one was to bet totals, and always take the over, over 1000 bets they should be around 50% winning percentage. If you start off at around 40-45% over the first 500 bets, it seems like regression to the mean would means a 55% or so avergae for the next 500 bets. Therefore if you double the amount you wager per bet in that second half, you should be pretty successful. Why would this not work?

[LT Profits]

Gamblers Fallacy, odds would always be 50/50 in each individual game regardless of previous unrelated results. Regression to the mean takes place over infinity, so 40-45% could potentially last through a million trials.

[steve227]

very well put,helps the mindless for chasing the due monster and going broke

[indio]

Gamblers Fallacy, odds would always be 50/50 in each individual game regardless of previous unrelated results. Regression to the mean takes place over infinity, so 40-45% could potentially last through a million trials.

Your post prompted me to satisfy my curiosity in regards to the approximate odds of results through 1 million 50/50 coin flips.

498,000 or less (49.80%) apx 31,250-1
497,900 or less (49.79%) apx 77,000-1
497,800 or less (49.78%) apx 200,000-1
497,700 or less (49.77%) apx 500,000 -1
497,600 or less (49.76%) apx 1,000,000-1

My binomial calculator wont go higher than 1 million to one. The odds of 45% or lower through 1 million trials would have to be so astronomically low that there might not be enough space on a full page to handle all the zeros next to the decimal point needed to figure it's probability.

Obviously, the original posters thought is gamblers fallacy, as 500 results will have no impact on the next 500 results, which could realistically vary. But although possible in "theory", getting 45% or less on 50/50 shots over 1 million trials is about as close to impossible as it gets. I would guess the chances of the Earth getting hit by the moon and mars at the same time are greater than getting 450,000 or less coin flips correct over 1 million trials.

[LT Profits]

Your post prompted me to satisfy my curiosity in regards to the approximate odds of results through 1 million 50/50 coin flips.

498,000 or less (49.80%) apx 31,250-1
497,900 or less (49.79%) apx 77,000-1
497,800 or less (49.78%) apx 200,000-1
497,700 or less (49.77%) apx 500,000 -1
497,600 or less (49.76%) apx 1,000,000-1

My binomial calculator wont go higher than 1 million to one. The odds of 45% or lower through 1 million trials would have to be so astronomically low that there might not be enough space on a full page to handle all the zeros next to the decimal point needed to figure it's probability.

Obviously, the original posters thought is gamblers fallacy, as 500 results will have no impact on the next 500 results, which could realistically vary. But although possible in "theory", getting 45% or less on 50/50 shots over 1 million trials is about as close to impossible as it gets. I would guess the chances of the Earth getting hit by the moon and mars at the same time are greater than getting 450,000 or less coin flips correct over 1 million trials.

That's hardly the point though, I think you got the gist of what I meant. You can substitute any number you want for a million, I just pulled a large number to illustrate the point.

[indio]

That's hardly the point though, I think you got the gist of what I meant. You can substitute any number you want for a million, I just pulled a large number to illustrate the point.

You stated that 40-45% could potentially last through a million results. That statement is just as ridiculous as the original posters idea. Your point illustration loses its original viable message when you make an analogy that's not remotely plausible.

[micase]

You stated that 40-45% could potentially last through a million results. That statement is just as ridiculous as the original posters idea. Your point illustration loses its original viable message when you make an analogy that's not remotely plausible.

40-45% could last over a billion results. The odds are astronomically small but there is still the possibility. He was using the million result example to illustrate that using chasing logic you could be in for a long wait before regression towards the mean shows any sort of positive results.

[BigDofBA]

Start with $100 and if you lose keep doubling it. See how that works out....

You'll end up risking thousands of dollars just to win your 100 back.

People go on losing streaks all the time. When you hit a losing steak you'll losing everything chasing that original unit. Basically, you will be risking thousands to get that $100 back. It's one of the worst ideas ever.

[LT Profits]

You stated that 40-45% could potentially last through a million results. That statement is just as ridiculous as the original posters idea. Your point illustration loses its original viable message when you make an analogy that's not remotely plausible.

You are getting too caught up in the less important part of my statement. If it bothers you that much, stop taking the "million" so literally and consider it to be more bets that the OP will probably make in a lifetime to justify his strategy. My point to him still stands and I know that you know that also.

[indio]

40-45% could last over a billion results. The odds are astronomically small but there is still the possibility. He was using the million result example to illustrate that using chasing logic you could be in for a long wait before regression towards the mean shows any sort of positive results.

Astronomically small? The odds of being 45% over a billion results is more minute than hitting powerball 10,000 times in a row. So using such ridiculous examples deter from the intended message.

[TheMoneyShot]

Gamblers Fallacy, odds would always be 50/50 in each individual game regardless of previous unrelated results. Regression to the mean takes place over infinity, so 40-45% could potentially last through a million trials.

And I will be the first to admit... My success rate on O/U's is probably around 30%.... that's why I never wager totals. It just shows you I'm not a good Roulette player. So in theory... depending when you wager an O/U your odds could be worse.

[indio]

You are getting too caught up in the less important part of my statement. If it bothers you that much, stop taking the "million" so literally and consider it to be more bets that the OP will probably make in a lifetime to justify his strategy. My point to him still stands and I know that you know that also.

Yes, of course I understand it, but I think when you use the ridiculous notion that someone could go 45% over a million tries, your message gets lost, because that's like saying, you should always look inside your trash can, because there is the slight possibility that an escaped bank robber left a sack of $500,000 dollars in there to hide (which statistically is more likely).

I think if you had said something more to the effect that there's a 1.45% chance (better than rolling back to back 9's on 2 dice rolls) of hitting 466 or less out of 1000, so if you only hit 225 the first 500 (45%) what's to say the next 500 wont be less than 50% also? There's an even stronger 3.3% likelihood that one will hit 715 or less over 1500 tries, so what's to say you wont get only 490 out of the next 1,000? I think that could show how long it might take for averages to "catch up" after a cold run, in a way that is more comprehensible.

I think that examples using his own logic with realistic possibilities to show how flawed his thinking is drives the point home better than incurring far fetched examples that would never happen.

But hey, thats just my opinion. I'm certainly no authority on the effectiveness of various analogies to drive home a point.

[gamblingisfun]

Martingaling is the best gambling system ever if these two criteria are met. One has infinite bankroll and there are no betting limits. But if you had infinite bankroll you wouldn't need to gamble to earn money. You can still hit 50% of 1000 flips, but it's the losing streaks, no matter how rare they are that will kill you. What if you win 20 in a row, then lose 20 in a row? You still hit 50%, but the results were in such a way that you couldnt survive a 20 loss streak with doubling up and chasing. Why would you want 1000's on the line just to win a few bucks?

[LT Profits]

Yes, of course I understand it, but I think when you use the ridiculous notion that someone could go 45% over a million tries, your message gets lost, because that's like saying, you should always look inside your trash can, because there is the slight possibility that an escaped bank robber left a sack of $500,000 dollars in there to hide (which statistically is more likely).

I think if you had said something more to the effect that there's a 1.45% chance (better than rolling back to back 9's on 2 dice rolls) of hitting 466 or less out of 1000, so if you only hit 225 the first 500 (45%) what's to say the next 500 wont be less than 50% also? There's an even stronger 3.3% likelihood that one will hit 715 or less over 1500 tries, so what's to say you wont get only 490 out of the next 1,000? I think that could show how long it might take for averages to "catch up" after a cold run, in a way that is more comprehensible.

I think that examples using his own logic with realistic possibilities to show how flawed his thinking is drives the point home better than incurring far fetched examples that would never happen.

But hey, thats just my opinion. I'm certainly no authority on the effectiveness of various analogies to drive home a point.

All true BUT technically, even is the example given has a .000000000000000000000000000000000000000 000000000000000000000000000000000000001% chance of happening, that is still good enough to disprove the OP's premise which assumes a full 100% chance that there will be regression. So in that sense, I think an humongous exaggerated number actually drives the point home more as a deterrent.

[mebaran]

Break it down as simply as possible; trial by trial.

If we assume that each bet is absolutely identical, and has a negative EV of $0.02 per $1 bet, then the more we bet on an individual trial, the more we lose. This would be akin to reverse optimal Kelly: if you have a negative EV for a trial, the proper bet is zero, so by increasing your bet size, you are getting progressively farther away from optimal. A $100 flat bettor, for example, is expected to lose $2 per bet, whereas someone who is on their 4th leg of a chase system would be betting $800 to win their initial $100 back, which is expected to lose $16.

[sploofdogg]

This is going to work my friend.

[HUY]

If one was to bet totals, and always take the over, over 1000 bets they should be around 50% winning percentage. If you start off at around 40-45% over the first 500 bets, it seems like regression to the mean would means a 55% or so avergae for the next 500 bets. Therefore if you double the amount you wager per bet in that second half, you should be pretty successful. Why would this not work?

I think some bans should be in order in the Think Tank, because quality has taken a nosedive lately. Garbage like the above should never see the light of day.

[Sawyer]

Chasing might be a good idea..

If you're chasing the steam

[skrtelfan]

op is misunderstanding regression to the mean. it means future results will cause the overall results to approach the mean, not that the future results will be better if the earlier ones were poor or vice versa. 50% picks that went 45% should go 50% going forward not 55%.

[Topo]

The problem is that you can never increase the likelihood of hitting a 50-50 wager above 50%. So even when you increase the stakes your expected return for the next few wagers is still zero. That means increasing your stakes only has one effect: increasing your risk of going bankrupt.

[dalogester]

Start with $100 and if you lose keep doubling it. See how that works out....

You'll end up risking thousands of dollars just to win your 100 back.

People go on losing streaks all the time. When you hit a losing steak you'll losing everything chasing that original unit. Basically, you will be risking thousands to get that $100 back. It's one of the worst ideas ever.

BigDofBA, I only think this is a bad idea if you are betting more than your roll can handle. I wouldnt suggest doing this past 3-4 bets. And also a big part of it is betting on games that aren't as likely to happen as if you were to bet on other games. For instance, if you used a betting system waiting for teams to win 3 games in a row then fade this team the next three games doubling up each bet up to 3 games and quit whenever you win you would have better odds to win and get a better price then just throwing darts and doubling any games until you win/lose it all. Does this make any sense?

[dalogester]

Chasing might be a good idea..

If you're chasing the steam

What is chasing the steam exactly? Ive heard of chasing bets through series.. Is this similar?

[matthew919]

BigDofBA, I only think this is a bad idea if you are betting more than your roll can handle. I wouldnt suggest doing this past 3-4 bets. And also a big part of it is betting on games that aren't as likely to happen as if you were to bet on other games. For instance, if you used a betting system waiting for teams to win 3 games in a row then fade this team the next three games doubling up each bet up to 3 games and quit whenever you win you would have better odds to win and get a better price then just throwing darts and doubling any games until you win/lose it all. Does this make any sense?

This is the flaw in your reasoning.

It's perfectly reasonable to assume that chasing offers no better odds than if you were throwing darts, and almost perfectly unreasonable to assume otherwise.

I say "almost" because it's not completely impossible that chasing offers profitability. However, it's only profitable if there exists an implied market bias due to the team's recent performance, which is strong enough to skew lines one way or another. The fact that chasing systems don't even attempt to quantify this perceived bias should cause you significant distress.

If you're really interested in investigating whether it holds, get your hands on some data and test your hypothesis. That's a ThinkTank-worthy post. Otherwise, this topic has no business being posted in a quantitative subforum.

[jv040]

only way to make it in flat betting is underdogs and choosing your games wisely

[NunyaBidness]

Will someone please change the forum title from ThinkTank to DumbDumbTank.

[Blax0r]

Will someone please change the forum title from ThinkTank to DumbDumbTank.

We've definitely had a bad run of threads...

[mcduggly]

Start with $100 and if you lose keep doubling it. See how that works out....

You'll end up risking thousands of dollars just to win your 100 back.

People go on losing streaks all the time. When you hit a losing steak you'll losing everything chasing that original unit. Basically, you will be risking thousands to get that $100 back. It's one of the worst ideas ever.

This.

[Sawyer]

People who think chase systems don't work are gay or they don't know baseball, lol.

Chase works only in baseball. Don't ask any other questions. You won't get any reply. Do your own homework, go research u lazy baztards!

[Iceman88]

Chasing isn't bad if you set a limit. I start with 100 account and I try to double it by betting $30 then chase if needed. Worst case I lose $100. Best case I keep flipping. Made $660 since feb working for me.

[Sawyer]

The downside of chase/martingale systems is, the huge risk on Leg-4 and afterwards. 6-7 Leg Chase systems can be very dangerous, you lose everything when you lost, lol. But if you limit your chase to 2 or 3 games, then a loss won't kill you. It will be just a stratch. Few years ago, I had a thread in another forum I had a great success in totals with a 3-Game Chase in baseball, record was 71-3. I was chasing for 3-games so every loss means -8.3 or -9 units.

[matthew919]

Chase systems = never-ending supply of coin flip money pumped into the pot. I am so thankful that things like this exist.

[statnerds]

I think some bans should be in order in the Think Tank, because quality has taken a nosedive lately. Garbage like the above should never see the light of day.

the title made me think something along the same lines. but there are some solid replies in here with good info delivered in a fashion that will not repulse people.

OP has been here only a few months, has under 50 posts, so it might be safe to assume he is new to gambling.

So if he posts something that makes the majority of us cringe, at least he is asking a question and seeking help. And are we here to help everyone, or just have a circle jerk between about 8 posters?

Labeling his original post 'garbage' doesn't help him. Doesn't answer his question. Just makes you sound like an asshole.

[a4u2fear]

LT stop making it seem like you or anyone should give a crap about a 1 in a trillion chance. Pretty weak neither of you could get past that point

[pokernut9999]

People who think chase systems don't work are gay or they don't know baseball, lol.

Chase works only in baseball. Don't ask any other questions. You won't get any reply. Do your own homework, go research u lazy baztards!

Makes no sense at all.