Hypothetical question in regards to win/loss streaks…

[NickTheGreek+]

This is just a hypothetical question, and I’m not exactly sure if this question can be resolved with some sort of a mathematical analysis.

Assuming one has a guaranteed 55% edge in picking winners, is it possible to predict by that percentage alone, what would be his longest losing streak? Would it be eight in a row, twelve in a row, or fifteen in row straight losses?

Not sure if this can be resolved, but it’s just a shot in the dark. Everyone’s feedback is greatly appreciated!

 

[Gridster]

It's an interesting question. The short answer is that anything can happen at any time, but your not likely to have an extended losing streak, say 20 games, until you've played a very long time. For instance, if you had a hat with 55 white balls and 45 black ones, you can define an "attempt" as picking a white ball after a black ball is pulled. So as those attempts build, you're more likely to hit a rare streak of consecutive balls. Likewise, if its 10-1 to put together a 3 game losing streak, then you'd think that it's even money you'd bump up against one in 11 tries, 11 wins following 11 losses. Ultimately, it's a function of how many plays you make.

 

[ATX]

you can get a distribution that shows the percent chance of losing x number of times in a row.

In other words something that gives:

"you will lose 9 times in a row at least x number of times for every 1000 wagers at a win rate of 55%."

"you will lose 10 times...."

etc.

That is what I had in my files until they were destroyed. I found this format more relevant than "there is a x% chance of losing 9 in a row" since I look at it from the standpoint that I WILL make 1000 wagers from today onward.

There is a big flaw in one of the contingencies of this distribution, IMO. The contingency that every wager has exactly the same (55%) chance of winning simply b/c you choose it, is not true. This gets complicated, but the way I see it there are 'right sides' and 'wrong sides' from a value standpoint on numerous events. At some points you are really seeing the off-numbers and choosing 60% winners, and then maybe the next week you make mistakes and during some of those stretches you may be a 40% handicapper. So the chances of losing a certain number in a row is constantly changing, just as some events hold more value than others due to public misconception etc, etc.

This is one of those topics that is difficult to put into words. The above example is a good example to illustrate random streaks. But be very careful using static, raw statistics to apply to this market. Backfitting is a huge error when trying to predict future events. If you won 55% of your last 10,000 wagers are you a 55% handicapper? Maybe. All those 10,000 wagers show is that you WERE a 55% handicapper. And the next 10,000 events may offer less value or maybe more as the game changes, players become more athletic and/or injury prone,etc. For a lot of 'human nature' reasons the math gurus usually don't do well with their computer models betting on sporting events.

I could go on and on about my opinion, but I think what you are getting at is trying to figure out how much you can risk without going broke on each wager given probable streaks. Maybe some sort of Kelly system which I would stay away from. The way I look at it is this: the more wagers that I make the more probable I will lose 100 in a row at some point. Although it is almost statistically impossible, if I live long enough it will happen. It might take a trillion wagers to lose 100 in a row, but never count on losing 10 in a row as impossible or even improbable given a 55% historical win rate. It will happen it's just a matter of time, and you never know when--it's usually after a really good streak, for me.

 

[NickTheGreek+]

ATX you wrote: "I could go on and on about my opinion, but I think what you are getting at is trying to figure out how much you can risk without going broke on each wager given probable streaks."

That is exactly what I'm trying to figure out. Great analysis, I should have been more clear. However I don't use the Kelly System. I just flat bet 5% of my bankroll, then double up my unit when my BR doubles as well. I know 5% is a little too risky, wagering 2 to 3% is fairly reasonable, but I'm a bit of a risk taker

And Gridster, I love your analody as well. Made a lot of sence to me. To be honest with you, I never really looked at it at that angle. Thanks guys!

 

[SixthSense]

Nick - Your chances, in a series of 10 bets each, of doing the following, assuming you are a 55% handicapper are:

0-10 .034%
1-9 .416%
2-8 2.23%
3-7 7.46%
4-6 15.96%
5-5 23.40%
6-4 23.84%
7-3 16.65%
8-2 7.63%
9-1 2.07%
10-0 .253%

Obviously, anything could happen but these percentages are the most likely to happen, if you really have a 55% edge or are a 55% handicapper. Having said that, I believe the chances are very good you will go 0-10 even though the percentages here would not make that claim.

 

[Gridster]

When I responded, Nick, I thought it was a hypothetical. I want to make it clear that I would never advise anyone to try to determine the maximum length of a streak. In real life its not balls in a hat. I try to be as analytical as possible when I cap a game, but when it comes to losing streaks I firmly believe that the old cliche applies: It's like a black cloud hanging over your head.

My advice regarding losing streaks is to manage them. Fewer bets, lower risk, until that cloud passes. Its been my experience that streaks are over when they are over. If you lose 12 in a row then hit 2 of your next 3 I think its safe to say the streak is over, and you can resume your normal betting activity.

 

[ATX]

Sixth, solid info as usual.

The reason that I dont have a similar chart any more is b/c it's inaccurate FOR ME personally. If I backfit my win% then it's highly unlikely that I lose 9 or 10 in a row, going forward. But I have those extreme losing streaks a little more often than what the statistics indicate. A little more often is quite a bit of money. For this and other reasons I feel that a person's win rate is constantly changing, and the higher volume of wagers that you place, the more it varies. This is also correlated to value offered on each day. Some sides have a 60% or better chance of landing, some are the opposite, I dont buy into the fact that all games offer equal value. If you are being objective and seeing clearly and have solid info then you will be on the right side of a lot of those higher +EV events, and this is directly related to my feeling of a constantly changing win rate, beer helps.

 

[jwunderdog]

I have won 17 in a row and lost 13 in a row, and been on several other double digit streaks in both directions. Over thousands upon thousands of bets you will go on a cold and hot streak in double digits. JUSTMYOPINION.

 

[Potter]

Streaks just happen & the more bets you have then the more likely you are to encounter longer ones(the law of large numbers??)

All the law of averages requires is that the proportions that an event happens,tends,over time towards that events true probability.

Say you get a large run of heads tossing a fair coin,tails don't have to catch up in the near future,if at all.The proportions will tend towards 50/50 even if heads & tails appear equally for the rest of time.


(If heads continue to forge ahead it might pay to run a few statistical checks to make sure is a fair coin.)

Once you've capped a game & given a team a 60% chance,then that team's still a 60% chance if you're on a run of 10 losers or winners(assumming your capping abilities remain intact).

Not entirely sure if you can treat ats streaks in the same way as the quotes are generated by the odds compiler,so he might anticipated a large demand for one side regardless of the numbers.

In the absence of any capping it would seem logical to bet against(or maybe with )a long ats streak.

It all boils down to individual games being largely independent events,if you're confident that you've capped them correctly & if you've hit a streak,it's just lady luck smiling or not.

P

 

[ATX]

I agree with Potter's statement regarding independent events for the most part.

There are times when I feel that the LVSC really struggles, BTW they are humans and do fine tune the lines a bit, so the events are not entirely independent. The LVSC does a good enough job OVER TIME.

The more you take cards, dice, and coins out of the equation as a source of comparison for sporting events the better off you will be. If I used those statistics I would get out of this market, I am so <due to lose> it's scary.

Keep in mind that my standpoint now is an attempt entirely aimed at +2% a week. I was much more aggressive in the past, but now there is no need for it. I must admit that I'm very biased towards this standpoint and will disagree with wagering on more than 1.5-2% per event. If I fall short and accomplish only half of that +2% weekly... The system I now use makes it almost impossible for me to lose more than 30% of BR, and even if I did I would still be up a staggering amount over the last few years. This is quite simply the highest risk end of my 'folio and one of the most profitable. If anyone out there can do simple math, and jot down some projections and forecasts, in 7 years you can really do some things with a small advantage. Why push the envelope? Do not worry about bonuses and cashback rewards on losses. If you are truly in this for the long term profit, at the top of your list will be reduced vig and favorable lines. I used to track this in the past as far as vig and good lines, and I made over 30% in one year just with those two very simple things. When something looks REAL GOOD, learn to look for value on the opposite side, there is a definite psychological angle developed into a lot of spreads. I'm drinking Guiness today, Winbet got me thinking that I dont drink enough.