#1
|
|||
|
|||
How big a sample do you need?
Sorry if this thread is already out there, but I couldn't find it ...
Say you've got a ROI of about + 8-9%, but the sample size is only 50-60 games. How big a sample size do you need before you could consider yourself to be consistent winner in sports betting? |
#2
|
|||
|
|||
Re: How big a sample do you need?
[ QUOTE ]
Sorry if this thread is already out there, but I couldn't find it ... Say you've got a ROI of about + 8-9%, but the sample size is only 50-60 games. How big a sample size do you need before you could consider yourself to be consistent winner in sports betting? [/ QUOTE ] Depends on how sure you want to be in your consideration. |
#3
|
|||
|
|||
Re: How big a sample do you need?
Good point ... how big an edge would you need to have over a period of say, 100 picks to be have have the confidence that you'll be an overall winner over a period of 500 picks 90 percent of the time.
I'm not a statistician, so bear with me if my questions/example are stupid ... |
#4
|
|||
|
|||
Re: How big a sample do you need?
In all honesty, you need like 10,000 trials to have a 95% confidence interval in anything that's going to be close to 50% and normally distributed.
Now, if you win 5000 bets and lose 2000 ATS, you can be pretty sure. But you'd be Bill Gates by then, I would think. |
#5
|
|||
|
|||
Re: How big a sample do you need?
So you're saying it's gambling and I should just have fun then ... holy christ, it would take FOREVER to have that many decisions, up or down ...
|
#6
|
|||
|
|||
Re: How big a sample do you need?
yep, takes alot of trials
there is a good chart in Wong's "Sharp Sports Betting" that most anyone can follow. |
#7
|
|||
|
|||
Re: How big a sample do you need?
Speaking of that book MyTurn, ever decipher his saying on page 212? He's speaking about home field advantage but says he doesnt want to give too much away since linesmakers are likely to read his book, but gives a 'hint' that would reveal a big factor or something of the like to consider.
"I can give you a hunt to narrow your search for factors that help explain home-field advantage: I live near the ocean, and high tides do not water my lawn". |
#8
|
|||
|
|||
Re: How big a sample do you need?
[ QUOTE ]
yep, takes alot of trials there is a good chart in Wong's "Sharp Sports Betting" that most anyone can follow. [/ QUOTE ] It shouldnt take a lot of trials should it if the numbers are heavily skewed? For example, lets say you play 30 events with a 50% chance. If you win 23 times the odds of that happening (23 wins or greater) randomly is .0026 if you are a break even player. So, one should conclude that there was very likely skill involved. What am I missing? |
#9
|
|||
|
|||
Re: How big a sample do you need?
Out of every 390 2p2ers on this forum, at least one of them will have that hot streak. It explains nothing.
Are you saying when Noobie McNewb bets 15 games week 1, ATS and O/U, and hits 23 right and 7 wrong, that it MUST be skill??? What am I missing? Naj |
#10
|
|||
|
|||
Re: How big a sample do you need?
[ QUOTE ]
"I can give you a hunt to narrow your search for factors that help explain home-field advantage: I live near the ocean, and high tides do not water my lawn". [/ QUOTE ] IMO he is saying the home field advantage isn't significant at all. In my picks I hardly ever go with the home team and they are good so far. (Except for last night [img]/images/graemlins/frown.gif[/img]) |
|
|