Useful metrics to determine skill

[MMANickAM]

Hey everyone
I just joined this site, look forward to interaction with this community!

I am currently trying to implement metrics that are useful in sports betting.
While I am tracking my progress for over two years now, I haven't been able to display metrics that I could use.
I tried implementing fund management variables (max drawdown, Sharpe ratio, Sortino ratio, volatily measuring risk and average). Because we are operating in two totally different universes finance vs sports betting, it wasn't of much use.

So what metrics are useful to track of in your spreadsheet? I imagine return on investment. Anything else?

Also how do you determine skill in sports betting?
My current objective is making 10% on a monthly basis: {(1+10%)^12}-1=214% thus doubling my capital each year.
Is that a good approach?

Any thoughts? I welcome all suggestions.

[gui_m_p]

So what metrics are useful to track of in your spreadsheet? I imagine return on investment. Anything else?

Also how do you determine skill in sports betting?
M

ROI or Yield is certainly a good metric, but this information will only be accurate after thousands of plays, and you don't want to wait so long to know if you are a good bettor.

The best data to determine skill in my opinion is closing line value.

CLV is determined by taking the difference between a line on a bet/pick when it is made and comparing it to what the line is when the game starts.

This "closing line" is considered to be the most accurate line, since all available information, opinions, and money bet have been factored in to it. In general bettors with the most past success and best resources are going to bet the most, and win the most, therefore influencing the line the most. If you are consistently getting a better line than what the closing line is, you are very likely to be a long term winner, simply based on the math value of what each point is worth.

If you calculate, you will also see that, while beating closing odds, most of your bets are "scalpable" (e.g., with odds greater than the odds of closing line with no vig). Example: if the line for a game -104)/-104, the chances are 50% for each side. So the equal odds disregarding the juice would be +100. If you bet at +104 you made a scalpable bet, which is for sure a sign of a winning performance in long run.

You'll see some people arguing that CLV is overrated (see the thread "the misunderstanding of beating the closing line"), but all long term winners I know consistently beat the market. Otherwise, see what the head of pinnacle answered on twitter:

"#AskPinnacle Is beating the @PinnacleSports closing line a good indicator of a value bet in soccer markets?
HoS: If you beat our closing line often enough you'll soon be a rich man".

[a4u2fear]

GUI, how do you use the scalpable terms you mentioned above when looking at a total? Like 45 -110 open and closes at 47-110

[AussiePunter123]

I find the quickest method to see whether someone's record is evidence of good knowledge in their chosen field, or down to chance, is to use this simple equation.

Add the person's wins and losses together. Divide this total by two, then add the square root of the total this total. If the number of wins is greater than this number, than the person's record is more likely to be down to skill, rather than pot luck.

Example.

40 wins 22 losses.

40 + 22 = 62

62/2 = 31

31 plus square root of 62 =
31 + 7.9 = 38.9

This person has 40 wins, which is greater than 38.9. They are likely to be knowledgeable in their chosen sport.

Tail.

[a4u2fear]

I find the quickest method to see whether someone's record is evidence of good knowledge in their chosen field, or down to chance, is to use this simple equation.

Add the person's wins and losses together. Divide this total by two, then add the square root of the total this total. If the number of wins is greater than this number, than the person's record is more likely to be down to skill, rather than pot luck.

Example.

40 wins 22 losses.

40 + 22 = 62

62/2 = 31

31 plus square root of 62 =
31 + 7.9 = 38.9

This person has 40 wins, which is greater than 38.9. They are likely to be knowledgeable in their chosen sport.

Tail.

Regardless of your equation, there is still a chance it is luck they got to that record. Need to look at the chances of that using a binomial probability distribution

[a4u2fear]

Just for example,, you posted 40-22, if odds are 50/50, there is about a 1-2% chance they got there from luck. Still good, but need to check this

[gui_m_p]

GUI, how do you use the scalpable terms you mentioned above when looking at a total? Like 45 -110 open and closes at 47-110

Search for a conversion chart, to see what would be the odds of 45 after the move to 47. SBR has one here, but not with total points:

http://www.sbrforum.com/betting-tools/spread-ml-converter/

After some use you'll decorate how much a 0.5 move represents. If I am right 0.5 total points in NFL represents nearly 0.05 in decimal odds. Or you can search one by one here:

http://www.oddsportal.com/american-football/usa/nfl/results/

[gui_m_p]

I find the quickest method to see whether someone's record is evidence of good knowledge in their chosen field, or down to chance, is to use this simple equation.

Add the person's wins and losses together. Divide this total by two, then add the square root of the total this total. If the number of wins is greater than this number, than the person's record is more likely to be down to skill, rather than pot luck.

Example.

40 wins 22 losses.

40 + 22 = 62

62/2 = 31

31 plus square root of 62 =
31 + 7.9 = 38.9

This person has 40 wins, which is greater than 38.9. They are likely to be knowledgeable in their chosen sport.

Tail.

This is for bets with -110 odds right?

One point: consider one person with a 33-29 record in those 62 bets. He's a 53,23% capper, so a winner at least in this sample. But your equation advice not tailing him.

[a4u2fear]

GUI,

so I used the odds for the superbowl to get these estimates:

Say the total is 47 @ -110 and you are playing the under (which I am in another handicapper think tank thread)
for each half point above 47, the juice adds about 8 pts, or u47.5 -118; and u49 -139

for each half point below 47, the juice lessens about 6 pts, or u46.5 -104; and u45 +115

Say I see the opening line at 45 and it is a play under my system, so I play it at 45 -110, but the game ends up 47 -110 (45 +115), what is the best case scenario here for what you call your "scalpable" bet?

The issue I have with this is I have a bookie and can only place bets 1 hr pregame, which is almost always the closing line, within a half point or point; and my system performed beautifully @ 61% over 100 plays. vs the opening line it was only 54.5%.

[JR007]

ROI or Yield is certainly a good metric, but this information will only be accurate after thousands of plays, and you don't want to wait so long to know if you are a good bettor.

The best data to determine skill in my opinion is closing line value.

CLV is determined by taking the difference between a line on a bet/pick when it is made and comparing it to what the line is when the game starts.

This "closing line" is considered to be the most accurate line, since all available information, opinions, and money bet have been factored in to it. In general bettors with the most past success and best resources are going to bet the most, and win the most, therefore influencing the line the most. If you are consistently getting a better line than what the closing line is, you are very likely to be a long term winner, simply based on the math value of what each point is worth.

If you calculate, you will also see that, while beating closing odds, most of your bets are "scalpable" (e.g., with odds greater than the odds of closing line with no vig). Example: if the line for a game -104)/-104, the chances are 50% for each side. So the equal odds disregarding the juice would be +100. If you bet at +104 you made a scalpable bet, which is for sure a sign of a winning performance in long run.

You'll see some people arguing that CLV is overrated (see the thread "the misunderstanding of beating the closing line"), but all long term winners I know consistently beat the market. Otherwise, see what the head of pinnacle answered on twitter:

"#AskPinnacle Is beating the @PinnacleSports closing line a good indicator of a value bet in soccer markets?
HoS: If you beat our closing line often enough you'll soon be a rich man".

I believe that but it become hard for guys to understand when you are beating the closer and are continuously losing, some guys on bad rolls in college bb at this time

[AussiePunter123]

Just for example,, you posted 40-22, if odds are 50/50, there is about a 1-2% chance they got there from luck. Still good, but need to check this

That equation I posted is one from statistics to give you a 95% confidence level.

[gui_m_p]

GUI,

so I used the odds for the superbowl to get these estimates:

Say the total is 47 @ -110 and you are playing the under (which I am in another handicapper think tank thread)
for each half point above 47, the juice adds about 8 pts, or u47.5 -118; and u49 -139

for each half point below 47, the juice lessens about 6 pts, or u46.5 -104; and u45 +115

Say I see the opening line at 45 and it is a play under my system, so I play it at 45 -110, but the game ends up 47 -110 (45 +115), what is the best case scenario here for what you call your "scalpable" bet?

The issue I have with this is I have a bookie and can only place bets 1 hr pregame, which is almost always the closing line, within a half point or point; and my system performed beautifully @ 61% over 100 plays. vs the opening line it was only 54.5%.

I don't know if I understood you properly. If you play the under at 45 and the closing line is 47 your bet isn't scalpable. On the contrary, people who bet the over 45 have it.

To find a scalpable bet you have to calculate the no-vig odds of the closing line (see a calculator here http://sportsbettingsites.org/bettin...ig-calculator/). If your pick was made with a better line, you have a scalpable bet.

If you can only bet 1 hour prior the game it don't apply to you since you won't be able to take the benefit of line movement. If your system performs better vs. closing line than opening this is intriguing, because normally the winners show opposite results.

[AussiePunter123]

This is for bets with -110 odds right?

One point: consider one person with a 33-29 record in those 62 bets. He's a 53,23% capper, so a winner at least in this sample. But your equation advice not tailing him.

Yes that is correct.

That equation applies to traditional capper odds.

If I am picking either an over or under handicap, then I have a 1 in 2 chance of picking the right one. I could easily get a 33-29 record by just randomly picking an over or under.

[marcoforte]

Calculate the z value of your record. If you're greater than 2.5, it's not by chance.

[gui_m_p]

Calculate the z value of your record. If you're greater than 2.5, it's not by chance.

How do you calculate it?