In this thread I answer your gambling questions:

mmaed

Member
By liquidity you mean the overall amount of money in the market, correct? I was kind of wondering how much money will actually make it into the mma markets of augur because it seems like a big loophole for the average person to go through to bet on mma.
 

NunyaBidness

Active Member
Oddsb Poker Champ
Yeah, thats why I don't expect Augur to work anyway. The sheer complexity of it is ridiculous.

The amount of money poured into the augur tokens will at least generate some level of motivation in the market however. But, I think its wildly overvalued.

So, basically I don't think it'll ever work, but if it does generate the volume that the the token sales predicted, then it should be gigantic. So I'm optimistically pessimistic about it I guess.
 

Chris M

Member
What do you mean by "high value wagers"? You mean different bets? I don't suspect we'll see as many bets offered there as we get on other books, just because I don't think the liquidity will be there to support it. I think we'll be looking at over/unders and MLs for the most part.
Sorry, by "High value wager" I meant wagers in which you calculate your edge as being large.
 

lucky

Active Member
@NunyaBidness or anybody. should I be using Poisson calculator on total bases props? If player A averages 1.5 per gm and player B 2; then what would be the line?
 

Chris M

Member
@NunyaBidness or anybody. should I be using Poisson calculator on total bases props? If player A averages 1.5 per gm and player B 2; then what would be the line?
You're talking about fair line on a h2h match up between the players? If you assume the total bases projection for each of the two players are independent events, then I believe poisson can be used. Looking at the poisson calculator for two variables on SBR, using expectations of 1.5 and 2, with a spread of zero, this yields push prob (.2162) and win probs for each side (.2903 & .4936). With these figures, I think fair line is simply .4936/(1-.2162) = .6298 = -170.13 / +170.13

FYI I never bet MLB, so yeah.
 

punisherind

Moderator
Foundation Member
doesn't poisson only work for situations that can only occur in multiples of 1? in other words, you wouldn't use it on total bases b/c a player can get multiple bases in one plate appearance.
 

Perra

Active Member
I'm trying a football strategy where I bet the under (to end in 0-0) in soccer games half time and close most of them with a profit. After 69 games I have the following stats:

Games ended 0-0 27
Games not ended 0-0 42
Average odds played 2,8
Average units played per game: 1 (always same betsize)
Total profit: 20,48 units
ROI: 28,8%

As I've closed out most plays and didn't let them run my profit is a bit higher than they would've been if I just would have let them run. How many games do I need, approximately to know if this is a sustainable strategy and rule out that I've just been lucky, 100, 200 , 300 , a thousand games?
 

punisherind

Moderator
Foundation Member
Any leagues in particular?

Are you betting every game indiscriminately, or focusing on certain criteria, e.g. only games where the 1h total is 1?

You'd probably want at least 500 games. Maybe back test past games.
 

Perra

Active Member
Any leagues in particular?

Are you betting every game indiscriminately, or focusing on certain criteria, e.g. only games where the 1h total is 1?

You'd probably want at least 500 games. Maybe back test past games.


Pretty much every league available and I have certain selection criteria regarding scored goals and the minutes they were scored.

Haven't backtested it, but doing my testing with small amounts of dollars. Would be nice with a database where I could check all these stats, but don't know where to find one.
 

Chris M

Member
doesn't poisson only work for situations that can only occur in multiples of 1? in other words, you wouldn't use it on total bases b/c a player can get multiple bases in one plate appearance.
Ah yeah, didn't consider that.

What if for an individual player you took projections for each type of hit/walk per game, multiplied each by their respective number of bases, and took the sum of those products (so if you expect the player to have 0.05 HR's per game, that adds (0.05 x 4) = 0.2 to their total bases projection). This sum should represent the player's total base projection against the league average pitcher, so the figure could be adjusted based on how much better/worse than average the opposing pitcher is of the specific game you're looking to bet. From there, you could add additional features to the model to fine tune it, like split up the player's hitting projections into "vs RHP" and "vs LHP" (this is assuming your data is split this way, and there's enough of it).
 

lucky

Active Member
This all started on mlb prop of donaldson vs Lindor. My friend was trying to say that if player A averages 2 and player B avg 1. That line should be 2-1. I said it should be higher. Which got me thinking about what line should really be. I said it would be somewhere between his 2-1 line and whatever number from Poisson calculator.

I ended up betting donaldson -115 and losing then pushing next gm. Then they switched matchups to encarnacion -110 vs lindor and I made a little back yesterday.
 

NunyaBidness

Active Member
Oddsb Poker Champ
doesn't poisson only work for situations that can only occur in multiples of 1? in other words, you wouldn't use it on total bases b/c a player can get multiple bases in one plate appearance.


This, plus average probably isn't a strong enough metric since it should vary do to pitcher strength and probably other things. You could probably include team totals as a variable since it should approximate most of that stuff.
 

NunyaBidness

Active Member
Oddsb Poker Champ
This all started on mlb prop of donaldson vs Lindor. My friend was trying to say that if player A averages 2 and player B avg 1. That line should be 2-1. I said it should be higher.

There isn't enough data there to determine if it should be higher or lower.

Throw out the idea of total bases for a moment and just consider a nebulous average of 2 vs an average of 1.

Can you construct an example where the person with the average of 2 should be a prohibitive favorite (-10000 or even greater) against the average of 1?

Can you construct an example where the person with the lower average should actually be the favorite in the matchup against the player with the higher average?
 
Top