pokerjoe Posted August 29, 2014 Report Share Posted August 29, 2014 I want to see this as I'm sitting on 1k+150% x10 Ask him to also include, as an estimable variable, the increased chance of a book going under as it correlates to it's bonus sizes. Quote Link to comment Share on other sites More sharing options...
GameBred Posted August 29, 2014 Report Share Posted August 29, 2014 Just a hint...It might also be easier for you to break it down into the 216 possible combinations and look at things that way. I can use all the help I can get :) Thank you. Quote Link to comment Share on other sites More sharing options...
pokerjoe Posted August 29, 2014 Report Share Posted August 29, 2014 Which has more value, a free point off the opener or a free point off the close? Rank these numbers in order of RMSE from the final score, for a game in which there has been no public news from open to close: raw opener (basically, the initial overnight line), smoothed out but still pre-broad market opener (say, by 7 am), the number just after the broader open (maybe 8:30am), the number 30 minutes before the game starts, the closing number. Would your ranking change for big games, MNF and such? Would it change if public faves have been really rolling lately? Would it change for periodic events, like the World Cup? Quote Link to comment Share on other sites More sharing options...
GameBred Posted August 29, 2014 Report Share Posted August 29, 2014 Which has more value, a free point off the opener or a free point off the close? Rank these numbers in order of RMSE from the final score, for a game in which there has been no public news from open to close: raw opener (basically, the initial overnight line), smoothed out but still pre-broad market opener (say, by 7 am), the number just after the broader open (maybe 8:30am), the number 30 minutes before the game starts, the closing number. Would your ranking change for big games, MNF and such? Would it change if public faves have been really rolling lately? Would it change for periodic events, like the World Cup? Great question. Interesting caveat with the "no public news" from open to close and I am assuming no market manipulation is taking place so I will go with: I'd take the free point off the close and as for my rankings, they would be from the bottom to the top of your list.... meaning I would rank the close as the having the lowest prediction error, then the # 30 minutes before the game, etc.. down to the raw opener having the largest prediction error. This is probably the cliche answer but it is the best I can do :) As for the change in rankings, perhaps I would make some adjustments for all three scenario's by not exactly sure how I would rearrange the rankings; sorry for the non-committal response. Quote Link to comment Share on other sites More sharing options...
pokerjoe Posted August 29, 2014 Report Share Posted August 29, 2014 How do you calculate a vig-free number? Suppose Pinny has an NBA game -7 -110/+100 (-110 for the fave). What is the no-vig line if CRIS, LVH, CTG, The Greek, etc (hereafter "broad market") has the same game 6.5 -110/-110? What if the broad market has it -7 -110/-110? What if the broad market has it -7.5? What if the line opened at -6 and went up to -7? Pinny still with the 7 -110/+100. What it it opened -8 and went down to -7? What if it went down but squares were generally betting the fave? What if it went up and squares were betting the fave, and sharps were on the sideline? IOW, how, if at all, do these variables affect your vig-free CLV estimate? Suppose a line opens -6.5 on a game in which the sharp consensus has the line at -7. The fave is a big public team and attracts a lot of action and the line goes up to -7.5. Assume a sport where a half point equals the juice, so that the sharps have no EV on the fave at -6.5 OR the dog at +7.5, and thus stay on the sideline. How much CLV edge would a bettor who laid -6.5 have, in this game which closed -7.5? Thanks in advance. Quote Link to comment Share on other sites More sharing options...
pokerjoe Posted August 29, 2014 Report Share Posted August 29, 2014 Consider a regular market consisting of X sharps and 10X squares. Every 4 years the market, for only a few games, explodes to X sharps and 1000X squares. Do the lines get sharper or duller? Quote Link to comment Share on other sites More sharing options...
pokerjoe Posted August 29, 2014 Report Share Posted August 29, 2014 Consider a line the sharp consensus makes 7. The game in one universe opens 6, and thus gets bet to 6' (not all the way to 7 because the half point = the juice and thus denies EV at that point). The game in an otherwise identical universe opens 8 and thus gets bet down to 7'. SO we have a swing of 6' to 7' on the same game with the same market estimation with the same size error in the opener. How do you account for that in making CLV observations? Quote Link to comment Share on other sites More sharing options...
cyberbabble Posted August 29, 2014 Report Share Posted August 29, 2014 Can your model out perform my lucky silver dollar? Quote Link to comment Share on other sites More sharing options...
GameBred Posted August 30, 2014 Report Share Posted August 30, 2014 And while we're on the topic of math...here is a quick quiz for the board... A casino offers up a new game, a dice game. It's played using three completely fair 6-sided dice, meaning there's a 1/6 chance that each face, on each die, shows up. The dealer tells you that you can bet on any one number from 1 to 6. You'll then roll the three dice. If none of the dice show the number that you bet on, you'll lose $1. If one shows the number you bet on, you'll win $1. If two of the dice show the number you bet on, you'll win $3. And if your number comes up on all three dice, you win $5. Who has the edge in this game, the player or the house? I’d like to solve the already solved puzzle. No Matches (5/6)^3 = 57.87 * -$1 = -57.87 1 Match 3 rolls * (5.6)^2 * (1/6) = 34.77 * $1 = 34.77 2 Matches 3 rolls * (5/6) * (1/6)^2 = .0692 * $3 = 20.76 3 Matches (1/6)^3 = .0046 * $5 = 2.30 Had 2W2P2S not provided the answers, I would not have known how many decimal places to go and I am still not positive. For instance, I would have came up with 2.50 (not 2.30) for the 3 matches in 3 rolls as I would have rounded .0046 to .005. Is it a mistake to ever round up in these situations? How do we know how many decimal places we need to go? Also, any reason why my numbers were slightly off from 2W2P2S’s, especially in the 2 matches scenario? My math was 3 * .8333 * .0277 * $3 = 20.77 (not 20.82). Any feedback would be appreciated. Thank you. Quote Link to comment Share on other sites More sharing options...
evade Posted August 30, 2014 Report Share Posted August 30, 2014 GB, stop rounding off. Just plug the above arithmetic expressions in a spreadsheet and see for yourself. The numeric values differ(slightly) from the ones you have got. Quote Link to comment Share on other sites More sharing options...
Deal With It Posted August 30, 2014 Report Share Posted August 30, 2014 My head hurts now looking at the math equations in this thread... Quote Link to comment Share on other sites More sharing options...
GameBred Posted August 30, 2014 Report Share Posted August 30, 2014 GB' date=' stop rounding off. Just plug the above arithmetic expressions in a spreadsheet and see for yourself. The numeric values differ(slightly) from the ones you have got.[/quote'] Thank you Evade. Plugged this into excel for the 1 match: =3*(5/6)^2*(1/6) and came up with 34.72 Plugged this into excel for the 2 matches: =3*(5/6)*(1/6)^2*3 and came up with 20.83 Plugged this into excel for the 3 matches: =(1/6)^3*5 and came up with 2.31 Much closer but still not an exact match. Hmmm…. Quote Link to comment Share on other sites More sharing options...
milwaukee mike Posted August 30, 2014 Report Share Posted August 30, 2014 Ask him to also include, as an estimable variable, the increased chance of a book going under as it correlates to it's bonus sizes. why would that one tiny variable (by itself) increase the chances of a book going under? if the rollover is large enough they could be better off in terms of juice, payout fees, etc...how many "new books" have come around in the last 5-10 years without being bonus shops, to use for comparable purposes? Quote Link to comment Share on other sites More sharing options...
evade Posted August 30, 2014 Report Share Posted August 30, 2014 GB' date=' stop rounding off. Just plug the above arithmetic expressions in a spreadsheet and see for yourself. The numeric values differ(slightly) from the ones you have got.[/quote'] Thank you Evade. Plugged this into excel for the 1 match: =3*(5/6)^2*(1/6) and came up with 34.72 Plugged this into excel for the 2 matches: =3*(5/6)*(1/6)^2*3 and came up with 20.83 Plugged this into excel for the 3 matches: =(1/6)^3*5 and came up with 2.31 Much closer but still not an exact match. Hmmm…. 3*(5/6)^2*(1/6) 0.347222222 3*(5/6)*(1/6)^2*3 0.208333333 (1/6)^3*5 0.023148148 Set the cell format as Number - general and you will get the above results (based on one try). The point I was trying to make was the error resulting due to rounding off of individual components (as three in this case) and then adding them. Another exercise for you - just add the three expressions using the fractions. :-)) Quote Link to comment Share on other sites More sharing options...
pokerjoe Posted August 31, 2014 Report Share Posted August 31, 2014 It isn't a tiny variable. % of books with awesome bonuses going under > % of books without awesome bonuses going under. Chance of collapse correlates to bonus size. I don't have a DB on the issue, just many years of observation, so, anecdotal, sure, but that doesn't mean I'm wrong. Here's another such variable: willingness of established books to do transfers. It used to be as simple as this: if Spiro didn't trust a book, I didn't trust a book. That's a little stringent, and may over-empower the establishment, but then again, it's hard to be too careful with post-up funds. Here's a third indicating variable: off-numbers. Books posting weak numbers either boot winners or face collapse. I'm not saying don't fund new outs. I'm saying understand that there is a countervailing risk to the obvious rewards of such gambles, and yes it is part of a proper EV calc. Work it backwards: what % chance of being stiffed negates a bonuses EV? By instinct (you don't have much else to go on) is that chance lower than your estimate of the book's chance of collapse? Quote Link to comment Share on other sites More sharing options...
milwaukee mike Posted August 31, 2014 Report Share Posted August 31, 2014 i see what you're saying pokerjoe, but just because it correlates doesn't mean it causes it. i could say that retailers with smaller bathrooms fail more often, that is probably true because smaller retailers fail more often, but their bathroom didn't cause the failure. just like smaller and newer sportsbooks have to give out large bonuses to attract customers. without an online casino, i imagine running a bonus shop would be a losing proposition. like the bars around here that make most of their money off of slot machines or taking bets. Quote Link to comment Share on other sites More sharing options...
Hedge Posted August 31, 2014 Report Share Posted August 31, 2014 1-Vice contest has a selected poster pick exactly 100 games. If he gets 64 or more right he wins $500. So what are the odds of an Average Capper (50% winners) getting 64 or more? Also, what are the odds of a Good Capper (55% winners) hitting 64 or more? Quote Link to comment Share on other sites More sharing options...
GameBred Posted August 31, 2014 Report Share Posted August 31, 2014 So what are the odds of an Average Capper (50% winners) getting 64 or more? ~ 0.33% Also' date=' what are the odds of a Good Capper (55% winners) hitting 64 or more?[/quote'] ~ 4.3% Quote Link to comment Share on other sites More sharing options...
GameBred Posted August 31, 2014 Report Share Posted August 31, 2014 Evade, I did that originally with Excel (Number, general) and still came up with those figures, which I guess is good because: 34.72 + 20.83 + 2.31 = 57.86 (2W2P@S's #). Thanks. No more rounding off :) Quote Link to comment Share on other sites More sharing options...
GameBred Posted August 31, 2014 Report Share Posted August 31, 2014 Great question. Interesting caveat with the "no public news" from open to close and I am assuming no market manipulation is taking place so I will go with: I'd take the free point off the close and as for my rankings, they would be from the bottom to the top of your list.... meaning I would rank the close as the having the lowest prediction error, then the # 30 minutes before the game, etc.. down to the raw opener having the largest prediction error. This is probably the cliche answer but it is the best I can do :) As for the change in rankings, perhaps I would make some adjustments for all three scenario's by not exactly sure how I would rearrange the rankings; sorry for the non-committal response. PJ, I was looking forward to your comment on this post. Agree? Disagree? Quote Link to comment Share on other sites More sharing options...
Hedge Posted August 31, 2014 Report Share Posted August 31, 2014 ~ 0.33% ~ 4.3% Thanks GameBred. So an Average Capper has a 1 in 300 chance of winning. Since this is a one-participant monthly contest, a winner will occur once in every 25 years!!! Quote Link to comment Share on other sites More sharing options...
ERBtheGREAT Posted August 31, 2014 Report Share Posted August 31, 2014 Thanks GameBred. So an Average Capper has a 1 in 300 chance of winning. Since this is a one-participant monthly contest, a winner will occur once in every 25 years!!! Yeah 60 would be much better. 64 is very tough. Quote Link to comment Share on other sites More sharing options...
TommyL Posted August 31, 2014 Report Share Posted August 31, 2014 Evade, I did that originally with Excel (Number, general) and still came up with those figures, which I guess is good because: 34.72 + 20.83 + 2.31 = 57.86 (2W2P@S's #). Thanks. No more rounding off :) When I did it for the first time, I just calculated it using fractions. I looked at it as "216 separate trials" each covering one of the possible combinations, so I didn't have any rounding. And then just added the possibilities together (ie 1 trial of +5, 125 trials of -1, etc.). Quote Link to comment Share on other sites More sharing options...
TommyL Posted August 31, 2014 Report Share Posted August 31, 2014 Thanks GameBred. So an Average Capper has a 1 in 300 chance of winning. Since this is a one-participant monthly contest, a winner will occur once in every 25 years!!! If someone were smart about it (especially during this time of year when we have these types of games), they'd be sure to play favorite/over or dog/under on some of these huge CFB spreads, to increase their chances of either going 2-0 or 0-2. It's obviously easier to hit 64% over 50 games than over 100 games. Quote Link to comment Share on other sites More sharing options...
GameBred Posted August 31, 2014 Report Share Posted August 31, 2014 When I did it for the first time, I just calculated it using fractions. I looked at it as "216 separate trials" each covering one of the possible combinations, so I didn't have any rounding. And then just added the possibilities together (ie 1 trial of +5, 125 trials of -1, etc.). I started out with the "216 separate trials" approach and got off to a flying start when I did 1/216 * 5 = $2.31 (payoff for all 3 #'s coming up) then got caught up when trying to complete the rest. Adding the other possibilities together just didn't come easy for me. That was fun as I love exercising the brain so if you have anymore, send them our way please. Quote Link to comment Share on other sites More sharing options...
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.