System is stable as expected, can't gather greater R/R or break 50% win percentage.

[Beetlejuice]

My system is very simple. So of all the college football games of ranked teams yesterday, making a total of 36 wagers, I won exactly half the outcomes at a fixed $2.00 position size. I made a total of .12 cents. I'm looking to drive a deeper edge. I like that I am not losing money, but now I need to tweak the system so I can make some. Suggestions are welcome.

I take the home team's ppg (at home) and average it out against the road team's point's against (on the road).
I then take the road team's ppg (on the road) and average it out against the home team's points against (at home). And then I make a revised point spread and look to see if it differs from the ACTUAL point spread.

So then the 5th grade math equation looks something like this....

USC -16 @ Colorado
HT PPG 32.0 - RT PA 25.3 = 57.3 / 2 = 28.75
RT PPG 33.3 - RT PA 30.15 = 63.45/ 2 =30.15

30.15 - 28.75 = USC favored by -1.4 NOT -16 as is listed in the betting markets.

Well that's nice. In this example, I took Colorado 15.5+ and won easily. I woulda took them at an even lowe number like +10 to jack up the risk reward, but I'm slowly discovering books are seldom offering me the opportunity to even bring the buy the line down at a lower number, like +120 or +150.

Totals follow a similar methodology, whereby we simply add up the numbers we just subtracted which would have been 58.9, which differed from the existing line 60.5 only slightly. Since this was a slight change it was a no play.

I'm PISSED I get stuck having to wager -105 or +100 or something, as the books wont offer +120 or +143 or something that I like even to play with the numbers to make the risk reward even better. That seems to be the trouble.

[Waterstpub87]

You could try a pleaser. Selling 6 points on two games pays out 6 to 1

[statnerds]

play at 5Dimes. not sure on the selling in college now though as i write this. check it right before kickoff i guess.

i know i sold NYG down to +4.5 +130, but that is NFL.

even if a book does allow you to sell down in college keep in mind the market is efficient and you are not going to be getting FV for the numbers you do sell

[dishwasher]

heritage sports will let you buy up or down

[Beetlejuice]

You could try a pleaser. Selling 6 points on two games pays out 6 to 1

I am leery of progressives, but the thought had crossed my mind. I ran the numbers from since I've been playing this way in October, and technically the system has a 51.3% win percentage. Not sure how progressives would jive with this.

play at 5Dimes. not sure on the selling in college now though as i write this. check it right before kickoff i guess.

i know i sold NYG down to +4.5 +130, but that is NFL.

even if a book does allow you to sell down in college keep in mind the market is efficient and you are not going to be getting FV for the numbers you do sell

I definitely prescribe to the efficient market principle, rather than random walk. I do bet at 5 dimes. I take advantage of the reduced juice whenever possible. I will be trying to put the bets in before kickoff whenever possible, as that does seem to help as your thoughtfully suggested. The 36 different outcomes in college that were wagered on back this on Saturday were placed in the very early morning hours that day, so I agree in the since that the betting market may not have, "matured," at that point, allowing less wiggle room to buy up or down.

heritage sports will let you buy up or down

I'm American. Thanks though.

[Waterstpub87]

Heritage allows US players. It is my go to book and I am an American.

Little bit of math, I am assuming that you are selling off 15 or something high, and not 7 because it makes the value of the half points worth more.

Call the value of a half point as 4 cents. For the two team college pleaser, assumption is 6 points per game for a payout of 6 to 1.

An individual game would start at -100 vig free, the sold 6 points would move the price to 148. 148 equates to a 40.32% chance of victory.

The parley value of this would be .4032 x .4032= 16.25% winning chance, x 6 = .975, for a house advantage of 2.5%

Consider instead that we start for your assessed winning % of 51.3, equivalent of -1.05. This would then push this to 143 as an end result.

This would equal a 41.45 winning %.

.4145*.4145 = .1718 * 6 = 1.03, which is 3% in your favor.

I am ignoring ties.

My values for half points are way off. Selling from 15 to 21 is 199.

Which is around 33%.

33% * 33% = 10.89 *6 = 65.34. Normally

Starting at -105 for your system is 194

34% * 34% = 11.56 * 6 = .6934

This is how I would value it. Anyone please let me know if my math is off.

So don't bet pleasers.

[evo34]

My system is very simple. So of all the college football games of ranked teams yesterday, making a total of 36 wagers, I won exactly half the outcomes at a fixed $2.00 position size. I made a total of .12 cents. I'm looking to drive a deeper edge. I like that I am not losing money, but now I need to tweak the system so I can make some. Suggestions are welcome.

I take the home team's ppg (at home) and average it out against the road team's point's against (on the road).
I then take the road team's ppg (on the road) and average it out against the home team's points against (at home). And then I make a revised point spread and look to see if it differs from the ACTUAL point spread.

So then the 5th grade math equation looks something like this....

USC -16 @ Colorado
HT PPG 32.0 - RT PA 25.3 = 57.3 / 2 = 28.75
RT PPG 33.3 - RT PA 30.15 = 63.45/ 2 =30.15

30.15 - 28.75 = USC favored by -1.4 NOT -16 as is listed in the betting markets.

Well that's nice. In this example, I took Colorado 15.5+ and won easily. I woulda took them at an even lowe number like +10 to jack up the risk reward, but I'm slowly discovering books are seldom offering me the opportunity to even bring the buy the line down at a lower number, like +120 or +150.

Totals follow a similar methodology, whereby we simply add up the numbers we just subtracted which would have been 58.9, which differed from the existing line 60.5 only slightly. Since this was a slight change it was a no play.

I'm PISSED I get stuck having to wager -105 or +100 or something, as the books wont offer +120 or +143 or something that I like even to play with the numbers to make the risk reward even better. That seems to be the trouble.

Not trying to be rude, but your system will have no edge. Even 10-15 years ago, lines were set using much more predictive statistics that average points scored/allowed. These days, a model needs to be quite sophisticated to have even a tiny edge.

[Beetlejuice]

Not trying to be rude, but your system will have no edge. Even 10-15 years ago, lines were set using much more predictive statistics that average points scored/allowed. These days, a model needs to be quite sophisticated to have even a tiny edge.

I don't think that's rude. Knowing that the system is slightly better than 50%, I'm now simply only working with a risk reward of 1 to 1 or better. No more -105 or -110. Last night's NBA was a good example. I took three wagers.
Pacers +6 (+100). WIN
Celtics +165. WIN
Thunder +135. LOSS
So, that's a net gain of 1.65 unit fixed position size. I will not accept any wagers that do not offer a one to one. I'll report back as the week progresses. I think this is the edge I need, as its the similar mindset I use while trading forex. I'm only a 50% trader, but the risk\reward keeps me in the green.

[Waterstpub87]

You don't know that your system is slightly better than 50%. Even at 1000 plays you are not a standard deviation away from the mean.

[Beetlejuice]

You don't know that your system is slightly better than 50%. Even at 1000 plays you are not a standard deviation away from the mean.

I think I have a large enough sample. Literally, hundreds of games, and the W/L rate is about half. \ I'm having trouble understanding what you're trying to communicate.

[dhristov211]

this is too basic

[agendaman]

in long run system will fail you will die broke