Ind+2.5 -105, +3 -115, +3.5 -125 or P +125

Ganchrow · 09-20-07, 11:45 AM

Originally posted by HedgeHog

This is simple to calculate using my <a href="#" onClick="OpenCalc(); return false;">Half-Point Calculator</a>.

On the top half of the calculator select the following:

League: "CFB".
Bet Type: "Spread"

Taking Pinnacle's Illinois -2½ -108/Indiana +2½ -102 as the fair market indicator (if you believe you have a better estimate of the fair market, you should of course use those figures instead):

Spread: 2.5
Fave Odds: -108
Dog Odds: -102

On the lower part of the calculator you shoudl now look at each possible bet:

IND +2.5 -105 -- Change the contents of the Dog +2.5 box from -102 to -105 and click elsewhere on the calculator. We see that this implies an expectation for IND +2.5 of -3.742%.
IND +3 -115 -- Change the contents of the Dog +3 box from -117.11 (Pinnacle's alternate odds at +3 are -120) to -115 and click elsewhere on the calculator. We see that this implies an expectation for IND +3 of -1.535%.
IND +3.5 -125 -- Change the contents of the Dog +3.5 box from -132.76 (Pinnacle's alternate odds at +3.5 are -137) to -125 and click elsewhere on the screen. We see that this implies an expectation for IND +3.5 of +0.244%.
IND PK +125 -- Change the contents of the Dog +0.5 box from +116.7 (Pinnacle's alternate odds at +0.5 are +120) to +125 and click elsewhere on the screen. We see that this implies an expectation for IND +0.5 of +1.380%.

So if you believe the push probabilities embedded in the calculator and believe that Pinnacle's primary point spread market is a fair indicator of the full market then your intuition was indeed correct -- of the bets listed, the money line provides the best value at +1.380%. (And of course if you don't trust the embedded push probabilities, you're certainly welcome to key in your own.)

HedgeHog · 09-20-07, 12:03 PM

Thanks, Ganch. Your calculator is quite handy!

Justin7 · 09-20-07, 12:44 PM

Random thought... The total on this game is 62. The median total in NCAAF is about 50. When I do point conversions for NCAAF, I typically multiply the push rate * 50/game total. In this case, the moneyline becomes even better.

HedgeHog · 09-20-07, 01:20 PM

Originally posted by Justin7

Random thought... The total on this game is 62. The median total in NCAAF is about 50. When I do point conversions for NCAAF, I typically multiply the push rate * 50/game total. In this case, the moneyline becomes even better.

Good point. Total has gone to 63 and with that much scoring the 2.5 points shouldn't mean much.

raiders72002 · 09-20-07, 04:31 PM

With a total of 63 I show the ML to be the best choice followed by +3.5.

raiders72002 · 09-20-07, 04:37 PM

I just saw that Ganch has the same thing. I guess it's always just best to go with Ganch.

HedgeHog · 09-20-07, 04:56 PM

Originally posted by raiders72002

I just saw that Ganch has the same thing. I guess it's always just best to go with Ganch.

Thanks Raiders. Sorry about the crap I gave you last night.

raiders72002 · 09-20-07, 04:57 PM

np- Let's have a big weekend.

raiders72002 · 09-20-07, 06:59 PM

SSB

Deriving "Key Numbers" all by Your Lonesome by Colin Caster
September 10th, 2004

Spend any amount of time on sports wagering message boards and you'll see that one of the most frequently asked questions has to do with the "value" of particular point spread numbers. For example, a poster will ask whether it's worth the extra vig to buy from +3.5 onto +4, or whether one should prefer a point spread of +2.5 to a money line of +120, or whether it's worth a mile walk down the Strip to lay -2.5 when a perfectly good -3.5 is available ten steps away from the most comfortable spot on the Sigma Derby machine.

The realization that every extra half point has some expected value is like the nascent sharp taking his first steps. Understanding that different point spread numbers are associated with varying degrees of "value" is a baby sharp's wobbly walk across the room. Learning to apply the value associated with differing point spreads is a confident stride across the playground. But a lot of sharps seem to stop there, for some reason. Having successfully and profitably learned how to apply a table of key numbers (like that provided for the NFL by Stanford Wong in his book, Sharp Sports Betting), many bettors profess an inability to take the next step in their development as a sharp: deriving for themselves such information. Well, you've had the drool off your chin for some time now, and you've learned to walk behind the stroller instead of sitting in it; now it's time to shed those poopy diapers.

Make no mistake: the ability to derive your own charts is an important skill for the sports bettor. If you are dependent upon someone else's chart you will be helplessly confined to the specific betting situations to which the chart applies. For example, a little while ago in the SharpSportsBetting.com College Football Box Seats, Stevie Y posted a historical table that shows how frequently college football games push against the spread for lines ranging from 1 to 41, home and away, in general and in conference games. It's a terrific tool, and it has a wide range of betting applications. But suppose you want to bet first half lines? What if the rules of the game change next year in some way that has implications for margin of victory? What about Canadian football? Unless you can put together your own charts for novel circumstances you're plain out of luck.

A common misconception is that such charts can only be constructed by a computer whiz with tons of historical data at his beck and call. Sure, it's nice to be spoon fed by an expert every now and then, but it's even more important to understand that most of the information you desire is already freely accessible... from sports books. The books, of course, need to know (or need to be told by an expert) what every point, half-point, goal, or "behind" is worth so that it can provide all the betting options that players demand. With a minimum of record keeping you can observe how the books value these scoring increments and thereby derive the charts used by the books. Sure, the results will often be rough, and an expert with a great database can frequently provide more finely-tuned charts, but for many, many betting applications you are in great shape if you know as much as the book does.

Let's consider as an example the chart that Stevie Y presented evaluating college football point spread numbers. Suppose we didn't have the benefit of his work and we wanted to construct a similar table. By observing how the sharpest sports books equate point spreads and money lines we can create a rough version of this chart.

To illustrate how this might be done, I recorded the fair money lines associated with a small sample of point spreads for college football games in the first two weeks of this season, and displayed the results in the table below.
Spread Money Lines Median Expected Win %
2.5 125, 128, 132, 135, 135, 135, 135, 137, 140 135 57.4
3.5 150, 160, 160, 172 160 61.5
4.5 180, 187, 193 187 65.2
5.5 none - -
6.5 225, 240 237 70.3
7.5 275, 280 277 73.5
8.5 275, 302 288 74.3

"Fair money lines" are calculated here as the midpoint between the "minus" line (favorite line) and the "plus" line (underdog line), so if a book offered -135 on a 2.5 point favorite and +115 on a 2.5 point underdog, I recorded the fair money line as 125.

In the table above I listed fair money lines for a select group of point spreads. To characterize the "typical" fair money line, I found the median (midpoint) of the fair money lines for each of the point spreads. Then I calculated the expected win percentage associated with that median fair money line (median fair line/[median fair line +100]).

As you can see, a point spread of 2.5 is typically associated at the sports books with an expected winning percent of 57.4%; a point spread of 3.5 is typically associated with an expected winning percent of 61.5%. In general, then, sports books consider a line of 3 in college football to have a value of 61.5% - 57.4% = 4.1%. By the same reasoning, a line of 7 is generally considered to have a value of 3.2%; a line of 8 is considered to have a value of 0.8%. I cannot compute a value for the numbers 5 and 6 because my sample had no instances of point spreads of 5.5. If I want an estimate of the value of these numbers I need to make additional observations.

How confident am I in the values of the other numbers? Well, it depends upon the distribution of fair money lines I obtained. Although there is some variability, the fair money lines associated with a point spread of 2.5 seem pretty clearly to center around 135. However, for some other spreads -- for example, 8.5, for which I have only two data points -- I am less confident that I have uncovered the most closely associated fair money line. For these spreads I will need to collect more observations from the sports books in the weeks to come. But by making observations of just two weeks' worth of odds, I have already come a long way towards creating my own chart -- without anyone holding my hand.

This is just one example of how one might make use of information provided freely by the sports books. If you keep your eyes open you can learn all sorts of things about the sports on which you wager.

The HG · 09-21-07, 02:13 AM

I don't follow football at all, but volatility is always a major factor to consider when buying or selling points in any sport. I would imagine that there are many different factors and angles that could be analyzed from a volatility standpoint in football that would affect the value of various bought or sold points.