Setting Your Own Spread

[Skating Tripods]

I've seen this talked about a lot with wiseguys and even some of the more serious sports bettors who don't do it for a living, and obviously BTB did all the CFB games.

But I want to know how. What criteria do you use to set a spread and what is the reason that you use certain stats/matchups over others? Do you adhere to the standard -3 for the home team in a pretty evenly matched game and then go from there?

Often times, guys will get a number in their head and then bet the line if it's off the number by a decent amount. How do you come to that personal number?

[Weebay]

(08-28-2011 08:35 PM)Skating Tripods Wrote:  I've seen this talked about a lot with wiseguys and even some of the more serious sports bettors who don't do it for a living, and obviously BTB did all the CFB games.

But I want to know how. What criteria do you use to set a spread and what is the reason that you use certain stats/matchups over others? Do you adhere to the standard -3 for the home team in a pretty evenly matched game and then go from there?

Often times, guys will get a number in their head and then bet the line if it's off the number by a decent amount. How do you come to that personal number?

They have their own power rankings and from there they can make a line. CFB home field advantage is different because 3 shouldnt be used as the standard home field edge. Some teams offer no edge at home, others having a much stronger home field edge.

[Skating Tripods]

(08-29-2011 06:19 AM)Weebay Wrote:  They have their own power rankings and from there they can make a line. CFB home field advantage is different because 3 shouldnt be used as the standard home field edge. Some teams offer no edge at home, others having a much stronger home field edge.

That's not necessarily how everyone does it. And even still, you didn't answer my question about criteria. If you don't make your own spread, the question wasn't really directed at you.

I've asked a wiseguy contact of mine but he asked that I wait until after football because it's a lengthy explanation and he's extremely busy. I'm eagerly awaiting his response.

I'd still start with a baseline of -3 for an evenly matched conference game and elaborate it from there based on all kinds of factors. Team they're playing against, how they play at home, travel for non-con games, etc.

I wouldn't say that there are any teams who operate without a homefield advantage. Even if they lose by 2 TDs at home, they'd probably lose by 3 or 4 on the road. It may not show up in ATS stats, but it can certainly make a game look closer for the program's sake.

[kreatture]

I can't type a lot right now as I'm restricted to typing with one hand while the other heals from a 'to the bone' cut on the knuckle, so it's hard to contribute here with lengthy posts.. Typing slowly with one hand takes too long and I end up losing my focus on what I originally started talking about, LoL.. but 2 categories used to create lines are Player Factor and Home Turf Factor. There is a general consensus among sportsbooks as to what certain players add to a spread, and the same for Home Fields. If you know someone at a Sportsbook, or someone who has access to what numbers the Sportsbooks use for this, then you'll be able to start creating your own lines a lot easier. (I'm sure BTB has access to this and maybe he'd be willing to share what he knows).

I know Peyton Manning is worth about 7pts to the spread and I believe is the highest of any NFL player. Star QB's are generally worth 3-6pts to the spread if they're out, while RB's are 1-4. These are numbers I've come across anyway and numbers I use myself.

I'm one of those guys who sees a number in my head and goes from there. Every day when I'm lookin at games, not once do I EVER go to the Sportsbook first and look at the lines. After I narrow the list down to the few games I want to look at more in-depth, I will spend time reading absolutely everything I can find on that game, from injury and weather reports to local beat writers to players themselves on Twitter (sometimes a player will talk about a party the team or players were at the night before until quite late which can be useful info.. stupid players.) and then I create a line in my head based off all the info I've received. THEN I will check the posted lines and flag the ones with big discrepancies and look to make plays based off of that.

It's taken me so long to write that I completely forget what else I was trying to say.. So I'm just gonna wrap up now and quit before I ramble on about nothing..

[Beyond the Bets]

Everybody does it differently, or at least I think they do, and every line or total can't be arrived at the same way. For instance, my approach to point spreads is much different than my approach to arriving at season win totals.

Some oddsmakers, I'm sure, rely primarily on data. Over time, that's probably for the best, but I think there's considerable danger involved when you blindly follow the numbers and are ignorant to public perception, a team's motivation, situational factors, weather, etc.

You can't arrive at the best numbers purely through quantitative means, and you can't arrive at the best numbers through qualitative means, either. That's why a mixed method approach — in which you apply the data to observation, finger-on-the-pulse type of stuff, etc. — is the best method, IMO.

My approach is very basic (see: Wannabe Oddsmaker) but I think I arrive at decent numbers anyway. My college football win totals, for instance, were put out for public view a month ahead of the sportsbooks and I thought mine were sharper numbers, for the most part. Now, did I have some stinkers? Absolutely. I totally underestimated West Virginia, over rated Virginia Tech (kind of), under rated Texas, and so on.

While I think you're most interested in point-spread setting, I'll explain how I arrived at win totals first. I'll post again about point-spread setting later today.

For me, it's a five-step process.

1. Power rank the teams, with particular emphasis on ranking teams in each conference. How does Team A stack up against Team B on a neutral field; how does Team B stack up against Team C or Team D; and so on.

2. Analyze the schedules. It's not as simple as saying, "OK, Team A plays four difficult opponents and two extremely difficult opponents." No, you have to get more specific than that and look at how the schedule is structured. If a team with a new coach has an overly-easy non-con schedule (with four likely wins) and two winnable home games to open conference play, I have to keep that in mind. That's useful to know. If a new coach gets buried with difficult games early, the team might fall apart by midseason. Oppositely, if a new coach is able to build trust and establish confidence with his team early by beating up on terrible opponents, the team might believe they are better than they really are. You have to do a little prognosticating and projecting when analyzing the schedule.

3. For each team, assign the percentage chance of them winning against each opponent. So, for instance, Texas might have a .95% chance of beating Rice; a 65% chance of beating Texas Tech; a 33% chance of beating Oklahoma. Convert all these percentages to decimals (i.e., .95, .65, .33, etc.) and then add them all up. That's the "rough" win total. By assigning the point spreads to every game like I did, there are multiple ways that you can go about creating a win total.

4. Next, you go back through the schedule again, only this time, you assign wins and losses. No percentage. So, instead of Texas has a 95% chance of winning against Rice," it's "Texas will beat Rice." It's also "Texas beats Texas Tech," and "Oklahoma beats Texas." When I'm done establishing the wins and losses, I add them up and get my number. I will then compare this number to the previous one from step 3.

5. OK, so now I have two numbers to work with. I must keep these in mind when making my final number. The final number is arrived at by factoring in the first two numbers, in addition to other factors. If you are trying to project Vegas's number, then you must include public perception, etc. If you are merely trying to project your own, then tweak and adjust however you want.

Compared to most, my way of doing it is extremely elementary. But the process doesn't matter as long as the final answer comes out correct — or close to correct, anyway.