Predicting how many wins WFU will have in football next year (a statistics lesson)

Wake_forest_football_medium

via deadspin.com

Today I came across an article on our sister site Tomahawk Nation, that is essentially a "gamblers" take on any given team. I would encourage everybody to take a look after the jump and come up what I believe to be a much better win forecast for Wake Forest football this year. 

I will post my own results in a day or so to give everybody a look at my thoughts without influencing their own thoughts on each game.

Click through the jump for everything you need to know about a more accurate way to predict a teams season. Thanks to Tomahawk Nation for the article idea.

The following quote blocks are excerpts from this Tomahawk Nation article last year about a better way to predict how a team will do in any given year. The main theory is assigning percentages to each game based on the line (or in this case, your predicted line for each game), rather than assigning just  "win" or "loss" to each game:

The error my friend is making is that he is making definitive all of nothing calls on games that are anything but locks.  Football games are not 100%/ 0% probabilities (typically).  Now, could my friend be right?  Absolutely, but his methodology is suboptimal at best.  

It just so happens I have a friend of mine who is a professional gambler.  He has a day job, but he makes his living betting sports like a day trader plays the stock market, complete with a healthy dose of arbitrage.  Every year since my freshman year of college, we've gotten together and broken down the upcoming season, and we have a method.  We call it proportional win shares.  You can call it whatever you want, but I think you'll see how it's more effective than simply assigning "win" or "loss" to the games. 

The idea is simple:  assign a winning % to each game.  For instance, you could say that FSU has an 80% chance of beating Maryland (0.80), of that they have a 20% chance to beat UF (0.20).

Here is the chart on how you take your perceived line for each game, and translate it in to a percentage:

Point Spread Money Line
-2 -120/+120
-2.5 -130/+130
-3 -145/+145
-3.5 -175/+175
-4 -185/+185
-4.5 -200/+200
-5/-5.5 -215/+215
-6 -240/+240
-6.5 -260/+260
-7 -290/+290
-7.5/-8/-8.5 -330/+330
-9/-9.5 -350/+350
-10 -385/+385
-10.5 -415/+415
-11 -445/+445
-11.5 -475/+475
-12/ -12.5
-500/+500
-13 -525/+525
-13.5 -550/+550
-14 -580/+580
-14.5 -610/+610
-15/ 15.5 -640/+640
-16/16.5 -665/+665
-17 -690/+690
-17.5
-715/+715
-18/ -18.5/ -19/ -19.5 -740/+740
-20 -770/+770
-20.5 -800/+800
-21 -815/+815
-21.5 -830/+830
-23 -845/+845
-24 -875/+875
-28
-920/+920

After you get your line for each game (and write it down or whatever), you do the following with each of the 12 numbers that you have:

You might be a bit lost right now, but that's okay.  Here's the idea.  A 3 point underdog, has a moneyline value of "+145".  In numerical terms, that comes out to 245, because you add 100.  Still with me?  If not, it's okay.  Keep following.  To determine a team's projected winning percentage, based on the vegas line, simply take the moneyline value in numerical terms, and divide it into one.

  • Team A is favored over Team B by 3 points.
  • Thus, Team B is an underdog of 3 points.
  • Team B can be referred to as +3
  • Using the chart above, we can see that an underdog of +3 is "+145" moneyline value.
  • Remember to add 100 to the "145", to get "245".
  • Then, divide 100 (always 100 here) by 245.
  • 100/245= 0.408, which we will round to 0.41
  • 0.41 is the same thing as 41%.
  • So, Team B (the 3 point underdog) has a 41% chance of winning this game.
  • If Team B has a 41% chance of winning, then Team A must have a 59% chance of winning (because these numbers must add up to 100%) 
  • After you get the percentage number for each game (you will have 12 numbers that look like the following: .99, .77. 55 etc.), you simply add them up, and that will give you a number that you can round to the nearest whole number. Voila, there is your win prediction for the year. You can also isolate the ACC games and just add those up to see how many ACC games you think we will win based on the percentages you assign to each game.

    Post your results below, or just keep them to yourself, but I found this very interesting and I actually hit right around my number. It does allow a good way to look at each game and not demand a "win" or "loss" for each team, which can often be inaccurate.

    Big props to Bud Elliott as always for his excellent writing and analysis of football.

    Also...here's the Wake Forest 2011 Schedule for easy access:

    2011 WAKE FOREST DEMON DEACONS SCHEDULE
    DATE OPPONENT RESULT/TIME RECORD/TICKETS
    Sat, Sept 3 8:00 PM ET ESPN3.com Tickets
    Sat, Sept 10 3:30 PM ET Tickets
    Sat, Sept 17 6:30 PM ET ESPN3.com Tickets
    Sat, Oct 1 TBD Tickets
    Sat, Oct 8 TBD Tickets
    Sat, Oct 15 TBD Tickets
    Sat, Oct 22 TBD Tickets
    Sat, Oct 29 TBD Tickets
    Sat, Nov 5 TBD Tickets
    Sat, Nov 12 TBD Tickets
    Sat, Nov 19 TBD Tickets
    Sat, Nov 26 TBD Tickets

    Enjoy!

    X
    Log In Sign Up

    forgot?
    Log In Sign Up

    Forgot password?

    We'll email you a reset link.

    If you signed up using a 3rd party account like Facebook or Twitter, please login with it instead.

    Forgot password?

    Try another email?

    Almost done,

    By becoming a registered user, you are also agreeing to our Terms and confirming that you have read our Privacy Policy.

    Join Blogger So Dear

    You must be a member of Blogger So Dear to participate.

    We have our own Community Guidelines at Blogger So Dear. You should read them.

    Join Blogger So Dear

    You must be a member of Blogger So Dear to participate.

    We have our own Community Guidelines at Blogger So Dear. You should read them.

    Spinner.vc97ec6e

    Authenticating

    Great!

    Choose an available username to complete sign up.

    In order to provide our users with a better overall experience, we ask for more information from Facebook when using it to login so that we can learn more about our audience and provide you with the best possible experience. We do not store specific user data and the sharing of it is not required to login with Facebook.

    tracking_pixel_9347_tracker