{"id":95,"date":"2011-01-10T16:45:12","date_gmt":"2011-01-10T22:45:12","guid":{"rendered":"http:\/\/godismyjudgeok.com\/DStats\/?p=95"},"modified":"2011-03-09T16:02:54","modified_gmt":"2011-03-09T22:02:54","slug":"auburn-oregon-title-preview","status":"publish","type":"post","link":"https:\/\/godismyjudgeok.com\/DStats\/2011\/ncaa-football\/auburn-oregon-title-preview\/","title":{"rendered":"Auburn-Oregon Title Preview"},"content":{"rendered":"

The study of football is much more tricky than that of basketball–there are more players on the field and fewer plays and games to study.\u00a0 The links in the sidebar list a few of the sites that have done good work studying football statistics.<\/p>\n

Since the players are hard to analyze, I have confined myself primarily to studying team-level play.\u00a0 I have developed two frameworks for analyzing and ranking teams.<\/p>\n

The first is a basic points-based approach.\u00a0 Who beat whom, and by how much?\u00a0 Of course, I didn’t just stick to the points.\u00a0 I analyzed game margins to determine how much to regress the games to be maximally predictive of the rest of the season.\u00a0 The approach I took was to reduce the points for large margins, but increase it for small margins.\u00a0 So a 1 point win is actually worth 6 points or so in my system, a 10 point win worth 10 points, a 20 point win worth about 16 and so on.\u00a0 Chase Stuart has used a similar approach in his Simple Rating System of NCAA Football for Pro Football Reference<\/a>.\u00a0 To generate my numbers, I fitted a cubic curve to predict the last 3 or 4 games of the year from the first 8 or 9, over the past several years.\u00a0 My “effective points” vs. actual margin are shown in the chart below:<\/p>\n

\"Effective<\/a>

Predictive Point Margin vs. Actual Margin<\/p><\/div>\n

To generate rankings of all of the teams in the NCAA, I simply took all of the games this year (thanks to College Football Reference<\/a> for the data) and recursively generated ratings by adjusting for opponent, then updating, and adjusting for opponent again using the new ratings until the results converged.\u00a0 This is the approach discussed as the “Simple Rating System”<\/a>.<\/p>\n

So now, to the results.\u00a0 These are the current ratings updated through the Sugar Bowl:<\/p>\n\n\n\n\n\t\n\n\t\n\t\n\t\n\t\n\t\n\t\n\t\n\t\n\t\n\t\n\t\n\t\n\t\n\t\n\t\n\t\n\t\n\t\n\t\n\t\n\t\n\t\n\t\n\t\n\t
Rank<\/th>Team<\/th>Points<\/th>SOS<\/th>\n<\/tr>\n<\/thead>\n
1<\/td>Oregon<\/td>28.8<\/td>4.3<\/td>\n<\/tr>\n
2<\/td>Stanford<\/td>26.3<\/td>6.6<\/td>\n<\/tr>\n
3<\/td>Boise State<\/td>25.6<\/td>-0.6<\/td>\n<\/tr>\n
4<\/td>Texas Christian<\/td>23.8<\/td>-1.0<\/td>\n<\/tr>\n
5<\/td>Alabama<\/td>21.7<\/td>4.5<\/td>\n<\/tr>\n
6<\/td>Auburn<\/td>21.0<\/td>5.1<\/td>\n<\/tr>\n
7<\/td>Ohio State<\/td>19.9<\/td>0.0<\/td>\n<\/tr>\n
8<\/td>Oklahoma<\/td>19.2<\/td>5.6<\/td>\n<\/tr>\n
9<\/td>Oklahoma State<\/td>17.4<\/td>2.6<\/td>\n<\/tr>\n
10<\/td>Wisconsin<\/td>16.9<\/td>0.5<\/td>\n<\/tr>\n
11<\/td>Arkansas<\/td>16.3<\/td>5.4<\/td>\n<\/tr>\n
12<\/td>Florida State<\/td>14.6<\/td>5.3<\/td>\n<\/tr>\n
13<\/td>Louisiana State<\/td>14.1<\/td>4.6<\/td>\n<\/tr>\n
14<\/td>Missouri<\/td>13.8<\/td>2.6<\/td>\n<\/tr>\n
15<\/td>Nevada<\/td>13.7<\/td>-3.7<\/td>\n<\/tr>\n
16<\/td>Virginia Tech<\/td>13.5<\/td>3.1<\/td>\n<\/tr>\n
17<\/td>Texas A&M<\/td>13.1<\/td>3.9<\/td>\n<\/tr>\n
18<\/td>South Carolina<\/td>12.6<\/td>6.9<\/td>\n<\/tr>\n
19<\/td>Southern California<\/td>12.6<\/td>8.6<\/td>\n<\/tr>\n
20<\/td>Utah<\/td>12.5<\/td>0.9<\/td>\n<\/tr>\n
21<\/td>Nebraska<\/td>12.3<\/td>1.8<\/td>\n<\/tr>\n
22<\/td>Arizona<\/td>12.0<\/td>8.2<\/td>\n<\/tr>\n
23<\/td>Arizona State<\/td>11.9<\/td>7.0<\/td>\n<\/tr>\n
24<\/td>North Carolina State<\/td>11.5<\/td>2.7<\/td>\n<\/tr>\n
25<\/td>Mississippi State<\/td>11.0<\/td>3.4<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n\n

Oregon blew people out, and so did Stanford.\u00a0 This points-based approach loves Oregon’s scoring, and penalizes Auburn for scraping out close wins.<\/p>\n

However…<\/p>\n

I also have another rating system.\u00a0 This other system looks at yards-per-play, interceptions, and fumbles, regressed appropriately.\u00a0 For this method, I got data from NCAA.org, which has all of the information you’ll ever need for each team.\u00a0 An example table of those I pulled is for Air Force’s Offense.<\/a><\/p>\n

So, how would one rank teams based on yards-per-play?\u00a0 Well, the basic rating would be simply to solve recursively for each team’s opponent-adjusted offensive and defensive yards-per-play.\u00a0 Then add in the value of each interception and fumble.<\/p>\n

I was a little more tricky.\u00a0 I ran fudged split-half correlations for yards-per-play, int\/play, and fumble\/play.\u00a0 I say fudged because I split it into halves of the season, rather than into, say, even plays and odd plays.\u00a0 I then took that correlation value and generated a regression rate base on it, following TangoTiger’s approach for baseball.\u00a0 See this headache-inducing post of his.<\/a><\/p>\n

The regression rates I got:<\/p>\n\n<\/col>\n\n\n\n
Off YPP<\/th>\nO Int\/Play<\/th>\nO Fum\/Play<\/th>\nDef YPP<\/th>\nD Int\/Play<\/th>\nD Fum\/Play<\/th>\n<\/tr>\n
205<\/td>\n1553<\/td>\n4638<\/td>\n182<\/td>\n1548<\/td>\n30469<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n

Okay, so what does that mean? Well, if you have the numbers of\u00a0 NCAA-average plays to add<\/strong> to regress properly.\u00a0 If I’ve seen 700 plays of Auburn’s, after I generate Auburn’s opposition-adjusted yards-per-play, I then add 205 plays of average offensive yards-per-play to their numbers to predict Auburn’s future<\/strong> yards-per-play.\u00a0 It’s just a mathematically-more-rigorous form of regression-to-the-mean.<\/p>\n

Next, one has to translate the interceptions and fumbles into yards.\u00a0 Interceptions are worth about 45 yards, I found, and fumbles about 15 (I’m using total fumbles, not fumbles lost, so that would be about 30 yards per fumble lost).\u00a0 Finally, the adjusted final yards-per-play ratings must be translated to points.\u00a0 I’m using 7.75 points per yard-per-play advantage.<\/p>\n

So, now I have run this yards-per-play rating system, and what do the numbers show?<\/p>\n\n\n\n\n\t\n\n\t\n\t\n\t\n\t\n\t\n\t\n\t\n\t\n\t\n\t\n\t\n\t\n\t\n\t\n\t\n\t\n\t\n\t\n\t\n\t\n\t\n\t\n\t\n\t\n\t
Rank<\/th>Team<\/th>YPP Mar<\/th>Points<\/th>Off YPP<\/th>O Int<\/th>O Fum<\/th>Def YPP<\/th>D Int<\/th>D Fum<\/th>\n<\/tr>\n<\/thead>\n
1<\/td>Alabama<\/td>2.89<\/td>22.4<\/td>7.0<\/td>1.1%<\/td>2.4%<\/td>4.4<\/td>1.8%<\/td>2.4%<\/td>\n<\/tr>\n
2<\/td>Boise St.<\/td>2.82<\/td>21.9<\/td>6.8<\/td>1.2%<\/td>2.4%<\/td>4.2<\/td>1.5%<\/td>2.5%<\/td>\n<\/tr>\n
3<\/td>Auburn<\/td>2.65<\/td>20.6<\/td>7.4<\/td>1.2%<\/td>2.3%<\/td>4.9<\/td>1.4%<\/td>2.4%<\/td>\n<\/tr>\n
4<\/td>Arkansas<\/td>2.47<\/td>19.1<\/td>7.3<\/td>1.5%<\/td>2.4%<\/td>4.9<\/td>1.6%<\/td>2.4%<\/td>\n<\/tr>\n
5<\/td>Oregon<\/td>2.37<\/td>18.3<\/td>6.7<\/td>1.1%<\/td>2.4%<\/td>4.6<\/td>1.8%<\/td>2.4%<\/td>\n<\/tr>\n
6<\/td>Stanford<\/td>1.88<\/td>14.6<\/td>6.6<\/td>1.2%<\/td>2.2%<\/td>5.1<\/td>1.9%<\/td>2.5%<\/td>\n<\/tr>\n
7<\/td>TCU<\/td>1.86<\/td>14.4<\/td>6.3<\/td>1.3%<\/td>2.3%<\/td>4.6<\/td>1.5%<\/td>2.4%<\/td>\n<\/tr>\n
8<\/td>Ohio St.<\/td>1.81<\/td>14.0<\/td>6.2<\/td>1.6%<\/td>2.2%<\/td>4.5<\/td>1.7%<\/td>2.5%<\/td>\n<\/tr>\n
9<\/td>South Carolina<\/td>1.74<\/td>13.5<\/td>6.4<\/td>1.5%<\/td>2.4%<\/td>4.7<\/td>1.4%<\/td>2.4%<\/td>\n<\/tr>\n
10<\/td>Oklahoma St.<\/td>1.72<\/td>13.3<\/td>6.5<\/td>1.4%<\/td>2.4%<\/td>4.9<\/td>1.6%<\/td>2.4%<\/td>\n<\/tr>\n
11<\/td>Georgia<\/td>1.69<\/td>13.1<\/td>6.4<\/td>1.2%<\/td>2.4%<\/td>4.9<\/td>1.6%<\/td>2.4%<\/td>\n<\/tr>\n
12<\/td>Nebraska<\/td>1.62<\/td>12.6<\/td>6.2<\/td>1.1%<\/td>2.8%<\/td>4.8<\/td>1.7%<\/td>2.4%<\/td>\n<\/tr>\n
13<\/td>Arizona<\/td>1.55<\/td>12.0<\/td>6.2<\/td>1.2%<\/td>2.3%<\/td>4.8<\/td>1.5%<\/td>2.4%<\/td>\n<\/tr>\n
14<\/td>LSU<\/td>1.54<\/td>11.9<\/td>5.8<\/td>1.3%<\/td>2.5%<\/td>4.5<\/td>1.8%<\/td>2.4%<\/td>\n<\/tr>\n
15<\/td>Iowa<\/td>1.41<\/td>10.9<\/td>5.9<\/td>1.2%<\/td>2.3%<\/td>4.8<\/td>1.8%<\/td>2.4%<\/td>\n<\/tr>\n
16<\/td>Florida St.<\/td>1.35<\/td>10.4<\/td>6.2<\/td>1.2%<\/td>2.5%<\/td>5.0<\/td>1.5%<\/td>2.4%<\/td>\n<\/tr>\n
17<\/td>Hawaii<\/td>1.31<\/td>10.1<\/td>6.7<\/td>1.5%<\/td>2.4%<\/td>5.5<\/td>1.8%<\/td>2.4%<\/td>\n<\/tr>\n
18<\/td>Virginia Tech<\/td>1.28<\/td>9.9<\/td>6.3<\/td>1.1%<\/td>2.3%<\/td>5.4<\/td>1.9%<\/td>2.4%<\/td>\n<\/tr>\n
19<\/td>Texas A&M<\/td>1.19<\/td>9.2<\/td>5.7<\/td>1.3%<\/td>2.5%<\/td>4.6<\/td>1.6%<\/td>2.4%<\/td>\n<\/tr>\n
20<\/td>Florida<\/td>1.18<\/td>9.1<\/td>5.5<\/td>1.3%<\/td>2.7%<\/td>4.5<\/td>2.0%<\/td>2.4%<\/td>\n<\/tr>\n
21<\/td>Miami (FL)<\/td>1.13<\/td>8.7<\/td>6.2<\/td>1.9%<\/td>2.4%<\/td>4.9<\/td>1.6%<\/td>2.4%<\/td>\n<\/tr>\n
22<\/td>Mississippi St.<\/td>1.11<\/td>8.6<\/td>5.7<\/td>1.4%<\/td>2.4%<\/td>4.7<\/td>1.6%<\/td>2.4%<\/td>\n<\/tr>\n
23<\/td>North Carolina<\/td>1.10<\/td>8.5<\/td>5.9<\/td>1.3%<\/td>2.4%<\/td>5.0<\/td>1.7%<\/td>2.4%<\/td>\n<\/tr>\n
24<\/td>Oklahoma<\/td>1.07<\/td>8.3<\/td>5.8<\/td>1.2%<\/td>2.3%<\/td>5.0<\/td>1.7%<\/td>2.4%<\/td>\n<\/tr>\n
25<\/td>Wisconsin<\/td>1.06<\/td>8.2<\/td>6.3<\/td>1.2%<\/td>2.3%<\/td>5.4<\/td>1.6%<\/td>2.4%<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n\n

Wow, that looks a little different than the points-based ranking!\u00a0 Now the SEC dominates the list, rather than the Pac 10.<\/p>\n

Comparing Oregon and Auburn:\u00a0 Yards-Per-Play has Auburn favored by 2 1\/2; SRS-Points has Oregon favored by 7 1\/2.\u00a0 Which is right?\u00a0 I don’t know.\u00a0 Other points-based rankings typically have Oregon ahead, like I do here.\u00a0 Sagarin<\/a> does; Team Rankings<\/a> does; so do the Massey Ratings<\/a>.\u00a0 Other play-by-play rankings (F\/+<\/a>) like Auburn better, as my Yards-Per-Play ratings do.<\/p>\n

Either way, this promises to be an exciting and balanced BCS championship game.\u00a0 The two best undefeated teams are here.\u00a0 Of course, Boise St. is probably their match, and Alabama (how DID they lose 3 games?) is also right there.\u00a0 But those 2 lost, so they are out.<\/p>\n

Oregon-Auburn, here we come!<\/p>\n

Read the full NCAA Pre-Title Game Ratings here<\/a>, showing both SRS-Points and Yards-Per-Play. P.S. They aren’t sortable, but they are full of pretty colors!<\/p>\n","protected":false},"excerpt":{"rendered":"

The study of football is much more tricky than that of basketball–there are more players on the field and fewer plays and games to study.\u00a0 The links in the sidebar list a few of the sites that have done good work studying football statistics. Since the players are hard to analyze, I have confined myself … [Read more…]<\/a><\/span><\/p>\n","protected":false},"author":3,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"ngg_post_thumbnail":0,"footnotes":""},"categories":[16],"tags":[59,23,22,24],"_links":{"self":[{"href":"https:\/\/godismyjudgeok.com\/DStats\/wp-json\/wp\/v2\/posts\/95"}],"collection":[{"href":"https:\/\/godismyjudgeok.com\/DStats\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/godismyjudgeok.com\/DStats\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/godismyjudgeok.com\/DStats\/wp-json\/wp\/v2\/users\/3"}],"replies":[{"embeddable":true,"href":"https:\/\/godismyjudgeok.com\/DStats\/wp-json\/wp\/v2\/comments?post=95"}],"version-history":[{"count":11,"href":"https:\/\/godismyjudgeok.com\/DStats\/wp-json\/wp\/v2\/posts\/95\/revisions"}],"predecessor-version":[{"id":349,"href":"https:\/\/godismyjudgeok.com\/DStats\/wp-json\/wp\/v2\/posts\/95\/revisions\/349"}],"wp:attachment":[{"href":"https:\/\/godismyjudgeok.com\/DStats\/wp-json\/wp\/v2\/media?parent=95"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/godismyjudgeok.com\/DStats\/wp-json\/wp\/v2\/categories?post=95"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/godismyjudgeok.com\/DStats\/wp-json\/wp\/v2\/tags?post=95"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}