I recently had the opportunity to have a nice talk with Zach Rosenfield, an analyst with AccuScore. AccuScore is a sports forecasting company, that uses computer simulations to attempt to correctly estimate how a game or season will play out. Obviously this is a very helpful thing for how crazy college football can get. Rosenfield is the analyst that deals mainly the NCAA Football. His analysis has been featured by Wall Street Journal, Yahoo!, USA Today, Los Angeles Times. He has also appeared on several radio programs on ESPN Radio, Sporting News Radio and others. I was able to ask him a view questions for the blog.
————————————————————————————————-
BCSWatch [BCSW]: Tell us a little about AccuScore and what you do there.
Zach Rosenfield[ZR]: AccuScore is a statistical game forecasting company that uses past performances and statistical history to forecast games. We built a simulation engine that simulates games one play at a time under real game conditions (110+ variables) and then re-simulate the game 10,000 times. We had 79% accuracy in College Football last year and 67% in the NFL, which made us champions of the ESPN NFL Expert Pick Em League.
AccuScore.com is a premium site that is designed to give the fan or sports investor every bit of information they might need. We forecast player statistics, as well as who will win, by how much and the total score. Last March, we went from being a free site to a paid site and it has been very successful.
Our current partners include ESPN, Yahoo!, Wall Street Journal, Sporting News, NBC Sports, Versus, NY Post, San Diego Union Tribune, as well as more than a dozen local television affiliates nationwide.
We are a small company which is a lot of fun because I get to wear many hats. I am an analyst for multiple sports, but College Football is my primary focus. I oversee AccuScore’s various media campaigns, our editorial content that gets licensed and/or syndicated to multi-media outlets, and work with Stephen Oh on the day-to-day business and strategies behind AccuScore.com.
BCSW: That sounds really cool! Does your simulation engine ever pick any upsets that really surprise you when you first see them, but then they actually happen? Can you give us an example?
ZR: Upsets are hard to predict with a computer, but we can identify games with upset potential based on the projections for a game. The biggest surprise we see in simulations is when the probability or score is not anywhere close to what you think it would be. We saw it last year with Oklahoma hosted Texas Tech. Everyone thought the game would be close, but we had Oklahoma winning 80% of the time. That was pretty significant for a one-loss team over an undefeated team that late in the season.
Another example is last year we had OU beating Texas by four points. In that simulation, Sam Bradford was not throwing an interception and was sacked twice. We re-simulated the game with Bradford throwing two interceptions and getting sacked four times and saw that jump Texas from a 45% underdog to a 52% favorite. Ironically, Bradford was sacked four times and did throw the two interceptions and Texas went on to win that game.
It is situations like that which get you excited about what you can extract from statistical analysis.
BCSW: How do you see the upcoming college football season playing out? Is it going to a crazy year like the past couple of seasons, or do you think it will be pretty clear who the top two teams are at the end of the season, like in 2005 with Texas and USC?
ZR: The 2009 Season looks to be the easiest season to dope out on paper since 2005. With the way the pre-season rankings shape up, the schedules of the elite teams and the players who play for those teams, we are confident that it will be the winner of the SEC versus the winner of the Big 12 South.
Based on PR, schedules and players, whoever comes out of the Big 12 round robin of OU-Texas-Oklahoma State should play for the National Championship. I also think whoever wins OU-TX will go undefeated. Our preliminary projections have Texas beating both OU and OSU.
The SEC is kind of boring this year. Florida doesn’t play Ole Miss or Alabama. They host Tennessee (beat down) and then go to LSU (who is an average team). Ole Miss has both Alabama and LSU at home and we are projecting the Rebels to win them both. So that sets up Florida and Ole Miss in the SEC Championship game.
BCSW: What do you think about the SEC’s dominance over college football? Do you think any other conferences can challenge them this year?
ZR: The SEC has been very fortunate the last few years. I think they are a solid conference, but I am not one of these guys who thinks the SEC is amazing. Last year, the SEC had two really good teams. The rest were in and around average. You can say the same thing for the SEC this year. In my opinion, Florida is best team in the SEC (and the country). They are followed by Ole Miss and maybe Alabama.
That being said, Ole Miss has question marks, but I think they will be answered and we think they will represent the SEC West. Expect Alabama to really struggle without John Parker Wilson. If you try to sell me on LSU, then you just lose all credibility. They are an 8-4 at best with a very poor game coach. Georgia and Tennessee are rebuilding, Auburn is a circus and South Carolina, Arkansas, Miss State and Vanderbilt are filler.
BCSW: Which conference would you say is the worst of the BCS conferences?
ZR: The Big East continues to disappoint. Eventually one of these years, one of the teams from that conference is going to back their way into the BCS Title Game and ruin it for everyone. At least two years ago, West Virginia was innovative and good. You cannot say the same thing for West Virginia this year. I think the conference is too small and lacks true name programs. It’s a sad state of affairs when our conference winner (Cincinnati) is projected to win the conference with a 5-2 record.
To further demonstrate this point, we took the projected third place finishers from the other five BCS conferences and ran individual simulations with them in the Big East. North Carolina (ACC), Oklahoma State (Big 12), Iowa (Big 10), Cal (Pac 10) and Alabama (SEC) would all win the Big East by a comfortable margin.
BCSW: What are couple games this year that you are especially looking forward to?
ZR: Oklahoma-Texas should be great. Both teams will be in the top 5, there is bad blood from last year, the amount of All-American’s on the field. Also, the winner has the inside track to the BCS National Championship Game. But it’s not the most important game in the Big 12.
The “most important” game in the Big 12 is Texas at Oklahoma State on October 31. This game has the opportunity to completely undo whatever happens in Dallas and you need to look at the domino effect it will have. OSU will most likely be 7-0 going into it and be in the top 5. If Texas beats OU and OSU, then they are in the BCS Championship Game. Period. But what if Texas beats OU and loses to OSU? Then you get relive all the fun of 2008 all over again. The OU-OSU game on Thanksgiving weekend would match a top-5 OSU team against a highly ranked one-loss OU team. The magnitude of that game, the national spotlight and the BCS points that come with it will propel the winner into the Big 12 Title game. OSU wins and they are in. OU wins and they will once again be the highest ranked BCS team in a 3-way tie breaker.
I am really excited to see Ole Miss host Alabama. I think Ole Miss is legit this year and think Alabama is a paper tiger. If Ole Miss wins this game, they will most likely be 10-0 when they play LSU. Both myself and the AccuScore computer think Ole Miss wins both of these games.
USC at Ohio State has huge national championship implications. If you have Big 12 or SEC rooting interests, you need to root for USC in this game. A USC victory virtually eliminates the Big 10 from the hunt. Ohio State will need to run the table to get in the BCS Title Game. If USC beats them, they will never recover in the polls. Ohio State is the team most likely to Penn State later in this year and will need to do so to eliminate Penn State (by the way, check out the horrible PSU schedule). It should be noted that I think USC has no chance at playing for it all. They have a new quarterback, three new linebackers and a horrible road schedule (@ OSU, @ Oregon, @ Cal, @ Notre Dame, @ Arizona State). If they are 80% favorites in all those games (and they won’t be anywhere near that), they would be 30% to undefeated.
BCSW: What do you think about the BCS? Is it good for college football?
ZR: First off, the BCS and the misery it’s created has been great for AccuScore. It makes our ability to simulate games even more valuable to the public. Additionally, the past two years we have created a “Simulated Playoff” that has been hosted (with great success) on ESPN.com.
However, beyond those plugs I have some real thoughts on this. Most importantly, Bowls are exhibition games. And if you think they matter, you are wrong. They don’t. They are high profile exhibition games. Excluding the National Championship Game, bowl games (and the practice leading up to them) are the only actions a team will take over the course of the year that are not in line with their goals.
Sure Bowl games are fun to watch, but don’t fool yourself into thinking they are big games. A big game is a game of consequence. Where the winner has something to gain and the loser suffers. Are Boise State and Utah truly better off because they won BCS games? Their programs might be higher profile, but how did that win help that team. Beyond a trophy, what would Alabama had gained if they won in the Sugar Bowl last year.
I am not a fan of the BCS or any system where subjectivity governs over championships. Anyone who says it is great for College Football is selling you a bill of goods. However, the BCS is the system and I am fully cooperative with it until it changes. People need stop furiously dialing talk radio and learn the rules that govern this sport. Fair or not, it is the system. That being said, I welcome the day where it is settled on the field.
BCSW: It is interesting that you bring up subjectivity in regards to the rankings as a bad thing. I do not necessarily believe that a little subjectivity is all that bad. In fact, I think it is one of the reasons that college football has such a passionate following.
ZR: College football would be just as fun if you took the subjectivity out of determining who gets to play for a championship. Any time you have more than one great team competing for a title, you have great subjective debates. However, to think that subjectivity is what fuels the passion of college football is something I don’t agree with it. People feel a kinship with their University, the reunion like tailgates, the pageantry of the sport and the competitive nature of the games. I am sure the same passion that is provoked from subjectivity could be re-purposed in other areas.
BCSW: If you were charged with creating a playoff for college football, how would your system work?
ZR: You have to take the subjectivity out of the game, or at least minimize it. All teams need to have a chance to play for the National Championship as they do in basketball. So, my proposal is simple. Make it a 10-team playoff with seven automatic bids going to the seven conference champs (SEC, ACC, Big 10, Big 12, Pac 10, Big East, MWC) and round it off with three at-large bids. The three at-large bids are determined by end of season rankings. The At-Large round would be #7 vs. #10 and #8 vs. #9 with the #7 and #8 having home games.
You then seed #1 through #6 by rankings, with the top four seeds getting home games in the next round (#1 vs. #8, #2 vs. #7, #3 vs. #6, #4 vs. #5). Home field advantage is a very big deal in College Football and the better teams deserve it.
After that round, go to neutral sites and you can use your bowl system. Although, do you really need to call the National Semi-Finals the “Sugar Bowl?” Don’t you think you would get just as big a game if had a catchy name like “The Final Four.” Nevertheless, you play the semifinals at two different sites around Christmas time. Make it a tradition. Two weeks later, play the Championship game.
BCSW: Is there a specific reason you chose the MWC as the conference you added?
ZR: I don’t think you can ignore what is going on in the Mountain West. Utah, BYU and TCU are schools that high school students consider playing for, rather than settle for. And that is a big deal. I also think the offensive lines and speed these teams field would be competitive across the entire college football spectrum. The same cannot be said for the MAC. The MAC is undersized and makes the Big 10 look explosive. I think it is a matter of safety on the field (just ask Hawaii). That being said, with three at-large entries, there would be room for a team like Ball State who made a strong run last year.
However, Ball State is the perfect example. The feather in their cap last year was beating Indiana. Please name me one other team in the top 30 who would place so much excitement on beating the Hoosiers. I am not saying a MAC team can’t rise up and win a game, but I don’t think you can carve out a guaranteed spot for them in a playoff. Watching Hawaii play Georgia was similar to watching the first half of the “Little Giants” play the Cowboys in that Rick Moranis movie.
———————————————————
A big thanks to Zach for taking the time to answer these questions. Be sure to check out AccuScore.com.