Every year, EA Sports produces several sports software packages including one for NCAA Football. This is a video game for the different game console formats, but it is also something else--a game simulator with an amazing track record. Since 2003 I've been following their projections and noticed something rather remarkable. In its projections before the college football season starts, NCAA Football has accurately predicted every Boise State regular season game. That's 36 in a row.
These projections are made by simulating a matchup several times to show the range of possible outcomes. The suggested minimum number is ten times, although the more times the simulation is done, the more accurate the results will be. The number of times that each team has won is then totaled up and you have the projection for the matchup. By dividing the team wins by the total simulations, you come up with the chance (in percentage) that each team will win. You can also get an average score for the game.
The only Boise State game EA missed (including bowl games) was the 2004 Liberty Bowl vs. Louisville, in which they projected a Boise State win by three. And if you think about it, a Boise State touchdown on their final drive would have produced a Bronco win by....three points. Perhaps that was what was meant to happen.
But disregarding all that, you have the one incorrect game in three years--still a pretty impressive record. I should also point out that the simulation has not just been accurate in good years for Boise State. Back in Boise State's early years in Division I-A, I ran simulations and they did not match up favorably at all with most of the teams they played. So the accuracy of their predictions is certainly not a new phenomenon. Knowing this, one certainly cannot discount NCAA Football given that accuracy. So what does the software forecast for the 2006 Boise State season? To give as accurate a picture as possible, I simulated the season 60 times. Does the software also accurately predict other schools? I don't know because I've never paid attention to that. Maybe it's just accurate as far as Boise State is concerned. But for fun, I ran ten season simulations of the other schools in the Western Athletic Conference.
Here are the simulation results for this year:
San Jose State is projected to have a 2-10 season. The software predicts a home win against Utah State, but little else. San Jose entertains Boise State, Fresno State, Louisiana Tech and Utah State while they travel to play Nevada, Hawai'i, New Mexico State and Idaho. In ten simulations, the Spartans did not win a game against Nevada, Hawai'i or Fresno State. In 60 simulations, they only won once against Boise State. They only won one of ten matchups against Louisiana Tech and two of ten against Idaho and New Mexico State. San Jose won six of ten simulations against Utah State. That adds up to a 1-7 record in the Western Athletic Conference. The Spartans have no chance against Washington, Stanford and San Diego State (going 0-10 against each), but will easily put away Cal Poly at home (10-0).
Utah State is not predicted to win any of its games, although Idaho won just six of the ten games. Utah State is projected to lose vs. Utah (0-10), BYU (1-9) and Arkansas (1-9) out of conference. In the WAC, the Aggies are forecast to lose to Boise State, (2 wins against 58), Nevada (0-10), Fresno State (2-8), Hawai'i (3-7), New Mexico State (3-7), San Jose State (4-6), Idaho (4-6) and Wyoming (4-6). Overall, that's 0-8 in WAC play and 0-12 overall.
New Mexico State had a dismal first year under Coach Hal Mumme last year. Will the Aggies improve this fall? Yes, but not much, says E/A Sports. New Mexico State should easily beat Southeastern (10-0) and Texas Southern (10-0), and are also forecast to win games against San Jose State (8-2) and Utah State (7-3). Losses are expected against Boise State (1 win against 59), UTEP (1-9) and New Mexico (4-6) in non-conference games and to Nevada (2-8), Idaho (2-8), Fresno State (3-7), Louisiana Tech (4-6) and Hawai'i (4-6). That makes New Mexico State 2-6 in the WAC and 4-8 overall.
Hawai'i showed last year they have the talent to compete near the top this fall, giving Boise State a tremendous battle before finally succumbing at home. Does the software bear this out--will Hawai'i be a force this year? Absolutely. While the Rainbow Warriors have a difficult non-conference schedule, playing at Alabama and hosting Purdue, Oregon State and UNLV, they do get Nevada, Louisiana Tech, Idaho and San Jose State at home. Coach June Jones' team must travel to play Boise State, Fresno State, New Mexico State and Utah State. They are predicted to lose to all of their non-conference foes this year--Alabama (0-10), Oregon State (1-9), UNLV (3-7) and Purdue (4-6). EA predicts a win over Nevada (6-4), New Mexico State (6-4), Utah State (7-3), Louisiana Tech (7-3), (Idaho (10-0) and San Jose State (10-0) and losses to Fresno State (4-6) and Boise State (6 wins to 54). Thus, although their non-conference performance would not do the league very proud, Hawai'i is still projected to have at least six conference wins, and seven if they can upset the Bulldogs in Fresno. In summary, EA is forecasting Hawai'i to go 6-2 in the WAC and 6-6 overall.
Fresno State--I must preface this by saying I hope EA's portrayal of Fresno State this year is not right. Boise State very much needs a strong Fresno State and the WAC itself is not served well by having a less than stellar Bulldog team. They have made a big name for themselves on the national radar and they need to stay right there to continue to improve the way that the WAC is thought of. That said, here are the scary numbers for Bulldog fans. Fresno plays Oregon and Colorado State at home and Washington and LSU on the road. That's one of the toughest non-conference schedules in the nation. In the WAC, Fresno State travels to Boise State, Utah State, Louisiana Tech and San Jose State and hosts Nevada, New Mexico State, Idaho and Hawai'i. The Bulldogs are projected to lose three of the non-conference battles--Oregon (0-10), LSU (0-10) and Colorado State (4-6). In the WAC, Fresno's is barely favored over Hawai'i (6 wins to 4), while the Bulldogs are forecast to defeat San Jose State (10-0), Utah State (8-2), New Mexico State (8-2), Idaho (8-2) and Louisiana Tech (6-4). They are predicted to lose to Nevada (4-6) and they won just five of 60 simulations against Boise State. Thus, Pat Hill's club is forecast to have a 6-2 league mark and 7-5 overall.
Nevada--As defending co-champions of the WAC, you would expect Chris Ault's team to be tough again. That seems to be the case. The Pack hosts Colorado State and Northwestern in Reno and plays Dirk Koetter and Arizona State on the road as well as an interstate matchup against UNLV. In league, Nevada hosts Boise State, San Jose State, New Mexico State and Utah State. EA Sports projects losses to Arizona State (1-9) and Colorado State (2-8) and narrow victories over Northwestern and UNLV. In the WAC, Nevada should easily beat San Jose State (10-0), Utah State (10-0), Louisiana Tech (10-0) and New Mexico State (8-2). The Pack is favored on the road against both Fresno State (6-4) and Hawai'i (6-4) but win only 11 of 60 simulations at home vs. Boise State. The final prediction for Nevada--a WAC record of 7-1 and 9-3 overall.
Louisiana Tech--The Bulldogs are often a difficult team to peg for most preseason national publications since their recruiting area is so different from the rest of the league. This fall, Tech plays at Nebraska, Texas A & M and Clemson and hosts North Texas State. They will travel to play Boise State, Hawai'i, San Jose State and New Mexico State while hosting Nevada, Fresno State, Idaho and Utah State. That's a pretty even league schedule--not too tough and not easy. EA Sports forecasts La Tech to defeat San Jose State (9-1), Utah State (7-3), Idaho (6-4) and New Mexico State (6-4) while losing to Boise State (4 wins in the 60 simulations), Nevada (0-10), Hawai'i (3-7) and Fresno State (4-6). For the non-conference schedule, they seem to be heeding the league's call to "Play Up", even if it does mean getting beat badly. Tech has a 10 percent chance of beating Nebraska, A & M and Clemson, winning one out of ten simulations against each. The Bulldogs are also forecast to lose to North Texas State (4-6). The summary would be a 4-4 WAC record and 4-7 overall.
Idaho--How much will a seasoned coach make a difference? Yes, a couple of games, according to EA Sports. Idaho has an unrealistic schedule of playing at Michigan State, Washington State and Oregon State with a lone home game against Division I-AA Idaho State. In league play, the Vandals host Nevada, San Jose State, Boise State and New Mexico State and travel to Utah State, Louisiana Tech, Hawai'i and Fresno State. NCAA Football forecasts Idaho to be beaten badly by Michigan State (0-10) and Oregon State (0-10), and in many of those games, season-ending injuries to Moscow players are suffered. Idaho is also expected to lose to Washington State (3-7) while beating Idaho State (8-2). In conference tilts, the Vandals will beat San Jose State (8-2) and New Mexico State (8-2) and they should also win against Utah State (6-4). But, Idaho is projected to lose to Boise State (1 win out of 60 simulations), Nevada (0-10), Hawai'i (0-10), Fresno State (2-8) and Louisiana Tech (4-6). This amounts to a 3-5 WAC record and 4-8 overall.
Boise State--A new coach and more returning starters than any other Division I-A college football team in the nation. How does that all blend in? The Broncos host Sacramento State and Oregon State and travel to play Mountain West foes Wyoming and Utah. In league, this is a very favorable schedule that has the Broncos hosting Hawai'i, Louisiana Tech, Fresno State and Utah State and traveling to Nevada, New Mexico State, Idaho and San Jose State. The software predicts...
Boise State will defeat Sacramento State (60 simulations to 0), San Jose State (60-0), New Mexico State (59-1), Idaho (59-1), Utah State (58-2), Wyoming (57-3), Louisiana Tech (56-4), Fresno State (55-5) and Hawai'i (54-6). There are three opponents that NCAA Football gives more than a 12% chance of winning. Boise State is forecast to defeat Nevada, winning 49 of the 60 simulations, and Oregon State, winning 37 of the 60.
By far the toughest Bronco game that I've ever seen for the software to predict is this year's Boise State-Utah game. The only game which was tougher to call was the Louisville game. No matter how many times the game is simulated, I could get no separation between the two. After 60 simulations, Boise State won 30 and Utah 30. Essentially what this says is that each team has exactly a 50% chance of winning and the software predicts a tie. Of course that can't happen, so the tiebreaker is average score. When this is calculated, Boise State scores an average of 27 while Utah scores an average of 25.
Here are the average scores after 60 simulations:
Boise State 47, Sacramento State 5
Boise State 28, Oregon State 24
Boise State 41, Wyoming 12
Boise State 43, Hawai'i 24
Boise State 27, Utah 25
Boise State 39, Louisiana Tech 15
Boise State 44, New Mexico State 16
Boise State 39, Idaho 14
Boise State 38, Fresno State 16
Boise State 48, San Jose State 8
Boise State 44, Utah State 10
Boise State 37, Nevada 17
Is the software predicting a 12-0 season? Not exactly. Remember the Utah game came out with 30 wins to each side. Better to say that it's a toss-up and either team can come out on top. The overall results are too close to definitively predict one team or the other.
The season simulations seemed to favor an 11-1 record with 10-2 also a good possibility. The Broncos were 12-0 11 different times, 25 times they were 11-1, 16 times they were 10-2, four times they had a 9-3 record and 4 times they were 8-4.
The difference between 10-2 and 12-0 is injuries--if Boise State can stay away from them, the software is saying that there is an 18 percent chance that the Broncos will go undefeated. It also says there is a 42 percent chance that Boise State will finish the year 11-1, although it does not favor any of the twelve opponents over Boise State.
It will once again be interesting to watch the season unfold and compare these results to the actual ones, especially where the other WAC games are concerned. Can EA Sports continue its amazing record of accurately predicting Boise State football games? That's the million-dollar question!