Tuesday, May 26, 2009

A statistician looks at the MWC

A football stats blogger called "LeftyLoon" has weighed in on the MWC. Using a stat called "SDPI", which he describes as a "look at how far above or below average (by standard deviations) a specific team is relative to their conference brethren," this blogger compared each MWC team's offensive output in yards. The stat correctly predicted the conference standings, except it put TCU ahead of Utah.

"How can this be? For starters, an eight game conference season is an extremely small sample size. If we replay the 2008 Mountain West season over a million times or so, Utah will probably win a rather high percentage of conference
titles. However, TCU will probably win more. . . . Secondly, for anyone with access to CBS College Sports, who actually watched the TCU/Utah battle that ultimately decided the conference title, you'd be hard-pressed to argue Utah was the better team. The Horned Frogs outgained the Utes by 141 yards and averaged more than a yard more per play. However, TCU kicker Ross Evans missed two makeable field goals of 26 and 35 yards (he would only miss two other field goals all season), Utah kicker Louie Sakoda made kicks of 49 and 39 yards, the Horned Frogs threw two interceptions, the Utes fell on both of their own fumbles, and it still took a touchdown pass with 48 seconds left to for the Utes to emerge victorious at home. Credit Utah, for you know, actually winning the game, which is why its usually played in the first place, but just remember they were extremely fortunate to do so."

That's a bold analysis-- and boldly wrong. I wasn't just watching the game on TV, but shlepped my then-east coast buns out to SLC, and by the time Utah finally managed to squeeze a touchdown out of the Frogs' winded D, I don't know how anyone could say one team was better than the other. The Utes' kicker clearly outclassed the Frogs' Evans, but aside from that disparity, the two teams looked very well matched to me.

"Looking ahead to next season, the prohibitive favorite should be...BYU. The number one offense in the conference last season returns a senior quarterback (Max Hall) and a 1000-yard rusher (Harvey Unga). The Cougars do lose their top receiver, Austin Collie, who split with a year of eligibility remaining, but returning tight end Dennis Pitta also had 1000 yards receiving last season. The main concern for the offense will be replacing four starters along the line. Even if the offense dips a little, it should remain one of the best in the conference."
Hard to argue there. I've already gone on record predicting BYU will win the conference; I love it when math backs me up. LeftyLoon is high on New Mexico and Wyoming. While agreeing that New Mexico looks like it's headed for a good year, I have to ask: how does one not look at Wyoming and weep for its prospects in '09?

"The Cowboys turnover margin in conference play last season was -22, easily good for last in the league. . . . There's a very good chance this turnover margin will improve significantly in 2009, . . . Fumble recovery is totally random, so the Cowboys should see their recovery perentage trend back toward 50% . . . . In 2008, Wyoming scored two non-offensive touchdowns. They were both interception returns and this feat is not entirley remarkable. However, opponents scored an amazing nine non-offensive touchdowns against the Cowboys (a punt return, two kickoff returns, four interception returns, and two fumble returns). . . . You can rest assured the Cowboys won't be seven touchdowns in the hole on non-offensive touchdowns in 2009."
This is lazy predicting. The 'Boys in Laramie are installing a 100% new offense, with players who just aren't cut out for it. Ask SMU, which is a year ahead of a nearly identical change of scheme: turnovers will plague Wyoming in its first Year of the Spread.

LeftyLoon doesn't like Utah's chances of repeating.
"The Utes were a very good team in 2008, but a great team they were not. Remember, they survived the worst Michigan team perhaps in history by two points, they scored a touchdown with under a minute to play to edge Air Force by seven, they needed a furious rally including a recovered onside kick to knock off Oregon State by three at home, and of course the aforementioned good fortune to get past TCU. As said before, Utah should be congratulated for winning all their games, but lady luck is a fickle mistress and she will probably not be as kind to the Utes in 2009. Add to that the loss of starting quarterback Brian Johnson and a trio of senior receivers and it certainly seems the offense will experience growing pains in 2009. The defense returns eight starters, but does lose perhaps its two best players in defensive end and quarterback nightmare Paul Kruger and corner Sean Smith. Both were second round picks in last year's draft and will be sorely missed. The Utes must also replace All-American kicker Louie Sakoda who made 22 of his 24 field goals last season."
By now, I'm wondering if LeftyLoon doesn't bleed Cougar-blue. The '08 Utes simply have to rank as an outstanding squad, especially when its chips were down. The blogger gets part of his "sell" prediction correct, though: the reason Utah won't match last season's highs in 2009 is personel. Too many very good players graduated.

LeftyLoon says says TCU and SDSU. We're well familiar with TCU here, but the blogger's take on SDSU is interesting.
"It took Hoke half a decade to even match the win total the team achieved the year before his arrival. . . I'm not here to pass judgement on Hoke's coaching acumen just yet, but San Diego State was clearly the worst team in the league last season. Expecting a coach who struggled to win in the MAC to come in and win immediately in a league with three very strong mid-major programs is an extreme leap of faith. San Diego State has some building blocks, such as . . . quarterback Ryan Lindley. . . If the program can double last season's win total, 2009 should be considered a success."
I'm tempted to say he's underselling the value of those Aztec building blocks, but it's probably prudent not to. Yet.

More from LeftyLoon in the next post.