Friday, September 25, 2009

Illinois at Ohio State, Preview

Four weeks into the season, I'm still not sure what we really know about this year's Illinois football team. In Week 1, high expectations were apparently scuttled in a 37-9 loss to Missouri. Since then, Illinois has defeated FCS opponent Illinois State 45-17 and taken a week off while other teams rested. Injuries were a big part of the meltdown against Missouri, and despite two weeks off to rest, the injuries continue to mount.

With the injuries and blowout loss against their only real competition, there is an understandable feeling that the Illini may be en route to a lost season. I could see things starting to snowball, but this is still a team with a lot of talent. If Arrelious Benn is healthy, they will move the ball down the field. The 9 point effort against Missouri was not good, but Illinois also ranked 20th in the country in offensive yards per play last year. They are not suddenly going to be a bad offensive team and racking up 548 yards against Illinois State was a good sign, even if it wasn't against the best competition. Of course, the lingering concern with a Juice Williams-led team is that nice yardage numbers will be undone by crippling turnovers and although Illinois wasn't great moving the ball against Missouri, the two turnovers made things a lot worse.

The main point of concern is what may be a terrible secondary. The secondary simply wasn't very good last year even with first-round draft pick Vontae Davis; and without him the early returns have been poor, to say the least, as a QB making his first career start for Missouri threw for 319 yards and 3 TDs and even Illinois State threw for 354 yards against the Illini (albeit on 52 pass attempts).

As such, it's not surprising that Illinois is a 14 point underdog going on the road to play Ohio State. The good news is that betting on Illinois would make for a good contrarian pick. Bettors are all over Ohio State and yet the line isn't moving up, suggesting that the Illini have a good shot at covering the spread.

Illinois has played well against Ohio State in recent years, but I am going into this one with low expectations. My best guess is that Illinois will have some success moving the ball through the air, but will stall out on key drives, and that Terrelle Pryor, lately maligned for his passing, will still be able to shred the Illinois secondary. Still, I think Illinois will be up for this game and keep things competitive, proving that this may not be a lost season after all, and gaining some momentum for next week's game against Penn State.

Pick: Ohio State 27, Illinois 17

Wednesday, September 23, 2009

Projecting Illinois Basketball, Part 2

About a month ago, I ran a post projecting how Illinois basketball might rank in the Pomeroy Ratings in the upcoming season. To my delight, Vegas Watch has updated his regression model to take into account incoming recruit ratings. This addition is relevant to Illinois given that we are bringing in a touted recruiting class this year.

First, a table of past projections based on this regression model and how the Illini have actually performed:





















YearPrev Pyth% ReturnProjectionActualDiff
2005.95392.0.976.982.006
2006.98240.9.828.940.112
2007.94065.7.887.932.045
2008.93257.9.896.900.004
2009.90067.8.875.911.036
2010.91157.0.865??????




Key:
Prev Pyth = Previous season's Pythagorean Win Percentage. Check out Ken Pomeroy's website for more info.
% Return = Percentage of previous season's minutes returning to the team.


As I noted last time, Illinois has outperformed its projection every single season. I credited this to the coaching of Bruce Weber, and he actually looks even better this time around, because Illinois' projections are generally lower after taking recruiting ratings into account.

Of course, recruiting is part of the coach's job too, and Weber has in the past not really delivered the way a coach at a top program should. Illinois has brought in at least one four-star prospect every year listed in the table (counting Calvin Brock as a recruit for 2006 because he took a redshirt year, and adding Alex Legion as a transfer last year), but no five-star prospects, and only one year with more than one four-star prospect (2008 - with Demetri McCamey, Mike Tisdale, and Rodney Alexander - not coincidentally, if you sub in Mike Davis for Alexander, the foundation of the current team). Thus Illini fans have become very excited about this year and next year's highly touted recruiting classes as indicators of a sea change in Illinois recruiting.

That said, by the Scout.com ratings, Illinois is actually only bringing in two four-star prospects this year (DJ Richardson and Brandon Paul), as Tyler Griffey and Joseph Bertrand rate as three-star prospects. Thus Illinois has a comparably low projection of .865 for the coming season. I have seen Griffey rated as a four-star prospect by other sources, so if I include him at four-stars, the team's projection would increase to .884. Either way, this projection is lower than that of the previous two seasons.

The good news is that even if Illinois posts a relatively disappointing Pomeroy rating of .865 this year, likely leaving them on the NCAA bubble, they project to bring back 90%+ of minutes and also add a 5-star and two 4-star recruits for the following season. With hypothetically, 92% of minutes returning in 2011, Illinois projects at .946 for next year, which would be by far their highest projection since 2005, when they were projected to be one of the premier national championship contenders. And that's a conservative estimate, considering that the team has outperformed their projection by an average of 2.6% each year under Bruce Weber, even throwing out the 2006 projection which was way too low. Obviously I want to be optimistic about my team, but although Illinois may appear to tread water this season, they will be stacked and likely a Top 10 team in 2011. Now that's what better recruiting can do for you.

Tuesday, September 22, 2009

In Defense of Jim Tressel

I'm a little late on this post but I figured now is as a good time as any to address coach expectations at the big programs. I just wish I had written this last week before USC lost to Washington.

After Ohio State blew a winnable game against USC last Saturday, there was a lot of criticism of Jim Tressel. I expect irrational fan outbursts after a disappointing loss but even some of my favorite bloggers were rather quick to pick up their pitchforks and torches. Now I'm not saying that Tressel doesn't deserve criticism for the loss. Tressel was conservative when he should have been aggressive (not going for it on 4th and 1 at the goal line, punting to USC from the USC 37 prior to USC's game-winning TD drive) and aggressive when he should have been conservative (passing near the end of the first half), and although Chris Brown's sentiment was extreme, his analysis was spot-on.

All that said, what Jim Tressel has done throughout his tenure at Ohio State is remarkable. He is 85-20 in 8 years and has won or tied for 5 Big 10 titles in that span. And oh yeah, he won a national title too. The problem for Tressel is that he is a victim of his own success. Because OSU has lost a lot of high-profile games, too many are quick to dismiss the fact that Tressel never loses the little games, barring a home loss to Illinois in 2007. And as witnessed by USC's seemingly annual loss in a game in which they are heavily favored, this feat is not as easy as it might seem. We like to think that the favorites should always win, but for various reasons, they don't.

The best coaching comparison for Tressel is Bob Stoops. When Bob Stoops took over in Oklahoma, the team had gone just 12-22 in the previous 3 seasons under John Blake. Yet in just his second year, Stoops won a national title. But with some recent bowl losses and two national championship game defeats including one embarrassing blowout, a la Tressel, Stoops has been subject to a lot of undue criticism while not properly being praised for getting his teams into the position to win national championships in the first place. If it was that easy to win at Oklahoma, why did John Blake fail so miserably? The same questions could be asked of Ron Zook at Florida, Paul Hackett at USC, John Mackovic at Texas, or John Cooper at Ohio State, for while these coaches experienced varying degrees of success, they never made it look as easy as their successors have.

My ultimate point is this: by any standards of success, what Tressel has done this decade ranks only behind Pete Carroll at USC and Urban Meyer at Florida, with Mack Brown at Texas and Bob Stoops at Oklahoma as his peers. I'm not saying that Tressel should be above criticism. But his accomplishments over the last eight years are remarkable and should be seen in that light, whatever disappointments there have been along the way.

Monday, September 21, 2009

College Football Pick 'em Review - Week 3

Week 3 turned out to be an outstanding one for me on College Football Pick 'em. Or was it? First, a review (note that as I cautioned, my picks did change somewhat from the initial posting):

Texas (10), W 34-24
California (9), W 35-21
Auburn (8), W 41-30
Clemson (7), W 25-7
Virginia Tech (6), W 16-15
North Carolina (5), W 31-17
Iowa (4), W 27-17
Cincinnati (3), W 28-18
Arkansas (2), L 41-52
Utah (1), L 24-31

Not only did I score 52 out of a possible 55 points, well surpassing my 39 and 45 point totals from Weeks 1 and 2, I also look very smart for getting my top 8 games right and only my least-confident games wrong. I also moved up from the 72nd to the 87th percentile.

I wound up making some alterations to my picks because I simulated how I would perform against a 'typical' public entry, as determined by the pick statistics provided by ESPN. I ran 100 random simulations for the outcome of the weekend's games based on odds derived from Las Vegas money lines, and then determined which picks would net me on average, the greatest victory margin over the public.

Looking back at my simulation results, I shouldn't get too puffed up about having scored 52 points. My expected value for my picks was just 40.7 and only 9 times out of 100 could I have expected to do as well or better than 52. So even though I planned well, I got lucky too.

And indeed it was a good week for the favorites in general, as I still only outperformed the average entry by 6 points. As it turns out, this was not that outstanding of a performance. It was above my average expectation, but I could have expected to do better or as well 45 out of 100 times.

In other words, despite a raw total that was far beyond what I would have expected on average, my margin over the public, though slightly above expectation, was not really all that special. Had ECU beaten UNC, or better yet, Arkansas beaten Georgia, I would have looked a lot better in comparison to the public. So I am grateful for 52 points but I feel like I may yet have an even more colossal week in my future.

Tuesday, September 15, 2009

College Football Pick 'em, Week 3

This year, I've been participating in ESPN's College Football Pick 'em contest for the first time. Through two weeks, I am ranked in the 72.4% percentile but am optimistic that I can continue to move up.

First, a review of last week:

BYU (10), W 54-3
Auburn (9), W 49-24
TCU (8), W 30-14
Tennessee (7), L 15-19
Georgia (6), W 41-37
Wisconsin (5), W 34-31
USC (4), W 18-15
Wake Forest (3), W 24-17
Notre Dame (2), L 34-38
Connecticut (1), L 10-12

A pretty good week on the whole, although I missed my 7 point game. I probably ought to have taken UNC, since even while being overvalued they were still 64.2% to win but it only cost me a point.

After a relatively easy Week 1 and Week 2, Week 3 provides a much tougher slate of games. There are only two double-digit spreads, which makes me nervous. Since it is only Tuesday, my picks are subject to change as lines move and ESPN users vacillate, but this is how I am calling it right now.

10: Texas over Texas Tech (-17.5)

Fortunately, the first slot is a relatively easy pick. Texas is favored by 17.5 points, putting them 4 points ahead of California and are at home, unlike the Bears. Although Texas struggled early against Wyoming last week, they rebounded for a 41-10 victory and dominated statistically. Mike Leach is notorious for being able to place any QB successfully into his system but it took his greatest Texas Tech team to beat Texas and he doesn't have that this year. The downside, of course, is that 97.1% of ESPN users are picking the Longhorns but the average confidence is just 6.2.

9: California over Minnesota (+13.5)

Although they are the only road favorite in this group, they are favored by a huge margin as Minnesota has struggled to put away Syracuse and Air Force. This is another public pick, but no other remaining match-up has a spread anywhere near this large.

8: Auburn over West Virginia (-7)

At 8 points I am already leery about my choice. But despite being selected by 84.4% of users, the average confidence in Auburn is just 3.5. Considering Auburn has the third-largest spread of any pick in the pool, that gives them a lot of value at 8 points.

7: Clemson over Boston College (-6.5)

Although users who have picked Clemson are fairly confident they will win (avg. confidence of 6.2), 28.2% of users are still picking Boston College, meaning this game is actually appropriately valued.

6: Iowa over Arizona (-5.5)

I'm dipping a little bit but although users are overrating the favorite in this match-up (as usual), they are also not all that confident in Iowa's eventual victory.

5: North Carolina over East Carolina (-7)

This game takes a big plunge on my board since a staggering 96.8% of users are selecting UNC, with an average confidence of 8.3, making them the most-bet team for the entire week along with California. This is the second week in a row that ESPN users have LOVED the Heels and they almost got burned big time for it in their game against U Conn. I think UNC will win here as ECU has struggled to start the season but there is about a 27% chance that the public will get hammered and I want to avoid that if possible.

4: Oregon over Utah (-4)

Interestingly, the public loves Utah here despite the Utes being a 4 point underdog. Clearly, Utah is overrated after their win over Alabama last year - this is probably not as strong a squad. Furthermore, I suspect the public is overlooking the fact that Utah will be playing on the road. All that said, I am nervous about taking Oregon considering their poor start to the season but they are more than capable of winning this game.

3: Virginia Tech over Nebraska (-4)

Not a lot of value here since most users don't think Nebraska is ready for a road upset over a ranked team yet. I agree but Tech has enough offensive troubles that it could be closer than you would expect.

2: Arkansas over Georgia (PK)

One of two toss-up games on the board and the public loves Georgia on the road (82.6%). That means Arkansas for me, but I feel good about it. Georgia is lucky to not be 0-2 and Bobby Petrino has the Razorbacks poised for a breakout season. Similar to Utah-Oregon, the public is correctly identifying Georgia as the stronger team but not properly valuing Arkansas' home field advantage.

1: Oregon St. over Cincinnati (PK)

Another road team the public is all over, although less so than with Georgia-Arkansas. I admit Cincinnati has looked impressive so far, but Corvallis is a tough place to win. Just ask USC.

My emphasis with these picks is on value. My competition is the ESPN public as a whole so I want to make picks that not only give me a good chance of getting the games right but also give me a leg up on the competition if they come through.

Monday, September 14, 2009

Illinois 45, Illinois State 17

I apologize for the long hiatus from this blog but my overly optimistic season preview appears to have been rendered moot by Illinois' 37-9 loss to Missouri. Fortunately, there is still hope for the season and the Illini rebounded with a 45-17 victory against the Redbirds of Illinois State.

Despite losing Juice in the first quarter to a thigh bruise, Illinois racked up 548 yards of offense, including 384 on the ground. I have no complaints about the Illini offense in general, as they averaged 9.6 yards per pass and a stunning 8.3 yards per rush. Eddie McGee overthrew a receiver and was intercepted in the end zone in the first quarter, but otherwise was quite solid. The offensive line overpowered the ISU defensive front and with so many talented RBs on hand, there was no stopping the Illini today.

Still, I have my concerns about whether the team's coaching is getting the most out of the talent on hand. Even in a blowout win, Illinois racked up 13 penalties for 115 yards. And they also allowed ISU QBs to throw for 354 yards, averaging 6.8 yards per attempt. Illinois won big because they dominated on the lines (allowing just 2.2 yards per rush), not because of coaching or playcalling. This reflects the talent Zook has brought in, but they won't be able to assert their superior talent so visibly against Big 10 competition.

I want to be a voice of moderation, but I am still skeptical about the direction this team is headed in after last week's loss to Missouri. There is a lot to like about Ron Zook, most notably his recruiting prowess and enthusiasm for his players. Yet he is also too stubborn about his way of doing things. He ought to admit that designing plays is not his strength, and delegate more responsibility to his coordinators, while focusing on motivating his players to give 100%. His coordinators do call the plays now, but they are locked in by the philosophy Zook espouses. Furthermore, Zook is always good for a boneheaded two-point conversion attempt or a fake punt that fools nobody.

The head coach does not have to have the last word on all decisions. Look at Penn State, where Joe Paterno's assistants run the show behind the scenes. Zook would be better off sticking to recruiting and motivation, while hiring some talented assistants to manage more of the in-game aspects. Unfortunately, that scenario will almost certainly never happen. I think Illinois is capable of beating Ohio State in two weeks, but it will take a great effort and a well-called game to get there.

Friday, August 7, 2009

Projecting Illinois Basketball, 2009-10

Today I am taking a break from my ongoing Illinois football preview and looking way ahead - to basketball. I was prompted to do this thanks to an interesting post over at Vegas Watch. Using Vegas Watch's projection system, I wanted to take a look at what we can expect from the 2010 Illini squad as well as how they performed against their projections in the past.




















YearPrev Pyth% ReturnProjectionActualDiff
2005.95392.0.979.982.003
2006.98240.9.882.940.058
2007.94065.7.913.932.019
2008.93257.9.891.900.009
2009.90067.8.894.911.017
2010.91157.0.876??????




Key:
Prev Pyth = Previous season's Pythagorean Win Percentage. Check out Ken Pomeroy's website for more info.
% Return = Percentage of previous season's minutes returning to the team.



I think this is pretty interesting stuff. The success of the 2005 Illini turns out to have been predictable, as they returned 92% of their minutes from a team that was already very good. The simplicity of the system shows up in the 2006 projection, which really undersold that team since it only returned 40.9% of minutes, but couldn't see that it was bringing back 2 excellent players in Dee Brown and James Augustine. Otherwise, the projections for the past 3 years have been more on the mark, albeit with a consistent trend of over-performance by the Illini. I am a big Bruce Weber fan, so I will speculate that Illinois' consistent ability to exceed their projections is attributable to his coaching skills.

So what can we expect in 2010? Probably a little bit of decline from last year, as Illinois returns just 57% of its 2009 minutes. The system projects a Pythagorean win percentage of .876, which last year would have put the Illini squarely on the NCAA bubble. However, even if you throw out 2006, where the projection for Illinois was clearly way too low, Illinois has out performed their prediction by .012 points per year (.021 if 2006 is included). Taking into account that bump puts their projection at .888, which would put them more squarely in the tournament, although they probably wouldn't be seeded as well as last year. Ultimately, it will depend on how quickly the highly touted freshman class can develop.

For Illini fans, the season to look forward to is 2011, where the team will likely return over 90% of its minutes. Even if their Pomeroy rating is a disappointing .888 this year, returning 92% of minutes (a la 2005) would result in a .941 projection for 2011. Giving that projection the same Bruce Weber bump yields a projection of .953, the same rating as this year's national runner-up, Michigan State. It's my job to be an optimist, but I think Illinois will return to the NCAAs next year and then be a legitimate Final Four contender in 2011.