Guessing the Seeds: Part Three

Same thing as the first two times. If you're participating in the contest, please send your picks, with the same entry name as before, to vegaswatch@gmail.com by noon EST on Saturday, February 13. Each correct pick is still worth 6-5-4-3 for the four lines, and this round will be weighted at 50%.

This Week's Links (2/1-2/5)

Posnanski on the greatest quote ever.

Please do not expand the field.

Pinnacle has odds on UNC and UConn making the tournament. Not looking good for either unless a format change comes significantly sooner than anticipated.

Majerus does the unthinkable. (Also: KenPom, now on Twitter!)

Pro-Football Reference breaks down the EV of each Super Bowl square.

The Most Ridiculous Super Bowl XLIV Props

Welcome to the third annual edition of our Super Bowl proposition bet coverage. I was hoping to wait out for BetUS and SportsBetting to post some interesting props before writing this, but they are really slacking. I'll have to include those in another post if they come up with anything good in the next couple days. For now it's mostly Bodog (as usual), with appearances from Sports Interaction and Bookmaker.

How long will it take Carrie Underwood to sing the National Anthem?

Bodog:
Over 1m 42s: +130
Under 1m 42s: -160

Sports Interaction:

Over 1m 38.5s: -125
Under 1m 38.5s: -115

It would be fun to go for the middle here, but the juice on U1:42 is just way too much. Looking at the variety of videos on YouTube of Underwood singing the Anthem, it looks like on average she comes in a bit under the Bodog total, but certainly not enough to pay the -165. Both of these lines seem about fair to me after considering the juice.

What Color will the Gatorade be that is dumped on the Head Coach of the Winning Super Bowl Team? (Bodog)

Lime Green: +600
Yellow: -125
Orange: +550
Red: +1250
Blue: +1250
Clear/Water: +160

Prior to last year, clear/water was riding a three Super Bowl winning streak, but that was snapped when the Steelers dumped yellow Gatorade on Tomlin. There may be some value in clear/water at +160, but I would hold off to see if BetUS posts anything, since last year they had that option at an absurd +700, which was a great price even though it lost.

Who will the Super Bowl MVP of the Game thank first? (Bodog)

God: -105
Family: +650
Teammates: +115
Coach: +1000
Does not thank anyone: +450

I think there's value in "doesn't thank anyone" here. Holmes didn't last year, and neither did Eli when they beat the Pats. At that point everything is just so disorganized, I think betting on chaos is the way to go. It's kind of strange that "doesn't thank anyone" is +450 given its success the last two SBs, but I get the feeling that most people don't pay a whole lot of attention to these things on a year-to-year basis.

How Many Times will CBS show Archie Manning on TV during the Game? (Bodog)

Over 4: -130
Under 4: EV

CNBC's Darren Rovell has had good coverage of this one, which actually opened at 2.5. According to Rovell's post, that opening number was based on a similar prop for Brenda Warner last year, but Bodog failed to consider Archie's Saints ties. They took a lot of action on O2.5, and eventually moved it all the way up to the number listed above.

I don't have any further analysis, just thought that was interesting.

Which CBS Show will get the most Promo’s during the game? (For a show to be counted it must appear in a picture box on the screen during the game broadcast only.) (Bodog)

Undercover Boss: -220
CSI: +625
CSI Miami: +450
NCIS: +525
NCIS LA: +1200
Big Bang Theory: +1500
How I Met Your Mother: +900
Two and a Half Men: +1500

Undercover Boss is a new CBS reality show that will be debuting immediately after the game. It actually opened at EV, but apparently received a lot of action at that price and moved significantly since then.

And with good reason, I think; the combination of being a new show and coming on immediately after the game has to give it a huge advantage. It's not like you have to leave this up to random chance either, which is always a problem when playing a big favorite. The CBS executives are in complete control of how many times they mention each show during the game, and you'd have to think that it'll be very frequent for a new show in that time slot.

Pinata Pick - Match Peyton's total pass yds and we'll pay 44/1 (Sports Interaction)

This would be a fun one to play, but there's no value here regardless of which number you pick. Here are Pinnacle's odds on his total passing yards:

0 to 100: +1000 (6.0%)
101 to 190: +600 (9.5%)
191 to 245: +400 (13.3%)
246 to 299: +180 (23.8%)
300 to 325: +220 (20.8%)
326 or more: +150 (26.6%)

When you consider how many yards are in each group, "300 to 325" is the sweet spot, with each of those single yards being about 125:1. Pinnacle also has the O/U on his passing yards at 308.5, with the under favored -118/+102. So if you're going to play this I'd go with 304 or so, but that's definitely -EV.

What will be the highest rated commercial on the USA Today Super Bowl Ad Meter?

Bodog:
Anheuser-Busch/Budweiser: -160
GoDaddy.com: +800
Career Builder: +800
Coca-Cola: +700
Family First: +1200
Doritos: +200
Other: +260

Bookmaker (under "Entertainment"):
Anheuser-Busch: +900
Pepsico's Frito-Lay (Doritos): +1650
33 other options: here

I think there is tremendous value in Anheuser-Busch and Doritos at Bookmaker here, and it goes beyond just comparing their odds to Bodog's. Busch won ten years in a row before Doritos broke their impressive streak last year. Even in defeat, Busch showed very strong, with their ads placing both second and third. In 2008 Busch placed first, fifth, and sixth, with Doritos coming in fourth. And in 2007 the two companies dominated the standings, monopolizing the top seven spots. Even at +900, Busch actually has the shortest odds of any of Bookmaker's 35 options; it's like BM was aware that they should be the favorite, but had now idea just how dominant they've been.

As for this year's game, the Clydesdales won't be appearing, but Busch has purchased five minutes worth of ads, and their non-Clydesdale commercials have scored well in the past. Doritos is running a similar contest to the one that landed them the top spot last year; they'll have three commercials during the game.

Rating the Power Conference Teams (and BYU)

A couple weeks ago I posted about adjusting the KenPom top 20, and thought that was a good start but wanted to continue to improve my approach. The main thing that was missing was the failure to factor in very recent results; a team's most recent game has almost no effect on their rating the way I was doing it before, since it's not factored into any of the lines.

I've tried to fix that issue in the ratings that follow by considering margin of victory in recent games. Since point spreads are far more stable than single game scores, these margins have to be heavily regressed; for example, if a team beat the spread by 14 in their most recent game, in this version that will add one point to their rating. The weights get smaller as the games become farther in the past, since at that point they're already factored into the more recent spreads.

Aside from that, these ratings are very similar to what I did last time. They take a team's Pomeroy rating and adjust it based on the spreads in their last six games, with the more recent spreads being given additional weight. For this iteration I looked at each major conference team, as well as BYU, which is the only non-power conference team in the KenPom top 20.

Here is my latest top 20, through all games of 1/31:

KenPom: Pomeroy rating translated into points.
Sp Adj: Adjustment based on recent spreads.
Rec Adj: Adjustment based on recent performance.
Rating: Final rating.

Kansas remains the clear #1. I was kind of hoping they would lose to K-State in OT, so they wouldn't return to the top spot in the polls and people would start to doubt them for whatever silly reason, but their position in my ratings was never in doubt. The Jayhawks were favored by four in that game, which was extremely impressive in itself; you will remember that Texas was a one-point underdog in Manhattan a couple weeks ago.

Duke is the easy #2 again, and I realize that looks very odd. But the fact is that they were favored by two at Georgetown on Saturday, which is exactly what KenPom had as the spread. They didn't exactly impress in that game, which shows in the column that considers their recent performance, but the gap between #2 and #3 is so huge in KenPom that it doesn't make much of a difference. I'll admit that I don't really buy the Blue Devils as the clear #2 in the nation myself, but we'll have to see how the oddsmakers treat them going forward. Duke hosts Georgia Tech on Thursday night.

It's after the top two that I think things really get interesting. I've suspected for awhile now that Purdue is still an elite team despite their three losses early in Big Ten play, and these ratings would suggest that that's correct. Purdue was an impressive 9.5-point favorite at home against Wisconsin last week, and don't seem to have lost much respect after their three-game losing streak. The true test of how good they are will come a week from Tuesday when they play at Michigan St. Either the Spartans will be a very short underdog in that game (these numbers suggest MSU -1.5), or I'm wrong about (at least) one of those teams.

The West Virginia-Syracuse comparison is an interesting one, and not something I think there's a conclusive answer to. WVU was a 5.5-point favorite when the two teams squared off in Morgantown a few weeks ago, but the Orange won that game by one. From that you'd think the 'Neers were about a point better than the Cuse. But then in the last week WVU played DePaul, was favored by 17, and won by 16. Pretty standard. After that I expected the SYR@DEP line to be 16-17, yet it closed at 14.5 and DePaul almost managed to to win outright. So I guess WVU being 1.5 points better is reasonable enough, but the gap may actually be even wider.

Ohio St. -- now there's a name you don't see in many top fives. They're actually only 15th in KenPom as I write this, although that hasn't yet incorporated Sunday's 22-point victory against Minnesota, or the games Turner missed. The Buckeyes were favored by seven on Sunday, which surprised me; the Gophers haven't been doing too well in terms of wins and losses, but they were favored by 1.5 at home against Michigan St. as recently as a week ago. Considering those two spreads and Sunday's dominant performance, I'm comfortable saying that the Buckeyes are the second best team in the Big Ten. Fifth in the whole country though? Definitely not a commonly held opinion, but I'm struggling to find a reason that they aren't.

There continues to not be any quantitative reason to think that KenPom is wrong about Kentucky. Their spreads come in right around where you'd expect, and their actual results have fluctuated but on average they haven't been extreme one way or the other. Maybe with all that talent they'll just "turn it on" in March and their lines will reflect that, but for now they stay at #7.

After writing this I feel like it'd be better to go through it by conference and then do the top 20 at the end. Next time.

The ratings for each team in the six major conferences can be found here. At some point I'll compare these ratings to the futures odds that are out there; hopefully I'll have a chance to do that some time this week.

This Week's Links (1/25-1/29)

The 2010 PECOTA depth charts were released, but don't put too much stock into them just yet. It is still January.

Great Hardwood Paroxysm post on NBA usage and PER at the league's halfway point.

Calcaterra fisks some more McGwire silliness.

One of the guys over at Lookout Landing put together an impressive collection of articles on a variety of sabermetric issues.

Super Bowl props! (Scroll to the bottom for the good ones.) More on these, and hopefully a bunch of others, next week.

Revisiting Our Projections: Missing High

This summer MoneyLine, Jonny, RTC and myself took on the task of giving preseason KenPom ratings to each team in the six majors conferences, as well as a few other relevant one. I'll obviously have a full analysis of how we did after the season -- I'd say it's gone decent to well so far -- but for now I wanted to look at some of the teams we really missed on. This week we'll discuss some of the teams that have been much worse than we expected.

North Carolina
Projected: 0.930
Actual: 0.862

VW: Okay, we may have been wrong about UNC overall, but I was correct about one thing:

"Now I doubt that John Henson:2010 :: Tyler Hansbrough:2006..."

Which really illustrates the main problem with projecting a team that lost 75% of its scoring; I didn't think Henson would be nearly as good as Hansbrough, but it was within the realm of possibility.

Davis has been as good as advertised, but Thompson hasn't, and the freshmen most certainly haven't; one of their five 4+ star recruits has an ORtg over 100, and only Strickland (who's been awful) has played more than 30% of possible minutes. Ginyard's injury definitely didn't help the already thin backcourt, but turnovers have been a problem on both ends of the floor with or without the senior guard.

RM: I wasn't around to vote on the ACC teams, but had I been, I would have definitely put the Heels above .900, and I would have also been wrong because UNC looks like an 8-8 team. The guard play of Drew the Second and Strickland is discouraging -- they've combined for 88 turnovers in eighteen games, or nearly five per contest. By comparison, Ty Lawson had 65 all of last season. Roy knows that it doesn't help to play uptempo basketball (+6.6 possessions vs. the average) if your primary ball handlers are giving most of those extra scoring opportunities away.

JP: After looking at the Tar Heels’ four factors two things really jump out: They turn the ball over too much, and they don’t get to the line. I don’t think comparing Drew II/Strickland to 2009’s most offensively efficient player is fair, but they clearly need to improve. By slowing it down and working the ball inside they can utilize the 3rd tallest team in the nation in effective height, and improve from here on out. Whether they do this, remains to be seen.

Washington
Projected: 0.908
Actual: 0.816

VW: Washington's biggest problem on offense has been horrendous shooting from deep; they're currently at 30.5% from three, which is 289th in the country. You'd think that will have to regress at some point, but they weren't great last year (33.7%, 179th), and they lost Dentmon, who shot 41.2%. It might be time to accept the fact that Isaiah Thomas isn't a very good shooter; he shot 29% on 134 attempts last year, and is 31/103 (30%) so far this season.

Washington has also fouled an incredible amount on defense, more than every other major conference team. Overton and Gaddy appear to be the biggest liabilities in this category, along with a couple of freshmen who don't play much but still manage to get their hacks in. I'm not quite sure what to make of this, but I know it's not helped by their opponents shooting 70% from the line.

RM: The biggest question mark I have about this team is what happened to Abdul Gaddy? The consensus #2 PG prospect behind John Wall has looked completely lost in a structured offense this year, and is bringing almost nothing other than spot minutes and poor shooting (41.5% eFG%) to the table.

Iowa
Projected: 0.674
Actual: 0.563

JP: Don't look now but progression has begun!

VW: It indeed has; Iowa's KenPom rating has risen nearly 100 points since I started this e-mail thread last week. That's another thing we need to keep in mind -- some of these gaps may be huge now, but they will inevitably get smaller as more games are played.

Regardless, that's a lot of red. It's good that they've been so strong on the defensive glass, because otherwise they've had one of the worst defenses in the country. We all knew they were going to be bad, but I'm not sure how much lower we could have gone with the projection.

Auburn
Projected: 0.800
Actual: 0.640

JP: Lucas Hargove, remove yourself from the 3-point line immediately. Thank you.

VW: Auburn's freshman class ended up being worse than one could've reasonably expected -- this is going to sound crazy, but it's inferior to even UNC's -- but I think this was a case where we overrated the gap between average seniors and 3* freshmen. Barber, Barrett, and Robertson weren't great, but they used up a lot of possessions without embarrassing themselves and played D. That's not going to win any championships, but it will keep you in the top 100 in KenPom.

RM: The one thing they do well (shoot twos at 53%), they don't do enough. Instead, they prefer to chuck up threes with reckless abandon, having six players tossing two or more a game at the rim. Beyond the stats, though, Barber was the heart and soul of that team. Once he left campus, Lebo's days were numbered.

Louisiana St.
Projected: 0.778
Actual: 0.493

VW: And here I was, thinking Georgia was clearly the worst team in the SEC; they're not even in the bottom three thus far. I think this is another example of us underrating the importance of depth, as I pointed out that they were going to have trouble finding production beyond Mitchell and Spencer, but apparently didn't realize just how difficult that would be.

One thing that would've been tough to predict is that Spencer would go from shooting 40% from three as a sophomore to 27% as a junior. They've seen a huge drop in 3P% as a team, going from 37% last year to 29% this year despite returning their best shooter; freshman Aaron Dotson's 6/38 (16%) has not helped matters.

RM: I thought Mitchell and Spencer could carry the Tigers so long as their freshmen stepped into vacated roles. Storm Warren has been a pleasant surprise to go along with that duo, but after those three, there's nobody doing much of anything here.

Oregon St.
Projected: 0.802
Actual: 0.482

VW: This is a strange one. The Beavers finished at .677 last year, and returned everybody except Ricky Claitt, who was very efficient but extremely low usage. I wasn't totally convinced that they had enough talent returning to take another big step forward, but I had no idea their offense would fall off a cliff like this (95th to 253rd).

RM: They also expected to have four-star recruit Roberto Nelson in the fold, but eligibility issues have sidelined him this season. Still, it is bizarre, because their offense has nearly fallen to 2008 levels (the 0-19 Pac-10 year). I'm not sure I've ever seen a high-major team get demolished by a mid-major on its home court like Seattle did to the Beavers, though -- that's a signal of more than just talent and execution being the problem.

JP: Craig Robinson needs to sit Joe Burton down and tell him he isn’t getting back on the floor until he agrees not to be an overly confident terrible offensive player. I know this is making a big deal out of someone playing less than half the game but his usage/offefficiency split is pretty disgusting. This team would benefit much more if they gave his possessions to senior Seth Tarver (as long as he agrees to not launch treys – of which he has not hit a single one in 2010).

UCLA
Projected: 0.907
Actual: 0.629

VW: I guess Darren Collison was an important part of their offense. Who knew? Everybody was wrong on the Bruins, and that certainly includes us. Five-star freshman Tyler Honeycutt has been useless, Dragovic's ORtg has fallen by almost 30 points, they're 333rd in the nation in FT%; it's really been an incredible fall from grace, and I'm not sure we could have seen it coming.

RM: Speaking of talent, Howland has a roster full of busts (among those who bothered to stick around). I'm still trying to figure out how on such a bad team J'Mison Morgan can only play 8 mpg and average 2 points and a single rebound. His five-star colleagues in the Class of 2008, Lee and Anderson, may not be the stars everyone thought they could be, but at least they're on the floor. Could that class be the worst #1 rated recruiting class of all-time?

Oklahoma
Projected: 0.878
Actual: 0.750

VW: Their offense has fallen off about as much as you'd expect considering what they lost, but it's been D that's really killed them. Griffin wasn't the greatest post defender ever, but he did pull in a ton of defensive boards, which is a factor we probably underrated. Their 3P defense has gotten a lot worse as well, which is something that should probably be expected when you have a 5'11 freshman playing over 80% of the time. Great name, though.

RM: Much of this has to fall on Willie Warren's inability to lead as a sophomore. His numbers are marginally improved (all but shooting %s), but it's no secret that Capel is unhappy with him. He's the kind of player who can carry a team for a while if he gets his head on straight, though, so out of the teams we've reviewed here I'd give OU the best shot to turn things around and make the NCAAs.

JP: I may just be my love for Tiny, but I agree that OU has a shot to turn this season around. Ryan Wright has been getting elevated minutes after a brief hiatus at the end of the bench, and has performed above expectations. If he can learn to play clean defense without making stupid ticky tack fouls, he could become a decent contributor for their stretch run. It also helps moving forward that OU has been quite unlucky defending 3’s (36.1%) and FTs (71.7%); both should regress.

Opening Super Bowl Line

Here's what Pinnacle opened it at right after the Saints game ended:

They have clearly taken a lot of action on the Colts since putting that number out there, as Indianapolis is currently favored -4.5 -116/+107. This clearly has something to do with how both teams looked this week, but is still a bit surprising; SportsBetting posted prospective lines last week and had the Saints at +3 -120 in this matchup.

The current money line, -210/+190, gives the Colts a 66.3% chance of winning. That's slightly shorter than the normal ML for similar spreads; I'm not sure how much of that is the books shading it because the underdog ML usually attracts a lot of action in the Super Bowl, and how much of the difference is caused by the high total (55.5).