Bucky free throws

The Badgers have been much better at getting to the line this season, and a big part of that has been the play of Nigel Hayes. He reminds me a lot of  Jared Sullinger when he was a freshman with the way he gets into defenders and gets his shot into their arms to force the officials call. Sullinger obviously was a bigger part of the offense as a freshman than Hayes, and he finished his freshman season with 267 FTA and 410 FGA.

According to this site Nigel is currently 48th in the nation, and 3rd in the Big Ten in free throw rate (free throw attempts/field goal attempts). The two Big Ten players ahead of him are lower usage players, Iowa's Anthony Clemmons and PSU's Geno Thorpe. Neither has more than 36FGA. Nigel is at 91 FGA, and 85 FTA. I couldn't find a way to compare Nigel to other players with similar usage, but perhaps when I have more time I'll investigate. I would guess he is near the top in the country for free throw rate for similar usage.

Friday Facts: Where's the home cookin'?

Time for another installment of Friday Facts, my weekly look at the Big Ten's per-possession stats, adjusted for opponent and venue. This week I break it down by home (HEM) and road (VEM), as well:

Not a whole lot different than last week. MSU, Iowa, and Michigan are clearly a cut above the rest of the league, and this shows in their records, their raw efficiency margins, and their adjusted efficiency margins. Wisconsin, OSU, and Minnesota are fighting for the coveted fourth seed. Everyone else is not so good, but good enough to pull some upsets (apparently). Northwestern is creeping up on Illinois, but is still dogged by those blowout home losses to Iowa and Wisconsin. 

I posted the other day about how Michigan State has been just an average Big Ten team at home, noting that they'd played the same teams at home as Nebraska had, except for Indiana, who came to visit Lincoln last night. Since MSU beat Indiana by five at home, I went ahead and predicted a five-point victory for Nebraska. BOOM!

Speaking of Nebraska, you might be surprised to see Nebraska actually dropped from -.04 to -.05 this week, despite beating both Minnesota and Indiana at home. The reason for this is that an average Big Ten team would have been expected to win those games by slightly larger margins. That said, Nebraska appears to be the team that the Friday Facts thought they were.

Anyhow, Michigan State is not alone in its home woes. Nearly every team—all but Minnesota—has performed better (compared to average) on the road. This shows up in a number of surprising home losses: Indiana and Wisconsin to Northwestern; Michigan State to Michigan; Ohio State to Iowa, etc. Simply put, home-court advantage hasn't been as big a deal this year in the Big Ten as it usually is. But maybe things will even out... 

Introducing the Torvik Big Ten Thrill Quotient

One of the side-benefits of my awesome spreadsheet that lets me produce adjusted efficiency margins for Big Ten teams is that it sort of automatically creates something akin to a "game quality score." The adjusted EM figure starts with a calculation of how each team in a particular game would do against an average Big Ten opponent in that venue. If there's a game with a good road team playing a good home team, it is sometimes the case that the average home team would be expected to lose, and the average road team would be expected to lose. So, the lower the expected EM of the average team is, the better the matchup.

For example, in the Iowa at Ohio State game, the average road team against Ohio State is expected to lose by .14 points per possession and the average home team is expected to lose to Iowa by about .02 points per possession. So you get a "game quality score" of 16. This score is basically just a straight function of the quality of the teams involved. So the highest quality score of the season so far was Iowa at Michigan, because those are the two best teams right now according to Kenpom. The lowest quality game so far was Purdue at Northwestern, at -22.

This number was more or less an organic creation of my adjusted EM spreadsheet, but it gave me an idea to create something similar to Kenpom's Thrill Score. So I did, and the result is Torvik's Thrill Quotient (which is a completely different thing, and not to be confused with, Kenpom's Thrill Score). The inputs to the Thrill Quotient are: (1) game quality score; (2) expected competitiveness (closer the better); and (3) pace (cuz people love the pace).

So, in case you were wondering, here are the most fan-tasitc games remaining in Big Ten play:

I would produce a chart of the least fan-tastic games remaining, but it's more efficient to just refer you to Northwestern's schedule. Northwestern is by far the worst team in the Big Ten (on a per-possession basis) and they play slower than anyone else in the conference. So there is nothing fan-tastic about watching their games, especially given that they have likely road blowout losses awaiting them at Wisconsin, Ohio State, Michigan State, and Minnesota. All those games have TQs in the 20s. 

What's up with Michigan State at home?

Michigan State is 18-2 overall, and 7-1 in conference, and they've put that record together despite significant injuries to pretty much every one of their important players. Pretty impressive.

But when you look at their performance so far, one thing jumps out: this team has not played well at home. Both their losses are at home: a disturbing blowout loss to North Carolina, and this weekend's loss to Michigan. By themselves, those losses probably wouldn't be cause for concern, given that MSU has won every other game. But they haven't exactly set the world on fire in their home wins, at least in conference play:

OSU, 72-68 (OT)
Minn, 87-75 (OT)
Ind., 71-66

Last week I debuted "Friday Facts" to show the "adjusted" efficiency margins in Big Ten play. At that point, Michigan State was number one in that metric. But even then, this was largely based on their impressive road performances (blowouts of lower-division teams Penn State, Indiana, Northwestern, and Illinois). One cool aspect of the adjusted EM is that I can break it down to home and road performances. Here are the splits, ordered by the home margin ("HEM"):

So Michigan State has just been treading water—an average Big Ten team—at home, on par with Nebraska. In fact, Nebraska's performance at home has been eerily similar against an almost identical schedule:

L, 71-70 to Michigan
W, 68-62 over OSU
W, 82-78 over Minnesota

A close loss to Michigan (check), and close wins over OSU and Minnesota (check, check). As luck would have it, Nebraska plays Indiana at home on Thursday, so after that game Nebraska and Michigan State will have played the same four teams at home. (Might I be so bold as to predict a 5-point win for the Huskers?)

So what's going on? There are at least two possibilities. First, it could be that this is just randomness, and Michigan State has just happened to have its poorer performances at home. Second, it could be Michigan State isn't quite as good as we might think they are. On the other hand, this could bode well for Michigan State's chances in the post-season, when there are no home games.

Michigan State has built up its impressive road margin against the bottom half of the league. Its remaining road games are against Iowa (tomorrow), Wisconsin, Purdue, Michigan, and Ohio State. It will be interesting to see how well their numbers hold up against that slate.

Going Pro

Sam Dekker started out this season projected as a possible mid to late first round pick in this year's NBA draft. He dropped a bit in the early season, but has been showing back up in the 20s again. This is all very premature as there is a lot of season left, and much of the player's draft status will be determined at the tryouts and camps after the season.

I would love to see Dekker play another season or 2 at UW, but I am of the opinion that any UW player who could be a first-round pick should go pro. With the new collective bargaining agreement, NBA first round picks are guaranteed 2 year contracts and will make over a million dollars even if they get cut and don't make their NBA roster. Even the 30th pick in the draft will get a minimum salary of about $750,000 each year. That is a lot of money. Too much money for a 20 year old kid to turn down. It's enough money to give you the opportunity to do whatever you want to do for the rest of your life, as long as that isn't buying drugs, cars, and other crap.

It's great for fans if players stick around, but there is just too much risk in not taking that money if they can be a first round pick. One blown out knee and there goes your million dollars.

The 2 foul question

First of all it's good to be back after a nice long hiatus overseas, but unfortunately my Badgers don't seem to be able to win when I am not around. They must have been hung over from the lack of my presence against Minnesota, but they returned to their winning ways yesterday. I promise I will not take anymore overseas trips this season. I also expect they will not lose another game until the dreaded back to back road games against Iowa and MI. Even I don't have the power to help them overcome a back to back road affair vs. ranked Big Ten opponents. On to the post.

Just one game after benching Kaminsky for the rest of the 2nd half against Minnesota due to picking up his 2nd foul in the first half, Ryan put him back in the game in the first half multiple times after doing the same against Purdue. There are pros and cons to either strategy, and I don't really feel either one is right or wrong. This is one of those issues much like fouling at the end of a game when up by 3, that some people seem to have strangely strong feelings about, when in reality it is about a 50-50 type of situation.

Matt Painter also put Hammonds back into the game after he collected his 2nd foul. Hammonds only played about a minute and a half into the 1st half before he got his 2 fouls, and then when put back in the game he gave Painter all of 30 more seconds before collecting his 3rd. Hammonds has been a hacker this season, so it is not that surprising that he picked up his 3rd right away. Kaminsky is not as foul prone, but on his first series on the floor after picking up his 2nd foul he almost climbed the back of a Purdue player going for an offensive board and got his 3rd. You could see him pulling back his arms as he realized the folly of what he was doing trying to get a rebound over an opposing player when he already had 2 fouls.

In general I like the idea of sitting a guy after he gets 2 fouls in the first half. The primary reason is that there is no better teacher to a young college player than spending some time on the pine. While some fouls are the result of bad calls or just being in the wrong place at the wrong time, the vast majority are the player's fault. Spending a little time watching your teammates play and possibly struggle without you is the best medicine to remind players that they need to mind their hands and move their feet. There is also the rather large risk that the player will keep picking up fouls and take the choice out of the coaches hands, as Hammonds did.

I don't like keeping a player on the floor with 2 early fouls because it messes up your defense. The player must play cautiously knowing they have 2 fouls, so they can't challenge shots, work over screens, or even box out as they would normally. It also effects the entire team defense because the player can't rotate as aggressively and the player can't help as they normally would. When you are defending a screen and roll and the other guy is basically useless because he can't close out on the shooter, hedge or really anything that might risk getting a cheap 3rd foul it screws up your whole defensive plan. It's not like you can hide the fact a guy has 2 fouls, there is that giant scoreboard glaring at you from center court with that big 2. Every good coach will make sure all his players know that, and whoever is matched up on that player should be attacking them at every opportunity.

Then there is the human element of the game. We have all played in games when your head is just not on quite right. Often these are the games when we get into foul trouble because we are making mistakes. Getting called for fouls is frustrating, especially when you think they were bad calls. Getting tossed back into the game after getting 2 fouls when you are already frustrated is a recipe for making another mistake.

All that being said, I don't really have a problem if a coach plays a guy after getting his second foul in the first half. There are players that just don't foul much, and if a coach played that guy in the 2 foul situation I wouldn't criticize him for it. I also think there are definitely circumstances that merit a change in coaching philosophy. For example, if you are playing in a win or go home game in the NCAA and your star player gets his 2nd early in the half. Your team starts to struggle and the game appears to be getting away from you, then put your star player back in and take your chances. If you are a bad mid major team and you are playing against a ranked team and are hanging with them and a star player gets his 2nd foul, take a chance and leave him in. You only get so many of these opportunities. Much like an NFL coach is more likely to go for it on 4th down in the playoffs, the risk-reward equation in any single game may be different than what you usually do on a normal basis.

Friday Facts

John Gasaway has come out with this season's first full-fledged "Tuesday Truths"—his indispensable weekly breakdown of the conference-only efficiency margins for every team in every conference worth caring about.

But the Tuesday Truths can be deceptive. The efficiency margin is calculated like so: points scored minus points allowed, divided by number of possessions. There is no accounting for strength of opponent. By the end of the season, this matters little, as we can expect things to more or less even out. But early in the season, differing schedules can make a big difference.

So I took a stab at producing an adjusted efficiency margin for the Big Ten. For each game, I calculate the difference between a team's actual efficiency margin and the expected efficiency margin of an "average" Big Ten team against that particular opponent at that particular venue. That difference is adjusted efficiency margin. Add up each game for each team, and you get their total adjusted efficiency margin.

To illustrate, let's look at Iowa versus Wisconsin, at Wisconsin. The average Big Ten team—that is, a team with adjusted offensive and defensive efficiencies at the average of all Big Ten teams—would be expected to lose to Wisconsin at Wisconsin by 17 points per 100 possessions. In fact, Iowa lost by four points in a 70 possession game, which comes out to 6 points per 100 possessions. Thus, Iowa's adjusted efficiency margin for that game was +11 per 100 possessions, or 0.11 per possession.

On the flip side, the average Big Ten team playing Iowa at home would be expected to lose by 1.3 points per 100 possessions, and Wisconsin won by 6 per 100. So Wisconsin was above average too—+0.07 per possession.

When you add up each team's performances from all their games you get their total adjusted efficiency margins:

There's no real surprise at the top or very bottom of this chart. Michigan State, Michigan, and Iowa have clearly performed the best in conference play so far, and that shows up in the raw efficiency margins. But the adjusted margins give Iowa credit for its close losses at Wisconsin and at Michigan, and so Iowa just barely trails Michigan in the adjusted ranking.

Minnesota also gets credit for its tougher schedule and better wins. Although its raw efficiency margin is negative overall—placing it behind Purdue—the adjusted margin gets Minnesota ahead of Purdue and into the conference's upper division, where it clearly belongs based on its play so far.

The other notable outlier is Nebraska, which jumps ahead of three teams (Indiana, Illinois, and Penn State) at the bottom of the standings. Nebraska's raw efficiency margin gets killed by its big blowout loss to Ohio State, and doesn't reflect any credit for its respectably close losses to Michigan and at Iowa, or for the impressiveness of its win over Ohio State. These results are why Nebraska has actually moved ahead of Illinois in the Kenpom ratings, and the Adjusted EM accounts for them. 

Title hopes dashed?

After two upset losses this week, the Badgers' chances of getting a share of the Big Ten title have taken a big hit. According to my spreadsheet, they now have only about a 10% chance of getting a share. Here are a couple looks at how the race has changed since Wednesday.

First, the contenders' chances of getting to various Big Ten win totals:

Second, in line graph form!

Okay, I'll stop now.

Update on the Big Ten race

Here's a chart showing each of the contenders' current chances of getting to various win totals:

Two things:

1) Last night's loss has devastated Ohio State's chances of winning a share of the title, which are now down to about 4%.

2) Michigan State has opened up a small but significant lead on UW. Prior to the Spartans' win over Northwestern, it was more or less a dead heat. But a combination of that win and some resulting improvement in the Kenpom ratings has lifted MSU's status to minor favorite.

Big Ten Record

Kenpom now projects the Badgers to go 15-3 in the Big Ten—a full two games better than Iowa and Michigan State.

I've been doing some simulations of the remaining Big Ten schedule, and should have a post up over at Bucky's 5th Quarter soon. As a teaser, here's a chart of the Badgers's chances of getting to various win totals in Big Ten play, using Kenpom win probabilities:

The Seth Davis "Buy" Curse

So far, things are looking pretty bad for Seth's portfolio. There were eight teams that Seth put a "buy" rating on that I told you to sell:

1) Butler, which lost to Depaul at home last night. Yikes.
2) Missouri, which lost to Georgia at home on Weds. Yikes.
3) Notre Dame, which lost to NC State at home on Tues. Yikes.
4) Oregon, which lost Cal at home last night. Yikes.
5) Colorado, which squeaked by Washington State in OT.
6) UMass, which squeaked by St. Joe's at home.
7) Oklahoma, which lost at home to Kansas. (No shocker, but nothing special either.)
8) Syracuse, which blew out Va. Tech at home.

So far, this is a brutal slaughter. Here's one way to put this in perspective: after just one game, four of these teams have fallen more than ten spots in the Kenpom ratings: Butler (-15), Missouri (-17), Notre Dame (-15), and Oregon (-10). On average, the eight teams have fallen about 7 spots. After one game.

Here's another way to put it in perspective. According to their Kenpom ratings at the time Davis rated a them a buy, the chance that Butler, Missouri, Notre Dame, and Oregon would all lose at home in their next game was 0.09%. That's not 9%, that's nine one-hundredths of a percent. In other words, one-in-a-thousand.

Chance cannot explain this. Something more sinister must be at work here. Is Davis pumping and dumping? Or does the Monkey have powers that cannot be explained by science or math?

Seth versus me

Just to put my money where my mouth is, I went ahead and made my own "buy, sell, hold" picks. Here's where I differ with Seth Davis:

Seth Davis versus a monkey

Seth Davis has his annual "Taking Stock" column out, where he advises whether to "buy, sell, or hold" imaginary stock in a bunch of college basketball teams.

Last year I made fun of his prognostication skills, and concluded that he did worse than you'd expect a monkey to do.

As promised, this year I let a monkey pick whether to buy, sell, or hold the same group of teams. Here are the teams where Seth and the Monkey disagree:

Cheer for the monkey!

(By the way, no surprise: Seth is selling his stock in the Badgers.)

StatWatch™: Hayes FTAs

Badgers' forward Nigel Hayes has been a really nice surprise this year. He's got some issues on defense, but he's shown an ability to play with his back to the basket and get to the rim that we haven't seen in a while—especially not from a freshman.

His hard work is resulting in a lot of free throws. He's got 40 free throw attempts in just the last four games, which is sensational. I took a look at the best four-game FTA stretches in the Bo Ryan era, and here is the result:

1) K. Taylor, 43: OSU, NU, Pur., IL in 2006-07 (50 over 5, 56 over 6)

2) Hughes, 41: Duke, GSU, GB, Marq. in 2009-10

3) J. Taylor, 41: MSU, Mich., Pur., Neb in 2011-12 (but had 0 against Michigan! And 42 over 5.)

4) Harris, 40: Minn, NU, Iowa, OSU in 2003-04 (0-1 against Iowa, 47 over 5, 57 over 6)

5) Harris, 40: first 4 games of 2003-04 (45 over 5)

6) Tucker, 40: Pitt, Iowa, MSU, MN 2005-06 (had 49 over 5)

Most of these performances were by point guards, probably largely inflated by late-game free throws. Alando in 05-06 is the only similar stretch by a Badger forward. And he was a starter who was getting a lot more minutes than Hayes.

Hayes's effectiveness will presumably go down a little bit as opposing coaches get more tape on him. But so far he has been an absolute beast.

Will the Badgers get to No. 1?

I did some investigating and posted it over at Bucky's 5th Quarter.

Comparing the Preseason Computer Ratings

Since we're about half-way through the college basketball regular season, I thought it'd be a good time to check in on the performance of the major preseason ratings systems: mine, Pomeroy's, Team Rankings (David Hess), and Dan Hanner's. Behold:

These are the current Top 50 in the Pomeroy ratings, and the first four columns after the team names are the where each system put them in the preseason. Cells shaded green show which teams each system predicted to be in the Top 50. The highest rated team that none of the systems predicted to be in the Top 50 is UMass.

The next four columns (labeled "off") show how much better or worse each system rated each particular team compared to the other systems. This is calculated by comparing how far off a particular system's prediction was, and then comparing it to the average of how far off the other three systems were. For example, No. 39 Florida State comes out as +46 for the T-Rank because the T-Rank was only 3 spots low in its prediction for FSU, but the other three systems averaged 49 spots low. This is by far the biggest "hit" for any system (for teams currently ranked in the top 50). 

The biggest miss was Pomeroy's projection for Arkansas.

One way to measure the relative performance of the systems against each other is to add up the "off" columns (which sum to zero when all four are put together), with higher totals being better. The result:

1.  T-Rank +54

2.  Hanner +38

3.  Hess -23

4.  Kenpom -69

There are other ways of analyzing the relative performances, but none of them are so flattering to T-Rank, so they will not be publicized at this time.

The ultimate test, of course, will be the NCAA tournament's S-Curve.

Badgers at Northwestern

I will be at tonight's game with my daughter Ivy.

The last time I went to the the UW-Northwestern game here in Evanston, Josh Gasser had a triple-double. And then the Packers beat the Bears in the NFC Championship game later that day.

That will be hard to top.

The Surprising and Impressive Badgers

I've got a post up over at Bucky's 5th Quarter that takes a look at college basketball's remaining unbeaten teams. In short, the Badgers have been both surprising and impressive.