Sunday, March 22, 2015

Day 6 Open Thread + Picking The Lines

Most sports narratives follow a pattern, and that pattern is easy to predict if you've followed the narratives in the recent past. A classic example of this is in college football, where the last place SEC team knocking off the first place SEC team is evidence of the league's depth, while the last placed Big Ten or Big 12 team knocking off the first place team in the league is evidence of how weak the league is. In reality, two teams from the same conference playing each other says almost nothing about the league's strength, and you'll see basically no change in conference computer ratings, but if you want to believe a narrative, then you will make the results fit that narrative.

In college basketball, when we talk about which conference is best, in the media and with casual fans this always becomes a contest for which conference has the best teams at the top. When comparing whether the Big East or ACC is better, nobody is asking whether DePaul is better than Virginia Tech, even though they should. It should still matter if those road games in conference play include dangerous teams like Kansas State and TCU rather than floor mats like Virginia Tech and Boston College. But since we don't care, we create a bias toward conferences with a lot of teams, as well as conferences that are top heavy. If your league has ten teams where almost all of them are good (like the Big East or Big 12), you will necessarily be underrated, while if your league has 15 teams and is extremely top heavy (like the ACC), you are necessarily going to be overrated.

Remember, I say every year that you cannot judge coaches, teams or programs by results in the NCAA Tournament. They are single game samples, and weird things happen. If "the better team" won every game then the tourney would be extremely boring. All you'd need to do is figure out which were the best teams and you'd know the result of every game in advance.

Now, we understand that previous paragraph when we want to. In years when Michigan State goes down early, we say "Gosh, you don't often see Tom Izzo teams upset early", and forget about it. When Georgetown goes down early we say "LOL, you can set your watch by JTIII teams going down early, amirite?"

Similarly, when UAB knocked off Iowa State, nobody used that as evidence that Iowa State didn't deserve their 3 seed, that UAB was better than Iowa State, or that Conference USA was better than the Big 12. We knew that would be stupid. But then when North Carolina State upset Villanova, we got this sentiment all across the media:
Of course, NC State is a high variance team. They beat Duke by 12, and won on the road at North Carolina and Louisville, while also losing to Wofford, Wake Forest and Boston College. If NC State knocked off Duke in the NCAA Tournament, nobody would feel the need to rewrite the narrative of Duke's entire season to prove that they actually sucked all along. But why should NC State beating Villanova prove that NC State is better than Villanova or that the ACC is better than the Big East when the same isn't true about UAB and Iowa State? Because when a result fits the narrative we want, we plop the result into that narrative, and when the result doesn't fit the narrative we let it pass.

Is NC State better than Villanova? No. Not even close. Villanova just happened to have a nightmare game at a terrible time (though despite their second worst shooting performance of the season they still had a shot to win in the final 30 seconds). Villanova outscored the Big East by 0.21 PPP while NC State outscored the ACC by 0.04 PPP. You can argue that the ACC was better than the Big East, but you can argue just as well that the Big East was stronger than the ACC. And hell, we'd have had to rewrite that whole narrative if Butler had knocked off the ACC tourney champ about an hour later. And did we forget that the team which finished 6th in the 10 team Big East league is still alive in the Sweet 16? All of those latter points would have come up if the media wanted to promote a "Big East > ACC narrative". But the media was driving the ACC bandwagon all season long, so they're going to find the pieces to fit into that.

Is the ACC the best conference this year? No, it's certainly not. The Big 12 was the best, by a landslide. Is the ACC the second best conference? You can make the case, though Sagarin and Pomeroy still both have the Big East second, with the Big Ten and ACC ever so slightly behind. It's close enough that you can make the case for the ACC being second. But please, if you want to argue that the ACC is better than the Big East, don't use NCAA Tournament W/L record. That's an asinine way to judge a bunch of teams that each played over 30 different games.

So enjoy the games today, and enjoy them specifically because the worse team from the worse conference can win over a 40 minute sample size of basketball. If the best team always won, the sport would be so boring that none of you would be bothering to read a blog post about it right now.

As always, feel free to shoot me a comment to this blog post, or tweet at me. I'll do my best to respond to everybody.

Let's get to the games:

Yesterday ATS: 6-2-0
2015 Tournament ATS: 26-18-0
2014 Tournament ATS: 40-23-4
2013 Tournament ATS: 36-30-1
2012 Tournament ATS: 30-35-2
2011 Tournament ATS: 40-26-1
2010 Tournament ATS: 35-25-3

Virginia (-4.5) over Michigan State: This is a fair line. Both of these teams match up well against each other, and there are no obvious mismatches. I give Virginia the edge, though, because Justin Anderson was a different player against Belmont than he was in the ACC tournament. He finally looks close to 100%, and a full strength Virginia team is probably the second best team in the nation. Michigan State is going to need all of their outside shooters to shoot well to pull this upset.

San Diego State (+9.5) over Duke: This is a very "public" line. San Diego State's offense is not good, but their defense is built to stop Duke. They have the Mountain West Defensive Player of the Year in 6'10" Skylar Spencer, and they were second in the Mountain West in defensive 3PA/FGA. In other words, they can leave Spencer on Jahlil Okafor and lock down on Duke's perimeter scorers. If Quinn Cook and Tyus Jones get hot from deep then the Aztecs won't be able to keep up, but I think there's a pretty good chance of San Diego State keeping this a low-scoring slog.

Wichita State (+1.5) over Kansas: I picked this upset in my bracket, for reasons described here.

Oklahoma (-4.5) over Dayton: If you think this is going to be a pseudo-home game for Dayton in Columbus, you might want to ride them here, but I'm concerned about a very undersized and thin Dayton team against a tough, physical Oklahoma defense. You might remember that I said the same thing about Providence's size, but I also said that if Dayton won the game it would be because they used their superior speed to cause problems for a foul-prone Providence team. And they had a 30-to-7 advantage in free throw attempts. Oklahoma isn't foul prone, though. They are much more quick defensively, and were 21st in the nation in defensive FTRate. Dayton's going to have to hit outside shots to win here.

Gonzaga (-6.5) over Iowa: Iowa completely dominated the paint against a totally undersized Davidson team. Gonzaga, however, is one of the few teams in the country that is actually taller and bigger than Iowa. The test for Gonzaga will be keeping a quick, aggressive Iowa front line off the glass. But if they can, they're the more efficient offensive team. Iowa tends to settle against two-point jump shots, while Gonzaga runs all of their offense through the paint. The 6.5 point spread is fair, but I don't think you can take the points unless you really think this game is even, and I don't.

Oregon (+12) over Wisconsin: These two teams played a classic in the Round of 32 last season, perhaps the best game in the NCAA Tournament. This year, Wisconsin is better and Oregon is worse, so I wouldn't expect this one to come down to the final possession again. But Joseph Young will pose a test for a Wisconsin defense that is vulnerable to speedy, penetrating guards, particularly if Frank Kaminsky gets in foul trouble. And even if Wisconsin wins this one easily, they could allow a late backdoor cover, as they did in the Round of 64 against Coastal Carolina. So I like the 12 points.

West Virginia (+1) over Maryland: With this line, just pick who you think is going to win the game. I took West Virginia in my bracket, for reasons explained here.

Northern Iowa (-2.5) over Louisville: I picked Northern Iowa to win this game for reasons detailed here, and unless a game is a total toss-up I wouldn't worry about a spread as small as 2.5 points.


Anonymous said...

I'm not sure I can articulate this well, but a semi-flaw in your logic when it comes to rating conferences (not teams), you say the NCAA isn't a fair place to evaluate and I agree to a point but when you say 1 game shouldn't trump the previous 30, that flies against what you said earlier that games in conference shouldn't change really when you are talking about evaluating leagues, it's not 30 games but more like 15 and even that is misleading because probably half of those game are 'buy' games against clearly inferior opponents. So leagues are more based on how they do against 5-7 games they fared against high opponents early in the season. Seems to me that adding in the end of season games in NCAA play on true neutral courts could swing things around on which league is stronger. You are adding a significant amount of meaningful games to the equation. It's not adding 1 or 3 or however many games to 30 but more like adding 1-3 games to the 5-7 early non-league games.

I also think depth is important to determining the 'best' league but having 6 top 25 type teams is more meaningful than having 10 top 100 type teams. This is obviously subjective though...

DMoore said...

"you cannot judge coaches, teams or programs by results in the NCAA Tournament"

Remember that when the coaching carousel starts this summer, and coaches advance to jobs paying 5 times as much because they had one star player who hit a big shot at the right time.

Jeff said...

Anonymous, you are conflating games played by one team vs games played by a whole conference. For example, let's take your assertion that teams can only be judged against other conferences by the 8 non-"buy" games (using your numbers).

Well, in that case, you've still got the fact that the ACC has 15 teams in it, so you've got 120 games against decent opponents out of conference to judge things on. So turning all of that over because of five or six NCAA Tournament games seems silly in that context. For example, the ACC played more games in the ACC/Big Ten Challenge than in the NCAA Tournament so far. Why doesn't that loss to the Big Ten matter?

And you have to go further, in judging opponents by how they do against expectations based on their opponent. The top of the ACC beat the top of the Big Ten in the ACC/Big Ten Challenge, but they lost the challenge because the bottom of the league got crushed. The ACC had their NCAA Tournament teams all get high seeds. We expect them to have a good record through the first couple of rounds. If we had slid Syracuse, Pittsburgh and Clemson into the NCAA Tournament, the ACC's W/L record would have taken a nosedive.

The computer ratings are iterative, and they take every game into account. There's a reason why a handful of NCAA Tournament games don't move the needle too much in the relative conference ratings.

Jeff said...

And DMoore, I think you meant to end that post with "*cough* Andy Enfield *cough*".

Bob said...

Virginia's offense was awful that game, very stagnant (although they did just miss some shots as well, credit to Michigan State's defense), and the game itself was nearly unwatchable with the amount of free throws on both sides in the second half.

Been following your picks for years now, appreciate the write ups man.

Anonymous said...

I feel like the Nova NC State game is a good indicator and talking point about watching the game and checking the score to determine how "good" a team is. I was at the game and the entire time my son and I were saying, "Villanova is better, better offense, better players, they will find a way to win this...oops Cat Barber just bounced one in off his shoe!" NC State's offense was in disarray the whole game they just made shot after shot and got 50-50 balls at a ridiculous rate. (not fans of either team - just a bball fan) To judge the conference or even the teams by that game is silly, if you actually watched and know what you are looking at.

Jeff said...

Yep. Weird things happen in small sample sizes. The same thing happened in Virginia/Michigan State, where good UVA shooters were bricking wide open threes while Travis Trice was hitting off-balance 24 footers with the shot clock about the expire.