Saturday, March 18, 2017

Day 5 Picking The Lines

People are complaining about how chalky the NCAA Tournament has been. It's the first year without a 13+ seed in the second round since 2007, and even the 12 seed that won (Middle Tennessee) was favored in Vegas. But despite that, we have had some really good basketball games, and the lack of upsets sets us up for some absolutely fantastic second round match-ups. First round upsets are all fun and games until you end up spending your Saturday or Sunday watching a 12 and 13 seed playing each other. This weekend will, instead, be packed with great match-ups.

There are no bad teams left. Every team tipping off today has a legitimate argument that they deserve to be here, and should expect to be competitive. Let's get to the match-ups:

Yesterday ATS: 9-6-1
2017 Tournament ATS: 18-16-2 (53%)
2016 Tournament ATS: 36-30-1 (55%)
2010-15 ATS: 220-167-11 (57%)

West Virginia (-2.5) over Notre Dame: Notre Dame's depth and front court size are going to be tested by Press Virginia. Matt Farrell has been perhaps the most improved player in the nation this season, and his ability to take care of the ball against relentless pressure will be key to the Irish potentially winning this game. One other advantage that Notre Dame has is their 79.5% free throw shooting considering how much West Virginia fouls. But aside from their match-up advantages, West Virginia has simply been the better team this season.

Villanova (-6) over Wisconsin: This game is an interesting contrast in styles. On one hand, Villanova is going to struggle to defend both Ethan Happ and Nigel Hayes in the post, as they just don't have much big man depth after Darryl Reynolds. On the other hand, Villanova's wings are so explosive and athletic that I'm not sure how Wisconsin can defend all of them either. Villanova is also a team that is awfully efficient offensively, while Wisconsin is a team that can get into a rut when their outside shots aren't falling (they were falling, of course, in that Virginia Tech game). In the end, it's a tough call on this spread. I'm going with Villanova because they hit 79.2% at the free throw line this season, so if they're winning at all in the final minute I think they'll hit enough free throws to cover.

Gonzaga (-11) over Northwestern: This is a big spread, but it's a fair spread. The difference in this game, more than anything, is Gonzaga's ridiculous size and height. In the first round, Dererk Pardon was able to defend 7-footer Luke Kornet by using his physicality and shoving the skinny Kornet out of the way, but that isn't going to fly against the monstrous Przemek Karnowski. Without much in the way out of outside shooting (Northwestern was 13th in the Big Ten with 30.9% three-point shooting in conference play), I expect Gonzaga's ownership of the paint to deliver a victory.

Xavier (+6.5) over Florida State: Florida State is the better team, but not by much, and this game could easily go down to the final possession. Florida State is a massive team, but they're vulnerable to teams that are aggressive in the paint. Xavier, who led the Big East in offensive rebounding rate and FTRate, fit that bill. That said, if there's a quick whistle for both teams, Florida State has significantly more depth.

Butler (-3.5) over Middle Tennessee: Middle Tennessee was a one point favorite against Minnesota, so it wasn't even an "upset" when they crushed Minnesota (that final score was deceptively close, to be honest). This Middle Tennessee team is awfully good, and it's a shame that RPI obsession would've relegated them to the NIT had they failed to win the Conference USA tournament. But Butler is not short-handed Minnesota, and their ability to pass the ball and executive on offense will challenge a Middle Tennessee defense that is not particularly strong inside. With a spread this small I'd only take Middle Tennessee if you really think they're going to win, and I expect Butler to pull this one out.

Saint Mary's (+5) over Arizona: I chose Saint Mary's to win this game outright in my bracket. Offensively, they are one of the most efficient teams in the nation, and can light up the scoreboard with 40% three-point shooting. Defensively, they have enough size to hang with Arizona. Jock Landale vs Lauri Markkanen will be a hell of a match-up. Despite the Vegas line, most computer ratings have this game significantly closer - basically a toss-up. I'll take the points.

Virginia (+2) over Florida: This is an Elite 8 game in the 2nd round. The two teams are 8th and 9th in the Pomeroy ratings at the moment, and both are among the best defensive teams in the nation. Considering that Florida will likely not be getting as many baskets as they usually get off of turnovers, this game will likely come down to whether or not Florida hits their outside shots at a high rate. This is definitely a game to enjoy, but it's not really one where anybody can have a strong opinion on who should be favored. Personally, I just think Virginia is a slightly better team than an Egbunu-less Florida.

Purdue (PK) over Iowa State: I took Purdue in my bracket simply because I think they're a really tough match-up for Iowa State. The size that they bring with Caleb Swanigan and Isaac Haas is going to be so hard for Iowa State to deal with. Outside of 6'8" freshman Solomon Young, Iowa State's only big man is 6'5" Deonte Burton. And defensively, I'm not sure that I buy that Iowa State will have big mismatches with Burton against defenders like Haas - if it is Burton and not Monte Morris who is Iowa State's primary playmaker then that is a huge problem for the Cyclones. Purdue's three-point defense is very strong as well.

12 comments:

Anonymous said...

Using 'the computers have this closer than the Vegas line' is such a ridiculous way to justify picks. More often than not, the Vegas line is factoring in additional factors that the computer ratings often don't account for at all such as rest days, injuries, athleticism, defensive styles (man, zone, press, etc.), altitude effects, and the list goes on. There's a reason the Vegas line is consistently the best predictor of games compared to public computer ratings as tracked at thepredictiontracker.com. If it were so easy to beat Vegas point spreads simply by taking the difference in two team ratings from KenPom, Sagarin or whatever public rating you choose, then any idiot should be able to win long term against the Vegas spreads just by doing that. Except that never happens because it's NOT that easy.

Jeff said...

I have not seen any analysis suggesting that the Vegas line is more accurate than KenPom, or any of the other top ratings. The reason why betting KenPom doesn't guarantee profits is because the Vegas line and KenPom line are almost always within 1 point of each other, and because you need to clear bets 55% of the time or so just to make back the vig. Even if KenPom is more accurate than the Vegas line, just betting that way on every game is not the path to profits.

Anonymous said...

These are the results for this season for college basketball. http://www.thepredictiontracker.com/bbresults.php

Not one system with a reasonable sample size is better than 52.4% ATS which is what you need to hit just to break even versus -110 lines. KenPom isn't tracked there this year, but it has been in the past and it didn't hit greater than 52.4%.

This is the case literally every year for all of the sports tracked there, primarily because of what I cited before in regards to how the line incorporates multiple factors that the computers don't factor in.

This comment you made is also nonsense:

"Even if KenPom is more accurate than the Vegas line, just betting that way on every game is not the path to profits."

If KenPom was sufficiently more accurate overall than the Vegas line by at least 1 point on average to overcome the vig, then betting that way on every game would yield profits long term.

You're obviously very uniformed when it comes to gambling so you should probably just stick to hoops.

Jeff said...

I never said that KenPom is consistently 1 point better than Vegas. Please read what I wrote rather than being trigger-quick to fire off personal insults.

Anonymous said...

Well you also said you never had seen evidence that the Vegas line is more accurate than any rating system and I showed you a link where the line has the lowest mean square error among multiple common power ratings. The information is out there if you took five seconds to google it.

Jeff said...

I don't see KenPom on that website. And I will ask you for a second time to please refrain from personal attacks. Stay polite and I will be happy to discuss this issue with you.

Anonymous said...

Tell the anonymous dude to leave! We appreciate your analysis Jeff. Keep it up!

What a loser he is!

Anonymous said...

I'm aware of that. I mentioned before that KenPom wasn't on there this year. It was a few years back when KenPom projections were available for free. I'm guessing it stopped being tracked there when KenPom went to a pay-model to see his projections. The site doesn't appear to keep archives for the college basketball results like they do for NFL and College Football.

KenPom is correlated extremely closely with Sagarin, so it's highly unlikely KenPom would have a better absolute error or mean squared error than the Vegas line in the long run. Occasionally a few systems will beat the Vegas line in those areas even with a decent sample size in 1 season, but no system consistently ranks higher each year than the Vegas line in the long run.

That doesn't mean there aren't a very small percentage of individuals who can win betting sports. It's just the people who are winning are using more sophisticated systems than KenPom, Sagarin, etc. which aren't publicly available and go beyond just assigning one general rating to each team to use as the basis for prediction. They're often using player level ratings like Real Plus Minus in the NBA and looking at how teams interact at a matchup level across various stats. In the NBA there are gambling groups who factor in officials as well, particularly when it comes to projecting totals. SportVu data can also be utilized to look at team tendencies in regards to frequency and success rates of shots from various spots on the floor.

Jeff said...

In the NCAA Tournament, the dominant reason why a line is off from the computers is not some advanced SportVu analysis, but because of public perception. Teams like Wichita State and Saint Mary's are going to trail their KenPom line by a couple of points simply because the public isn't going to believe that they're really as good as KenPom has them. Similarly, Duke typically trends a point or two ahead of their KenPom line.

And that was, of course, my point above. Saint Mary's is going to be good gambling value in the NCAA Tournament because the Vegas line is always going to be a "public" line.

sumon roy said...

NCAA Tournament Bracket

NCAA Tournament Bracket Live

NCAA Tournament Bracket Live Stream

NCAA Tournament Bracket Live Streaming

NCAA Tournament Bracket Live Online

Watch NCAA Tournament Bracket

Watch NCAA Tournament Bracket Live

Watch NCAA Tournament Bracket Live Online

NCAA Tournament Bracket 2017

NCAA Tournament Bracket 2017 Live

NCAA Tournament Bracket 2017 Live Stream

NCAA Tournament Bracket 2017 Live Streaming

NCAA Tournament Bracket 2017 Live Online

Watch NCAA Tournament Bracket 2017

Watch NCAA Tournament Bracket 2017 Live

Watch NCAA Tournament Bracket 2017 Live Online

2017 NCAA Tournament Bracket

2017 NCAA Tournament Bracket Live

2017 NCAA Tournament Bracket Live Stream

2017 NCAA Tournament Bracket Live Streaming

2017 NCAA Tournament Bracket Live Online

Watch 2017 NCAA Tournament Bracket

Watch 2017 NCAA Tournament Bracket Live

Watch 2017 NCAA Tournament Bracket Live Online

NCAA Tournament

NCAA Tournament Live

NCAA Tournament Live Stream

NCAA Tournament Live Streaming

NCAA Tournament Live Online

Watch NCAA Tournament

Watch NCAA Tournament Live

Watch NCAA Tournament Live Online

NCAA Tournament 2017

NCAA Tournament 2017 Live

NCAA Tournament 2017 Live Stream

NCAA Tournament 2017 Live Streaming

NCAA Tournament 2017 Live Online

Watch NCAA Tournament 2017

Watch NCAA Tournament 2017 Live

Watch NCAA Tournament 2017 Live Online

2017 NCAA Tournament

2017 NCAA Tournament Live

2017 NCAA Tournament Live Stream

2017 NCAA Tournament Live Streaming

2017 NCAA Tournament Live Online

Anonymous said...

The public has very little impact on betting lines outside of maybe the Super Bowl. Sharp money far outweighs public money in college basketball and sportsbooks respect action from accounts that have been profiled as sharp more than Joe Public bettors. Pinnacle and CRIS, (which are the two major offshore books that every Vegas book copies lines from) profile their users and accounts that win on a consistent basis will move the line more than a lifetime loser account even if the loser is betting more money. Again, you really have no idea how any of this works so should you should quit while you're ahead.

If St Mary's really should have been favored by 1 point over Arizona as KenPom suggested, then there would have been multiple bets from sharps at 10K a pop (since Pinnacle/CRIS allow you to bet again and again once the line moves) to move the line to where it should have been. The average person is betting like $20 a game and it would take 500 people just to equal 1 bet from a betting syndicate at that rate. But keep thinking lines are set based on public perception. If that was the case then any idiot could make a fortune fading "public" teams like Duke, Dallas Cowboys, Lakers, etc. But if you check the long term cover record for Duke, they are 51.4% ATS since 2003. If you had bet $100 against Duke every game since 2003, you would be down $3880.

Jeff said...

I have no problem debating topics with people who want to argue constructively. If you insist on constantly telling me how stupid I am, I have little incentive to want to engage.