College Basketball & The Computers
Analytics have changed the way we can look at sports. Fortunately, most who make a living analyzing sports have embraced this change. Some never will and I always get a laugh at the college football experts in particular who seemingly "throw their hands up" each week lamenting the computer element of the BCS when at the same time you have a poll that comes about as a result of coaches that don't even take the time to watch half the teams they are voting on.
The "analytical revolution" began in baseball and has worked its way over into basketball as well. It's not yet as prevelant at the NBA level, but in College (where the weekly polls matter far less than in football), I find them to be quite interesting. Particularly this time of year when those of us in the handicapping community are busy prepping for the NCAA Tournament.
The oldest metric that exists in College Basketball is the RPI (Ratings Percentage Index), which factors in a team's won loss-percentage along with strength of schedule. It is 25% comprised of a team's win percentage, 50% opponents' win percentage and 25% the winning percentage of those opponents' opponents.
The RPI was never perfect (doesn't factor in margin of victory) and few to this day understand it, so it was only natural that modifications were made by other entities and/or individuals for judging/ranking college basketball teams. ESPN trotted out its own metric, the very uncreatively named BPI (Basketball Power Index), which supposedly builds in more factors by considering strength of schedule, margin of victory and the absence of key players in any given game. The BPI was hardly the first power ranking to come into play post-RPI. Ken Pomeroy has his own ranking system, which I personally value very highly.
For my first NCAAB handicapping article here at Covers Experts, I wanted to take a look at a handful of teams, whose ranking in the BPI or at KenPom differs significantly from their ranking in human polls and see why.
Three teams I have labeled as "underrated" and three more as "overrated." I've presented both lists below.
1. Pittsburgh - The Panthers just might be the most underrated team in the country, at least according to the computers. Pitt is ranked fourth at KenPom and sixth in the last Sagarin rankings. ESPN's BPI has them 10th. In last week's human polls, they were 16th and 17th. As of a couple weeks ago, they were UNRANKED in the human polls. What gives?
The Panthers did not have the most impressive non-conference schedule, which was rated 295th by the RPI. However, in the other computer rankings, margin of victory is Pitt's best friend. Half of their wins have come by 20 or more points. They had recently turned it on in Big East play with three consecutive 10-point victories before suffering a loss in Saturday's matchup with Marquette, where they were actually a two-point favorite. Despite that loss, you're starting to see that the computers were well out in front of the humans as it pertains to Pittsburgh and it looks as if they were right.
2. Minnesota - The Big 10 is undoubtedly the "toughest conference in America" this season and the Gophers have definitely found this out the hard way. After starting the season 15-1 straight up (only loss was to Duke), they have since dropped seven of ten conference games. They are unranked in both human polls, yet in the Top 15 of both the BPI and KenPom. Why? Four of their losses are to teams currently ranked in the top 20 (not even including Duke) and other than Michigan State (whom the Gophers beat at home earlier in the season), no team had beaten the Gophers by more than eight points. That was until Sunday when they suffered their worst loss of the season at Iowa. This is a team where you can throw won-loss record out. I see a lot of value on the Gophers down the stretch, particularly a week from Tuesday when they will be a home dog against #1 Indiana.
3. Colorado State - This is the most unknown commodity on the underrated list. The Rams play in a deep Mountain West Conference where no clear top team has emerged. New Mexico has occupied first place in the MWC Standings for much of the season, but the Lobos are actually overrated according to the some of the power rankings, who instead have CSU as the best team in the league. At least that's the case at KenPom where the Rams are rated 15th compared to 33 for New Mexico.
Analytically speaking, Colorado State is the best rebounding team in the nation. They have only four losses, none of them by double digits and one coming in overtime. Right now, this team is projected as a 6 or 7 seed in the NCAA Tournament. That would make them a very dangerous second round opponent for some higher seed.
1. Kansas State - The Wildcats came into last Monday's contest with Kansas as the higher ranked team. We all saw on the court how foolish that was as the Jayhawks snapped a three-game losing streak by rolling KSU 83-62 at Allen Fieldhouse. Had you been following some of these computer rankings, you would have known to be on KU. That's because even despite the Jayhawks' losing streak, they are still 8th in the BPI and 12th at KenPom. Compare that to Kansas State, who came in 10th in the human polls and not in the Top 40 in either computer ranking! I realize the Wildcats did bounce back from the loss to Kansas with arguably their best game of the season, beating Baylor by 20. But too much of the Wildcats' resume is reliant on a win over Florida. They are just 2-4 vs. the AP Top 25. Depending on who they are matched up against, it could be a an early exit for Bruce Weber's team come NCAA Tournament time.
2. Butler - What? How could America's favorite darling be on the overrated list? Well, it's pretty simple. The Bulldogs' three best wins - over Indiana, Gonzaga and Marquette - are by a combined FOUR points. They have no wins by 20 or more points, though they recently beat St. Bonaventure by 19. Yet since that time, they have two very narrow wins over George Washington and Fordham, both of whom have losing records, and an outright loss to Charlotte. While nationally ranked (#11 in last week's poll), like Kansas State, Butler is not in the Top 40 in either the BPI or KenPom. In fact, they have dropped out of the Top 50 at KenPom! Although they just got back center Andrew Smith from injury, the Bulldogs are the most overrated team in the country according to the computers.
3. Notre Dame - The Fighting Irish always get some slack from AP voters due to the name on the front of the jerseys. But in addition to being perennially overrated on the football field, they have been a recent 'early out' in past NCAA Tournaments as well. Consider that they have made the Big Dance every year but one (2009) going back to 2007 and not a single time did they make it past the first weekend. Three times they were one and done. In and out of the human polls all season, KenPom currently has Notre Dame as the 60th best team in the nation, which hardly sounds Tournament worthy. The Fighting Irish definitely made the computers "look smart" over the weekend as they lost by 17 at Providence. Their previous three wins have all come via overtime, including a 5 OT marathon over Louisville in South Bend on February 9th. If they were to get a six-seed in the Big Dance as was projected last week, they could be one and done yet again.
- Marc Lawrence
- May 25, 2016 - 8:10 PM
- Offered at:
- bookmaker @ 125 Baltimore
Play - Baltimore w/Wilson vs.McHugh.
Edges - Orioles: Tyler Wilson 4-1 away career team starts; and 2-0 away with 3.75 ERA and 1.08 WHIP this season. Astros: Colin McHugh 0-3 team starts versus A.L. East foes this season, and 5.13 ERA and a 1.59 WHIP this season.
With Houston just 11-22 at night this season, and the Orioles 21-10 under the lights at night this campaign, we recommend a 1* play on Baltimore. Thank you and good luck as always.
> It doesn’t get any better than this - Marc’s red-hot top rated 10* NBA Playoff Game Of The Week and it goes Wednesday night in the Cavs-Raptors Eastern Conference finals showdown. If you’re serious about winning you know exactly what to do!