Gaming The System

Notre Dame coach Mike Brey pays close attention to the RPI.

“Once we get into January, it’s like the stock market, the first thing I look up every morning is the RPI,” he said earlier this week. “I don’t pull up anything else, except the RPI. I see where we are – what’s our stock?”

Brey’s interest in the ranking system (RPI stands for ratings percentage index) is pragmatic, not philosophical. He’s not sure the RPI is a good measurement – and certainly not the best ranking system for college basketball. But it is the measurement that carries the most weight with the NCAA Tournament Selection Committee.

It’s not that the committee is a slave to the RPI rankings. But the committee pays a lot of attention to things like a team’s Top 50 and Top 100 wins and its strength of schedule. And how are those factors determined? The RPI.

“It’s the Dow Jones, man. ‘Where are we today?’” Brey said. “I’ve been excited to see us rise from the 60s or the 70s in November to the teens right now. Top 50 and Top 100 RPI wins are something that the committee really looks at. I’m always looking at the people we play and I’m kind of rooting for them to get in the Top 50 or the Top 100. It’s amazing how us coaches root on nights when we’re not playing.”

Indeed, Brey has to pull for Northwestern (No. 42) to keep winning and stay in the Top 50, just as he’d love to see Notre Dame victim VPI (No. 54) move up a little bit to give the Irish another Top 50 win.

Ironically, Northwestern’s recent victory over Iowa both helped and hurt the Notre Dame resume – Northwestern solidified its Top 50 spot, while the Hawkeyes (another Notre Dame victim) dropped out of the Top 100 with the loss.

But everything is in flux with the rankings until Selection Sunday. Brey – and dozens of other coaches – will be checking the RPI every morning to see how it changes.

That might change in the future.

Dan Gavitt, the NCAA’s senior vice president of basketball, and Jim Schaus, the athletic director at Ohio State and a member of the tournament selection committee, will meet with a number of statistical gurus Friday in Indianapolis to discuss other systems for ranking teams. On hand will be Ken Pomeroy (KenPom.com), Jeff Sagarin, Ben Alamar (ESPN’s BPI), and Ken Paiga (KPI).

Depending on the outcome of the meeting, a new metric (or several new metrics) could be added to supplement the RPI or to replace it entirely – as early as next season.

The NCAA’s outreach to other statisticians is in response to a request by the board of the NABC (National Association of Basketball Coaches). Both N.C. State’s Mark Gottfried and Brey are on the board.

“We talked about that in our May meeting with the basketball committee,” Brey said. “I think this is really smart by Danny Gavitt to explore a number of things. Can you get a little more current as far as evaluating?

“I think there is probably a better way. I don’t know what it is yet, but there is a consensus to try and investigate this. So I’m glad they are doing it.”

Several coaches I talked to had a variety of opinions about a new ranking system. But two ideas seem to be unanimous – one positive and one negative.

“We’ve used the RPI effectively, but there are more things you can value,” Miami coach Jim Larranaga said. “One example is a metric that takes into account road wins. I think it is much more difficult to win on the road than at home.”

Several other coaches mentioned the home/road factor. Technically, the RPI has an adjustment for home/road records, but it’s obviously not enough for the coaches I talked to.

On the other hand, there is one metric that no coach wants to see added:

“The only thing I’d be opposed to is any metric that uses margin of victory as a factor,” N.C. State’s Gottfried said, echoing almost every other coach I talked to.

“I think the RPI is a pretty good measuring tool,” Gottfried said. “It does force people to play better non-conference teams.”

However, a number of coaches noted that it’s possible to “game” the RPI to achieve an artificially high ranking. The problem is that there are around 350 Division I teams (actually 351 this season) and they are all ranked 1-to-351. But while there is a huge gap between No. 1 and No. 100, the gap between No. 100 and No. 200 is small – as is the gap between No. 200 and No. 351. Teams that can consistently schedule teams in the 100-200 range fare much better than teams that play too many 200-plus teams – a difference that is not really reflected on the court.

“When I was at George Mason, a mid-major school, we had a plan for doing our scheduling and it worked very well,” Larranaga said. “Our RPI in particular seasons – ’06 and ’11 – we were in the Top 25 in the RPI. I knew that we didn’t play the kind of schedule that teams in the ACC did. So, we took advantage of the guidelines.”

Oddly, Larranaga did not take advantage of the guidelines this season with a team that could easily end up on the NCAA bubble. It’s not that his Miami Hurricanes have played just four Top 50 and six Top 100 opponents so far. That hurts. But six of his non-conference games were against teams that ranked between 217 and 334 in the RPI. His overall strength of schedule will not be bad after Larranaga gets through playing 18 ACC games, but his non-conference strength of schedule – a metric the committee looks at closely for bubble teams – is not going to get much better than the current 294.

For the record, Duke’s non-conference strength of schedule is a solid 42 – four Top 50 opponents and just three from the plus-200 range. There’s a good chance that Duke’s overall strength of schedule – which will continue to climb as the Devils make their way through the murderous ACC – will climb from the current (as of Tuesday) ranking of 18 into the top 10.

THE RACE FOR 11 BIDS

Speaking of the murderous ACC, it’s interesting to watch the league strive to maximize its NCAA potential.

Before the season, it seemed possible that the ACC could land 11 NCAA Tournament bids – matching the record set by the Big East in 2011.

That’s not going to happen now.

It’s fascinating to compare what happened in the Big East in 2011 with what seems to be happening in the ACC this season. To begin with, both are extremely deep, loaded with talented teams. Connecticut, which finished 9-9 in the league and was seeded No. 9 in the conference tournament in 2011, ended up winning the national championship.

On the other hand, the other 10 teams to get bids that year went a miserable 7-10 in NCAA play. Marquette was the only other Big East team to reach the Sweet 16 that year.

Buzz Williams, who coached the Warriors in 2011, is in the ACC now. Notre Dame’s Mike Brey, Louisville’s Rick Pitino and Syracuse’s Jim Boeheim who were also part of the 2011 Big East tidal wave.

They all have teams that are bidding to be part of a 2017 ACC tidal wave, although Williams at Virginia Tech and Boeheim at Syracuse will have to scramble to make it happen.

Here’s the thing. The 2011 Big East had one big advantage over the 2017 ACC – the bottom of the league.

The Big East had 16 teams in 2011 (one more than the ACC this year). Eleven of those 16 teams were somewhere between solid and very good. Three teams (13-18 Seton Hall, 15-17 Providence and 15-17 Rutgers) were pretty weak. And two teams (10-23 South Florida and 7-24 DePaul) were downright godawful.

The top 11 teams fattened up on the bottom five.

DePaul was 0-14 against the top 11 teams in the Big East. South Florida was 0-12. Rutgers was 1-11 (upsetting Villanova at home). Providence was 2-10. Seton Hall was 3-10.

That means that the Big East’s 11 NCAA contenders went 6-57 against the five bottom feeders. That really helped the league’s bubble teams earn their bids.

Before the season, the thinking was that the bottom of the ACC might be nearly as bad. Boston College was DePaul-ugly last season and new coach Josh Pastner was inheriting a mess at Georgia Tech. Wake Forest was coming off a 2-17 (counting the tournament loss) ACC season and had lost its best player.

Unfortunately (for the league’s chances of getting 11 NCAA bids) it hasn’t worked out that way. Both Boston College and Georgia Tech have shown surprising strength.

The Eagles have already beaten Syracuse and N.C. State -- two teams with strong NCAA aspirations. The Yellow Jackets have also beaten N.C. State (on the Pack’s home floor!), plus Clemson and ACC contender North Carolina.

That’s five conference wins so far the ACC’s two patsies – just one less than the Big East’s bottom five teams managed in all of 2011.

Now, North Carolina can shrug off its loss to Georgia Tech, but Clemson, Syracuse and especially N.C. State (with two losses to the bottom feeders) have seen their NCAA chances hurt badly.

To get 11 NCAA bids, the ACC would need FOUR patsies for the rest of the league to fatten up on. Wake Forest us following the scrip (0-4 against the teams with a chance at the NCAA Tournament). But that’s not enough.

As it stands today, the ACC has six of the top 16 teams in the RPI. It would be an earthshaking shock if Duke, Louisville, UNC, Virginia, Notre Dame or Florida State missed the NCAA field. All six will be in – very likely with strong seeds.

But who else?

For all their good work so far, neither Boston College nor Georgia Tech appears to be in the NCAA hunt. Wake has a surprisingly strong RPI at No. 26, especially since the Deacs don’t have a Top 50 win and are 1-7 versus the Top 100. At 10-7 overall and 1-4 in the ACC, the Deacs are hardly a strong NCAA candidate.

Who does that leave?

Well, Pitt and Syracuse – two schools that were part of the 2011 Big East wave – are not making a very good case, Syracuse is No. 125 in the RPI with one Top 50 win – and that’s over Pittsburgh, which has lost five of its last six games and as of today, sits at dead last in the ACC.

Miami, Virginia Tech, Clemson – I’m not sure any of them are in the field as of today, but all three are still close enough to make it with a little surge. For Pitt, Syracuse and N.C. State, it’s going to take a lot more than “a little surge.”

Amazingly, Joe Lunardi’s latest bracket for ESPN has 10 ACC teams in the field – adding Clemson, Virginia Tech, Pitt and Miami to the certain six. Personally, I think that’s a bit optimistic.

Still, it’s a wild and wacky league so far and things could change drastically. Or we could have a wide open tournament as the Big East did in 2011 – won by the No. 9 seed. Whoever winds up ninth in the ACC this season (at the moment, Duke is listed ninth in the standings) could win the tournament and earn an automatic bid. That would certainly be a path for a regular season underachiever such as Syracuse or N.C. State to salvage a disappointing season.

I’m fairly certain the ACC won’t match the Big East’s 11 bids in 2011. I have no idea whether a mid-level ACC team will make a serious national title run as UConn did six years ago, but I’d be willing to bet that no matter how many teams the ACC puts in the tournament, it will end up with a better NCAA record than the 2011 Big East (13-10).

For now, I’ll do as Mike Brey does and check the RPI every morning to see what changes the previous night’s games will bring. Hopefully a year from now, we’ll have a better metric to follow.

But for now, the RPI – as flawed as it is – remains king.