For what it's worth, Zags have an RPI problem
01-03-18

A TV analyst -- I believe it was Sean Farnham -- weighed in on Gonzaga’s post-season prospects on a recent broadcast. He said he thought that, playing well, the Zags were a “second-weekend” NCAA-tournament team, meaning one that could get to the Sweet 16 under those circumstances.

I can’t argue with that.

If, on the other hand, you were to judge GU’s outlook solely on the RPI computer -- the Ratings Percentage Index -- you’d be debating whether the Zags belong in the tournament.

Fortunately for them, a lot more goes into it -- more than ever, in fact.

Wednesday, the RPI has the Zags at No. 66 -- up from 70 earlier in the week, probably because of Florida’s pillaging of Texas A&M Tuesday night -- and if you took only that cold, hard number, Gonzaga would probably be on the outside looking in at the field.

At the risk of rendering the GU faithful apoplectic, we’ll point out here that Washington is No. 56 in the RPI and Saint Mary’s is No. 43 -- no matter that Gonzaga smoked the UW on the road, and that any sane observer would conclude that GU has more wins of quality than Saint Mary’s.

Scheduling is usually cast as tricky business, and the Zags seemingly missed the mark this year as regards their lower-end opponents. They beat two dead-weight, 300-plus RPI teams -- Incarnate Word at 335 and Howard at 331 -- as well as North Dakota (273) and IUPUI (267). (Never mind that No. 273 just about came into the Kennel and beat the Zags a few weeks ago.)

To you and me, playing somebody like Maryland-Baltimore County seems the same as playing Howard. But UMBC is No. 198, while Howard’s lowly ranking has the effect of being one more heavy drag on Gonzaga’s RPI.

In other words, all stiffs are not created equal.

The bad news for the Zags is that the RPI has the West Coast Conference ranked a miserable 14th, immediately behind scofflaws like the Colonial (13), Summit (12) and Mid-American (11). In its better years, the WCC has had a single-digit ranking.

Indeed, that 101-52 win over Santa Clara last weekend came against the RPI’s No. 345 team. This week brings Pepperdine, which is No. 321. All of it seems to say to me that if anything, the Zags’ RPI is apt to worsen, unless Gonzaga does something dynamic -- let’s say, lose only once between now and the end of the WCC tournament.

How important is the RPI? Well, the NCAA basketball committee has steadfastly maintained over the years that it’s only one tool in the selection/seeding process. I think it’s safe to say it’s true that a team’s RPI is only a modest consideration, but on the other hand, we constantly hear quoted a team’s record against the RPI top 50, the top 100, or how many (inconsequential) wins it had against teams 200 or lower.

Fortunately for Gonzaga, the trend is for the committee increasingly to use other metrics. I ran the question by David Worlock, the NCAA’s director of media coordination and statistics, and he responded that the NCAA will use KenPom, Sagarin, KPI (that was a new one on me) and ESPN’s BPI.

In marked contrast to the RPI, both KenPom and the BPI have Gonzaga at No. 9.

Meanwhile, as it concerns the RPI, the committee will use an altered formula starting this year. Instead of assessing a flat number of top-50 or top-100 wins, it will assign value to any game result and put it in one of four quadrants. Quoting directly from Worlock’s e-mail:

“We have altered the team sheets that the committee uses to help evaluate the teams, and those place a greater emphasis on winning away from home. Whereas in the past, each team’s schedule was divided into four quadrants sorted solely by the RPI (1-50, 51-100, 101-200, 201-351), now they are sorted by the game location and the RPI. The new quadrants are as follows:

“Quadrant 1: Home games against 1-30, Neutral games against 1-50 and Road games against 1-75.
“Quadrant 2: Home games against 31-75, Neutral games against 51-100 and Road games against 76-135.
“Quadrant 3: Home games against 76-160, Neutral games against 101-200 and Road games against 136-240.
“Quadrant 4: Home games against 161-351, Neutral games against 201-351 and Road games against 241-351.

“In other words, a road win over a team ranked 73rd in the RPI is in the same quadrant as a home win over 28. In the past, that road win would have been in quadrant 2 even though statistically it is just as hard to beat 73 on the road as it is a team ranked 28th at home.”


By my reckoning, Gonzaga is 2-2 in Quadrant 1, 2-1 in Quadrant 2, 0-0 in Quadrant 3 and 8-0 in Quadrant 4.

What’s it all mean for the Zags? There’s no doubt it’s a good thing respected metrics like KenPom are in use. The more of them that tie a good number to Gonzaga, the more it might cast the RPI as an outlier. In that vein, Sagarin has the Zags at No. 10.

On the other hand, the fact the committee is refining the way it looks at any result, but still relying on the RPI to inform those quadrants, shows that the RPI is still alive and kicking. And that 8-0 in Quadrant 4 could be a concern for Gonzaga, especially as the number prospectively balloons in the soft WCC.

As always, it’s about winning games. For the Zags, anything that keeps them away from seeds like 8 and 9 is a good thing.