Music to the ears, by the guys' Mercer coach

[DobbsRover2]

From Bob Hoffman, coach of the men's Mercer team that ousted Duke in the first round game.

“You don’t want to get me started. I would tell you what I think about the R.P.I. I don’t know how you decide, but to me, that’s a flawed instrument.”​

He noted how the weaker mediocrities from the P5 conferences get selected for at-large bids over the good mid majors because RPI skews the decisions against them. The arrogant Duke coach K had sounded off after the selections about how mid majors like the men's Atlantic 10 teams were getting too many selections, so the Devils going down to Mercer was really, really, really sweet, and Hoffman was glad to put K in his place on his delusional potty. Sorry K, it's a new era, and the men's mid majors have taken their place near the top.

Now there are some vocal guys on the BY who swear by the RPI crystal meth and cite it every chance they get. And there are those who say it's a well-known given that RPI is useless, and why even mention it? But the fact is, throughout February and March we have the ESPN experts and much of the other commentators using the RPI in their references. Granted the women's mid majors are not at the level of the men's mid majors (partly because of the one-and-done careers of top male players), but the BYU win and close losses by FGCU and Fordham and quite respectable performances by teams like Western Kentucky and Chattanooga and South Dakota (and St. Johns is now officially a mid major) all indicate that the mid majors could be on the rise and RPI is ever less of a real marker of teams' values.

Now if one of those teams from one of the mid major conferences can win the NC this year, everything will be really peachy and we can all be on cloud 9 here.

 

[VAUConnFan]

It goes without saying that an overplayed narrative can become ingrained in one's mind and skew one's perception. However, I fail to accept the notion that the one and done basketball players are having this tremendous impact on college basketball other than adversely skewing graduation rates. Given the relatively low number of one and one basketball players in relation to the number of college basketball players, even the pundits, the so-called experts have never, ever fully explained their positions on the one and doners and their impact on the game. They throw out the comments because everyone else does and lead us all to that believe that it is a detriment to the game. Now, I will accept that familiarity and team chemistry built over the course of years is great for the game, but I do not believe the game suffers when teams try to win with youth. Individual teams might suffer, but the game, itself, goes on; And it gets better, adjustments are made, attitudes change, the players come in more ready to compete at the college level. Even Coach Auriemma's attitude toward starting freshmen has changed. Quite possibly, it changed out of necessity. Even if this is the case, has it adversely affected the results? Has the quality of UConn's play on the floor shown any tangible displays of deterioration? Does anyone doubt that if Coach Auriemma was compelled to start a freshman laden team that his teams would still compete for a national championship? Caron Butler left for the NBA after his sophomore season and UConn still won the title two years later. Maybe it can be said that Butler's departure cost them back to back titles, but UConn continued to put a quality product on the floor.

And this attitude change can be seen in the other sports. It used to be NFL head coaches would not dream of starting a rookie quarterback: had to learn the game/get seasoned on the sidelines a few years. Now, rookie QBs are given the playbook and the keys to the practice facility before they are even drafted.

I refuse to buy the notion that the one and done players are having this significant an impact. The mid majors are gaining strength because everyone is getting better. The recruiting pool is bigger than it ever was. The result - Weekends like this one in the NCAA Tournament. Given time, the women's game will get there also.

 

[DobbsRover2]

The one-and-done does not apply to WCBB, and clearly you can ride a freshman to an NC on both the men's and women's side. It is also possible that a game like the Mercer-Duke game where a band of seniors defeated a freshman-led Duke team whose star got the tourney jitters is a singular event that skews perceptions of how important the one-and done factor is. The number of players going to the NBA after one or two seasons is still small, but the cumulative seasons that they would have played for their teams in the old days is much larger. And some of these mid majors can build a veteran corps that get just enough advantage from experience to knock off a Duke. Many of the P5 teams are always in a bit of an experiential flux season after season.

But right, I'm not sure how you would design a study to analyze the issue with real numbers, or whether this is just an excuse for the former powers that the pundits throw out there.

 

[VAUConnFan]

The RPI, like the SAT, is an indication, a measuring gauge for future success. Are they infallible? Of course not, but there has to be a way to measure strengths and weaknesses. The fact that the odds are overcome in the respective athletic and academic arenas make it necessary for those making respective decisions to evaluate other criteria. But, we cannot also pooh pooh the RPI, then analyze a team's performance with comments like," their schedule did not prepare them for this tough game" when they are so closely related. Plus, having the RPI compels teams to beef up their schedules which, in turn, creates more compelling contests. Plus, how often does the RPI actually get turned on its head? Certainly not enough to reduce its weight in the decision making processes.

 

[DobbsRover2]

Not sure if you were being tongue-in-cheek about the RPI encouraging teams to "beef up their schedule," as we've had numerous threads about the kind of scheduling the RPI does encourage teams to do, which is basically to game the system with a lot of teams you're pretty confident of beating at home and avoid the tough home games. Basically just go for wins against mediocre OOC competition so you can finish off the first part of the season at 13-0 like Arkansas and then hope you can go at least 7-9 in the SEC and likely get an NCAA bid with your 20 wins. Didn't quite happen for the Hogs, but they gave a game (the system) effort.

As the Mercer men's coach says about the brain-dead system that had UConn at #4 at one point in the season when Massey had the Huskies schedule rated #1, there's got to be a better way to rate the teams, and to encourage them to play a better schedule.

It's problematic to figure out how well the RPI judges the women's NCAA tournament because of the huge home court advantage for so many of the favorites, and a system would have to be really really bad to screw up a rating of a UTenn vs Northwestern State or UConn versus Prairie View and most of the first and second round games. The RPI favored team in the NC did win last year, but in the semis, both of the favored teams lost.

But the Mercer coach was talking about the seedings on the men's side when he was attacking the RPI. What percentage of your predictions should be correct if your system is accurate? At least 80%? You get at least the picks in the 1-16 and 2-15 basically free in the round of 64, and the 3-14 and 4-13 slots really shouldn't be that hard to judge. The middle ground of 16 games in the 5 through 12 seed are tougher. RPI favored teams won 23 of 32 games in the round of 64, or 71.8%, so not a great result. In the round of 32, the RPI favored teams won 11 of 16, or 69.8%, so the system had worse results when you go another round. How many of the top 16 RPI top teams should be left heading into the Sweet 16? Again, I would say at least 3/4 if you have a system that is working. Only 8 of them did win through though, with the #3 through #5 teams gone, and 50% just seems pretty bad to me. I'm sure the RPI proponents would say batting 50% is great results, but I'd say that's a cop-out for a system that works to distort the seedings and just has poor results in rating the teams.

 

[intlzncster]

The RPI, like the SAT, is an indication, a measuring gauge for future success. Are they infallible? Of course not, but there has to be a way to measure strengths and weaknesses. The fact that the odds are overcome in the respective athletic and academic arenas make it necessary for those making respective decisions to evaluate other criteria. But, we cannot also pooh pooh the RPI, then analyze a team's performance with comments like," their schedule did not prepare them for this tough game" when they are so closely related. Plus, having the RPI compels teams to beef up their schedules which, in turn, creates more compelling contests. Plus, how often does the RPI actually get turned on its head? Certainly not enough to reduce its weight in the decision making processes.

But it can be so easily gamed. I don't know why more teams don't do it. Cut all 200-300 RPI teams from your schedule. All cupcakes should be 150-200, with approximately 50% being 100-150. UConn men are going to have to game it a bit in the future(and I hope they do), because the bottom of the AAC drags down the RPI so egregiously.

Realistically, any team in the top 100 should never schedule a team 200-300 if they can help it. That's counter to desire for a 'regional rivalry' thing and unfair to those teams.

I think there should be a heavy emphasis on Top 50 games, followed to a lesser extent by 50-100, regardless of the bottom end teams you play (given you didn't play all 200+ games). JMO

 

[DobbsRover2]

On the men's side I would extend the focus to the top 100 teams, as a team like Mercer is a prime example of a decent team that might not fit in any top 50 ratings (#80 in RPI anyway). Massey has a system that does weigh a more balanced ratings system tagged to each team's level, and thus giving UConn's games against top teams far more weight than most other systems, but of course it does use margin of victory to some extent, which is a n0-no with the selection committee, which just wants wins and losses no matter who you played. A win at home over Middle Tenn St. wasn't worth that much this year in Massey, but in RPI it was worth far more than wins over FSU or MSU, two teams still cooking in the tourney.

 

[VAUConnFan]

A baseball player who hits safely three out of ten times is considered a good hitter. Despite the fact that he fails 7 out of ten times, he is still worthy of adulation. While the threshholds of success predictions stated above are reasonable and logical, who is to say that the powers that be and/or the RPI creator have not set threshholds at batting average levels?

The fact is we are still dealing with athletes playing the games, humans officiating the games, and any number of factors that come into play on any given day at any given time for which the RPI cannot be factored. The biggest thing for which the RPI certainly cannot account are matchups. Basketball is the ultimate game of matchups. Duke, despite being a perennial major power, has, in recent years, wrapped some ugly losses to lower seeded teams around a national championship; Lehigh, VCU, and recently, Mercer. Losses suffered in situations where they did not necessarily play badly, but locked horns in more equitable matchups.

 

[DobbsRover2]

Sure, but that type of logic can take you anywhere. A 10% success rate isn't usually considered that great, but if I win the megamillion jackpot 10% of the time I play, I can buy court-side seats for the UConn games the rest of my life. You do have to have a base level of competence for a system you employ or you may as well be flipping a coin, and getting 50% of the Sweet 16 picks right makes me think the coin is a more logical tool to use than the convoluted RPI system. Matchups can indeed affect results, but there might well be relatively poorer but well RPI regarded teams that employed favorable matchups to get to the Sweet 16. At some point you have to stop making excuses for a system and say, bottom line, this is the results it needs or it fails.

I would have loved to hit .500 in Little League, but getting a 50 on my science exam would have been a real downer. In my book, the RPI gets an F. Or with the 69.8% R32 results, maybe a D+.

 

[UcMiami]

Any rating system that doesn't account for relativity of teams rated is bound to be pretty bogus. 50% of the RPI rating is based on your opponents winning percentage, so playing the top teams in really bad conferences is much better for your RPI than playing good teams in competitive conferences. All you have to do to understand the problems with RPI is to look at the bottom of the SEC - those teams played dreadful OOC but racked up gaudy records in doing so meaning that 50% of their SEC brethren's RPIs was based on teams like Arkansas having an overall 19-11 record or MissSt with 19-13 record or Georgia's 20-11.
And for a top team - the challenge is playing a team ranked 100 vs a team ranked 325 especially on the women's side is exactly the same but the effect is astronomical. As for using point spread, not sure exactly how Massey or Sagarin handle it, but I am pretty sure it is not literal but a sliding scale - <5 pt or overtime win/loss, 6-12 point win/loss, 13-20 win/loss, >20 win/loss or something like that which I think could elevate some of the NCAAs concern - drop out the top end would work as well - I don't think anyone is comfortable with a 12 point lead until the last minute anyway so it wouldn't effect how teams play. AND they are doing it anyway - you cannot tell me that if Louisville had lost all three games to Uconn, but one was in overtime and the other two were decided by 2 points and 5 points that they would have been seeded as a 3.