Here is a look at the biggest differences between the RPI vs Massey rankings. There is a very clear East vs West bias between the two rankings. Any other comments? Code: School " W L T Massey Conf RPI Diff San Jose St " 6 5 3 58 MPSF 120 -62 San Diego St " 9 3 2 54 P12 96 -42 Missouri St " 6 6 2 73 MVC 115 -42 San Francisco " 5 6 2 87 WCC 128 -41 Cal Poly SLO " 6 6 2 39 BWC 78 -39 Bowling Green " 7 4 3 96 MAC 134 -38 Villanova " 5 6 4 99 BE 137 -38 Missouri KC " 5 6 1 127 SL 164 -37 Air Force " 8 5 2 53 MPSF 89 -36 Santa Clara " 5 5 3 94 WCC 130 -36 " Adelphi " 6 7 1 135 ASC 101 34 Yale " 6 5 2 71 Ivy 35 36 NC State " 6 8 2 141 ACC 103 38 La Salle " 4 6 3 140 A10 99 41 Quinnipiac " 5 7 0 153 NEC 112 41 Hofstra " 6 8 0 107 CAA 64 43 Navy " 4 7 3 151 PL 107 44 Tulsa " 5 10 0 134 CUSA 88 46 Clemson " 5 8 0 105 ACC 58 47 Columbia " 5 7 1 119 Ivy 70 49 (note Massey and the NCAA use slightly different teams which I did not adjust for)
Very interesting. To add perspective, can someone explain the difference in the formulas the two ratings systems use that allow them to come up with these differences.
Sure. The NCAA RPI is basically a forced, arithmetical device that weights the home, away and neutral field record of a particular team to obtain a "rating". The RPI ranking of a team is the order of its rating in the group. It's the stuff of spreadsheets, and can get tweaked by arbitrarily adjusting the weighting factors, including last 10 games, etc. The Massey system is a least squares, linear-regression, statistical analysis that primarily considers the margins of victory for a team, and the margins of victories for the team's opponents, adjusted for home/away/neutral fields. Thus, the NCAA RPI essentially reflects the order of teams with the best records (particularly away), while Massey essentially reflects the order of teams with the best margins of victory by comparison to their opposition's. The Massey rating (ranking) is arguably the more useful of the two for describing the quality of a particular team. The numerical difference between the RPI and Massey rankings is not particularly relevant, except to indicate the differences between the two approaches.
Thanks for the information. However, isn't RPI more about who you schedule than how well you do? How is Yale at 35 (ie into the NCAAs) at 6-5-2 and San Jose State with the same record at 120? For example how is La Salle at 99 while San Diego State is at 96? Yale's wins are against terrible teams and their some of their losses are against very average ones.
Yes it is. RPI is a relative measure of record against the weighted, aggregate record of your opponents (and thus schedule). The Massey rating is more about the game scoring results.
The 10.30 verson of the biggest differences: Code: School " W L T Massey Conf RPI Diff San Jose St " 7 5 3 51 MPSF 113 -62 CS Sacramento " 8 6 2 56 MPSF 102 -46 E Illinois " 7 6 1 115 MVC 155 -40 Cal Poly SLO " 6 7 3 47 BWC 86 -39 San Diego St " 10 4 2 62 P12 101 -39 Longwood " 9 6 1 108 ASC 147 -39 Edwardsville " 7 7 3 105 MVC 143 -38 San Francisco " 6 7 2 96 WCC 133 -37 Missouri KC " 5 7 2 142 SL 178 -36 CS Bakersfield" 10 3 4 11 MPSF 44 -33 Missouri St " 7 7 2 73 MVC 106 -33 " St Louis " 6 9 1 109 A10 76 33 Princeton " 4 8 1 132 Ivy 99 33 UNC Wilmington" 3 12 1 176 CAA 140 36 Kentucky " 8 9 1 118 CUSA 80 38 Towson " 2 11 1 187 CAA 149 38 Clemson " 6 8 1 79 ACC 35 44 Hofstra " 6 10 0 133 CAA 89 44 Columbia " 6 7 1 106 Ivy 59 47 MD Baltimore Co 7 7 3 127 AEC 79 48 Quinnipiac " 5 7 0 153 NEC 96 57
While there are flaws in the RPI, you have to be skeptical of a ratings system that has Bakersfield 11th.
Clemson has wins over Maryland (2), Boston College (9) and South Carolina (22). They also have wins over Gardner-Webb (123), @NC State (98), Adelphi (128) and @Wofford (148, though the game was played Monday and didn’t factor into this week’s RPI.) and a tie @ East Tennessee (69). Seven of their losses were to teams likely to get NCAA Tournament bids: @North Carolina (8), Charlotte (12), @UAB (13), Duke (24), Furman (25), Virginia (29) and @Wake Forest (42). Their only “bad” loss was @ No. 109 Elon (which is 5-10-2 yet is somehow 2-0-1 vs. the ACC, having also beaten Wake and tied Duke yesterday.) That’s 10 games against teams rated 42 or higher. You play a tough schedule, you get rewarded. They play @146 Virginia Tech to finish the season and even if they win their RPI will likely drop a bit because the RPI of their last two wins will be so low.
For the record, I agree that Bakersfield is too high in Massey, but I also think it's too low in the RPI. They've posted a good record against mostly solid opposition, and my own first attempt at a rating system put Bakersfield as one of the better teams in the west. What's incredible is how differently the two rating systems view strength of schedule for the two teams being discussed. Massey: Bakersfield #14, Clemson #26 RPI (via GauchoDan): Clemson #1, Bakersfield #86
This SOS is the heart of why the two are so different. Massey's SOS is based on his actual rankings, where rpi SOS is based primarily on opponents record, so a team that rpi thinks is worse than another may strengthen your schedule more, if they have a better record than a team that is higher rated, with a worse record
Quite frankly where is the tough schedule? All you did was name teams that are in Clemson's conference. Woffard, Elon,Adelphi, Charlotte, Furman is NOT a tough schedule. Your 6-8-1, your 6-8-1 I dont give a shit who you play!!! Losing record period!!! There are weaker conferences where teams are required to play Friday and Sunday back to back weekends...that is no cake walk especially when you are talking about traveling!
I'm sorry, but this post makes you look like an idiot. First off, Charlotte and Furman are very good this year. That's why they have a 12 and a 25 next to them in parenthesis, because that's where they're ranked in the RPI. Their combined record is 25-5-6. They also played OUT OF CONFERENCE GAMES against UAB (13) and South Carolina (22). That's four non-conference games against teams in the Top 25 of the RPI. How many other schools have four non-con games against teams in the Top 25 of the RPI? (I don't know the answer but I know it's not a lot.) You can quibble all you want with the RPI but to suggest Clemson hasn't played a difficult non-con schedule means you're either disingenuous and have an agenda... or you're an idiot.
That is only because you underestimate the quality of teams in the West. There are no Patriot Leagues in the West where Ivy League schools can play very easy to beat teams with good win/loss percentages. Instead they much more difficult ones that are often below .500. Look at Akron. They were 1-1-1 against WCC teams that are just 5 games above .500 and 11-1-2 against the rest of the country. One interesting thing about the differences between the two rankings is how West schools with more modest budgets have the biggest differences in rankings while schools that likely have more money such as Stanford and Cal actually have RPI ratings well above their Massey ratings.
Hard to read Akron's early trip and how they would do against those same teams now. Akron weren't exactly complimentary of perceived treatment at Fullerton, so not sure if results like that are indicative of stronger or weaker teams or represent a game plan in that particular game that worked. Again over the long haul the RPI for good is a better rule of thumb than a 3 game stretch. I also would not assume that bigger schools have bigger budgets and smaller schools less. They may, but I have certainly seen the opposite. I recall that when Bobby Clark was at Stanford the field and practice facilities were less than adequate for such a big school.
listen dickhead no need to throw the word idiot around. the fact is furman and UAB are new on the block this season, they both just appeared in the top25 for the first time together (nscaa). And I am not gonna go by the RPI because I would be arguing versus my own point. ill give you charlotte (even though there schedule has 4 nonconference cupcakes too). Safe to say south carolina sucks too losing 4-0 to NC. in the end 2 tough opponenets non conference doesnt make your schedule tough.
Actually Akron was 1-1-1 vs Big West Conference. They did not play any West Coast Conference schools.
The 11.8 verson of the biggest differences: Code: School " W L T Massey Conf RPI Diff San Jose St " 7 6 3 54 MPSF 119 -65 San Diego St " 10 4 3 59 P12 118 -59 Edwardsville " 8 7 3 89 MVC 140 -51 W Illinois " 10 5 1 63 SL 108 -45 Longwood " 10 6 1 108 ASC 148 -40 Iona " 15 3 0 21 MAAC 57 -36 Valparaiso " 8 5 4 86 HL 122 -36 E Illinois " 7 7 1 135 MVC 171 -36 Missouri KC " 5 8 3 150 SL 184 -34 Cal Poly SLO " 6 9 3 56 BWC 89 -33 Niagara " 9 4 4 71 MAAC 104 -33 Cleveland St " 9 8 1 105 HL 138 -33 " St Louis " 6 10 1 117 A10 86 31 Stanford " 4 9 2 145 P12 114 31 Columbia " 8 7 1 87 Ivy 51 36 William & Mary" 10 8 0 67 CAA 29 38 NC State " 6 10 2 137 ACC 98 39 Quinnipiac " 6 10 0 169 NEC 130 39 UNC Wilmington" 4 12 1 171 CAA 132 39 Yale " 7 7 2 88 Ivy 47 41 Tulsa " 5 12 0 143 CUSA 102 41 Towson " 3 11 2 175 CAA 133 42 Hofstra " 7 10 1 118 CAA 73 45
It's not at all clear what your point is, besides to highlight the difference between the two rating systems. Take SLU for instance. The RPI places their relative W/L/T percentage at 86th among D1 schools, while Massey places their relative margin of victory as 117th among D1 schools. So, as a practical matter, SLU's relative record doesn't reflect a particularly large relative margin of victory. So, it's entirely probable that two schools can have a virtually identical RPI, while having a significantly different Massey rating, and vice-versa. It's an apples/oranges comparison.
I also don't see how many teams, outside of Iona, who won't make the NCAA Tournament, would suddenly be on the right side of the bubble in the NCAA used Massey instead of RPI. Again, other than Iona, none of the teams with a large Massey/RPI discrepancy are NCAA Tournament candidates so I don't think there's much that suggests changing to Massey would right all these perceived injustices caused by using the RPI.
As we approach tournament time, it might be more interesting to see a list of the kind of discrepancies you describe. Looking at the rankings, there seem to be plenty of them. Clemson and William & Mary are beneficiaries of the RPI, while Bakersfield and Washington stack up much better under Massey.
Well, the Big West has 5 of the top 39 teams in Massey, but only 3 of the top 40 teams in RPI, though the Big West's problem is more going to be teams overall records. The other two are still in the borderline area. It will be interesting to see what happens in the Big West Tournament today. If Irvine and UCSB win, the big west probably only gets two or three teams in. If Riverside and Davis win, you could still make a decent argument for 4 (or possibly 5, since CSUN is at .500). Is the eligibility rule over .500 or at least .500? I think the right number is 3, regardless of what happens today, but it will be very interesting to see which way it goes for Riverside and Davis if UCSB or Irvine win the Big West Tournament. Needless to say, all the other bubble teams should be rooting for UCSB and Irvine today.
Here is a look by conference. A couple teams that are not in both systems are removed. Code: Conf # Massey RPI Diff A10 14 137.6 138.4 -0.7 ACC 9 56.7 42.2 14.4 AEC 8 129.9 126.0 3.9 ASC 14 145.2 147.7 -2.5 B10 7 60.4 65.7 -5.3 BE 16 70.5 67.0 3.5 BSo 9 144.2 146.0 -1.8 BWC 7 27.6 42.7 -15.1 CAA 12 90.4 69.3 21.2 CUSA 9 68.6 52.8 15.8 HL 9 114.0 123.4 -9.4 Ivy 8 91.8 72.5 19.3 MAAC 10 129.7 134.0 -4.3 MAC 7 105.0 110.4 -5.4 MPSF 8 69.9 92.5 -22.6 MVC 8 88.4 98.9 -10.5 NEC 11 123.1 108.5 14.5 P12 6 80.5 87.0 -6.5 PL 8 128.3 129.0 -0.8 SC 8 90.6 93.0 -2.4 SL 6 146.5 168.5 -22.0 WCC 7 91.0 100.9 -9.9
I certainly don't think it will fix everything but right now it is too easy to game the system if your conference is in the right location. But people blamed Irvine last year for their schedule. But the reality is how many team would want to do what Akron did this year and play a lot of teams from the big west. They go 1-1-1 playing very difficult teams around .500 that play teams around .500. Why not go to the ACC where they could play easier teams that have better win-lose percentages and have opponents with better win-loss percentages. As a result Akron is looking at road games in the NCAA instead playing at home. And if you think their hit was bad now, imagine if Big West lost more non-conference games? The other issue is too many people look at RPI as a reflection of schedule difficulty when it really isn't.