jpschmack

What res and I are referring to is the Conference Effect on RPI. Because we all play each other, each of our total W-L counts on each other's SOS. Since we're guaranteed to go .500 against each other, the only way to have a good SOS is by winning OOC games.

Yes, when two bubble teams are compared in March, SOS, OOC SOS, Marquee Wins, etc are compared.
But this came up because of St. Bona, who (as a Bona alum I lovingly say) probably isn't going to be competing for an at-large.

The goal is to have A-10 games feature teams with records like: 20-10 vs 20-10, and not 15-15 vs 15-15.
If you're not going to finish in the top six of the A-10, OOC wins are more important than OOC RPI.
If you are going to finish in the top six of the A-10, you need to get some marquee wins OOC, but pick and choose and don't go 3-7 vs 10 really good teams to do it.

Yeah, but:
-- ESPN isn't going to stop acting that way until they stop giving hundreds of millions of dollars to the Big East and ACC.
-- ESPN doesn't pick the field.
-- We can't control ESPN, we CAN control some aspects of RPI math.

antboy

Exactly, which is why it's such crap that they don't factor in home-away corrections to the SOS. You could play the same 15 teams at home (typical BCS-team sched) versus playing those same 15 teams on the road (typical Southland conference sched) and have the exact same rpi SOS. But if (big if, I know) both teams win all games, the Southland team would have a much higher rpi than the BCS team does, because the rpi does factor in home-away. And anyone looking at those two schedules would conclude this too.

This is of course why the NCAA selection committee stopped using the team's individual rpi, but continues to use the rpi SOS. One more excuse to select more BCS conference teams.

The change happened after the MVC caught on to the game and had a good year, so they got a lot of bids. The talking heads all raised a fit and said they "gamed the system", overlooking (intentionally I think) the fact that those MVC teams were a) pretty good and did fairly well in the tournament, and b) just negating the unfair advantage teams in conferences that schedule lots of home games get.

One other related gripe is that it's my understanding the committees still include the conference rpi after conference play, not only the OOC rpi. Since the committees say they also use conference strength to choose among teams, inflated home records further inflate differences between conferences = more bias.

jpschmack

Extremely well put.

The other ridiculous aspect of the RPI is that all you hear about in March is "They were X-X vs the Top 50 of the RPI and X-X vs the Top 100 of the RPI" and anyone over 100 in the RPI is a "bad loss."

If everyone who's 100+ of the RPI is a bad loss, and NCAA bubble teams should be kicking the crap out of them, why are we differentiating between WHICH of those 245 other teams each school played down to FOUR DECIMAL PLACES???

So the teams in weaker conferences have their RPIs dragged down by playing the bottom teams and winning by 40 and you have another excuse to take BCS teams!

Title_BU

1. 7 - Butler
2. 23 - Temple
3. 34 - La Salle
4. 38 - Charlotte
5. 39 - Virginia Common
6. 43 - Saint Louis
7. 57 - Massachusetts
8. 60 - Xavier
9. 78 - St. Joseph's (P
10. 90 - Dayton
11. 109 - Richmond
12. 132 - St. Bonaventure
13. 143 - Rhode Island
14. 169 - Duquesne
15. 190 - George Washington
16. 304 - Fordham (WTF?)

1. Big Ten
2. Big East
3. Big-12
4. SEC
5. Mountain West
6. ACC
7. A-10
8. Missouri Valley
9. Conference-USA
10. Pac-12 (LOL)

http://statsheet.com/mcb/rankings/RPI?id=2012-12-10

KYFlyer

Title_BU I would start using a different source for your RPI as that site is not up to date. Doesn't match with numerous other RPI sites : http://www.cbssports.com/collegebask...ngs/rpi/index1, http://www.udpride.com/images/rpi.htm, http://www.rpiforecast.com/live-rpi.html

Title_BU

"StatSheet is the only site on the web to compute RPI/SOS metrics multiple times daily.
That's one reason our numbers may vary from other sites--we are more up-to-date."

KYFlyer

I'm sorry what were those games that were played between midnight and noon today? Dude, if multiple sites agree and they don't, they are wrong.

FlyingElvis

There is an RPI calculator on UDPride that the guy who runs that site put together. It seems to get updated very quickly after games. I'm not sure if it is linked somehow to external sources. But the guy who runs the site has actually caught mistakes in the "official" calculations. He once contacted the NCAA because an away game loss for UD was mistakenly entered as a neutral site game and therefore hurt our RPI worse than it should have. They corrected the error and UD was a bubble team that got in the tournament that year. I don't know if a small RPI jump made the difference, but I guess it could have.

Here it is.

http://www.udpride.com/images/rpi.htm

A10 teams are in royal blue. Teams in light blue are teams that UD plays this year.

__________________
"US Bank Arena can't be that bad"

Title_BU

The nominal differences are utterly irrelevant, especially in December, but are you actually arguing that there is more gravitas in the calculations done by UDPride.com than StatSheet.com (USA Today)?

xu1990

Here is how you manipulate the RPI

2 Big East .5780 103-27 0-9 2-15 0-0 103-27 509-548 .5188 9 73-6 11-6 19-15 103-27 51-1

103-27 overall, but 0-9 vs top 25 and 2-15 against top 50. Play a million home buy games (73-6) and win them. Play some tough OOC games and lose them all. End up 2nd best conference in RPI.

__________________
4 out of 5 sweet sixteen's : Kansas, Michigan State, North Carolina and Xavier.
8 schools have made last 7 NCAAs: Duke, Gonzaga, Kansas, Marquette, Michigan State, Texas, Wisconsin and Xavier.

owlfan2272

Sorry thought you were describing Temple for a second. I hate our ooc schedule this year.

__________________
Temple Owls have won a lot.

Medford

As I understand it, RPI is a simply calculations. Win, losses, where the games are played. There is no date to manipulate, there is no "personal preferrences" to persuade. therefor, it should not matter if we are in nov, Dec, Feb or March, the rpi from all sites should be identical if they are indeed accurate.

The statement that Statsheet is the only site to update the rpi multiple times a day is simply not true. I've seen Chris' rpi at UDpride updated many, many times a day on a saturday. I can't say its always up to the minute or not, but if there is a UD game at 12:00, which ends at 2:00, and I check after the game, its going to be updated by 3 or 4 generally. Perhaps there is a key game late in the season involving a flyer OOC opponent, or someone above/behind them in the rpi that happens later, it gets updated later. I have no idea how relavent UD's games are to how often its updated, but its often updated several times a day.

What flyingelvis said above is true, Chris' rpi calculator has been as accurate as any I've come across and as he described, he's caught several mistakes that the NCAA has missed in their official RPI.

UMass87

Medford - discrepencies in RPI calculations occur for two reasons: 1) mistakes 2) because of varying interpretations of what is considered a home game.

SFspidur

RPI Forecast updates every five minutes I believe. I'm not sure whether they're spot on in terms of accuracy, but it's the best free one I've found for following up to the minute.

twisted3829

This is RPI from ESPN which claims to be the same as the NCAA ranks:

7 Butler
26 Charlotte
28 Temple
31 La Salle
39 St Louis
43 VCU
65 St Joe's
67 Dayton
76 Xavier
79 UMASS
85 Richmond
111 St Bonnie's
149 Duquesne
152 URI
160 GW
291 Fordham

I can't see how stat sheet interpreted the schedule like the other site but my guess is the way they do home away and neutral (I have seen some sites use semi-home and semi-away which would drastically change the RPI)

__________________
2010-2011 Lock of the Day Contest NIT Champion