The Holy Land of Hoops

rpiforecast.com projects where each team will finish in its RPI rating. Based on that formula, the Big East is ranked 4th among all conferences. Here's where each team is ranked nationally, based on that formula, including games through yesterday:

7. Villanova
23. Creighton
34. Georgetown
42. Xavier
61. Butler
70. Marquette
80. St. John's
92. Providence
125. Seton Hall
150. DePaul

Bill Marsh wrote:rpiforecast.com projects where each team will finish in its RPI rating. Based on that formula, the Big East is ranked 4th among all conferences. Here's where each team is ranked nationally, based on that formula, including games through yesterday:

7. Villanova
23. Creighton
34. Georgetown
42. Xavier
61. Butler
70. Marquette
80. St. John's
92. Providence
125. Seton Hall
150. DePaul

So this is the projection of where each team's ENDING RPI will be, at the end of the year, based on where all the teams are right now, right? If that all holds true, I think that looks good for potentially five bids -- Butler kind of on the edge, but Butler has looked good this year and I think they're going to end up higher than that.

OutlawWales wrote:

Bill Marsh wrote:rpiforecast.com projects where each team will finish in its RPI rating. Based on that formula, the Big East is ranked 4th among all conferences. Here's where each team is ranked nationally, based on that formula, including games through yesterday:

7. Villanova
23. Creighton
34. Georgetown
42. Xavier
61. Butler
70. Marquette
80. St. John's
92. Providence
125. Seton Hall
150. DePaul

So this is the projection of where each team's ENDING RPI will be, at the end of the year, based on where all the teams are right now, right? If that all holds true, I think that looks good for potentially five bids -- Butler kind of on the edge, but Butler has looked good this year and I think they're going to end up higher than that.

It's done by projecting all the games, and then doing the RPI. Butler at 61st isn't good at all-- you have to remember only about top 50-54 are in- the rest are the auto bids. 61 is almost always not even 1st 4 out, it's 2nd 4 out.

If anyone was curious about specifics, RPI Forecast uses Sagarin for its future-game predictions.

stever20 wrote:

OutlawWales wrote:

Bill Marsh wrote:rpiforecast.com projects where each team will finish in its RPI rating. Based on that formula, the Big East is ranked 4th among all conferences. Here's where each team is ranked nationally, based on that formula, including games through yesterday:

7. Villanova
23. Creighton
34. Georgetown
42. Xavier
61. Butler
70. Marquette
80. St. John's
92. Providence
125. Seton Hall
150. DePaul

So this is the projection of where each team's ENDING RPI will be, at the end of the year, based on where all the teams are right now, right? If that all holds true, I think that looks good for potentially five bids -- Butler kind of on the edge, but Butler has looked good this year and I think they're going to end up higher than that.

It's done by projecting all the games, and then doing the RPI. Butler at 61st isn't good at all-- you have to remember only about top 50-54 are in- the rest are the auto bids. 61 is almost always not even 1st 4 out, it's 2nd 4 out.

Still, there is plenty of time for Butler to move up and hopefully Marquette can make some noise as well.

XUFan09 wrote:If anyone was curious about specifics, RPI Forecast uses Sagarin for its future-game predictions.

Just curious, if you are anyone has data on how accurate RPI forecasts have been in previous seasons this early in the season. It seems so many variables are unaccounted for.

Are the ratings still weighted from last year? I forget the # of games that it drops last season's results.

XU85 wrote:

XUFan09 wrote:If anyone was curious about specifics, RPI Forecast uses Sagarin for its future-game predictions.

Just curious, if you are anyone has data on how accurate RPI forecasts have been in previous seasons this early in the season. It seems so many variables are unaccounted for.

Thing is, we are now 14/31 into the season for most teams. That's 45%. That's a lot of data already in the system.

It's a decent guess as to what the final rpi's will look like. For one it takes into account the remaining schedules and their impact.

stever20 wrote:

XU85 wrote:

XUFan09 wrote:If anyone was curious about specifics, RPI Forecast uses Sagarin for its future-game predictions.

Just curious, if you are anyone has data on how accurate RPI forecasts have been in previous seasons this early in the season. It seems so many variables are unaccounted for.

Thing is, we are now 14/31 into the season for most teams. That's 45%. That's a lot of data already in the system.

It's a decent guess as to what the final rpi's will look like. For one it takes into account the remaining schedules and their impact.

As interested in statistics as you seem to be, I would think you could do better than a "decent guess." For example, what percent of teams finished within 10 of their RPI forecast at this stage of the season? It may be "a lot of data already in the system," but without such a calibration, the "decent guess" is meaningless.

You may ask, why don't I do it myself? Answer: Because I don't feel like it and because I enjoy college basketball a lot more if I don't over analyze it.

It's a decent guess because it shows what the RPI's would be if there are no upsets. Obviously you can't take into account those. If the predictions are 100% right, it shows what the RPI's would look like. You also can look at that and get a general idea of what RPI each team would have if they finish with specific records.

Take Butler for instance.

At 17-12(remember, D2 game doesn't count)- their RPI is 66.3.
But if they got to 18-11, it's now at 55.5
19-10 it's at 45.6

It's a great tool because you have factored into things all the other schedules from the other conferences It's not guaranteed, but it's a very good educated guess as to what the numbers would look like.