• Welcome to The Valparaiso Beacons Fan Zone Forum.
 

How To Increase RPI Today

Started by Just Sayin, December 22, 2015, 08:57:47 AM

Previous topic - Next topic

0 Members and 3 Guests are viewing this topic.

Valpower

Quote from: ValpoHoops on January 03, 2016, 08:14:34 PM
As for RPI, it won't matter. Our opponents will go 1-1 in that game.

RPI is based solely on winning percentages.

For our resume, it would make the win over OSU look a bit better.
But the loss at Oregon worse?  :(

vu72

At least for the Sagarins it may help.  It may propel OSU into the top 50 so we may pick up a win over the top 50.  Oregon was at about 30 and OSU was just below us at about 52 or so. We'll see.  Seems to me that we want teams that we beat to beat teams we lot to but then again, I don't have a clue... :crazy:
Season Results: CBI/CIT: 2008, 2011, 2014  NIT: 2003,2012, 2016(Championship Game) 2017   NCAA: 1962,1966,1967,1969,1973,1996,1997,1998 (Sweet Sixteen),1999, 2000, 2002, 2004, 2013 and 2015

talksalot

vu72 nails it!

RPI Falls to 44; on account of the other games involving MVC and Summit league teams, as "our opponents' opponents" slipped.

Sagarin did move OSU up to 45 so we are 1-1 against top 50; our Sagarin rating is 49

Oregon is 40

Horizon League is 1-16 against top-50 teams  (0-11 against top 25),

Northern Kentucky is 0-3 against top-25 teams: Michigan, Xavier and West Virginia.


agibson

Quote from: HC on January 03, 2016, 10:46:21 AM
Edit: myself and realtime RPI think a lot alike.

In the past, the predictions at RPI forecast have seemed more reasonable.

talksalot

to answer the Subject line:  "Don't Lose to Youngstown State at home"

Valpo and UIC getting the night off... the HL finals are in, a lot of other games still out...but here are the RPI's... and Oakland took the pounding...

Valpo 44
Detroit 110
Oakland 131
Green Bay 145
Wright State 164
Milwaukee 178
Cleveland State 179
Youngstown State 219
Northern Kentucky 299
Illinois Chicago 345

talksalot

One of these days, someone is going to explain this to me... which would be better for Valpo's RPI:

Oklahoma State:  Loser to Missouri State whom we beat
Baylor:  Loser to Oregon who beat us


Gonna cheer for Baylor anyway...

agibson

Quote from: talksalot on January 05, 2016, 05:56:18 PMOne of these days, someone is going to explain this to me... which would be better for Valpo's RPI:

Oklahoma State:  Loser to Missouri State whom we beat
Baylor:  Loser to Oregon who beat us

It's basically a wash.  They're both opponents of our opponents, their win percentage each shows up in the third piece of our RPI. It is opponents' opponents' winning _percentage_, so, I guess, if one of OK State or Baylor is going to wind up with fewer games then we'd prefer them to win, it'll have a (slightly) bigger effect on the percentage.

talksalot

Baylor won it 77-62...and Valpo's RPI is back to 40  (thanks to Rhody)

Just Sayin

Quote from: agibson on January 05, 2016, 09:43:45 PM
Quote from: talksalot on January 05, 2016, 05:56:18 PMOne of these days, someone is going to explain this to me... which would be better for Valpo's RPI:

Oklahoma State:  Loser to Missouri State whom we beat
Baylor:  Loser to Oregon who beat us

It's basically a wash.  They're both opponents of our opponents, their win percentage each shows up in the third piece of our RPI. It is opponents' opponents' winning _percentage_, so, I guess, if one of OK State or Baylor is going to wind up with fewer games then we'd prefer them to win, it'll have a (slightly) bigger effect on the percentage.


Not sure it's a wash. Isn't the formula for RPI 25% our WL% + 50% Opponent WL% + 25% OO W-L%? Close but not a complete wash given unequal weighting of O and OO. 

a3uge

Quote from: Just Sayin on January 06, 2016, 11:36:16 PM
Quote from: agibson on January 05, 2016, 09:43:45 PM
Quote from: talksalot on January 05, 2016, 05:56:18 PMOne of these days, someone is going to explain this to me... which would be better for Valpo's RPI:

Oklahoma State:  Loser to Missouri State whom we beat
Baylor:  Loser to Oregon who beat us

It's basically a wash.  They're both opponents of our opponents, their win percentage each shows up in the third piece of our RPI. It is opponents' opponents' winning _percentage_, so, I guess, if one of OK State or Baylor is going to wind up with fewer games then we'd prefer them to win, it'll have a (slightly) bigger effect on the percentage.


Not sure it's a wash. Isn't the formula for RPI 25% our WL% + 50% Opponent WL% + 25% OO W-L%? Close but not a complete wash given unequal weighting of O and OO.
http://www.rpiforecast.com/teams/Valparaiso.html

Both Oklahoma State and Baylor account for ~.16% of our RPI, but that will drop to ~.03% by the end of the year. Perhaps Oklahoma state is .034 while Baylor is .031. Both teams won't really impact our overall RPI one way or the other by the end of the year.

agibson

Quote from: Just Sayin on January 06, 2016, 11:36:16 PM
Quote from: agibson on January 05, 2016, 09:43:45 PM
Quote from: talksalot on January 05, 2016, 05:56:18 PMOne of these days, someone is going to explain this to me... which would be better for Valpo's RPI:

Oklahoma State:  Loser to Missouri State whom we beat
Baylor:  Loser to Oregon who beat us

It's basically a wash.  They're both opponents of our opponents, their win percentage each shows up in the third piece of our RPI. It is opponents' opponents' winning _percentage_, so, I guess, if one of OK State or Baylor is going to wind up with fewer games then we'd prefer them to win, it'll have a (slightly) bigger effect on the percentage.


Not sure it's a wash. Isn't the formula for RPI 25% our WL% + 50% Opponent WL% + 25% OO W-L%? Close but not a complete wash given unequal weighting of O and OO. 


Hm? OK State and Baylor are both OO, right?

ValpoHoops

I posted this last spring, but wanted to bring it back to refresh everyone's memory on how RPI works.

--------------------------------------------------------------------------------------------------------------------

Quick RPI Tutorial

The RPI is a tool that the NCAA Selection Committee uses to determine at-large bids and seeding for the NCAA Tournament. It is not the only tool they use, and many believe that it is being used less and less as years pass and more analytics are available.

It is a measure that is based simply on winning percentages of your team, your team's opponents and your team's opponents' opponents. The last two of these is where the strength of schedule is calculated.

Only games against Division 1 teams are included in the calculations.


The Math (Simple Version) – See below for why some numbers may not always match:

25% of team winning percentage
50% of opponents winning percentage (games against your team are excluded in this area only)
25% of opponents' opponents winning percentage

Thus, if your team is 8-2, their winning percentage is .800. Multiply that by 25% and you get .200

If your opponents are all 7-3, their combined record would be 70-30. Subtract 8 losses and 2 wins because you take out results against your team, and that record changes to 68-22, good for a winning percentage of .756. After multiplying this number by 50%, you get .378.

If all of your opponents' opponents are 4-6, this is a total of 400-600 (ten games, ten teams, ten opponents). The winning percentage of these teams is .400, of which you take 25%, good for .100.

In this scenario, .200 + .378 + .100 would give your team an RPI rating of .678.

Notice in this example (and any you could create), that it DOES NOT MATTER which teams your team beats or their opponents beat, only the winning percentages. So, it's not a matter of who you beat for the RPI, just how many. (The who comes into play when you get their winning percentages factored in). Thus, there is no such thing as a "win that will really help our RPI".

Understand that my saying "does not matter which teams" only applies to a team's own schedule (ie: 15-15 is the same winning percentage, regardless of which of their games they win and lose). It ALWAYS matters how good the teams you play are.


The Math (Complicated Version Details):
The formula that the NCAA uses has a few tweaks to it. Over time, research has shown that the home team wins about 70% of the time, so the NCAA chose to adjust their actual formula.

Home Wins: .6 wins
Home Losses: 1.4 losses

Road Wins: 1.4 wins
Road Losses: .6 losses

Neutral site games did not change.

I won't go back through an entire example, but you can see how it would affect a team's record. To the NCAA RPI formula, Valpo has a 8.6-1.8 record thus far (1/7/2016)

agibson

Nice summary.

One clarification: apparently the home/away tweaks only affect the _first_ factor of the RPI, your win-loss percentage.  Not the other two factors.

Quote from: ValpoHoops on January 07, 2016, 10:27:29 AMThus, there is no such thing as a "win that will really help our RPI".

Right. Opponents come into play in a scheduling sense, "A game that would really help our RPI if we could get it on our schedule." And, once your schedule's set, any one win is the same as any other win. You could lose to your 15 best opponents, or to your 15 worst, and it would give you the same RPI.

Quote

The RPI is a tool that the NCAA Selection Committee uses to determine at-large bids and seeding for the NCAA Tournament. It is not the only tool they use, and many believe that it is being used less and less as years pass and more analytics are available.

It's also been said that your _opponents_ RPI matters more than your RPI.  That the summaries the selection committee sees on the team sheets (top 50 W-L, 51-100 W-L, or whatever) are persuasive, but that the team's own RPI is relatively less visible, relatively less important.

And, of course, there _is_ such a thing as a "win that could really help our selection/seeding" in the sense that I'm sure the committee cares about quality wins, and they do show up on the team sheet.

But, given a fixed W-L record, which teams you beat, and which beat you does not affect your own RPI.  (You'd have to decide, I guess, whether quality wins are more important than bad losses, for its effect on your seeding or selection.)


Just Sayin

Quote from: agibson on January 07, 2016, 09:08:16 AM
Quote from: Just Sayin on January 06, 2016, 11:36:16 PM
Quote from: agibson on January 05, 2016, 09:43:45 PM
Quote from: talksalot on January 05, 2016, 05:56:18 PMOne of these days, someone is going to explain this to me... which would be better for Valpo's RPI:

Oklahoma State:  Loser to Missouri State whom we beat
Baylor:  Loser to Oregon who beat us

It's basically a wash.  They're both opponents of our opponents, their win percentage each shows up in the third piece of our RPI. It is opponents' opponents' winning _percentage_, so, I guess, if one of OK State or Baylor is going to wind up with fewer games then we'd prefer them to win, it'll have a (slightly) bigger effect on the percentage.


Not sure it's a wash. Isn't the formula for RPI 25% our WL% + 50% Opponent WL% + 25% OO W-L%? Close but not a complete wash given unequal weighting of O and OO. 


Hm? OK State and Baylor are both OO, right?



Oops, you are correct. I confused Ok St. with Oregon State. :-X

ValpoFan

So had we lost to Oregon State and beat Ball State, our RPI would have been exactly the same.
Can you tell that I am unable to get over that stupid loss?

ValpoHoops

Quote from: ValpoFan on January 08, 2016, 10:44:55 AM
So had we lost to Oregon State and beat Ball State, our RPI would have been exactly the same.
Can you tell that I am unable to get over that stupid loss?

Yes.

Winning percentages are all that matter...

ValpoHoops

Some numbers to make some people realize that "scoreboard watching" might be a waste of time.

This is a bit simplified, but that's only to make the numbers easier to digest. It assumes that all teams play exactly 30 games.

------------------------
The RPI Formula ..... (.250)WP + (.500)OWP + (.250)OOWP

The "PERFECT" RPI is 1.000.
------------------------

If every team started at .000 and got "points" added to their RPI each time they, their opponents or their opponents' opponents won, each game would be worth the following:

WP = .250/30 = .008333 points per win
OWP = .500/900 = .000555 points per win
OOWP = .250/27000 = .00000925 points per win


So, while each game does matter...sitting on the couch and following along with the Portland State/Montana State game and rooting for Portland state because Oregon played them is probably a waste of your time.




justducky

Quote from: ValpoHoops on January 07, 2016, 10:27:29 AMThe Math (Complicated Version Details):
The formula that the NCAA uses has a few tweaks to it. Over time, research has shown that the home team wins about 70% of the time, so the NCAA chose to adjust their actual formula.

Home Wins: .6 wins
Home Losses: 1.4 losses

Road Wins: 1.4 wins
Road Losses: .6 losses

Neutral site games did not change.

I won't go back through an entire example, but you can see how it would affect a team's record. To the NCAA RPI formula, Valpo has a 8.6-1.8 record thus far (1/7/2016)
Looking back to our scheduling it is no mystery why we chose to play Chicago St on the road (if we needed another game). Agreeing to start series with Indiana St and Rhode Island on the road could also have been partially rationalized by the expected at large nudges that road victories could produce.  Or is this just another case of me rooting for Portland St over Montana St?  :)

agibson

Quote from: justducky on January 08, 2016, 11:37:20 AM
Looking back to our scheduling it is no mystery why we chose to play Chicago St on the road (if we needed another game). Agreeing to start series with Indiana St and Rhode Island on the road could also have been partially rationalized by the expected at large nudges that road victories could produce.

Those affect the highest "effect per game" section of the RPI.  So, they have a noticeable effect, for sure.

Quote
WP = .250/30 = .008333 points per win
OWP = .500/900 = .000555 points per win

And, even for OWP, the effect of individual games is not exactly negligible.  0.001 of RPI can sometimes mean a couple of places in the RPI ranking (at the moment #35 is 0.0005 below us, #33 is 0.0011 above us).  So, a couple of OW one way or the other, compared to alternate possibilities, can move you a few places in the RPI.  This seems roughly anecdotally consistent with the kinds of changes talksalot reports.

So, at best a way to pass idle time, to get engaged with the games a bit, sure. But, it can have _some_ observable effect.  (individual OO games? obviously less so).

valpotx

The Summit is holding at #10 in conference RPI, while the HL has moved down to #15.  The MVC is just at #14, so it is an off year for them. 
"Don't mess with Texas"

Valpower

Quote from: ValpoHoops on January 08, 2016, 11:05:17 AM
So, while each game does matter...sitting on the couch and following along with the Portland State/Montana State game and rooting for Portland state because Oregon played them is probably a waste of your time.

An NASCAR fan who looked at only their driver's car going around the track would be a fan watching a car go around in circles.  Thinking beyond one's own team and rooting for other participants is one of the major benefits of the RPI-watching silliness.

justducky

A good way to increase RPI is to beat Oakland on the road convincingly! Up to 28 on Warren Nolen. Pomeroy down to 20!

talksalot

I see before the game tonight, RealTimeRpi has us running the HL table...

01-08   at  Oakland   9-6 (1-1) 135  69-68 W - Scouting   
  01-10   at  Detroit   6-5 (2-0) 116  70-69 W - Scouting   
  01-14     Milwaukee   9-5 (2-1) 178  79-59 W - Scouting   
  01-16     Green Bay   8-5 (3-0) 140  78-60 W - Scouting   
  01-18   at  Youngstown St.   4-10 (1-2) 233  72-65 W - Scouting   
  01-22   at  Wright St.   5-8 (1-1) 169  71-67 W - Scouting   
  01-24   at  North. Kentucky   2-9 (0-2) 303  73-63 W - Scouting   
  01-28     Cleveland St.   3-11 (0-3) 185  79-58 W - Scouting   
  01-30     Youngstown St.   4-10 (1-2) 233  80-57 W - Scouting   
  02-06   at  Ill. Chicago   0-10 (0-1) 343  74-61 W - Scouting   
  02-11     North. Kentucky   2-9 (0-2) 303  81-55 W - Scouting   
  02-13     Wright St.   5-8 (1-1) 169  79-59 W - Scouting   
  02-16   at  Cleveland St.   3-11 (0-3) 185  71-66 W - Scouting   
  02-19     Oakland   9-6 (1-1) 135  78-60 W - Scouting   
  02-21     Detroit   6-5 (2-0) 116  77-61 W - Scouting   
  02-26   at  Milwaukee   9-5 (2-1) 178  70-67 W - Scouting   
  02-28   at  Green Bay   8-5 (3-0) 140  70-68 W - Scouting   

funny, tho... they predict us to finish with a record of 18-1 in the league.  Don't know how to do that.

agibson

Quote from: talksalot on January 08, 2016, 10:27:49 PMfunny, tho... they predict us to finish with a record of 18-1 in the league.  Don't know how to do that.

Probability and statistics. Say you're a 90% favorite.  90% chance to win one game.  But only (90%)^2 = 81% to win two in a row. Only 18.5% chance to win 16 in a row.

Before Oakland, RPIForecast had us favorites to win every game.  Only 53% at Oakland, 65% or better for every other game. But only a 4% chance to run the table.  (Should rise to 8%, or maybe 10% or so if our Sagarin improves, after the win).

Running the table's not going to be easy, for sure.

VULB#62

But wait. We only play 18 games. How can we be 18-1?