Cajuns' RPI Projection

[GoneGolfin]

The following is a projection of the Cajuns' RPI with the results of the final series vs. ULM as the variable. Everything else is held constant. This means that ...

1) Cajun opponents continue to average the same overall winning percentage (all Cajun opponents in aggregate, not the same overall winning percentage per team). This means that the Cajuns' OWP remains constant.

2) Other than the results of the Cajuns' series vs. ULM, all Cajun opponents' opponents aggregate winning percentage remains constant (OOWP). I did include the results of the Cajuns' series vs. ULM in calculating each of the Cajuns' opponents' OWP (which is the Cajuns' OOWP).

Obviously, the above two assumptions will not hold exactly. But the variation should be minor.

RPI projections are listed below. Note that Adjusted RPI and Base RPI will be the same unless Northwestern State climbs into the Top 75 Base RPI. If they do, .0012 can be added to the Base RPI to arrive at the Adjusted RPI.

Cajuns' current RPI:
Adjusted RPI: .5699
Base RPI: .5699

WP: .6538
OWP: .5554
OOWP: .5148

1) Cajuns sweep ULM 3-0
Adjusted RPI: .5689
Base RPI .5689

WP: .6727
OWP: 5429
OOWP: 5170

This is a fall of .0010.

Using the current ordinal RPI rankings, an Adjusted RPI of .5689 would place the Cajuns in the #38 spot.

2) Cajuns win 2/3 from ULM
Adjusted RPI: .5643
Base RPI .5643

WP: .6545
OWP: 5429
OOWP: 5167

Using the current ordinal RPI rankings, an Adjusted RPI of .5643 would place the Cajuns in the #44 spot.

3) Cajuns win 1/3 from ULM
Adjusted RPI: .5597
Base RPI .5597

WP: .6364
OWP: 5429
OOWP: 5164

Using the current ordinal RPI rankings, an Adjusted RPI of .5597 would place the Cajuns in the #53 spot.

4) Cajuns are swept by ULM (0-3)
Adjusted RPI: .5550
Base RPI .5550

WP: .6182
OWP: 5429
OOWP: 5161

Using the current ordinal RPI rankings, an Adjusted RPI of .5550 would place the Cajuns in the #55 spot.


With ULM losing 6 consecutive games and 12 of their last 14, their RPI has plummeted recently. But more importantly, their W/L percentage has plummeted ... as this becomes part of the Cajuns' all-important OWP (50% of the RPI formula). Due to ULM's losing streak, this weekend series has similarities to that of the UNO series (but not quite as bad).

And to illustrate the silliness of the RPI, once again I will provide the example of playing Southern in a three game series versus having to play ULM instead.

ULM has an Adjusted RPI rank of 188. Southern has an Adjusted RPI rank of 233. That is, Southern is ranked 45 spots lower than that of ULM. However, simply playing ULM this weekend results in a considerably lower exit RPI than if the Cajuns were to play Southern. Here is an example where in scenario #1, the Cajuns sweep ULM (as stated above). In scenario #2, the Cajuns sweep Southern.

1) Cajuns sweep ULM 3-0
Adjusted RPI: .5689
Base RPI .5689

WP: .6727
OWP: 5429
OOWP: 5170

This is a fall of .0010.

Using the current ordinal RPI rankings, an Adjusted RPI of .5689 would place the Cajuns in the #38 spot.

2) Cajuns sweep Southern 3-0
Adjusted RPI: .5742
Base RPI .5742

WP: .6727
OWP: 5565
OOWP: 5110

This is a gain of .0043.

Using the current ordinal RPI rankings, an Adjusted RPI of .5742 would place the Cajuns in the #33 spot.

Folks, this is a significant difference. We are talking .0053 RPI points of differential ... for playing quite arguably a weaker team (at least the RPI says so). The ordinal rank difference is 5 spots. The W/L results above do not matter. It is pulling the opponent's WP into the Cajuns' OWP that matters. As stated in prior writeups, the reasoning is that the RPI formula overweights OWP relative to WP and OOWP and it does not account for the opponents' opponents of your opponents (OOOWP). In other words, the formula does not go deep enough to arrive at a reasonable result.

Intelligent scheduling can take advantage of this information (stacking the RPI in your favor) as well as a few other tricks not mentioned.

Brian

[Cajun90]

The following is a projection of the Cajuns' RPI with the results of the final series vs. ULM as the variable. Everything else is held constant. This means that ...

1) Cajun opponents continue to average the same overall winning percentage (all Cajun opponents in aggregate, not the same overall winning percentage per team). This means that the Cajuns' OWP remains constant.

2) Other than the results of the Cajuns' series vs. ULM, all Cajun opponents' opponents aggregate winning percentage remains constant (OOWP). I did include the results of the Cajuns' series vs. ULM in calculating each of the Cajuns' opponents' OWP (which is the Cajuns' OOWP).

Obviously, the above two assumptions will not hold exactly. But the variation should be minor.

RPI projections are listed below. Note that Adjusted RPI and Base RPI will be the same unless Northwestern State climbs into the Top 75 Base RPI. If they do, .0012 can be added to the Base RPI to arrive at the Adjusted RPI.

Cajuns' current RPI:
Adjusted RPI: .5699
Base RPI: .5699

WP: .6538
OWP: .5554
OOWP: .5148

1) Cajuns sweep ULM 3-0
Adjusted RPI: .5689
Base RPI .5689

WP: .6727
OWP: 5429
OOWP: 5170

This is a fall of .0010.

Using the current ordinal RPI rankings, an Adjusted RPI of .5689 would place the Cajuns in the #38 spot.

2) Cajuns win 2/3 from ULM
Adjusted RPI: .5643
Base RPI .5643

WP: .6545
OWP: 5429
OOWP: 5167

Using the current ordinal RPI rankings, an Adjusted RPI of .5643 would place the Cajuns in the #44 spot.

3) Cajuns win 1/3 from ULM
Adjusted RPI: .5597
Base RPI .5597

WP: .6364
OWP: 5429
OOWP: 5164

Using the current ordinal RPI rankings, an Adjusted RPI of .5597 would place the Cajuns in the #53 spot.

4) Cajuns are swept by ULM (0-3)
Adjusted RPI: .5550
Base RPI .5550

WP: .6182
OWP: 5429
OOWP: 5161

Using the current ordinal RPI rankings, an Adjusted RPI of .5550 would place the Cajuns in the #55 spot.


With ULM losing 6 consecutive games and 12 of their last 14, their RPI has plummeted recently. But more importantly, their W/L percentage has plummeted ... as this becomes part of the Cajuns' all-important OWP (50% of the RPI formula). Due to ULM's losing streak, this weekend series has similarities to that of the UNO series (but not quite as bad).

And to illustrate the silliness of the RPI, once again I will provide the example of playing Southern in a three game series versus having to play ULM instead.

ULM has an Adjusted RPI rank of 188. Southern has an Adjusted RPI rank of 233. That is, Southern is ranked 45 spots lower than that of ULM. However, simply playing ULM this weekend results in a considerably lower exit RPI than if the Cajuns were to play Southern. Here is an example where in scenario #1, the Cajuns sweep ULM (as stated above). In scenario #2, the Cajuns sweep Southern.

1) Cajuns sweep ULM 3-0
Adjusted RPI: .5689
Base RPI .5689

WP: .6727
OWP: 5429
OOWP: 5170

This is a fall of .0010.

Using the current ordinal RPI rankings, an Adjusted RPI of .5689 would place the Cajuns in the #38 spot.

2) Cajuns sweep Southern 3-0
Adjusted RPI: .5742
Base RPI .5742

WP: .6727
OWP: 5565
OOWP: 5110

This is a gain of .0043.

Using the current ordinal RPI rankings, an Adjusted RPI of .5742 would place the Cajuns in the #33 spot.

Folks, this is a significant difference. We are talking .0053 RPI points of differential ... for playing quite arguably a weaker team (at least the RPI says so). The ordinal rank difference is 5 spots. The W/L results above do not matter. It is pulling the opponent's WP into the Cajuns' OWP that matters. As stated in prior writeups, the reasoning is that the RPI formula overweights OWP relative to WP and OOWP and it does not account for the opponents' opponents of your opponents (OOOWP). In other words, the formula does not go deep enough to arrive at a reasonable result.

Intelligent scheduling can take advantage of this information (stacking the RPI in your favor) as well as a few other tricks not mentioned.

Brian

So our out of conference schedule should include the best teams from some of the weakest conferences. In this scenario (i.e. Southern in the SWAC) they will more than likely have a good WP but we would still have a favorable advantage and more than likely an easy win.

This has potential to help our RPI a lot more than playing a middle of the pack CUSA team with a decent but not great RPI.

[raginsaints]

How does all this crap change our regional projections? Do we now have to sweep and win some games in the tourny to be in a regional?



Math makes my head hurt.

[GoneGolfin]

_ So our out of conference schedule should include the best teams from some of the weakest conferences. In this scenario (i.e. Southern in the SWAC) they will more than likely have a good WP but we would still have a favorable advantage and more than likely an easy win.

This has potential to help our RPI a lot more than playing a middle of the pack CUSA team with a decent but not great RPI. _

Yes.

You want to schedule a certain amount of bad teams that have a great chance at putting up a decent winning percentage because they play very weak schedules (both OOC and in Conference).

By the same token, avoid at all costs stronger teams that play difficult schedules and are likely to have weak winning percentages.

This takes some guesswork, but you can scour past history to put yourself on the good side of the percentages.

But you do not want to schedule exclusively this way. The selection committee also frowns upon too many games against low RPI opponents. Also, if you were to lose to one of these teams at home, you may suffer an RPI penalty. Just enough of these games to give your school an advantage is what you want.

And as far the CUSA example is concerned, let's take Tulane. If the Cajuns were to take 2/3 from Tulane this weekend (as opposed to 3/3 from Southern), the Cajuns would exit these games with a considerably lower RPI than if the Cajuns had swept Southern. In fact, a sweep of Tulane would only be marginally better than a sweep of Southern (about .0010 better). What do you think would be harder to do?

Brian

[GoneGolfin]

_ How does all this crap change our regional projections? Do we now have to sweep and win some games in the tourny to be in a regional?

Math makes my head hurt. _

The Cajuns need to win the series this weekend and not go 0-2 in the conference tournament to feel comfortable. A non-sweep of ULM and 0-2 in the conference tournament puts it into the hands of the selection committee. It could be dicey, but my guess is that they would still be in (depends on other conference tournament results). While a conference RPI of 5/6 and 2nd/3rd place finish really helps the Cajuns, a fall into the 50's (RPI) makes the Cajuns not a lock. While the Sun Belt was not quite as strong in 2006 as in 2010, the Cajuns had a 2nd place finish (one game behind Troy) and a #53 Adjusted RPI. They were passed over for a higher RPI team that finished behind them in the Sun Belt standings (South Alabama).

A sweep of ULM and 0-2 in the conference tournament will be fine.

Brian

[raginsaints]

Thanks a lot Brian, much easier to understand. We need good results this weekend. Clutch hits need to happen, can't leave runner on 2nd with 0 outs in the 9th again.

[CajunZ1]

How will OOOOOOOOWP come into play?

igeaux.mobi

[GoneGolfin]

_ How will OOOOOOOOWP come into play?

igeaux.mobi _

You do not need that many levels to complete the graph.

Brian (seriously)

[J-Town Cajun]

My guess for the remainder of the season:

0 wins and the Cajuns stay home. That would be 6 straight losses to end the year

1 win and we are most likely not going to get in.

2 wins and we are about 50/50 to get in

3 wins and we are most likely in

4 wins and its a lock

[lxa656]

Sounds like you need to be hired to be a consultant on scheduling games!