Cornell 1 vs. Merrimack 0 1/7/17

[Dafatone]

This means that wins are interchangeable.  Sure, had we beaten CC, we'd be better off.  But we'd be in exactly* the same spot had we beaten CC and lost Saturday to Merrimack.

who was proposing that we should have lost to merrimack twice instead

No one in particular.  I just find it a strange wrinkle of RPI that we may end the year going "if only we had beaten CC" when a win against any other opponent would have the same impact.

Back in the pre RPI-only days of the pairwise, specific games would have more of an impact.

[Jeff Hopkins '82]

This prompts the question: How high do they deserve to be?

"You are what your record says you are."  John has worn me down to the point where I accept KRACH as my personal savior.  We "deserve" to be 16th.

Would that be Jesus KRACH?

[abmarks]

This means that wins are interchangeable.  Sure, had we beaten CC, we'd be better off.  But we'd be in exactly* the same spot had we beaten CC and lost Saturday to Merrimack.

who was proposing that we should have lost to merrimack twice instead

No one in particular.  I just find it a strange wrinkle of RPI that we may end the year going "if only we had beaten CC" when a win against any other opponent would have the same impact.

Back in the pre RPI-only days of the pairwise, specific games would have more of an impact.

Road wins and quality win bonus DO make a difference.  Quality Win bonus is for teams in the top 20 in PWR.


We  have three remaining road games against teams in the PWR top 20 (assuming pwr stays the same for the sake of argument)



Beating #9 Union is the *best* RPI result and beating #14 St. Lawrence is the second best with a win at Quinny 3rd best.  All three are better than a neutral site win vs CC

[Swampy]

This means that wins are interchangeable.  Sure, had we beaten CC, we'd be better off.  But we'd be in exactly* the same spot had we beaten CC and lost Saturday to Merrimack.

who was proposing that we should have lost to merrimack twice instead

No one in particular.  I just find it a strange wrinkle of RPI that we may end the year going "if only we had beaten CC" when a win against any other opponent would have the same impact.

Back in the pre RPI-only days of the pairwise, specific games would have more of an impact.

Road wins and quality win bonus DO make a difference.  Quality Win bonus is for teams in the top 20 in PWR.


We  have three remaining road games against teams in the PWR top 20 (assuming pwr stays the same for the sake of argument)



Beating #9 Union is the *best* RPI result and beating #14 St. Lawrence is the second best with a win at Quinny 3rd best.  All three are better than a neutral site win vs CC

How do these compare with beating Harvard at home?

[Trotsky]

This means that wins are interchangeable.  Sure, had we beaten CC, we'd be better off.  But we'd be in exactly* the same spot had we beaten CC and lost Saturday to Merrimack.

who was proposing that we should have lost to merrimack twice instead

No one in particular.  I just find it a strange wrinkle of RPI that we may end the year going "if only we had beaten CC" when a win against any other opponent would have the same impact.

Back in the pre RPI-only days of the pairwise, specific games would have more of an impact.

Road wins and quality win bonus DO make a difference.  Quality Win bonus is for teams in the top 20 in PWR.


We  have three remaining road games against teams in the PWR top 20 (assuming pwr stays the same for the sake of argument)



Beating #9 Union is the *best* RPI result and beating #14 St. Lawrence is the second best with a win at Quinny 3rd best.  All three are better than a neutral site win vs CC

How do these compare with beating Harvard at home?

Not nearly as satisfying.

[abmarks]

Road wins and quality win bonus DO make a difference.  Quality Win bonus is for teams in the top 20 in PWR.

We  have three remaining road games against teams in the PWR top 20 (assuming pwr stays the same for the sake of argument)

Beating #9 Union is the *best* RPI result and beating #14 St. Lawrence is the second best with a win at Quinny 3rd best.  All three are better than a neutral site win vs CC

How do these compare with beating Harvard at home?

Good question, and with an answer that surprised me.

TL;DR Summary:  Beating Harvard at home improves RPI more than road wins at QU, UN or STL.

 W Home v HVD: RPI impact is + .0475 QWB +.028 home win = 0.076

 W road v UN: RPI impact is + .0300 QWB +.041 road win = 0.071

 W road v STL: RPI impact is + .0175 QWB +.041 road win = 0.059

 W road v QU: RPI impact is + .0075 QWB + .041 road win = 0.049


Not sure of the PWR change since I don't know how to calculate PWR comparisons over the course of the full year not yet played.





--If you are interested in why those are the numbers, here is the lopnger version with calculations:

All based on the PWR explanation at CHN http://www.collegehockeynews.com/info/?d=pwcrpi
   - RPI for each team is tweaked based on a number of factors.   Home win vs. Harvard would be adjusted for home/away and for a quality win bonus.
   - RPI is made up of three factors: Team win %, average of opponents winning %, and average of opponents opponents winning percentage:

        RPI = (WP * 0.25) + (OWP * 0.21) + (OOWP * 0.254)

When comparing the relative value of different wins or losses, CU WP is the most impacted since OWP and OOWP will change much less given the volume of games played used for each stat.

Home/Road weighting:
 
For purposes of calculating a final RPI, games are weighted based upon whether they are home or road games. Road wins and home losses are weighted by a factor of 1.2, while home wins and road losses are weighted by 0.8. Unlike basketball, all components of the RPI are weighted. This weighting system was introduced in 2013-14.

This adjustment is calculated in our winning percentage.  Win % is ((wins + .5 * ties) / total games played), including only NCAA games.  Simply put, a road win is worth effectively 0.4 more wins than a home win when calculating winning percentage.  (When adding up wins, you'd use .8 for a home win or 1.2 for a road win.  So a harvard home win adds .8 to the numerator while the road wins would add 1.2)

29 NCAA games through the end  of the regular season. Based on the WP formula, over 29 games the WP = (wins +.5*ties) / 29.

-Therefore a harvard win at home would change WP by .8/29 = 0.028  
-A road win against QU, UN, STL etc would change WP by 1.2/29 = 0.041
 > Road win improves RPI by 0.041-0.028= 0.014 more than a home win


The second adjustment is for quality wins.

Quality Win Bonus (QWB):
 A "Quality Win Bonus" was added for the 2013-14 season. For any win against the top 20 of the RPI, a team is awarded "bonus points" on a sliding scale from 1-20. In other words, a team is given a .050 RPI bonus for defeating the No. 1 team, sliding down to .0025 bonus for defeating the 20th team. The total bonus for the season is divided by the amount of games played (weighted for home-road), to give a final bonus figure.

Without showing the math, here are the QWB amounts for the games mentioned, assuming today's PWR.

  W v #2 HVD  +.0475

  W v #9 UN   +.0300

  W v #14 STL +.0175

 W v #18 QU  +.0075



So to answer the original question, here are the RPI impacts of the wins.


  W Home v HVD:  RPI impact is + .0475 QWB +.028 home win = 0.076

  W road v UN:  RPI impact is + .0300 QWB +.041 road win =  0.071

  W road v STL: RPI impact is + .0175 QWB +.041 road win =  0.059

  W road v QU:  RPI impact is + .0075 QWB + .041 road win = 0.049

[ugarte]

Road wins and quality win bonus DO make a difference.  Quality Win bonus is for teams in the top 20 in PWR.

We  have three remaining road games against teams in the PWR top 20 (assuming pwr stays the same for the sake of argument)

Beating #9 Union is the *best* RPI result and beating #14 St. Lawrence is the second best with a win at Quinny 3rd best.  All three are better than a neutral site win vs CC

How do these compare with beating Harvard at home?

Good question, and with an answer that surprised me.

You answered this question with math but the real answer was in your heart all along.

[upprdeck]

this also assume all games after that are never played true?

so if by beating harvard then they lose 7 in a row doesnt the value get decreased?

[abmarks]

this also assume all games after that are never played true?

so if by beating harvard then they lose 7 in a row doesnt the value get decreased?

Incorrect.  

As I said above, I was looking at the RPI and more specifically our winning percentage.  WP is wins + .5 (ties)  divided by games played.  I used a full season games played number of 29, not total games to date.  


However, to be able to do the math, I had to use current PWR ranking numbers for opponents- those of course will change as the season plays out.  So I'm making the simplifying assumption of using current PWR numbers. Quality win bonus figures will change depending on what the final PWR rankings are at season's end.

[Swampy]

this also assume all games after that are never played true?

so if by beating harvard then they lose 7 in a row doesnt the value get decreased?

Incorrect.  

As I said above, I was looking at the RPI and more specifically our winning percentage.  WP is wins + .5 (ties)  divided by games played.  I used a full season games played number of 29, not total games to date.  


However, to be able to do the math, I had to use current PWR ranking numbers for opponents- those of course will change as the season plays out.  So I'm making the simplifying assumption of using current PWR numbers. Quality win bonus figures will change depending on what the final PWR rankings are at season's end.

So if I understand you correctly, this confirms the axiom, "Just win, baby!"

[Trotsky]

this also assume all games after that are never played true?

so if by beating harvard then they lose 7 in a row doesnt the value get decreased?

Incorrect.  

As I said above, I was looking at the RPI and more specifically our winning percentage.  WP is wins + .5 (ties)  divided by games played.  I used a full season games played number of 29, not total games to date.  


However, to be able to do the math, I had to use current PWR ranking numbers for opponents- those of course will change as the season plays out.  So I'm making the simplifying assumption of using current PWR numbers. Quality win bonus figures will change depending on what the final PWR rankings are at season's end.

So if I understand you correctly, this confirms the axiom, "Just win, baby!"

An axiom cannot be "confirmed."