Ultimate College Softball
Register Calendar Latest Topics
 
 
 


Reply
  Author   Comment   Page 2 of 2      Prev   1   2
CajunAmos

Registered:
Posts: 967
Reply with quote  #31 
Quote:
Originally Posted by Prowler
Not just from this thread but similar discussions over the past few years, it's almost like some people have an idea of which teams should and shouldn't be in -- and how they should be seeded -- and want to find a formula or tweak the RPI to make it match their preconceived notions. Which sounds like a bass-ackward approach. A big one is the 'bonus for road win, less value for home win' adjustment to the RPI that some people want. I just did a spot check on Florida's losses from 2014-16. It's a small sample, three seasons, and only one team -- but it's a hard team to beat, as hard as any in college softball. By my quick count (and some correct me if I messed it up): UF has lost 13 games at home, 8 on the road and 4 at neutral sites (counting the SEC Tournament and WCWS). So that tells me Florida is more likely to lose at home than it is to lose on the road, or roughly even if you lump "non-home" with road and neutral together. Now, yes, Florida does have a winning home record (and that includes some cupcakes, but even against better teams Florida is more likely to win than lose at home). But Florida also has an equally impressive road record over the same span, moreso really if you look at only losses. To change the way things are done just because someone perceives "it's harder to win on the road than at home" without any imperical evidence or a close study of where losses and wins occur makes no sense. I believe it's harder for a team from east of the Mississippi to win games at the Mary Nutter than it is for Pac-12 teams (or anyone west of the Rockies), but I don't have anything to back that up. A road game played an hour or two's drive away maybe shouldn't equal one that involves a whole day on a plane if we're going to go down that road -- should all road games be equal? I saw a Minnesota fan indignant over the idea that road wins don't count for more, but two years ago the Gophes were 9-3 and home and 20-3 on the road. So of course he/she wants the 'bonus.' But with about a dozen home games per year, Minnesota loses three of those each season. It's not radically different statistically from the road record, and could easily be accounted for by showing a tougher strength of schedule (playing better teams in some cases) on the road. So to me it sounds simply like a way to try to fix the system to favor some teams based upon perception and not reality.


Just like looking at one team to make a determination on whether winning on the road or at home is more difficult. When it was proposed to the NCAA committee at the time of the last RPI changes that were enacted there was plenty of data supplied at the same time to back it up. The person supplying the data had a much more detailed system to at least make an attempt to better balance the moving target of RPI. If he's still monitoring the board, he could possibly give you some idea of the difficulty involved.
3leftturns

Registered:
Posts: 10,458
Reply with quote  #32 
I DO think that, to some degree, road wins should have a bonus attached to them
Kurosawa

Registered:
Posts: 2,529
Reply with quote  #33 
Quote:
Originally Posted by Prowler
A big one is the 'bonus for road win, less value for home win' adjustment to the RPI that some people want. I just did a spot check on Florida's losses from 2014-16. It's a small sample, three seasons, and only one team -- but it's a hard team to beat, as hard as any in college softball. By my quick count (and some correct me if I messed it up): UF has lost 13 games at home, 8 on the road and 4 at neutral sites (counting the SEC Tournament and WCWS). So that tells me Florida is more likely to lose at home than it is to lose on the road, or roughly even if you lump "non-home" with road and neutral together. Now, yes, Florida does have a winning home record (and that includes some cupcakes, but even against better teams Florida is more likely to win than lose at home). But Florida also has an equally impressive road record over the same span, moreso really if you look at only losses. To change the way things are done just because someone perceives "it's harder to win on the road than at home" without any imperical evidence or a close study of where losses and wins occur makes no sense. I believe it's harder for a team from east of the Mississippi to win games at the Mary Nutter than it is for Pac-12 teams (or anyone west of the Rockies), but I don't have anything to back that up. A road game played an hour or two's drive away maybe shouldn't equal one that involves a whole day on a plane if we're going to go down that road -- should all road games be equal? I saw a Minnesota fan indignant over the idea that road wins don't count for more, but two years ago the Gophes were 9-3 and home and 20-3 on the road. So of course he/she wants the 'bonus.' But with about a dozen home games per year, Minnesota loses three of those each season. It's not radically different statistically from the road record, and could easily be accounted for by showing a tougher strength of schedule (playing better teams in some cases) on the road. So to me it sounds simply like a way to try to fix the system to favor some teams based upon perception and not reality.


First you say it is a "big one" (change) and then that its "not radically different statistically".

The reason the NCAA went to weighing road wins/home losses in baseball was due to weather cancellations that created a large discrepancy between northern and southern schools in home and total games. That was on top of northern schools already scheduling fewer OOC home games to avoid early season weather.

Softball, on the other hand, is not as prone to weather cancellations as baseball, due to shorter games (easier to get in) and being able to continue playing in worse conditions (softer ball, easier to grip). Note, for instance, over the last five seasons, Washington has had four home and six road conference games cancelled due to weather, although four of the road cancellations were last season. Seattle tends to get drizzle rather than monsoons, meaning that delays happen more often than cancellations. Note, however, that northern schools in the midwest (Minnesota, Michigan, etc.) can be subject to more weather extremes.

So, while softball might not generally have a significant north-south bias, like baseball, that doesn't mean that RPI doesn't have a well-documented east-west bias, with scheduling home games against "RPI cookies" (opponents with high win-loss records from weak conferences) much easier for eastern schools than western schools, for instance. One school of thought is that home/road game weighting would introduce a qualitative component to an algorithm that is otherwise entirely quantitative, in that winning on the road is, indisputably, more difficult than at home.

OldWiseOne

Registered:
Posts: 301
Reply with quote  #34 
So since I brought up Kentucky in 2015 and someone else brought up Stanford in 2014 I wanted to compare them so people could understand what all goes into the different rankings.

Category   UK          Stanford
D1 W-L     29-24      30-25
non-conf   24-4        25-6
D1 SOS     2             28
NC SOS     42           129
D1 opp sos 8            17
NC opp sos 46          69

Kentucky was 2-2 vs. non-conference top 25 RPI (W over Florida St, Arizona, L to Florida St, Oklahoma), 5-1 vs. non-conference 26-50 RPI (W over WKU, FAU(2), SDSU, Louisville, L to Louisville) and did not lose a game to RPI 100+.  In addition, 7 of Kentucky's conference opponents were top 25 RPI teams and 1 was in the 26-30.  All of their conference opponents had records +.500.

Stanford was 0-3 vs. non-conference Top 25 RPI (L to Baylor, Tenn, LSU), 3-2 vs. non-conference RPI 26-50 (W over NC St, Northwester, SDSU, L to USF, NDSU) and they lost to 126 Central Conn State. In addition, only 5 of Stanford's conference opponents were in the top 25 RPI while 3 (Utah, Cal and Oregon State) were ranked 51-100 with Cal and Oregon State having losing records.

When you look at the details of each team it's easy to see how Kentucky was ranked that much higher then Stanford. 
jayrot

Registered:
Posts: 16,785
Reply with quote  #35 
Quote:
Originally Posted by olddawg


You are an SEC guy.  What reason, if any, do thy give for not leveling the playing field?  Do you buy whatever they give?  Do you have an opinion on the question?


I'm the furthest from knowledgeable about the SEC's decision.  I can speculate which is that they will never require the SEC to take on 12 3-game series.  36 conference games would be unfair to the conference members (when other conferences wouldn't be coming close to that number).  Also, if you drop it down to 2-game series and make it 24 conference games, then you run the risk of devaluing a series.  Who would've been happy with LSU/AU splitting the series 1-1?  Who would've been happy with the UF/Bama series last year ending in a 1-1 tie with no rubber match?

I buy it in that it was the easiest/quickest answer, and I hope they are reevaluating it, but knowing that this is softball I doubt they care to reevaluate.

I still feel they need to find a way to do divisions, even though it's unequal divisions.  I think the fun of the baseball setup is that the top seeds from each division get to go 1 and 2 in the conference tournament no matter what the records.  I honestly don't know, but it does need to be addressed.  And Bama_CF gave a good recommendation about allowing for flexibility to reschedule conference games based on standing from previous year.
Bama_CF

Registered:
Posts: 1,379
Reply with quote  #36 

From the OP:  Where teams and conferences as a whole misunderstand how to schedule is based on a few misconceptions:


1.  Playing a tough schedule early will earn me a high RPI - this is true ONLY if you win those games.  If you play 30 non-conference games and 15 are against top 50 teams but you lose 12 of those, your winning percentage is so low that it hurts you more than playing those teams and their winning records because of the number of variables.  The perceived gain by playing those teams with winning records is negated by the losses because of the number of variables in each category.
______________________________________________________


The item shown above in the OP as a "misconception" about the RPI is not really a misconception at all IMO. I disagreed with this when I first read it but didn't respond. Actually, playing a tough schedule helps your RPI period ... a lot! Certainly it helps more if you also win the games. It is correct that the definition of those opponents that help your RPI is their winning percentage, not their RPI rank. But a strong schedule playing a bunch of games against teams with a high winning percentage does help your RPI - even if you lose a significant portion of them.   

Perfect examples in the current RPI are UCLA (RPI #10 with 8 losses) and Wisconsin (RPI #32 with 2 losses). The key factor is UCLA has played against a lot of teams with a lot of wins and that is a boost to 50% of the formula even if your lose. Wisconsin has played a bunch of teams who don't have a poor winning %, and that hurts your PRI even if you win. 

__________________

 

3leftturns

Registered:
Posts: 10,458
Reply with quote  #37 
Yes, I showed deference, too, Bama_CF, but totally agree with you

11 non-conference Top 25 RPIs for UCLA, only 2 for the team with half as many overall losses that just swept them and is TEN slots behind them
OldWiseOne

Registered:
Posts: 301
Reply with quote  #38 
Quote:
Originally Posted by Bama_CF

From the OP:  Where teams and conferences as a whole misunderstand how to schedule is based on a few misconceptions:


1.  Playing a tough schedule early will earn me a high RPI - this is true ONLY if you win those games.  If you play 30 non-conference games and 15 are against top 50 teams but you lose 12 of those, your winning percentage is so low that it hurts you more than playing those teams and their winning records because of the number of variables.  The perceived gain by playing those teams with winning records is negated by the losses because of the number of variables in each category.
______________________________________________________


The item shown above in the OP as a "misconception" about the RPI is not really a misconception at all IMO. I disagreed with this when I first read it but didn't respond. Actually, playing a tough schedule helps your RPI period ... a lot! Certainly it helps more if you also win the games. It is correct that the definition of those opponents that help your RPI is their winning percentage, not their RPI rank. But a strong schedule playing a bunch of games against teams with a high winning percentage does help your RPI - even if you lose a significant portion of them.   

Perfect examples in the current RPI are UCLA (RPI #10 with 8 losses) and Wisconsin (RPI #32 with 2 losses). The key factor is UCLA has played against a lot of teams with a lot of wins and that is a boost to 50% of the formula even if your lose. Wisconsin has played a bunch of teams who don't have a poor winning %, and that hurts your PRI even if you win. 


Yes....playing a tough schedule CAN help you but you're looking at 2 extremes here with a limited scope.  If you look at the team sheets in detail, Wisconsin does have a 20-2 record but their non-conf SOS is 146 while UCLA's is 4.  Wisconsin went to far to the soft schedule extreme which cost them.  But simply playing a touch non-conference schedule doesn't boost you up unless you win.  SFA's non-conf SOS is 42 but they're 9-17 overall and 134 RPI.  Purdue has a non-conf SOS of 26 but is 10-21 and 133 in RPI.  Idaho State's non-conf SOS is 20 but at 6-12 theyire RPI is 115. 

So there are a lot of factors to look at beyond just playing a tough schedule.  I stand by what I wrote as Wisconsin is not even close to being in the same league talent wise as UCLA and if they played the same schedule they would have lost a lot more than 2 games and been a lot lower in RPI.  However, with the weak Big Ten RPI (avg RPI of 83) if they can win a lot of games in conference they've set themselves up for an at-large berth which is what they need to do to make post-season.  However, the weak RPI's of the conference overall may drag them down too far to be on the bubble.

As for the Utah/UCLA argument, again you have to look at more of the details.  Utah's non-conf SOS is 152 compared to UCLA's 4.  Again it's 2 extremes of the spectrum you're trying to compare.  In addition, Utah has 2 bad losses to UNO and Nevada while UCLA has none and UCLA has 6 wins over top 25 and 3 over 26-50 while Utah only has the 2 wins over top 25.  If Utah had UCLA's SOS they might not have won as many games and had a lower RPI.  As it stands now, Utah is pretty much assured an at-large berth since they enter PAC 12 play against teams with an avg RPI of 27 that will keep them near 20 and if they win a bunch may push them a little higher. 

At the end of the day, it comes down to playing a schedule that you can win and if EVERYONE in the conference is doing the same thing and winning those games the teams at the top will continue to rise as they feed off of each other's high winning percentages.
3leftturns

Registered:
Posts: 10,458
Reply with quote  #39 
Good stuff.... always love these discussions
BlueSky

Registered:
Posts: 521
Reply with quote  #40 
RPI guys, see post 142 in Arizona 2017...help me understand. Thanks.
__________________
 
Previous Topic | Next Topic
Print
Reply

Quick Navigation: