(04-27-2013 10:06 PM)waltgreenberg Wrote:  

(04-27-2013 09:57 PM)NicevilleWRC Wrote:  

Kendall Wrote:Arkansas has an RPI of 61, but actually otherwise has a solid resume. For instance, the Hogs are 7-7 vs. RPI Top 25, 11-9 vs. RPI Top 50 and 16-13 vs. RPI Top 100. As for the Owls, they currently sit second in Conference USA with a 9-6 record. They also have an RPI of 43, along with an 0-2 mark vs. RPI Top 50 and 13-10 record vs. RPI Top 100.

This type of analysis irritates me to no end, and Kendall isn't the only one guilty of it. If you know a team's RPI it doesn't matter what they did against the RPI Top 50, 100, etc. All of that is already taken into account in RPI, so breaking it down reveals no new information about the team's ability (unless you're curious how they got there)!

In fact, it only makes the analysis less accurate because a team's record vs the Top 50 could be completely misleading - maybe one team played 5 teams in the Top 10, and another played only teams in the 40s. Who knows? That's why you have RPI in the first place, because it takes all of that information into consideration.

And this isn't even going into all of the problems with the actual method of calculating RPI...

Sorry, but that's not true. No question RPI is a flawed measure, with definite southeastern bias inherent to the algorithm. However, record vs. Top 25/50/100 is NOT already captured within the RPI. Sure, it impacts it, but not that much. For example, you can be 0-6 against the Top 25 and just .500 vs. Top 50/100, but have a Top 25 RPI due to SoS. RPI and resume vs. Top 25/50/100 are not overlapping measures...and they're reported by Kendall and others because both are key criteria used by the Selection Committee.

I think the point was that RPI already awards bonuses for wins over 1-25, 26-50, and 51-75.

Quote:Kendall Rogers ‏@KendallRogersPG
The more I look at it, the more I think #Rice sneaks in and grabs a host site. #Owls up to 39 in RPI, w/ extremely favorable remaining sked.

(04-28-2013 02:53 PM)dragon2owl Wrote:  

Quote:Kendall Rogers ‏@KendallRogersPG
The more I look at it, the more I think #Rice sneaks in and grabs a host site. #Owls up to 39 in RPI, w/ extremely favorable remaining sked.

I'm guessing our RPI goes back up based on the "extremely favorable remaining sked."

UAB 189
Memphis 113
Marshall 218

Our W-L is only 25% of the equation

(04-28-2013 02:53 PM)dragon2owl Wrote:  

Quote:Kendall Rogers ‏@KendallRogersPG
The more I look at it, the more I think #Rice sneaks in and grabs a host site. #Owls up to 39 in RPI, w/ extremely favorable remaining sked.

I think Kendall's dreaming. Yes, very favorable remaining schedule, BUT, save for Lamar, against woeful competition that will not help our RPI much even if we win out. And, unless USM goes into a nosedive at the finish (highly unlikely given how well they've been pitching of late with Pierce and Drehoff both lights out since we played them mid-March), we're not going to win the regular season conference title in a down year for the conference.

(04-28-2013 03:15 PM)grol Wrote:  I'm guessing our RPI goes back up based on the "extremely favorable remaining sked."

UAB 189
Memphis 113
Marshall 218

Our W-L is only 25% of the equation

RPI is not recursive and the winning percentage component has the highest variability.

What does recursive mean?

(04-25-2013 04:31 PM)waltgreenberg Wrote:  Kendall has us at #2 seed in the Tempe, AZ regional, but his analysis of the proojections breakdown is spot-on, IMO...

http://www.perfectgame.org/Articles/View...ticle=8333

Quote:How about the curious cases of Arkansas and Rice? Both of these clubs have notched some nice wins this season. Yet, they're not in the hosting mix right now because of one chief reason -- RPI. Yeah, that mean old RPI that everyone talks about is actually hurting a pair of power programs this spring. There's no doubt the Hogs have elite potential, as evidenced by their high ranking in the latest PG College Top 25. However, some bad midweek losses, including a weekend loss to Pacific, have caught up with them, perhaps putting them in such a hole it could be impossible to climb out of down the stretch. With that said, let's delve into the two resumes. Arkansas has an RPI of 61, but actually otherwise has a solid resume. For instance, the Hogs are 7-7 vs. RPI Top 25, 11-9 vs. RPI Top 50 and 16-13 vs. RPI Top 100. As for the Owls, they currently sit second in Conference USA with a 9-6 record. They also have an RPI of 43, along with an 0-2 mark vs. RPI Top 50 and 13-10 record vs. RPI Top 100. The NCAA committee will have a tough time dealing with these two teams. On one hand, the RPI must be held against them, but just how much? We'll find out sooner rather than later. Stanford is another team in the same boat as the Razorbacks and Owls.

The flaws in RPI are really being highlighted this year as there are a large number of second tier conference team with Top 35 RPis, but few of the pundits believe any of them are legitimate Top 30 caliber clubs.

The RPI has its flaws, but do you really think the pundits do a better job? It Aurora me as wholly unlikely that the pundits could know the top, say, 100 programs well enough to develop a better determination of the top 30 than an objective ranking system can. In my mind, in order of likelihood of accuracy, it goes ISR > RPI > pundits' opinions.

Also, your numbers don't make sense. You're saying the RPI Top-35 includes teams that pundits say don't belong in the top 30. Is that supposed to mean something?

(04-28-2013 03:17 PM)waltgreenberg Wrote:  

(04-28-2013 02:53 PM)dragon2owl Wrote:  

Quote:Kendall Rogers ‏@KendallRogersPG
The more I look at it, the more I think #Rice sneaks in and grabs a host site. #Owls up to 39 in RPI, w/ extremely favorable remaining sked.

I think Kendall's dreaming. Yes, very favorable remaining schedule, BUT, save for Lamar, against woeful competition that will not help our RPI much even if we win out. And, unless USM goes into a nosedive at the finish (highly unlikely given how well they've been pitching of late with Pierce and Drehoff both lights out since we played them mid-March), we're not going to win the regular season conference title in a down year for the conference.

Don't think the regular season title is gone just yet. But, we have to go to Birmingham and get a sweep. Given our recent history there, it may not be likely, but we can't afford to fall any further behind. At least USM's last 6 conference games are on the road, while we have 3 on the road and the last 6 at home. That will make their road a bit tougher. Have to get some help from ECU and/or UH.

As far as regional host, I agree. I just don't see it. Not yet anyway.

(04-28-2013 04:31 PM)JOwl Wrote:  

(04-25-2013 04:31 PM)waltgreenberg Wrote:  Kendall has us at #2 seed in the Tempe, AZ regional, but his analysis of the proojections breakdown is spot-on, IMO...

http://www.perfectgame.org/Articles/View...ticle=8333

Quote:How about the curious cases of Arkansas and Rice? Both of these clubs have notched some nice wins this season. Yet, they're not in the hosting mix right now because of one chief reason -- RPI. Yeah, that mean old RPI that everyone talks about is actually hurting a pair of power programs this spring. There's no doubt the Hogs have elite potential, as evidenced by their high ranking in the latest PG College Top 25. However, some bad midweek losses, including a weekend loss to Pacific, have caught up with them, perhaps putting them in such a hole it could be impossible to climb out of down the stretch. With that said, let's delve into the two resumes. Arkansas has an RPI of 61, but actually otherwise has a solid resume. For instance, the Hogs are 7-7 vs. RPI Top 25, 11-9 vs. RPI Top 50 and 16-13 vs. RPI Top 100. As for the Owls, they currently sit second in Conference USA with a 9-6 record. They also have an RPI of 43, along with an 0-2 mark vs. RPI Top 50 and 13-10 record vs. RPI Top 100. The NCAA committee will have a tough time dealing with these two teams. On one hand, the RPI must be held against them, but just how much? We'll find out sooner rather than later. Stanford is another team in the same boat as the Razorbacks and Owls.

The flaws in RPI are really being highlighted this year as there are a large number of second tier conference team with Top 35 RPis, but few of the pundits believe any of them are legitimate Top 30 caliber clubs.

The RPI has its flaws, but do you really think the pundits do a better job? It Aurora me as wholly unlikely that the pundits could know the top, say, 100 programs well enough to develop a better determination of the top 30 than an objective ranking system can. In my mind, in order of likelihood of accuracy, it goes ISR > RPI > pundits' opinions.

Also, your numbers don't make sense. You're saying the RPI Top-35 includes teams that pundits say don't belong in the top 30. Is that supposed to mean something?

??? First off, there is no such thing as an purely "objective" ranking system. Though I agree that ISR has proven over the years to be superior to RPI, it has a Western regional bias inherent to it's algorithm just as the RPI has a heavily favored southeastern bias inherent to it's algorithm. I dare you to take a look at the Top 40 RPI teams as of this morning, and agree those are the Top 40 programs in the country. Do you honestly believe South Alabama is the 14th best team in the country at the moment, or ULaLa is the 24th best team? Conversely, are you telling me you think Stanford is the 93rd best team in the country (20 places below UH), or that Arkansas (with a 29-15 record, 10-11 vs. Top 50, 16-13 vs. Top 100) is only the 53rd rank team in the nation, with an RPI virtually tied with that of 21-19 UT (who are dead last in a lousy Big 12)?

The major national college baseball pundits (Kendall, Aaron Fitt, Eric Sorenson and Mark Etheridge covering the broad southeastern region) do very closely track top 150+ programs and do so on a game-by-game, week-by-week basis. Are they biased at times (based on teams they've seen the most or coaches who are responsive to them)? Of course, as they're human...BUT they know the relative merits of the Top 100 - 150 programs in the country far better than the coaches or the writers (who are also biased), and understand that rankings should be based on more than any one measure (e.g., W-L record, RPI, ISR, SoS, record vs. Top 25/50/100, weekend series W-L, road record, record over past 10 games), but rather an overall assessment of all these measures taken together.

So, to answer your question-- yes, there is no doubt in my mind the "pundits" (and even the Selection Committee) do a much, MUCH better job of ranking the Top 50 teams than either RPI or ISR alone.

(04-28-2013 05:16 PM)waltgreenberg Wrote:  So, to answer your question-- yes, there is no doubt in my mind the "pundits" (and even the Selection Committee) do a much, MUCH better job of ranking the Top 50 teams than either RPI or ISR alone.

Walt ... you may be right, but that tends to be the same argument we bemoan come NCAA MBB tournament time (the dreaded "eyes" test).

(04-28-2013 05:16 PM)waltgreenberg Wrote:  Rankings should be based on more than any one measure (e.g., W-L record, RPI, ISR, SoS, record vs. Top 25/50/100, weekend series W-L, road record, record over past 10 games), but rather an overall assessment of all these measures taken together.

So, to answer your question-- yes, there is no doubt in my mind the "pundits" (and even the Selection Committee) do a much, MUCH better job of ranking the Top 50 teams than either RPI or ISR alone.

This is simply incorrect. If you already know record and strength of schedule then how a team did against the Top 25/50/100, weekend series, road records, past 10 games, etc provides no additional information. It's already included in the record and strength of schedule (which should include adjustments for home/away).

If two teams have the same record and same strength of schedule but one has a better record against Top 100 teams, all that means is that same team was worse against the non-Top 100 teams. If one has a better road record, then it has a worse home record. If it's better in the past 10 games, it was worse over the rest of the season. If it won more weekend series, then it lost more weekday games or never swept a weekend.

And there's no way a human can keep track of the Top 50 teams in college baseball better than a computer can, if for no other reason than you have to track all 300 teams to be able to decide who are the Top 50. If you didn't follow all 300 how would you tell that a loss to Prairie View is worse than a loss to Seattle, which is worse than Hawaii, which is worse than Missouri? How would you know Stanford should be in the Top 50 but not Seton Hall?

Finally, ISR is NOT biased towards the West coast teams. Yes, it rates western teams higher than RPI does but that's because RPI is biased and inaccurate due to scheduling differences. The truth doesn't have to be in the middle. If I said a foot is 12 inches long and my friend says it's 14 inches, that doesn't inherently mean we're both off by an inch.

(04-28-2013 08:44 PM)NicevilleWRC Wrote:  

(04-28-2013 05:16 PM)waltgreenberg Wrote:  Rankings should be based on more than any one measure (e.g., W-L record, RPI, ISR, SoS, record vs. Top 25/50/100, weekend series W-L, road record, record over past 10 games), but rather an overall assessment of all these measures taken together.

So, to answer your question-- yes, there is no doubt in my mind the "pundits" (and even the Selection Committee) do a much, MUCH better job of ranking the Top 50 teams than either RPI or ISR alone.

This is simply incorrect. If you already know record and strength of schedule then how a team did against the Top 25/50/100, weekend series, road records, past 10 games, etc provides no additional information. It's already included in the record and strength of schedule (which should include adjustments for home/away).

If two teams have the same record and same strength of schedule but one has a better record against Top 100 teams, all that means is that same team was worse against the non-Top 100 teams. If one has a better road record, then it has a worse home record. If it's better in the past 10 games, it was worse over the rest of the season. If it won more weekend series, then it lost more weekday games or never swept a weekend.

And there's no way a human can keep track of the Top 50 teams in college baseball better than a computer can, if for no other reason than you have to track all 300 teams to be able to decide who are the Top 50. If you didn't follow all 300 how would you tell that a loss to Prairie View is worse than a loss to Seattle, which is worse than Hawaii, which is worse than Missouri? How would you know Stanford should be in the Top 50 but not Seton Hall?

Finally, ISR is NOT biased towards the West coast teams. Yes, it rates western teams higher than RPI does but that's because RPI is biased and inaccurate due to scheduling differences. The truth doesn't have to be in the middle. If I said a foot is 12 inches long and my friend says it's 14 inches, that doesn't inherently mean we're both off by an inch.

Let"s just agree to disagree. I stand behind my previous post.

(04-27-2013 10:57 AM)waltgreenberg Wrote:  2. Oklahoma is only projected as a #1 seeded host at the moment because they're running away in a lousy Big 12 (which is arguably no better than CUSA this year, and that's not saying much). We're already virtually tied with them in RPI, so we could very possibly pass them in RPI by next weekend. Oklahoma has played a total joke of an OOC schedule. As bad as our strength of schedule (currently #82)has been (due to unexpected horrid seasons by Hawaii, Louisiana Tech and Harvard), Oklahoma's is considerably worse (currently #141).

Texas Tech beat Oklahoma, 9-3, Sunday in the final game of their weekend series in Norman. (The Sooners took the series, 2-1.)

(04-28-2013 10:10 PM)Almadenmike Wrote:  

(04-27-2013 10:57 AM)waltgreenberg Wrote:  2. Oklahoma is only projected as a #1 seeded host at the moment because they're running away in a lousy Big 12 (which is arguably no better than CUSA this year, and that's not saying much). We're already virtually tied with them in RPI, so we could very possibly pass them in RPI by next weekend. Oklahoma has played a total joke of an OOC schedule. As bad as our strength of schedule (currently #82)has been (due to unexpected horrid seasons by Hawaii, Louisiana Tech and Harvard), Oklahoma's is considerably worse (currently #141).

Texas Tech beat Oklahoma, 9-3, Sunday in the final game of their weekend series in Norman. (The Sooners took the series, 2-1.)

That's gotta help our RPI, right?? Awesome. As well as damaging OU's hosting chances, of course. Win-win. Thanks, ITT Tech!

Texas is currently last in the Big 12 after being swept this weekend at Baylor. They have Kansas State and at TCU to close out their conference schedule (and three SWAC teams which should help their RPI lol). And they're now a game behind Texas Tech (and behind them in the tie-breaker) for the 8th and final Big 12 spot for that conference's post-season tournament.

It will be interesting to watch to see what Augie Garrido does at the end of the season after missing out on the NCAA's for the second straight year. Already he's pulling out the I'm not ready to retire quotes but I'm not so sure about it. I would imagine the UT fans would want one of their big three coaches to go after a horrible season in their main three sports - and Garrido is their final option after Mack Brown and Rick Barnes made it through their respective seasons.

http://www.statesman.com/news/sports/col...ing/nXSWW/

(04-29-2013 05:35 AM)Fort Bend Owl Wrote:  Texas is currently last in the Big 12 after being swept this weekend at Baylor. They have Kansas State and at TCU to close out their conference schedule (and three SWAC teams which should help their RPI lol). And they're now a game behind Texas Tech (and behind them in the tie-breaker) for the 8th and final Big 12 spot for that conference's post-season tournament.

It will be interesting to watch to see what Augie Garrido does at the end of the season after missing out on the NCAA's for the second straight year. Already he's pulling out the I'm not ready to retire quotes but I'm not so sure about it. I would imagine the UT fans would want one of their big three coaches to go after a horrible season in their main three sports - and Garrido is their final option after Mack Brown and Rick Barnes made it through their respective seasons.

http://www.statesman.com/news/sports/col...ing/nXSWW/

We actually lost .001 yesterday, and still at #39, but we're a win Wednesday vs. #49 Lamar from #32 or #33 in the RPI rankings. To say a win Wednesday is all-important would be an understatement. Hopefully, we can shake the rust off in our usual one or two intrasquads today and tomorrow. A win vs. Lamar gives us a chance to finish up with a winning record vs. Top 50 (we're currently 1-1), and they're our final Top 100 team before the post-season. Unfortunately, it's just as critical a game for Lamar, and you can bet they'll be up for the game. Wouldn't be surprised at all if they start a weekend starting pitcher against us.

BTW, Oklahoma has already fallen behind us a couple spots in RPI, but are still considerably ahead of us in the any hosting consideration given their resume vs. Top 25/50/100 and their Big 12 leadership.

(04-28-2013 08:44 PM)NicevilleWRC Wrote:  This is simply incorrect. If you already know record and strength of schedule then how a team did against the Top 25/50/100, weekend series, road records, past 10 games, etc provides no additional information. It's already included in the record and strength of schedule (which should include adjustments for home/away).

And there's no way a human can keep track of the Top 50 teams in college baseball better than a computer can, if for no other reason than you have to track all 300 teams to be able to decide who are the Top 50. If you didn't follow all 300 how would you tell that a loss to Prairie View is worse than a loss to Seattle, which is worse than Hawaii, which is worse than Missouri? How would you know Stanford should be in the Top 50 but not Seton Hall?

Again, IMO, your first statement above is simply not true. RPI, ISR or any other computerized measure simply "homogenizes" numerous different measures (each getting at a unique perspective) into one. Knowing SoS and overall W-L record does not tell us how a team has done of late (over past 10-15 games) or how it has done on the road or in weekend series. And while it is true that RPI or ISR does factor in strength of schedule and W-L record, both give significant weight to your opponents SoS and W-L record and, consequently, gives far less weight to each team's own record vs. Top 25/50/100.

I'll take the major national college baseball pundits' assessment of teams over simply a single RPI or ISR measure any day of the week. Not only do they follow almost all the conferences and D-I level programs, but they factor in trending and the nuances in their resumes into their rankings. Again, do you honestly believe that #14 in RPI South Alabama or #24 ULaLa is a MUCH better team this year than #53 ranked Arkansas? Given Arkansas' pitching, I'm willing to bet the Hogs would take at least 2 of 3 from either of them.

(04-28-2013 03:17 PM)waltgreenberg Wrote:  

(04-28-2013 02:53 PM)dragon2owl Wrote:  

Quote:Kendall Rogers ‏@KendallRogersPG
The more I look at it, the more I think #Rice sneaks in and grabs a host site. #Owls up to 39 in RPI, w/ extremely favorable remaining sked.

I think Kendall's dreaming. Yes, very favorable remaining schedule, BUT, save for Lamar, against woeful competition that will not help our RPI much even if we win out. And, unless USM goes into a nosedive at the finish (highly unlikely given how well they've been pitching of late with Pierce and Drehoff both lights out since we played them mid-March), we're not going to win the regular season conference title in a down year for the conference.

(04-28-2013 04:16 PM)Gravy Owl Wrote:  

(04-28-2013 03:15 PM)grol Wrote:  I'm guessing our RPI goes back up based on the "extremely favorable remaining sked."

UAB 189
Memphis 113
Marshall 218

Our W-L is only 25% of the equation

RPI is not recursive and the winning percentage component has the highest variability.

So these posts last night inspired me to do a little figurin' today.

First off, it's important to note the changes to the RPI formula this year: http://www.ncaa.com/news/baseball/2011-0...013-season

There are no longer bonus points for wins vs. top 25/50/75. Figuring your WP has become quite a bit harder since not only the numerator but the denominator as well change depending on whether a particular W/L was at home or on the road.

I decided to do some 'bracketing' of where our RPI could possibly land, based on winning out the remaining 12 games vs. going 0-12.

Our current RPI is 0.553 per Boyd. Plugging in our remaining opponents WP (assuming that ex-Rice, they stay approximately constant the rest of the way) reduces our RPI to 0.547 which will drop us to about #51 if our WP stays constant. [ Show the work: Current opponents' record: 865-810. Remaining opponents' record: 245-283 (UAB & Marshall are killing us here). OWP drops from 0.5164 to 0.5039. The 0.012 point drop is weighted by 50%. ]

So starting from 0.547, if we win out, that raises our WP from its current 27.4/41.7=0.657 to 38.8/53.1=0.731. This 0.074 point increase is weighted by 25% to give an RPI increase of 0.018, bumping us up to 0.565, or roughly #25. That's our rough ceiling this year.

On the flip side, if we were to somehow lose out, the revised WP would be 27.4/54.3=0.505, for a drop of 0.152*25%=0.038, dropping us to 0.509 (#136 & out of the tourney)

Taking a more 'realistic' case, let's look at what happens if Rice finishes up 5-2 at home vs. Lamar, Memphisx3, and Marshallx3; and 4-1 on the road at UABx3, LaTech, and TexSt. That puts our WP at 36.1/53.7=0.672, or a modest 0.015*25%=0.004 RPI boost (#41, or right about where we are now).

As far as rooting for non-con opponents (or conference opponents in OOC games) - our opponents will play roughly 2850 games. Therefore, for each 6 games our opponents play above 0.500, that boosts our RPI by 0.001. [Note that a Hawaii win counts as 4 wins, Stanford as 3, etc. in that analysis] Nice to have, but not going to move the needle too much.

Net, net, (as some would say) - Rice probably needs to lose no more than 4 games the rest of the way pre-tourney to keep RPI in that 'at-large' sort of range (or we can just win the tourney and not have to worry. :) ). Hosting's pretty much out the window.

Given the bizarrely down season in Texas and the mid-south, I could see us with a very outside chance to host if:

* We come back and win the conference regular season
* Our RPI, somehow, ends up in the 20s

As Walt said .... long shot at best.

(04-29-2013 03:29 PM)Orange County Owl Wrote:  Given the bizarrely down season in Texas and the mid-south, I could see us with a very outside chance to host if:

* We come back and win the conference regular season
* Our RPI, somehow, ends up in the 20s

As Walt said .... long shot at best.

Sure, if both of those things occur, I do think we cop the last hosting spot, BUT....that would require an enormous amount of help from both USM (near total collapse) and a number of teams (5 - 10 of them) ahead of us in RPI.