ADVERTISEMENT

Warren Nolan's College Baseball RPI site - interesting observation

AuroraHawk

HB Heisman
Dec 18, 2004
7,750
11,111
113
Wrote earlier that I found the "Impact Games" tab to be an addicting feature on his site.

That written, his "Predicted Results" tab has me scratching my head. Someone earlier posted that his prediction for Iowa had listed certain results (can't remember exactly what was posted or who posted it). After I read it, I thought . . . "weird. That's not what I saw earlier in the day."

For sh-ts and giggles, I have clicked on the "Predicted Results" tab several times today and received three markedly different sets of outcomes. For example, I've seen predictions that Iowa would sweep OSU and I've seen predictions that OSU would take 2 of 3 and I've seen that OSU would sweep Iowa. Can't help but wonder if his site is constantly running an algorithm and it constantly changes predictions. Iowa's predicted final RPI has ranged from 56 to 87.

Odd.
 
  • Like
Reactions: Couppee
Wrote earlier that I found the "Impact Games" tab to be an addicting feature on his site.

That written, his "Predicted Results" tab has me scratching my head. Someone earlier posted that his prediction for Iowa had listed certain results (can't remember exactly what was posted or who posted it). After I read it, I thought . . . "weird. That's not what I saw earlier in the day."

For sh-ts and giggles, I have clicked on the "Predicted Results" tab several times today and received three markedly different sets of outcomes. For example, I've seen predictions that Iowa would sweep OSU and I've seen predictions that OSU would take 2 of 3 and I've seen that OSU would sweep Iowa. Can't help but wonder if his site is constantly running an algorithm and it constantly changes predictions. Iowa's predicted final RPI has ranged from 56 to 87.

Odd.

It has to be. It had us sweeping UCI last I Looked and that’s a tall task.
 
Wrote earlier that I found the "Impact Games" tab to be an addicting feature on his site.

That written, his "Predicted Results" tab has me scratching my head. Someone earlier posted that his prediction for Iowa had listed certain results (can't remember exactly what was posted or who posted it). After I read it, I thought . . . "weird. That's not what I saw earlier in the day."

For sh-ts and giggles, I have clicked on the "Predicted Results" tab several times today and received three markedly different sets of outcomes. For example, I've seen predictions that Iowa would sweep OSU and I've seen predictions that OSU would take 2 of 3 and I've seen that OSU would sweep Iowa. Can't help but wonder if his site is constantly running an algorithm and it constantly changes predictions. Iowa's predicted final RPI has ranged from 56 to 87.

Odd.

Maybe it’s a Monte Carlo simulation but it’s only displaying one trial instead of a probability distribution
 
Merely for comparative purposes at a later time:

As of 11:30 a.m. on Saturday, Nolan's "predicted results" model shows Iowa losing to OSU today, tomorrow, beating WIU and Northwestern, getting swept by UCI, sweeping MSU and then losing out the balance of the games (losing 1 to WIU and getting swept by Maryland). Finishing regular season at 29-24 and 14-11 in B1G. Final RPI of 99.
 
Merely for comparative purposes at a later time:

As of 11:30 a.m. on Saturday, Nolan's "predicted results" model shows Iowa losing to OSU today, tomorrow, beating WIU and Northwestern, getting swept by UCI, sweeping MSU and then losing out the balance of the games (losing 1 to WIU and getting swept by Maryland). Finishing regular season at 29-24 and 14-11 in B1G. Final RPI of 99.

LOLz last night after the game was completely different
 
Examples of strength of opponent difference in RPI (per Nolan site):

If Michigan beats Rutgers today, it gains 11 RPI points
If Iowa beats Ohio State today, it gains 33 RPI points
If Illinois beats Nebraska today, it gains 34 RPI points
If Nebraska beats Illinois today, it gains 39 RPI points

If Michigan loses to Rutgers, it loses 67 RPI points
If Iowa loses to Ohio State it loses 32 RPI points
If Illinois loses to Nebraska, it loses 21 RPI points
If Nebraska loses to Illinois, it loses 31 RPI points

Even if both Iowa and Michigan win/lose today, there's a good chance that Iowa catches Michigan in the RPI rankings.

And, yet again,
-2.3 RPI points if Nebraska wins
-3.1 RPI points if Illinois wins
Given that Illinois has a higher RPI than Nebraska and Iowa swept them while Iowa only won 2 of 3 against Nebraska, this seems counterintuitive.
 
Yeah I don’t exactly get the college baseball RPI. Seems like it needs tweaked due to home wins/loses and even road. I don’t think you should win 2/3 at home and drop.
 
Merely for comparative purposes at a later time:

As of 11:30 a.m. on Saturday, Nolan's "predicted results" model shows Iowa losing to OSU today, tomorrow, beating WIU and Northwestern, getting swept by UCI, sweeping MSU and then losing out the balance of the games (losing 1 to WIU and getting swept by Maryland). Finishing regular season at 29-24 and 14-11 in B1G. Final RPI of 99.

As of 2:45 p.m., Nolan's "predicted results" model shows Iowa finishing with a final RPI of 58. Instead of getting swept by UCI, it has Iowa sweeping UCI. While it certainly appears that the official NCAA RPI site mirrors Nolan's RPI rankings (which suggests his "Impact Games" analysis is accurate), his "predicted results" model appears to be far from what it portends to be.
 
As of 2:45 p.m., Nolan's "predicted results" model shows Iowa finishing with a final RPI of 58. Instead of getting swept by UCI, it has Iowa sweeping UCI. While it certainly appears that the official NCAA RPI site mirrors Nolan's RPI rankings (which suggests his "Impact Games" analysis is accurate), his "predicted results" model appears to be far from what it portends to be.
yeah, when did he have UCI sweeping Iowa?

and then when did he do a 180 and have Iowa sweeping UCI?
 
I love the live RPI. I love seeing the impact games. But, the predicted results is crazy. When I went to bed last night our predicted final record was 31-22. At 6:45 this morning it was 29-24. At 8:05 it’s now 35-18 with a loss today, one loss vs Irvine and Maryland and the rest wins. As someone said, it’s a head scratcher.
 
I love the live RPI. I love seeing the impact games. But, the predicted results is crazy. When I went to bed last night our predicted final record was 31-22. At 6:45 this morning it was 29-24. At 8:05 it’s now 35-18 with a loss today, one loss vs Irvine and Maryland and the rest wins. As someone said, it’s a head scratcher.

Notably, under that predicted set of circumstances, Iowa’s RPI is predicted to be 46 and 17-8 in the conference.

If (and I stress IF) it played out that way and if Iowa has even a decent showing at the B1G tournament, the NCAA Selection Committee will be discussing Iowa’s resume.
 
Notably, under that predicted set of circumstances, Iowa’s RPI is predicted to be 46 and 17-8 in the conference.

If (and I stress IF) it played out that way and if Iowa has even a decent showing at the B1G tournament, the NCAA Selection Committee will be discussing Iowa’s resume.

"Predicted Results" model has already changed again . . . with no games having been played.
 
Those predicted results are useless. It's got NW winning 5 of 6 at home vs Nebraska and Minnesota, 2 out of 3 on the road vs Rutgers to finish 6th in the conference.
 
ADVERTISEMENT
ADVERTISEMENT