Bill Connelly's SP+ rankings - Printable Version
+- The CardBoard (https://thecardboard.org/board)
+-- Forum: C-House! (https://thecardboard.org/board/forum-4.html)
+--- Forum: The CardBoard (https://thecardboard.org/board/forum-5.html)
+--- Thread: Bill Connelly's SP+ rankings (/thread-23215.html)
+- The CardBoard (https://thecardboard.org/board)
+-- Forum: C-House! (https://thecardboard.org/board/forum-4.html)
+--- Forum: The CardBoard (https://thecardboard.org/board/forum-5.html)
+--- Thread: Bill Connelly's SP+ rankings (/thread-23215.html)
Bill Connelly's SP+ rankings - msqueri - 05-23-2022
Bill Connelly updated his SP+ rankings based on new information from the comings and goings of the transfer portal. He now pegs Stanford as having the #60 offense and #90 defense, which puts us as the #77 team. That's a nine spot drop compared to the beginning of the off-season. Baylor, Arizona State, UTSA, Wyoming, and James Madison are the only teams that have dropped more in Connelly's estimation (James Madison because of an adjustment in how he views their recruiting, the others because of transfer portal carnage). Here are how Pac-12 teams/teams on Stanford's schedule now rank according to Connelly:
7) Notre Dame
12) Utah
24) Oregon
25) BYU
41) UCLA
42) Arizona State
50) Oregon State
51) USC
54) Washington
69) Washington State
74) Cal
77) Stanford
96) Arizona - Stanford doesn't play
100) Colorado - Stanford doesn't play
N/A) Colgate
Connelly notes that USC and LSU present conundrums for his model and he acknowledges USC will certainly be a preseason top ten team.
If you'll allow me the liberty of conflating metrics/methodological approaches, one can get a sense of win probability according to Connelly using Sagarin's ordinal team rankings, numerical ratings, standard three point advantage, and Las Vegas norms for calculating probability based on point spread. To wit:
Colgate - 100 percent
USC - 47.5 percent
@ Washington - 29.7 percent
@ Oregon - 14.9 percent
Oregon State - 47.5 percent
@ Notre Dame - 0 percent
Arizona State - 45.8 percent
@ UCLA - 24.75 percent
Washington State - 52.5 percent
@ Utah - 5.1 percent
@ Cal - 38.5 percent
BYU - 26.2 percent
TOTAL WIN PROBABILITY: 4.3245
RE: Bill Connelly's SP+ rankings - 2006alum - 05-23-2022
Thanks for this analysis. A bit grim that the only game we'd be favored in is versus a liberal arts college with fewer than 3,000 undergrads...
RE: Bill Connelly's SP+ rankings - BostonCard - 05-23-2022
(05-23-2022, 10:23 AM)msqueri Wrote: TOTAL WIN PROBABILITY: 4.3245
I think you mean expected number of wins, because one can’t have a probability greater than 1. Also, 4 significant digits?
Four seems like a reasonable estimate, but to be honest, everything from 1 to 6 is probably in play. There are a lot of “toss ups” in there, though those are hardly independent. There’s a decent likelihood that Connelly overestimates the team, especially on defense, which I expect will be atrocious. Thus, I would take the under on four, but who knows.
BC
RE: Bill Connelly's SP+ rankings - msqueri - 05-23-2022
(05-23-2022, 05:36 PM)BostonCard Wrote:(05-23-2022, 10:23 AM)msqueri Wrote: TOTAL WIN PROBABILITY: 4.3245
I think you mean expected number of wins, because one can’t have a probability greater than 1. Also, 4 significant digits?
Four seems like a reasonable estimate, but to be honest, everything from 1 to 6 is probably in play. There are a lot of “toss ups” in there, though those are hardly independent. There’s a decent likelihood that Connelly overestimates the team, especially on defense, which I expect will be atrocious. Thus, I would take the under on four, but who knows.
BC
Good corrections/points. I was typing quickly and indeed misspoke in saying probability rather than expected wins. The digits was just what my back of the envelope math came up with, not trying to make a comment on significance of the digits. To me this just generally reinforces other attempts I've done at expected wins. Four seems like a fair guess.
I am not sure if it's more likely Connelly has overestimated us or underestimated us, but I think it's easier for me to see upside for the offense vis-a-vis his projection than it is to see downside for the defense. It's really not that tough for me to imagine Stanford having a top 30 offense, which would be a big difference compared to a #60 projection. I think the reasons for skepticism that lead to the #60 projection are valid, but if the OL "gels" as hoped and Smith is anything like Shaw talks about (and can stay reasonably healthy) it's not that hard to see a good offense. Not to say I'm optimistic about the defense, but pegging us as the #90 defense is already baking in a healthy disdain for what Lance Anderson will trot out there.
The 3-5 win range seems most likely to me. As down as I am on Stanford football a 1-2 win season strikes me as unlikely. Six seems more likely. More than that would require a lot of us being wrong and 2019/2021 really truly being aberrations and/or hangovers we've sobered up from.
RE: Bill Connelly's SP+ rankings - CTcard - 05-24-2022
(05-23-2022, 05:36 PM)BostonCard Wrote:(05-23-2022, 10:23 AM)msqueri Wrote: TOTAL WIN PROBABILITY: 4.3245Also, 4 significant digits?
4??