Jump to content

S&P+ rankings: Overperformance or overcorrection?

satyanash

By satyanash
in Football

 Share

Recommended Posts

  • Replies 1.1k
  • Created
  • Last Reply

Top Posters In This Topic

Popular Days

Top Posters In This Topic

Popular Days

Popular Posts

Dahobbs
Dahobbs

Fozzz is a fucking idiot. 

Magus Ossis
Magus Ossis

A model that drops the winner and sustains the loser of a game needs adjusting.

NeverMarryAStripper
NeverMarryAStripper

Exactly.  Just because you beat someone doesn't mean than you can beat them.

Posted Images

Valmy77

Valmy77 13226

Posted
Posted (edited)
1 hour ago, Bevo14 said:

Well that just proves Mizzou didn't want to be there. Thanks, Bill.

Missouri was way down entering the 4th Quarter and got a few TDs to make it closer than it really was. But I guess they were far more efficient (WTF is that exactly?) in falling behind 34-17 after 3 quarters than Wyoming was getting ahead 34-17.

I mean they gave up almost 300 rushing yards against Wyoming. You have to be pretty efficient to do something like that.

Edited by Valmy77
Link to comment
Share on other sites
texasstrong12

texasstrong12 4668

Posted
Posted (edited)

S&P + preseason rankings- 

Missouri #13 (lost to Wyoming)

Tennessee #26 (lost to Georgia State)

South Carolina #17 (lost to UNC who finished 2-9 last year)

It was obvious from the start these teams weren't going to be very good but don't tell Bill. His system is perfect. 

Edited by texasstrong12
  • Like 2
  • Haha 2
Link to comment
Share on other sites
Zavala

Zavala 1384

Posted
Posted
12 hours ago, HtownHorn said:

Maybe Bill should actually take into account Wins and Losses. Usually they are a pretty decent indicator of how good or bad a team is.

His SP+ SEC rankings took a bath today.

Imagine the tweaking he had to do to satisfy his homerism.

  • Like 1
Link to comment
Share on other sites
sushihorn

sushihorn 3233

Posted
Posted

Hypothesis: SP+ has a significant pro-SEC bias.

Design: Bias should manifest itself most strongly in assessment of group members who are being judged based on group identity rather than individual accomplishment.  That should impact the bottom teams most as few bother to analyze them as individual programs.  We will measure the lower ranked teams in the SEC by result vs SP+ projection.

Kentucky (+7.3 SP+) vs Toledo (-1.5) = projected MOV 8.8 pts + 3.5 home field = +12.3 final projection (+14 actual); +1.7 deviation

MS St. (+20.3) vs UL-Lafayette (-6.9) = +27.2 MOV + 3.5 HF = +30.7 projection (+10 actual); -20.7 deviation

Ole Miss (+8.5) vs Memphis (+10.9) = -2.4 MOV - 3.5 HF = -5.9  projected (-5 actual); +0.9 deviation

Tennessee (+10.9) vs Georgia St. (-15.6) = +26.5 MOV + 3.5 HF = +30.0 projected (-8 actual); -38.0 deviation

South Carolina (+15.3) vs UNC (+0.7) = +14.6 MOV + 3.5 HF = +18.1 projected (-4 actual); -22.1 deviation

Missouri (+18.9) @ Wyoming (-1.2) = +20.1 MOV - 3.5 HF = +16.6 projected (-6 actual); -22.6 deviation

Arkansas (+4.8) vs Portland St (not rated): no projection for SP+ but stat systems generally got this wrong e.g. Sagarin projected a 30 point Arky victory.

 

For the lower half SECx3 teams that played OOC this week we had 6 results that could be predicted using SP+.  2 of those were quite close to the predicted result: Ole Miss and Kentucky.  The other 4 were all HUGE misses.  They were all in the direction of overrating the SEC team by an average of 25.8 points.  In no case did SP+ underrate a lower tier SEC team.

Conclusion: Within the sample set SP+ appears to be significant biased towards the SEC.  The data set is not large but it is fully 25% of the relevant data we will get on this group during the season.  The bias affecting the bottom half of the conference has strong indirect effects on the upper tier through SOS and similar calculations.

  • Like 1
Link to comment
Share on other sites
Fritz Lanza

Fritz Lanza 2

Posted
Posted

I’ve been back and forth with BC a few times on twitter. His inability to accept that game situations have tactics, or that they could effect his formula is frustrating. Turn overs have significant outcomes on games, YPP, explosiveness, and even Points per scoring opportunity.

Also, his labeling of turn overs as luck is ridiculous. They might have elements of randomness, but one can certainly have players that have ball security flaws. Remember, in BC’s system a touchdown off a tipped pass is a sign of skill, but a cornerback stepping in front of out route is partially luck.

 

  • Like 1
  • Fuck You 1
Link to comment
Share on other sites
WBT

WBT 20565

Posted
Posted
4 hours ago, sushihorn said:

Hypothesis: SP+ has a significant pro-SEC bias.

Design: Bias should manifest itself most strongly in assessment of group members who are being judged based on group identity rather than individual accomplishment.  That should impact the bottom teams most as few bother to analyze them as individual programs.  We will measure the lower ranked teams in the SEC by result vs SP+ projection.

Kentucky (+7.3 SP+) vs Toledo (-1.5) = projected MOV 8.8 pts + 3.5 home field = +12.3 final projection (+14 actual); +1.7 deviation

MS St. (+20.3) vs UL-Lafayette (-6.9) = +27.2 MOV + 3.5 HF = +30.7 projection (+10 actual); -20.7 deviation

Ole Miss (+8.5) vs Memphis (+10.9) = -2.4 MOV - 3.5 HF = -5.9  projected (-5 actual); +0.9 deviation

Tennessee (+10.9) vs Georgia St. (-15.6) = +26.5 MOV + 3.5 HF = +30.0 projected (-8 actual); -38.0 deviation

South Carolina (+15.3) vs UNC (+0.7) = +14.6 MOV + 3.5 HF = +18.1 projected (-4 actual); -22.1 deviation

Missouri (+18.9) @ Wyoming (-1.2) = +20.1 MOV - 3.5 HF = +16.6 projected (-6 actual); -22.6 deviation

Arkansas (+4.8) vs Portland St (not rated): no projection for SP+ but stat systems generally got this wrong e.g. Sagarin projected a 30 point Arky victory.

 

For the lower half SECx3 teams that played OOC this week we had 6 results that could be predicted using SP+.  2 of those were quite close to the predicted result: Ole Miss and Kentucky.  The other 4 were all HUGE misses.  They were all in the direction of overrating the SEC team by an average of 25.8 points.  In no case did SP+ underrate a lower tier SEC team.

Conclusion: Within the sample set SP+ appears to be significant biased towards the SEC.  The data set is not large but it is fully 25% of the relevant data we will get on this group during the season.  The bias affecting the bottom half of the conference has strong indirect effects on the upper tier through SOS and similar calculations.

It doesn't change the conclusion but miss st and scar were neutral site correct?

Link to comment
Share on other sites
ClubWhatever

ClubWhatever 12140

Posted
Posted
On 9/1/2019 at 1:22 AM, Valmy77 said:

Missouri was way down entering the 4th Quarter and got a few TDs to make it closer than it really was. But I guess they were far more efficient (WTF is that exactly?) in falling behind 34-17 after 3 quarters than Wyoming was getting ahead 34-17.

I mean they gave up almost 300 rushing yards against Wyoming. You have to be pretty efficient to do something like that.

Watched this game.  Wyoming has a RS freshman QB who looks like he has a chance to be a draft darling in a few years.  Big, strong arm, good athlete.  Only a 3 star, recruited by mid tier programs out west.

Link to comment
Share on other sites
Fozzz

Fozzz 152

Posted
Posted (edited)
16 hours ago, Magus Ossis said:

60% ATS is good. I wonder what his rate would be including the FCS v FBS games (I know— can’t calc because the data aren’t  entered into his system from before). If he fixes much of what is wrong with his model, he might be able to start influencing lines. He may already anyway.

It was 54% ATS week 1 of last season.  I think Vegas caught up to him last year and likely will this season as well, but you may be able to make some money making a number of bets week 1 2020 using his predictions.  

Edited by Fozzz
  • Fuck You 1
Link to comment
Share on other sites
Hozz

Hozz 1017

Posted
Posted

Hopefully this is the year that the media figures out that the SEC is Bama and the 13 dwarves.  Per segment on SVP, 3 double digit SEC favorites lost this week, first time that has happened in over 20 years.  The East is laughably bad and Georgia is going to sleep walk into their title game.

Link to comment
Share on other sites
gmr548

gmr548 16580

Posted
Posted
16 minutes ago, Hozz said:

Hopefully this is the year that the media figures out that the SEC is Bama and the 13 dwarves.  Per segment on SVP, 3 double digit SEC favorites lost this week, first time that has happened in over 20 years.  The East is laughably bad and Georgia is going to sleep walk into their title game.

I think the SEC west has a chance to be very good this year - Bama, obviously, but LSU is obviously looking strong and Auburn did take down Oregon. Aggy and Mississippi State have some pieces but the jury is out on them. Piggy and klan aggy suck but the SEC West was probably the best division in college football last year, and is right up there in that discussion most years. Good lord the SEC East is trash though. That division is a G5 outside of Georgia.

Link to comment
Share on other sites
ztejas

ztejas 43425

Posted
Posted (edited)
8 minutes ago, gmr548 said:

I think the SEC west has a chance to be very good this year - Bama, obviously, but LSU is obviously looking strong and Auburn did take down Oregon. Aggy and Mississippi State have some pieces but the jury is out on them. Piggy and klan aggy suck but the SEC West was probably the best division in college football last year, and is right up there in that discussion most years. Good lord the SEC East is trash though. That division is a G5 outside of Georgia.

I'm not sure that Oregon is worth a shit but we will see.

LSU is good. Writing them off as a "dwarf" is silly. 

Oh and quit negging everyone in this thread Fozz you fucking cunt 

Edited by ztejas
  • Fuck You 1
Link to comment
Share on other sites
Hozz

Hozz 1017

Posted
Posted
1 hour ago, gmr548 said:

I think the SEC west has a chance to be very good this year - Bama, obviously, but LSU is obviously looking strong and Auburn did take down Oregon. Aggy and Mississippi State have some pieces but the jury is out on them. Piggy and klan aggy suck but the SEC West was probably the best division in college football last year, and is right up there in that discussion most years. Good lord the SEC East is trash though. That division is a G5 outside of Georgia.

I think the SEC West being the best division in FBS and the SEC as a whole being overrated can both be true.  I think divisions are an outdated construct at this point.  SEC East is easy mode path to their title game.  Flip UGA into the West for LSU and they likely aren't getting any SEC title game appearances, while LSU all of sudden is.  Missouri has two title appearances by virtue of being placed in the East and not the West.  Same with the weak division in the Big 10.  Northwestern doesn't sniff the title game if they are in the tough division.  This shit affects the NFL playoffs every year too and in the NBA the West has been the dominant conference (conferences sitting in for divisions here) for 20 years, resulting in a much easier path to the Finals for the East.

1 hour ago, ztejas said:

I'm not sure that Oregon is worth a shit but we will see.

LSU is good. Writing them off as a "dwarf" is silly. 

Oh and quit negging everyone in this thread Fozz you fucking cunt 

Hyperbole.  My point is, Bama's last 10 years have been an incredible run in the most competitive era of the sport and the rest of the SEC is riding their achievement in the form of elevated rankings, perception, TV ratings, NFL draft picks and live rights contract money.

Link to comment
Share on other sites
Tex Long

Tex Long 4394

Posted
Posted
7 minutes ago, Hozz said:

...  My point is, Bama's last 10 years have been an incredible run in the most competitive era of the sport and the rest of the SEC is riding their achievement in the form of elevated rankings, perception, TV ratings, NFL draft picks and live rights contract money.

Left out recruiting.

Also, part of the ratings boost is due to 4 OOC every year. Another part is scheduling the penultimate week patsies. 

Link to comment
Share on other sites
Hozz

Hozz 1017

Posted
Posted
1 hour ago, Tex Long said:

Left out recruiting.

Also, part of the ratings boost is due to 4 OOC every year. Another part is scheduling the penultimate week patsies. 

Absolutely agree.  The scheduling loopholes they exploit need to be closed by the CFP.  There needs to be scheduling requirements across the P5 in order to be eligible for a CFP selection.  8 conference games instead of 9 and multiple FCS teams on the schedule that are functionally simultaneously bye weeks and game reps for depth.

Link to comment
Share on other sites
Rickylovesweed

Rickylovesweed 14299

Posted
Posted
15 hours ago, Texas007 said:

I was just wondering why this person was spending so much time defending Connelly and negging everyone who bashed him... it all makes sense now.  

It's definitely odd. He spends most his time here defending Connelly and S&P+. I was just joking but I'm somewhat convinced it might actually be Connelly. 

 

  • Like 2
  • Fuck You 1
Link to comment
Share on other sites
Magus Ossis

Magus Ossis 2799

Posted
Posted
2 hours ago, Rickylovesweed said:

It's definitely odd. He spends most his time here defending Connelly and S&P+. I was just joking but I'm somewhat convinced it might actually be Connelly. 

 

Public figures that post here

Bill Connelly- Fozzz

Greg Davis- CTJ

Dan Quayle- Iconoclast

John Sharp- banned

Mike Leach- Danger Snacks?

Huff- ?

Matthew McConaughey- who else could act well enough to impersonate an orthopedic surgeon?

Hal Mumme- HPSlugga

Others?

 

 

  • Like 3
  • Haha 1
Link to comment
Share on other sites
sushihorn

sushihorn 3233

Posted
Posted
4 minutes ago, Magus Ossis said:

Public figures that post here

Bill Connelly- Fozzz

Greg Davis- CTJ

Dan Quayle- Iconoclast

John Sharp- banned

Mike Leach- Danger Snacks?

Huff- ?

Matthew McConaughey- who else could act well enough to impersonate an orthopedic surgeon?

Hal Mumme- HPSlugga

Others?

 

 

Machinator = Skynet

  • Like 3
Link to comment
Share on other sites
Fozzz

Fozzz 152

Posted
Posted
11 minutes ago, texifornia said:

Lol he dropped OK State of all teams?

Texas is now 23, between #22 Mizzou (fucking lol) and #24 Oregon (sure, they actually looked decent in that loss)

 

Without the five factor box scores it's not going to be pretty difficult to guess who is likely to move up and down each week.  It's not like you can quickly glean success rate from a regular box score.  

Link to comment
Share on other sites
texifornia

texifornia 56945

Posted
Posted
9 minutes ago, Fozzz said:

Without the five factor box scores it's not going to be pretty difficult to guess who is likely to move up and down each week.  It's not like you can quickly glean success rate from a regular box score.  

Hi, Bill! Your Mizzou Tigers may have shit the bed in Laramie, but at least they still look good in SP+

  • Like 4
  • Haha 2
  • Fuck You 1
Link to comment
Share on other sites
Burt Macklin

Burt Macklin 19023

Posted
Posted (edited)
4 minutes ago, texifornia said:

 

Love how he points out SEC falling to try to preemptively counteract the SEC bias arguments, when all that does is show his preseason numbers systematically overrated the SEC by a drastic margin. 

impressively, most of the SEC teams are still overrated even after dropping a lot this week. 

Edited by Burt Macklin
  • Like 2
  • Fuck You 1
Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

×
 Share
Go to topic listing


Football ... Basketball ... Baseball ... Other Sports ... Futbol ... 🤫995🤫 ... Gambling ... Movies & TV ... Music ... Hobbies ... Lulz ... Food & Travel ... Daily Texan ... Business and Markets ... Cloak Room ... Help ... For Sale ... Board Discussion ... Subscribe!... Donate!... Advertise... COOKIE MONSTER!

×
×
  • Create New...