WAC overachievers and underachievers via the confusing metrics of math
by Kevan Lee on Jul 20, 2010 12:36 PM PDT
Email
Print
Applying advanced math to WAC previews is a little like applying reading comprehension to a Bleacher Report article. I mean, why?
Fortunately, I don't let little questions like "why" stand in the way of my writing about Boise State football. Today, I wanted to take a look at Pythagorean wins and how they might impact the 2010 WAC standings. Get your thinking caps on, everyone.
Pythagorean wins defined by people smarter than I:
Pythagorean Theorem: Measurement of wins based on points scored and allowed. Used to predict future success. "Points For" squared divided by the sum of "Points For" squared and "Points Against" squared. A coefficient of 2.37 is used for football. Original formula by Bill James for baseball analysis.
In English words, Pythagorean wins basically compare points scored with points allowed, assuming that these numbers are the best predictor for actual team performance. If you consistently blow people out, Pythagorean wins will love you. If you consistently get blown out but win a handful of games by three points, Pythagorean wins will hate your guts. With that in mind, here's the WAC:
| Team | Points For | Points Against | Pythagorean Wins | Actual Wins | Difference | Overachiever rank |
|---|---|---|---|---|---|---|
| Idaho | 425 | 468 | 5.76 | 8 | 2.24 | 1 |
| NMSU | 149 | 411 | 1.08 | 3 | 1.92 | 2 |
| Boise State | 591 | 240 | 12.52 | 14 | 1.48 | 3 |
| Hawaii | 296 | 383 | 4.57 | 6 | 1.43 | 4 |
| SJSU | 165 | 414 | 1.22 | 2 | 0.78 | 5 |
| Fresno State | 440 | 369 | 7.84 | 8 | 0.16 | 6 |
| Nevada | 497 | 371 | 8.67 | 8 | -0.67 | 7 |
| USU | 349 | 408 | 4.90 | 4 | -0.90 | 8 |
| LaTech | 350 | 309 | 6.88 | 4 | -2.88 | 9 |
Let's draw some conclusions.
(Keep in mind that exceeding Pythagorean wins means you got lucky and probably won't do it again. Failing to reach your Pythagorean projection means that you lost a lot more often than you should have and will probably be better next year.)
Idaho is terrible, this is a fact
The Vandals won eight games in 2009 when they should have won six, tops. Their Season of a Lifetime was so tenuous that they actually got outscored by 43 points over the course of the year. Do you know what kind of teams do that? Overachieving ones. And do you know what kind of teams do that and are then picked to do it again the next year? Overrated ones.
New Mexico State is worse than expected, this is depressing
Somehow the Aggies found a way to win a full two more games than they should have. Football God must owe DeWayne Walker money or something.
Louisiana Tech should have been bowl eligible, this is why Derek Dooley is coaching Tennessee?
According to Pythagorus, the Bulldogs scored and prevented scoring well enough to win seven games last year. According to the standings in my Phil Steele, the Bulldogs won four games. Where did things go wrong? I don't know, but perhaps you should ask the man in cha-... oh, that's right, he's coaching Tennessee now.
Allow me to put Boise State's Pythagorean wins in perspective
First off, you are not supposed to win 14 games in a season. It just doesn't happen. So when you do win 14 games, it is understandable that the gap between Pythagorean wins (expected wins) and actual wins should be enormous.
It isn't for Boise State. The Broncos managed to outscore their opponents at a high enough rate to make 14 wins seem entirely plausible. Their win difference only ranked third in the conference and was a full three-quarters less than frontrunner Idaho. Pythagorus would spin around in his geometric grave if he saw this.
Your turn
Got anything to add? Glad there is a measurement to prove just how bad Idaho is? Share your thoughts in the comments if you'd like.
HT to SBN blog Roll Bama Roll for the chart idea
This content was not created by OBNUG and therefore may not meet our standards. On the contrary, it probably exceeds them.
29 comments
|
0 recs |
Do you like this story?
Comments
Idaho Should Have Finished 5.76-7.24
They will this fall. Including a 56.49-6.77 beat down by BSU in the KD.
And that's another Bronco... FIRST DOWN!!
by FirstDown on Jul 20, 2010 1:11 PM PDT reply actions 2 recs
Where's Loque when I need a "mathlete"?
I have no idea what any of that meant whatsover. I sort of stopped at math. Or maybe even metrics. Bear in mind that for me, college math was “Theory”. Yay.
OBNUG Resident Mom and Bronco Nation Podcast Field Reporter.
The Pythagorean Wins estimates expected wins ...
based on how many points your scored and how many points you gave up. Its formula is somewhat comparable to RMS and other “Root Mean” averages. If you don’t care about the formula, the important thing to note is that winning more than your Expected Wins then you either overachieved or got lucky, and winning less than your Expected Wins means you underachieved or were unlucky. Over time, most teams perform very closely to Expected Wins metrics, thus the assertion by Kevan that teams that overachieved will likely perform worse next year.
Ooooo see?
There’s that word again. Pythagorean. I’m instantly transported back to HS Algebra and all I can remember is passing notes to the cute football player who sat behind me and cheating off of the future West Point grad who sat next to me. Yay!
OBNUG Resident Mom and Bronco Nation Podcast Field Reporter.
It's just a statistical prediction of how the season SHOULD have gone for a team.
For instance, I would really like to see how Iowa fared. They gotta be over 3-4 games, right? For you, look at the Difference column. This is what Kevan is referring too. Idaho won 2 and a quarter games than they should have. So, instead of having 8 wins, they should have only had 5 or 6.
For a real world representation:
Say you eat at Taco Bell. Given your tenure and circumstances, there will be days that you leave Taco Bell full of energy, ready to head down the road to hit the Dairy Queen (because lets face it, those caramel empanadas suck). There will be other days where you eat the cinnamon twists your kids leftover added with the fire flakes from your Volcano Taco™ (which they forgot to upgrade to a supreme .. again), and you barely survive screaming at your kids on the car ride home and succumb yourself to downing 3 chalky wafers of generic Tums™ (berry flavor!) before hitting the sack.
This idea of Pythagorean Wins is to determine how many times you should be winning and how many times you’re feeling the pains of acid reflux. If your body tries to compete with the metabolic levels of your children and you attempt to go on in the night, on the outside, it looks as if you have won. But who are we kidding? We’re getting older, growing shorter, and losing more hair (not specifically you, Crissie). We are who we are. This formula just compares how we should have fared with how we did fare, ans that’s that.
In most cases, it shows you who a fraud is on both sides of the spectrum: did you over achieve or underachieve? In Boise State’s case, we will always overachieve because we’re just not going to lose… and math hates that.
But what I think this really says about Boise State, is that Coach Petersen doesn’t run up the score and plays conservatively to do what he needs to do: win! If you’re Iowa, though.. this can and should be really exposing (I just hope that I’m right about that, because I haven’t seen Iowa’s wins).
"Everyone counted us out. I don't know why they keep doing that." -- Kyle Wilson
What this says about Nevada is:
they run up the score on the dregs of the WAC, and get shutout by Notre Dame.
Iowa's PW score
I did the math. Here it is:
Iowa’s Pythagorean wins: 9.42
Iowa’s actual wins: 11
Difference: 1.58
I’d say it is a pretty significant difference, about the same as Boise State (but keep in mind that Boise State’s is over a 14-game schedule and Iowa’s is a 13-game). Maybe D_Summit could help with the diff %.
"That was very sneaky of you." - Josh
% Diff = 16.8%
About 5% more of an over-achiever than Boise State (11.8% overall).
Insert clever signature here.
Can someone do the maths just for WAC conference games?
We want to build a university our football team can be proud of. -- Dr. George Lynn Cross
Oh lord
NMSU and SJSU really are awful, aren’t they? In-conference, out-of-conference, doesn’t matter. Just miserable.
We want to build a university our football team can be proud of. -- Dr. George Lynn Cross
I'm not a statistician, but
I’d venture to guess that “% Difference” between Pythagorean Wins and Actual Wins is a more valuable metric for over-achievement. I took the liberty of running those numbers and found that Boise State over achieved less than originally thought (Duh, they were quite dominant) and NMSU was actually the biggest over-achiever.
Team % diff
NMSU 177.8%
SJSU 63.9%
Idaho 38.9%
Hawaii 31.3%
Boise State 11.8%
Fresno State 2.0%
Nevada -7.7%
USU -18.4%
Can you explain percent difference?
I’m smart, but not that smart.
"That was very sneaky of you." - Josh
It's pretty simple:
“% Difference” = “Difference”/“Pythagorean Wins”
So for Boise State, % Difference = 1.48/12.52 = 0.118 = 11.8%
The reason I think it matters is that I think a team that is projected to win 1 game wins 3 has over-achieved more than a team expected to win 11 games but wins 13. Looking at “% Difference” gives a ratio that seems to reveal the “achievement level” a little bit more clearly.
That's nothing
Did you know that if you reduce 100 by 25% you get 75, but if you increase 75 by 25%, you only get 93.75? I think maths got some explainin’ to do.
We want to build a university our football team can be proud of. -- Dr. George Lynn Cross
by marktgarten on Jul 21, 2010 12:08 PM PDT up reply actions
Whoa... my head hurts
I’m going back to work now…
- He once punched a magician... that's right, you heard me!
Dude
Where’s the video?? Or Mr. Fiskers or something??? If OBNUG is hoping to attract more chicks, this is leaps in the wrong direction. Illustrations, please? hahahhah
OBNUG Resident Mom and Bronco Nation Podcast Field Reporter.
by Crissie on Jul 20, 2010 3:41 PM PDT up reply actions 1 recs
"That was very sneaky of you." - Josh
by Kevan Lee on Jul 21, 2010 8:42 AM PDT up reply actions 4 recs
Couldn't it be boiled down to this?
You can’t “overachieve” if it’s an expected outcome. Boise State was expected to dominate…and did, however, them going undefeated wasn’t necessarily expected, thus a moderate level of overachievement.
The Vandals were expected to suck and kinda didn’t. They survived by the skin of their teeth, and gave up more points than they scored yet still managed a winning record. This bucking of the odds represents a much more significant overachievement on their part.
Got it now.
Thanks for pulling through for me Drew. xoxo
OBNUG Resident Mom and Bronco Nation Podcast Field Reporter.
Wow...
What you’ve just said is one of the most insanely idiotic things I have ever heard. At no point in your rambling, incoherent response were you even close to anything that could be considered a rational thought. Everyone in this room is now dumber for having listened to it. I award you no points, and may God have mercy on your soul.
Signed,
Billy Madison’s Principal
by 4EverBleedBlue on Jul 21, 2010 10:44 AM PDT reply actions 2 recs
UofI really got lucky in their non conference games.
Their difference in the conference wasn’t all that bad, .75 more wins then their were supposed to get is respectable. But if you subtract the difference from their whole schedule from their schedule in the WAC, they overachieve by a game and half. Overachieving against Northern Illinois, Colorado State, San Diego State, Bowling Green, and Washington isn’t that impressive at all. With a 1 point victory over BG (virtually a home game), 1 point win over La Tech at home, 2 point victory over Colorado State at home, and a 3 point victory over at No. Illinois. Seems like the vandals are destined to have a 5-7 season this year, mark it down!
Comments For This Post Are Closed
Secondary Sidebar
User Tools
FanPosts
Community blog posts and discussion.
Recommended FanPosts
Recent FanPosts
FanShots
Quick hits of video, photos, quotes, chats, links and lists that you find around the web.
Recommended FanShots
Recent FanShots
2012 Boise State Football Schedule
@ Michigan State  |
Fri 08/31 | 6:00 p.m. MT |
| Miami (OH) | Sat 09/15 | 2:00 p.m. MT |
| BYU |
Thu 09/20 | TBA |
| @ New Mexico | Sat 09/29 | TBA |
| @ Southern Miss | Sat 10/06 | TBA |
| Fresno State | Sat 10/13 | 1:30 p.m. MT |
| UNLV | Sat 10/20 | 1:30 p.m. MT |
| @ Wyoming |
Sat 10/27 | 1:30 p.m. MT |
| San Diego State | Sat 11/03 | 8:30 p.m. MT |
| @ Hawaii | Sat 11/10 | 5:00 p.m. MT |
| Colorado St. | Sat 11/17 | 1:30 p.m. MT |
| @ Nevada | Sat 12/1 | TBA |
