All posts by Elena Curtoni

Beside The Points For Thursday, Jan. 4, 2017

Things That Caught My Eye

Water is technically snake oil, I guess?

Tom Brady is really great at being a football player, but his provenance as a nutritionist or health expert is suspect. In the spirit of “just because you’re a bird doesn’t mean you’re an ornithologist,” Christie Ashwanden took a serious look at the TB12 method, and the reality is that drinking lots of water isn’t the same as putting on sunscreen, drinking too much water is actually pretty dangerous, and “pliability” could have less to do with Tom Brady’s continued efficacy behind center than, say, a continually successful offensive line that’s been adept at protecting him from injury. On the other hand, as I’m currently on day four of The Gronk Cleanse, I am pretty hammered. [FiveThirtyEight]

The Bills are IN! But…

The Buffalo Bills have snapped the longest active streak in the NFL and made the playoffs after 17 years of not doing that. The organization is in heaven right now. Still, making the playoffs can have a lot more to do with luck than skill, and this Bills group is pretty middle-of-the-road talent-wise when compared to those 17 other teams. While they finished 9-7, teams that got outscored by 57 points over the course of the year usually get fewer breaks in their direction and finish with six or seven wins. [FiveThirtyEight]

Jordan Greenway with a first

Of 1,690 men’s Division I college hockey players, 13 identified themselves as black. One is Jordan Greenway, a Boston University forward who became the first African-American hockey player to represent Team U.S.A. at the Winter Olympics. [The Undefeated]

UCF is committing to the bit

The University of Central Florida is claiming that they’re the national champions of college football. Despite not making the college football playoff, UCF argues that, since they went undefeated and beat Auburn in the Peach Bowl, they get to be champs. This stunt could be a lot of things, but it’s not a half measure by any stretch: Athletic Director Danny White said that the coaching staff is getting their national title bonuses and all. [SB Nation]

Try out our interactive, Which World Cup Team Should You Root For?

They get a parade!

Cleveland went 0-16 this season, joining the Detroit Lions as the only franchises to lose 16 regular season games. Thousands of Cleveland fans will reportedly gather this Saturday to watch a parade commemorating the season of “perfection.” []

Four out of five top broadcasts!

Off the top 100 most viewed telecasts in 2017, 81 of them ended up being sports events. Live sports continues to be one of the main attractions on the television. Put in a historical context, even the NFL is doing alright: While Fox telecasts of NFL games were down from 2016 and 2015, viewership is still up compared to 10, 15 and 20 years ago, according to Sports Business Daily. [Austin Karp]

Make sure to try your hand at our fun NFL game: Can you beat the FiveThirtyEight predictions?

Big Number

21 percent

Going into the NFL Wildcard weekend, New England is the overwhelming favorite to win the Super Bowl, with a 31 percent chance of doing so. Minnesota, Philly, and Pittsburgh have 19, 15 and 14 percent chances of doing so according to out Elo model, so we’d expect about a 79 percent chance one of those teams wins the championship. Now, you may recall none of those teams are playing football this weekend. Among the wild card teams, the ones with the best shots are New Orleans, Kansas City, and Atlanta, with a 5 percent, 4 percent and 4 percent chance of winning the Super Bowl. Still, there’s only a 21 percent chance one of the teams this weekend wins it all. [FiveThirtyEight]

Leaks from Slack:


Rare D-I women’s quadruple-double recorded

quadruple double!


10 steals 👀


@neil is it easier to get 10 steals or 10 blocks?


Way easier to get the blocks (at least in the NBA)
There are only 20 games of double digit steals in Basketball-Reference’s database
There are 143 double digit block games


Oh, and don’t forget
Bills fans were very generous after Andy Dalton won them a playoff berth

This Buffalo Bills Team Is Worse Than Most Bills Teams This Century

After the “Music City Miracle” knocked the 1999 Buffalo Bills out of the postseason, Bills fans suffered through the longest playoff drought in American professional sports. But with only a few hours left in 2017, the streak came to a glorious end. After beating Miami on Sunday, Buffalo desperately needed Cincinnati to eliminate Baltimore, and Bengals quarterback Andy Dalton complied — finding Tyler Boyd for a game-winning touchdown on 4th-and-12 in the final minute. The play sent the Bills locker room into hysterics, while fans back home braved bitterly cold weather past midnight on New Year’s Eve to greet the team plane at the airport.

New head coach Sean McDermott has led the franchise out of the regular-season desert, but there’s an irony to his triumph: This Bills squad is actually worse than many of those that failed to make the playoffs since that 1999 season. There are Bills fans old enough to buy beer who can’t remember what a Bills playoff team looks like — but even they ought to be able to tell that this squad is something less than one of the 12 best teams in the NFL.

“You are what your record says you are,” Hall of Fame coach Bill Parcells famously asserted. But ESPN’s Brian Burke estimates that randomness accounts for about 42 percent of NFL outcomes, and the NFL’s playoff structure builds in even more inequity. The 2008 New England Patriots won 11 games but missed the playoffs, while the 2014 Carolina Panthers got in with a 7-8-1 record.

This year’s Bills team finished 9-7, but based on Football Outsiders’ estimated wins and Pythagorean wins metrics,Per Football Outsiders, Pythagorean wins are based on pure point differential, while estimated wins use a “Forest Index” to emphasize consistency and performance in high-leverage situations.

‘>1 it only earned 6.8 and 6.3 wins, respectively. Put another way, an NFL team that gets outscored by 57 points on the year (as the Bills did) mathematically ought to win about six or seven games.

But it’s not just about wins. Pick a metric of team strength, and the Bills are typically at or below the median: 21st in scoring differential, 15th in FiveThirtyEight’s Elo ratings, 23rd in Simple Rating System,Pro Football Reference’s Simple Rating System is derived from scoring differential and strength of schedule.

‘>2 21st in Defense-adjusted Value Over Average.Football Outsiders’ DVOA metric breaks down every NFL play and determines its value when compared to a league baseline based on situation (i.e., down and distance, field position, score).


These Bills don’t stack up well against the 17 iterations that came before them, either:

The 21st century has not been good to Buffalo

The Bills’ past 18 seasons by Defense-adjusted Value Over Average, Pythagorean wins and estimated wins

Season DVOA PYTHagorean WINS ESTimated WINS Actual WINS
2004 31.3% 11.1 12.4 9
2014 10.5 9.6 9.0 9
2015 2.7 8.5 8.8 8
2006 -2.1 7.7 8.0 7
2007 -5.0 4.9 8.0 7
2016 1.0 8.5 7.4 7
2000 0.6 7.0 7.3 8
2002 -8.0 7.5 7.1 8
2013 -3.3 6.7 7.1 6
2011 -9.7 6.4 7.1 6
2003 -7.3 6.8 7.0 6
2017 -9.8 6.3 6.8 9
2008 -8.4 7.8 6.7 7
2009 -10.4 5.9 6.6 6
2012 -12.1 5.7 6.5 6
2005 -17.8 5.2 5.8 5
2010 -21.3 4.3 5.5 4
2001 -22.0 3.9 4.2 3

Source: Football Outsiders

The best Bills team since 1999 was the 9-7 2004 squad. At 31.3 percent better than average, it was the NFL’s No. 3 team by overall DVOA. Buffalo’s defense, led by Pro Bowlers Takeo Spikes and Nate Clements, ranked No. 1. The team’s 12.4 estimated wins and 11.1 Pythagorean wins show they were actually much stronger than what their record said.

The 2004 Bills lost their first four games, three by a combined margin of just 8 points and the other a 31-17 loss to the (eventual Super Bowl champion) New England Patriots. They then went 9-2 over their next 11 games, including a six-game win streak during which they outscored their opponents 228-89. Needing one last win to secure a playoff berth, they tripped up against a Pittsburgh Steelers team resting some of its starters. Football Outsiders deemed the 2004 Bills the best team to miss the playoffs since 1986 — as far back as its data goes.

Despite racking up the same number of real-world wins, 2017’s Bills are a whopping 41.1 percentage points worse in DVOA. They also produced 5.6 fewer estimated wins and 4.8 fewer Pythagorean wins. Remember that 42 percent randomness? That this year’s Bills got to the same 9-7 record as the 2004 team suggests these two squads approached the upper and lower bounds of “lucky” and “unlucky” NFL results.

The 2017 Bills aren’t the best Bills team since 1999, or anywhere close — they’re actually no better than seventh-worst in any of these metrics. Outside of running the ball with LeSean McCoy, these Bills don’t do anything well: They’re 29th in offensive yardage and 26th in yardage on defense. They keep their heads above water with a per-drive turnover rate that’s a little higher on defense (13.1 percent, 10th-best) than offense (8.9 percent, eighth-worst).

2012 is the most recent year the Bills weren’t better across the board than they are now — and that group went 6-10.

None of this can take away the joy Bills fans everywhere felt when the Ravens’ Week 17 collapse handed their team a wild-card berth, of course. But there’s a reason the Bills organization sent the Cincinnati Bengals thank-yous: This trip to the postseason was much more about being lucky than good.

Check out our latest NFL predictions.

Senator Mitt Romney?

The 2018 midterms got a bit more exciting on Tuesday. The longest serving Republican senator in American history, Orrin Hatch of Utah, announced that he would not run for re-election. His retirement is unlikely to provide Democrats with much of an opportunity to win another Senate seat, but the door is now wide open for Trump nemesis and 2012 GOP presidential nominee Mitt Romney, who has reportedly told people he would enter the race if Hatch retired.

So, let’s break down Utah’s newly open Senate race …

1. Utah is really, really red.

Pretty much by any measure you look at, Utah comes out as one of the most Republican-leaning states in the country. The GOP holds a massive 4-to-1 advantage over Democrats in voter registrations. According to Gallup, only Wyoming gives Republicans a larger lead in party affiliation. Republicans have won 15 consecutive Senate elections in the state dating back to 1974. Democrats haven’t even won a governor’s race in Utah since 1980. And even though Trump won the state by a smaller margin than every Republican presidential nominee except oneGeorge H.W. Bush won the state by 16 percentage points in 1992.

“>1 dating back to 1964, he still was able to win by 18 percentage points.

Though Hatch has been unpopular in the state, he would have been a favorite for re-election. A November survey gave him a 15-point lead over potential Democratic nominee Jenny Wilson. Democrats are desperate for another Senate seat to contest in 2018 — in addition to Nevada and Arizona — but Utah probably isn’t the best place for them to look.

2. But Trump is unusually unpopular for a Republican.

Trump managed to carry Utah handily in 2016, but that likely had more to do with his party affiliation than Trump himself. Trump won only 45 percent of the vote in a race that featured independent conservative Evan McMullin as well as Democrat Hillary Clinton. Trump won just 14 percent of the vote in Utah’s Republican caucus, which was far less than the 45 percent he won nationally during the nomination process. To boot, his net approval rating has been teetering around break even in the state. For a Republican president in Utah, that’s bad.

Will Utah’s dislike of Trump affect the 2018 Senate election? Well, maybe not. Utah is red enough that almost any Republican candidate will be a clear favorite there despite Trump’s poor numbers. That’s especially true if Romney — who is well-known and has a clear anti-Trump resume — wins the nomination. That said, if the GOP nominates a more Trumpian candidate, it’s possible to imagine him or her having some problems. After all, Trump won only a plurality (as opposed to a majority) of the Utah vote in 2016.

3. Anti-Trumpism has won in Utah already.

Last year, there was a trial run of sorts for anti-Trump Republicanism in Utah. Former Provo Mayor John Curtis ran in a special election to replace Rep. Jason Chaffetz in Utah’s 3rd Congressional District. During the primary, he admitted that he didn’t vote for Trump, and he won the primary against two candidates who did. That race was fairly close, however, with Curtis winning only 43 percent of the vote. Curtis then went on to win the general election by 32 percentage points, which was significantly greater than Trump’s 24-point margin in the district. Curtis’s win is particularly interesting given the large swing against Republicans in most 2017 special elections.

These numbers suggest that even if Romney decides not to run, a GOP candidate in his mold could win a primary and would be a favorite in the general.

4. Romney would be a heavy favorite

If Romney does enter the race, it’s hard to see him losing — the primary or the general. In a survey taken in November, Romney was rated among the most popular politician in the state. His net favorability among all voters was +47 percentage points — far higher than Trump’s. He also scored a +73 net favorability among Utah Republicans. (By contrast, Romney’s popularity among GOP voters nationally plummeted when he spoke out against Trump during the 2016 primaries.) The same poll found Romney with an astounding 51 percentage point lead against potential Democratic nominee Wilson.

In short: If Romney runs, he’d be as big a favorite as favorites get.

5. A candidate can take the argument directly to the voters.

One last logistical point: Republican candidates used to have to win or place second in a statewide convention to get on the primary ballot. That might have been a problem for a Romney-esque candidate. Longtime Sen. Bob Bennett was defeated in such a convention in 2010. But the state law was changed in 2014; now, candidates can get on the ballot via the convention or by gathering signatures.There is a lawsuit pending against the current system.

‘>2 Curtis, for example, was beaten by the more conservative Chris Herrod in his district’s convention in 2017, but he got on the ballot for the primary and won anyway.

While it is unclear who will be strongest in a convention setting in 2018, widespread appeal is now more important for a candidate to win the nomination. If the polls are correct, Romney has the popularity necessary to win a nomination through a primary.

College Football Playoff Preview: Georgia’s Balance vs. Oklahoma’s Offense

If one College Football Playoff showdown is a matchup of old, familiar playoff foes (Clemson vs. Alabama, Part 3) the other is stocked with relative neophytes: Oklahoma and Georgia. It’s the Bulldogs’ first-ever appearance in the playoff since it began in the 2014-15 season, and Oklahoma is making only its second playoff bid after losing to Clemson in the 2015 Orange Bowl. Here’s what to look for in the Rose Bowl on New Year’s Day between the Bulldogs and Sooners:

Oklahoma’s defense is terrible by playoff standards. Will it matter?

Among this year’s four playoff teams, three are extraordinarily balanced: Alabama, Georgia and Clemson each rank among the nation’s top 10 in efficiency on both offense and defense.All stats and rankings in this story are as of the end of conference championships, since none of the playoff teams have played a bowl game yet.


Oklahoma, on the other hand, is a study in imbalance.

Not only do the Sooners have the best offense of 2017, but the difference between their offensive efficiency and the second-ranked offenses (Alabama and Oklahoma State) is about the same as the difference between No. 2 and No. 10 Central Florida’s. Since the playoff started four seasons ago, the only offense remotely close to being as efficient as Oklahoma’s belonged to Oregon in 2014 — and the Ducks weren’t really that close to the Sooners.

At the same time, the Oklahoma defense is easily the worst of any playoff team. The Sooners allowed 25 points and nearly 385 yards of total offense per game this season. They rank 59th in the country in defensive efficiency. It’s safe to say that Oklahoma has the most one-dimensional profile of any team to ever make the College Football Playoff.

That doesn’t necessarily mean the Sooners are primed for a playoff letdown. Oklahoma has the best quarterback in the country (Heisman Trophy winner Baker Mayfield) and a host of other weapons that can make opposing defenses pay. Their stat sheet could give any defensive coordinator a heart attack. (This is, after all, a team that surpassed 600 yards of total offense in more than half of its games!)

But it’s also worth noting that in three years of playoff action, the more efficient defense won 67 percent of its games, while the superior offense won only 56 percent. Even if those numbers are skewed by Alabama’s success as a defensive juggernaut, the Tide have won with defense for a reason. Oklahoma will have to buck that trend if they want to prove that a great offense can win, too.


Can Georgia keep winning from ahead and complete its rise?

Roughly one year and two months ago, the Georgia Bulldogs were 4-4 under new coach Kirby Smart after losing to the Florida Gators. Their Elo rating was +9.1, only the 33rd-best in the country. And although Georgia perennially hauled in great recruits, there was little to suggest it would be sitting two wins away from a national championship on New Year’s Day in 2018.

From that point on, though, the Bulldogs embarked on one of the great overnight improvements in playoff history. Since that day late last October, they went 16-2 and added 22 points to their Elo rating. Georgia now ranks second behind Clemson in Elo (+31.3). Since the College Football Playoff started in 2014, only one playoff team — Washington in 2016 — added more to its Elo rating between its low point of the previous season and its rating heading into the playoff:

The most improved playoff teams

Among College Football Playoff teams, the biggest improvements in Elo rating between the team’s low point in the previous season and the playoff

Elo Rating
Season Team Before Playoff Prev. Season Low Diff.
2016 Washington 29.4 4.8 +24.6
2017 Georgia 31.3 9.1 +22.3
2015 Oklahoma 29.5 8.8 +20.7
2017 Oklahoma 30.0 9.7 +20.3
2014 Florida State 32.1 12.0 +20.1

The College Football Playoff began in 2014

The Bulldogs are an exceptionally well-rounded team, ranked fifth nationally in offensive efficiency and second in defensive efficiency during the regular season. All-SEC running back Nick Chubb surpassed 1,100 yards for the third time in his career and No. 2 rusher Sony Michel nearly broke the 1,000-yard mark as well. Freshman quarterback Jake Fromm stepped in to instantly become one of the most efficient passers in the country.

FiveThirtyEight: Oklahoma needs help on defense to win the Rose Bowl

Perhaps the only concern about the Bulldogs is if they find themselves in the unfamiliar territory of playing from behind. Fromm has been much better when the Bulldogs have a lead. During the regular season, only three Power Five conference QBs had a bigger split between their Total Quarterback Rating while the team was leading and while it was trailing:

Granted, Fromm’s split comes in a small sample because Georgia seldom trailed during its 12-1 campaign. (Georgia ran only 16 percent of its plays while behind on the scoreboard, nearly half of which came against Auburn in the Bulldogs’ 40-17 November defeat.) But perhaps Oklahoma’s best chance against the Bulldogs is to use its dominating offense to jump out to an early lead, then hope its defense can force Fromm into freshman mistakes while playing from behind. Between Georgia’s impressive balance and Oklahoma’s shaky defense, however, that might be a task easier said than done.

Beside The Points For Thursday, Dec. 28, 2017

Things That Caught My Eye

One game more

The 0-15 Cleveland Browns play the 12-3 Pittsburgh Steelers this Sunday, and a loss will make them the second 0-16 team in NFL history. Browns Coach Hue Jackson has promised to make good on a vow he made to swim in Lake Erie if the Browns went 1-15 yet again. Overall, Cleveland is 1-30 under Jackson. Elo suggests that the Steelers have a 95 percent chance of winning on Sunday, and their bye week position in the playoffs is already locked up. [ESPN, FiveThirtyEight]

Okay, so this one settles it.

Surprise: Clemson and Alabama will play in the College Football Playoff again, their third such meeting in three years. That pair of games was a dead heat: each took a national championship, and the aggregate scoring was 76-75, Alabama. Stakes are high, but these aren’t the same exact teams we’ve seen in the playoff before: Alabama’s once ironclad defense has been deemphasized, but it has a stronger offense to compensate. [FiveThirtyEight]

The case for Gurley as MVP comes down to how many voters played Fantasy Football

Todd Gurley, the Rams star running back, has been making a highly persuasive case for why he should be the NFL MVP over other contenders, like water pitchman Tom Brady. One component of the pro-Gurley argument is his incredible December performance, in which he has scored eight touchdowns in three games. This comes as no surprise to anyone playing Fantasy Football, where Gurley’s 107.1 fantasy points gave him the best postseason ever. [ESPN]

Georgia D vs. Oklahoma O

Oklahoma has the best offense in the college football playoff, and on New Year’s Day they’ll play Georgia’s remarkable defense in the Rose Bowl. The winner goes on to face one of the two most recent national champions. Oklahoma’s quarterback, Baker Mayfield, has a completion rate of 53 percent when under pressure compared to 71 percent overall. [ESPN]

Try out our fun new interactive, Which World Cup Team Should You Root For?

Manziel Desired By Football Team

The Canadian Football League approved Johnny Manziel — former Heisman Trophy winner, Cleveland Browns quarterback and cautionary tale — for a 2018 contract. His CFL rights are owned by the Hamilton Tiger-Cats, who will now be able to sign him or trade his rights. [ESPN]

Sisters on opposite sides of the Olympics

Marissa Brandt and Hannah Brandt, two sisters from Minnesota, will be playing ice hockey in the Olympics, but not on the same squad. Marissa will be playing for South Korea, the country of her birth, while Hannah made Team USA. [NBC Sports]

Make sure to try your hand at our fun NFL game: Can you beat the FiveThirtyEight predictions?

Big Number

61 percent chance

There are six teams competing for three remaining spots in the NFL playoffs, but none of them are playing each other so it’s going to be a very weird Week 17: a bunch of teams with lots on the line are playing a bunch of teams with not a lot on the line. In the AFC, the contenders are the Ravens (94 percent chance of making the playoffs), Titans (58 percent), Chargers (31 percent) and Bills (17 percent), all playing squads with nothing on the line. In the NFC, it’s the Falcons (70 percent) and Seahawks (30 percent) fighting for the last spot. At least Atlanta is playing Carolina, a playoff-bound team still fighting for seeding. Atlanta has a 61 percent chance of winning that game. [FiveThirtyEight]

Leaks from Slack:


This 0-13 Basketball Team Is A Favorite To Make The NCAA Tournament

loved this, and wondered if there’s a ranking for us of the small-conference teams that most lean-in to their underdog nature. would help us see if texas southern is really the harshest


What a cool idea.


Yeah this is kind of amazing, they’re easily #1 in average opponent Elo so far this year
(non-conf only)


Oh, and don’t forget
Some hockey team had a weiner dog race on ice.

What If College Football Hadn’t Wasted Decades On Polls And Just Used A Stinking Playoff?

The College Football Playoff has transformed the way teams and conferences build their schedules — and created plenty of controversy along the way — in the four seasons since it debuted. And even if the system could stand to make some improvements, it’s also been a relatively successful experiment in adding legitimacy to a championship that used to be determined through such opaque measures as media voting and computer ratings. For all the debate over “who’s in,” at least the eventual champion can say it won the title by beating two top-ranked opponents on the field.

The benefits of a four-team bracket got us thinking: What if the current playoff structure had been in place before 2014? Who would likely have won the championship each year? (Would it have been different from the consensus champs of old?) And which schools would have gained — and lost — the most titles under a playoff system?

Let’s answer those questions. (If you’re not interested in how we’re answering those questions, skip down to the first table.)

First, we’ll need a way to determine which teams would have made the playoff each year. Unfortunately, over the first four years of the actual playoff’s existence, neither the AP poll nor our Elo ratings (which are designed, in part, to predict the playoff selection committee’s tendencies) have completely nailed the playoff field with their four highest-ranked teams going into the bowls. But a combination of bothSpecifically, I added together a team’s rank in each list and re-sorted by that combined ranking, using Elo as the tie-breaker.

“>1 has been a perfect 16 for 16 in terms of predicting the real-life playoff teams.

So we’ll use that Elo/AP combo to pick the four playoff teams in each historical season.A more complicated version of this process might have involved using our full CFP prediction algorithm to produce probabilistic playoff odds for more than the top four teams, but we’ll save that can of worms for another day.

‘>2 (Our Elo ratings can be calculated going back to the 1988 season, so that’s when our hypothetical exercise will begin.) I also found that, once the playoff field is set, the pre-bowl AP rankings alone have done the best job of matching the committee’s seeding for the teams, so we’ll set the seeds that way in our mythical playoffs.

Next, we’ll need a way to play out the theoretical playoff games themselves. For that, we’ll use Elo, which provides a probabilistic forecast for any given game based on the two teams’ pregame ratings. In most cases, we’ll use each team’s pre-bowl Elo ratings to give us the chances of each team winning both its semifinal game and the championship game (conditional on making it that far). The only exception is when a slated matchup happened in a real-life bowl that season, in which case we’ll use the actual result for that semifinal or final matchup.

A great example of this came in 2003, when both of our hypothetical semifinal games — No. 1 USC vs. No. 4 Michigan and No. 2 LSU vs. No. 3 Oklahoma — actually played out in the Rose Bowl and Sugar Bowl, respectively. In that case, the Trojans and Tigers would automatically advance to the title game, where each would have almost exactly a 50-50 shot at winning the championship, according to Elo.Technically, LSU would be the slim favorite at 50.47 percent.

“>3 At least one of these real-world matchups happened every year from 1988 to 2013 — except in 1989, when conference bowl tie-ins kept each of the four teams in our playoff field from actually playing one another.

After following all of the rules laid out above, here’s how each season since 1988 would look if a playoff had been in place instead of the system that was used at the time:

What 26 extra years of playoffs would have looked like

Hypothetical College Football Playoff fields for the 1988-2013 seasons based on Elo ratings and AP poll rankings

Playoff Teams w/ Championship Odds
Year Team % Team % Team % Team %
1988 Notre Dame 44 Miami 33 Nebraska 12 W. Virginia 11
1989 Miami 31 Florida St. 26 Notre Dame 24 Michigan 20
1990 Colorado 37 Miami 26 Florida St. 25 Notre Dame 13
1991 Miami 34 Washington 32 Florida 25 Michigan 9
1992 Alabama 44 Florida St. 23 Miami 18 Notre Dame 15
1993 Florida St. 48 Notre Dame 23 W. Virginia 22 Nebraska 8
1994 Nebraska 38 Penn St. 27 Florida 25 Miami 10
1995 Nebraska 53 Florida 18 Tennessee 17 Northwestern 12
1996 Florida 50 Nebraska 17 Arizona St. 17 Florida St. 16
1997 Nebraska 50 Michigan 38 Florida St. 12 Tennessee 0
1998 Tennessee 62 Florida St. 14 Florida 13 Ohio St. 11
1999 Florida St. 40 Nebraska 30 Alabama 24 Va. Tech 6
2000 Oklahoma 51 Miami 23 Florida St. 16 Florida 10
2001 Miami 46 Oregon 39 Florida 16 Colorado 0
2002 Ohio St. 39 USC 22 Georgia 21 Miami 18
2003 LSU 50 USC 50 Oklahoma 0 Michigan 0
2004 USC 51 Auburn 26 California 13 Oklahoma 10
2005 Texas 61 USC 15 Ohio St. 14 Penn St. 10
2006 Florida 42 Ohio St. 20 Oklahoma 19 Michigan 19
2007 LSU 33 USC 29 Va. Tech 27 Ohio St. 11
2008 Florida 62 USC 15 Alabama 13 Oklahoma 11
2009 Alabama 57 Florida 24 Texas 10 Cincinnati 8
2010 Auburn 56 Arkansas 19 TCU 15 Oregon 11
2011 Alabama 55 LSU 24 Okla. St. 11 Oregon 10
2012 Alabama 50 Stanford 20 Florida 19 Notre Dame 11
2013 Florida St. 40 Alabama 22 Stanford 22 Auburn 16

Actual champions (or co-champs) are listed in bold. In 1990, Georgia Tech was co-champion but is not projected to have made the playoff that season.

Playoff selection is based on pre-bowl Elo ratings and AP polls. Playoff games are simulated using Elo, except in cases where a matchup actually took place during bowl season (in which case the actual result was used). Certain teams are listed with a 0 percent championship probability because they lost a real-life game against a fellow playoff team.


The good news for the old system(s) is that each year’s real-world national champ — or at least the co-champ — would be the favorite to win the playoff as well. (The only time a historical national champ didn’t make our theoretical playoff was in 1990, when Georgia TechMy alma mater, it should be noted.

‘>4 claimed the national title in the coaches’ poll but missed the top four in our rankings after entering the bowls seventh in Elo.) But the fact that the real-world champ tended to be the favorite in our hypothetical playoffs is no guarantee those seasons would have played out the same way: Even after including real bowl results when they happened, the championship favorite in any given year had only a 47 percent chance of winning the title on average.

The most uncertain year of our hypothetical playoffs might have been the aforementioned 1989 campaign; without any real-life bowls to help guide us, our system gives all four teams at least a 20 percent chance of winning the national championship. And among years that featured at least one actual bowl result to work with, the wacky 2007 season — in which playoff favorite LSU would have only a 33 percent of replicating its real-world championship — probably would have kept providing us thrills well into January. But with a playoff in place, many seasons would likely have had different endings than the ones we’ve set to memory over the years.

How different? Here are all the schools that would have made at least one playoff appearance under our hypothetical system,Plus Georgia Tech!

“>5 along with their projected and actual national championships won:

How a playoff would have changed college football history

Most FBS college football championships by school under a hypothetical four-team playoff system, 1988-2013

Hypothetical Playoff Results
Team Appearances Semifinal Wins Champs Actual Champs Net Diff.
Florida 11 5.27 3.04 3.0 +0.04
Alabama 7 3.64 2.64 4.0 -1.36
Florida St. 10 5.38 2.59 3.0 -0.41
Miami 9 4.86 2.38 2.5 -0.12
Nebraska 7 4.03 2.08 2.5 -0.42
USC 6 3.60 1.80 1.5 +0.30
Notre Dame 6 2.87 1.28 1.0 +0.28
LSU 3 2.24 1.08 1.5 -0.42
Auburn 3 1.70 0.97 1.0 -0.03
Ohio St. 5 2.25 0.96 1.0 -0.04
Oklahoma 5 2.08 0.92 1.0 -0.08
Michigan 5 2.03 0.85 0.5 +0.35
Tennessee 3 1.09 0.79 1.0 -0.21
Texas 2 1.27 0.72 1.0 -0.28
Oregon 3 1.80 0.60 0.0 +0.60
Stanford 2 0.96 0.42 0.0 +0.42
Colorado 2 0.50 0.37 0.5 -0.13
Penn St. 2 0.84 0.36 0.0 +0.36
Virginia Tech 2 0.90 0.33 0.0 +0.33
West Virginia 2 1.05 0.33 0.0 +0.33
Washington 1 0.49 0.32 0.5 -0.18
Georgia 1 0.50 0.21 0.0 +0.21
Arkansas 1 0.36 0.19 0.0 +0.19
Arizona St. 1 0.40 0.17 0.0 +0.17
TCU 1 0.42 0.15 0.0 +0.15
California 1 0.34 0.13 0.0 +0.13
Northwestern 1 0.30 0.12 0.0 +0.12
Oklahoma St. 1 0.40 0.11 0.0 +0.11
Cincinnati 1 0.43 0.08 0.0 +0.08
Georgia Tech 0 0.00 0.00 0.5 -0.50
Total 104 52 26 26 +0.00

Playoff selection is based on pre-bowl Elo ratings and AP polls. Playoff games are simulated using Elo, except in cases where a matchup actually took place during bowl season (in which case the actual result was used).

In 1990, Georgia Tech was co-champion but is not projected to have made the playoff that season.


Aside from Alabama, which won the most real-life championships (four) of the 1988-2013 era but would project to have about 1.4 fewer under a playoff system, every other school’s projected title tally is within about a half-championship of its actual count, playoff or not. The anti-Bama might be Oregon, who made only one BCS title game in the years we’re covering (losing to Cam Newton and Auburn in the culmination of the 2010 season) but would figure to make three playoff bids under our hypothetical system — and probably would have given Miami more of a fight than Nebraska did in 2001. All told, the Ducks would figure to have won 0.6 more championships with a playoff than under the actual system.

Over about 25 years, a handful of national titles is about the best you can do (see Bama’s four). So even a half-championship gain is a lot. And the more marginal differences further down the list matter, too. Imagine the effect on the fan bases at Oklahoma State, Cincinnati or Northwestern (!!!) if their teams had managed to get hot during the playoff and take home the championship. In general, you can see a pattern emerge in the table above: Under a four-team playoff, the long-term effect is to take titles away from many of the top programs and give extra chances to the next tier of teams. As counterintuitive as that sounds, given the way a program like Alabama has dominated the CFP since its inception, the addition of an extra semifinal game introduces more randomness to the system, which helps teams down the list.It may have also been easier for non-powerhouse teams to crack the top four in a given season during the previous era of college football, given that the teams making the playoff since 2014 have uniformly been stellar programs. Or maybe after four seasons, we still don’t have enough of a sample yet to know for sure.


I once wrote that the BCS wasn’t any worse at picking champs than the College Football Playoff would be, and in a certain sense, that’s not wrong. (Again, the real-life champs each season above would have also been the favorites to win the playoff.) But the more we’ve seen teams get a chance to prove their championship merit on the field against top competition, the more appealing it is. Now I only wish college football had the current system in place for the past quarter-century instead of the confusing mismash of arrangements that preceded it.

What Harry Got Wrong In 2017

It’s almost 2018. For some, that means looking forward to the new year — what are the big political storylines to watch in the coming year? But I can’t do that until we finish one last end-of-year tradition: remembering all the times I screwed up!

That’s right: It’s time for my annual mea culpa column. Happily, this year’s column isn’t nearly as painful as 2016’s, and certainly not 2015’s, the year I laughed off Donald Trump’s entrance into the presidential race. Still, there’s always stuff that I wish I had written better, written differently or not written at all.

First, I think I undersold the chances of Republicans passing a health care bill. Of course, they didn’t pass a health care bill. The various bills they considered were super unpopular and the GOP’s math in the Senate was always difficult, so we were skeptical that Republicans would pass anything. Probably a little too skeptical, however.

Take a look at a couple of these headlines: “The GOP Health Care Bill May Have Found A Better Way To Fail” and “Trump’s Health Care Bill Won Over The Freedom Caucus — But Risks Losing Everyone Else.” For one, it didn’t alienate everyone else — it passed the House. And later, it was just one John McCain thumb away from passing the Senate too. So, I don’t think the GOP health care bill was destined to fail. After all, only recently Republicans passed another unpopular bill.

Screw-up No. 2 came in my preview of Virginia’s gubernatorial primaries. That article included a “survey” from “CSP Polling.” It shouldn’t have. I had never heard of CSP Polling, which showed up in May. And, as I wrote later — in “Fake Polls Are A Real Problem” — I could never find any person willing to go on the record with his or her actual name to stand behind the work. That’s obviously a big problem. We tend to take an inclusive attitude toward pollsters here at FiveThirtyEight, but any legitimate pollster should have real, identifiable people behind it.

Including one stray suspect poll might seem minor, but it’s not. A lot of people come to FiveThirtyEight for guidance on polling. We let them down in this instance. In 2018, we need to be extra vigilant on this front. There are probably going to be a ton of competitive House races. And most of them are unlikely to be surveyed by “gold-standard” pollsters. In other words, it’ll be tempting to grasp at any poll that pops up in a lot of these districts. Unfortunately, many will probably be suspect or shoddy. I put together some advice on how not to fall for these polls, and I would do well to heed it.

OK, screw-up No. 3 is a little wonky but still important. In early January, I wrote “Registered Voters Who Stayed Home Probably Cost Clinton The Election.” The main point of this piece relied on a SurveyMonkey poll. That data was fine, and I don’t think anything in the article was wrong, really. But it was thin. I should have used a voter file too.

Some pollsters use random digit dialing to do surveys — call a random phone number and ask that person if they’re registered to vote. Other pollsters use voter files, lists of registered voters with a bunch of information about each one, to construct their samples. I think a mixed approach to surveying (e.g. combining random digit dialing and a registered voter list) is probably best. And that goes for a lot of analyses too. That article on how much registered voters who stayed home hurt Hillary Clinton should have used more than self-reported voting status, which can be unreliable at times. I wish I had supplemented it with voter file data.

One of the big questions for the 2018 midterm elections is what will turnout look like. The evidence so far suggests Democrats will do far better turning out voters than they did in the 2010 or 2014 elections. How much better? We don’t know. So let’s not make too many assumptions about who will turn out in 2018 based on previous midterm elections. Voter lists will help us understand just how different the turnout patterns are in 2018.

Screw-up No. 4 is an error of omission. In July, I wrote, “Red-State Senate Democrats Haven’t Drawn Strong Opponents — Yet.” That was true at the time. But it’s not true anymore, and it hasn’t been for some time. In fact, the GOP has a number of fairly good candidates running in red states with Democratic senators, including Indiana and Missouri. You wouldn’t know that, though, if you were just reading FiveThirtyEight. That article deserved a follow-up.

Often, it’s not the bad articles you write that get you in trouble. Instead, it’s the articles you don’t write.

So that’s that. I’m sure there were other things I messed up in 2017, but these are the errors that still gnaw at me. As is tradition, however, let’s close out this column on a more hopeful note. Here are a few things I think I got right:

  • I never lost faith in the laws of political gravity. After Trump’s election, it became fashionable to say that “nothing matters,” that the normal rules of politics don’t apply anymore. I never thought that was the case. And, as we’ve seen in all types of elections in 2017, it’s not.
  • I’m now 2-for-2 — or 3-for-3, depending on how you count — in pulling out the “normal polling error” card. The idea here is that many people treat polling as an exact science. It’s not. Instead, it’s perfectly normal for polls to miss the result by a lot. So, just before the 2016 election I pointed out that Trump was only a normal polling error behind Clinton. And just before the special Alabama Senate election, I wrote that Democrat Doug Jones was just a normal polling error behind Republican Roy Moore.
  • Finally, here’s an article I’m proud of: “Fake Polls Are A Real Problem.” After a bunch of media outlets cited a Delphi Analytica poll of the 2018 Michigan Senate race showing Kid Rock winning, I spent weeks sifting through the survey’s data and trying to track down the people behind Delphi Analytica. The resulting article showed just how careful journalists need to be in citing pollsters they’ve never heard of.

Thanks for reading. I’ll see you all in 2018!

Beside The Points For Thursday, Dec. 21, 2017

Things That Caught My Eye

Five years is kind of a hard cap

Marvin Lewis was head coach of the Cincinnati Bengals for 15 years and never managed to win a Super Bowl. Reality is, coach-quarterback combos that work tend to be the ones that click early. Every coach/starting quarterback combination who won a Super Bowl did so within their first five seasons of working together. [FiveThirtyEight]

Yeah I had a nasty case of the “Vegas Flu” once and learned I probably shouldn’t play roulette anymore

The Las Vegas Golden Knights are doing really well, and here’s one theory as to why their opponents appear off their game: They play hockey in Las Vegas, a fun place with booze and gambling and late nights on the town before hockey games. “Hockey flus” have historically spread in certain cities, be they “Roxy Flu” in Vancouver thanks to the Roxy club, or “Philadelphia Flu” because the team was a bit punchier than a typical squad. Vegas has played seven teams this season that as of Tuesday appear playoff bound, and all those good teams lost in Vegas. House always wins. [ESPN]

Fiesta Bowl pivots to Overwatch

There’s a groundswell of colleges and universities fielding competitive esports teams, and Thursday the Fiesta Bowl organization announced they’ll be partnering up with Blizzard Entertainment, the company behind Overwatch, to put on the Fiesta Bowl Overwatch Collegiate National Championships in February. It’s the first time a collegiate bowl organization and a games publisher have collaborated, and comes just as the Overwatch League, a competitive league with the backing of several NFL owners, is about to kick off. [ESPN]

40 bowls, each slightly better

This year there are 40 collegiate football bowls, the first season of bowl contraction since 1995, thanks to the Poinsettia Bowl folding. Yes, this year has what an analysis of Elo scores indicates is the worst bowl matchup in the past 30 years, Georgia State vs. Western Kentucky. But overall, quality is up based on the average matchup Elo score compare to last year, which had some rough bowls. [FiveThirtyEight]

Try out our fun new interactive, Which World Cup Team Should You Root For?

Slash & burn

Scoring is up in the NHL this season, but it’s not the fault of the goalies. A number of rule changes — including higher enforcement of slashing penalties — means an average of 3.3 power plays per game, which kickstarts scoring opportunities and plausibly scoring. [FiveThirtyEight]

Figure skating in Detroit

A program spun off from Figure Skating in Harlem offers ice skates, uniforms, coaching and instruction to girls who commit to practicing two hours a day four days a week and maintaining a B average or higher in the city of Detroit. For a sport that can cost into the tens of thousands for high-level instruction, the program — which seeks to recruit 300 girls by year end 2018 — is trying to be an on-ramp for those who aren’t born into affluence. [The Undefeated]

Make sure to try your hand at our fun NFL can you beat the FiveThirtyEight predictions? game!

Big Number

0.69 first downs per target with Gronk

Rob Gronkowski gives the Patriots offense a massive kickstart. When Gronkowski is off the field, the team this season notched 72 receptions, 771 yards, 100 targets and 0.32 first downs per target. When Gronk was on the field — even if he’s not the one getting the ball, just if he’s there and needing to be covered — the team notched 277 receptions, 3,429 yards, 400 targets and 0.69 first downs per target. [FiveThirtyEight]

Leaks from Slack:


College basketball rankings: OU’s Trae Young has stats better than Steph Curry’s

cbs going hard


:fire: :fire: :fire: :fire:
In fairness, these numbers are pretty insane


Oh, and don’t forget
Wes, you ham

This Season’s Bowl Games Are Less Terrible Than Usual

As a college football fan, I love bowl season. Yes, the sheer number of bowls has exploded out of control in recent years, and yes, the economics of bowls is often problematic, and yes, some of the names have become downright ridiculous. But I’m still a bowl-game enthusiast. What can I say? It takes me back to the holidays of my childhood, when there was simple joy in consuming as much football as possible. It’s not Christmas without the CarQuest Bowl.

Which at various times has also been known as the Blockbuster Bowl, the Micron PC Bowl, the Tangerine Bowl, the Champs Sports Bowl and the Russell Athletic Bowl. Also: It’s now called the Camping World Bowl.


As SB Nation’s Bill Connelly recently pointed out, each of these games holds similar meaning for a large group of people, easy as that is to forget sometimes. (It doesn’t hurt that even low-level bowls draw decent TV audiences, too.) And after years spent climbing toward Peak Bowl Season, this season’s crop of games might even be toning down the “participation trophy” element of going bowling — if ever so slightly. Mediocre bowls are unavoidable in today’s game, but there are also more good ones this year than usual.

Last winter gave us 41 bowl games, which tied with the previous bowl seasonyear before for the most ever. It also provided some of the worst bowl matchups in history, the byproduct of including 17 different .500 teams and three more that finished with more losses than wins. (Notably, both North Texas and Mississippi State went into their bowls with seven losses apiece!) According to FiveThirtyEight’s Elo ratings, the 2016-17 bowl season had the lowest average matchup qualityWhich I calculate by taking the harmonic mean of pregame Elo ratings for the two teams in a matchup. (I used the harmonic mean to ensure that both teams in a game were of a certain quality — in this case, a low one.)

‘>2 of any season since at least 1988-89, the earliest season we track Elo.

Last year’s bowl slate was the extension of a two-decade trend toward postseason games of ever-decreasing quality. But it may have also been a culmination of sorts. After zero seasons of bowl contraction from 1995 to 2016, this year’s schedule has been sliced by a game, to 40 (the Poinsettia Bowl folded in January).The Miami Beach Bowl also ended its run, but it was relaunched as the Frisco Bowl.

‘>3 And although the current slate included Saturday’s Cure Bowl between Georgia State and Western Kentucky — the single worst bowl matchup of the last 30 years, according to Elo — the average quality of all bowl games has (slightly) increased for the first time in three years. After a period of runaway expansion, there might be signs the new-millenium bowl frenzy is finally leveling off.

Because of the College Football Playoff, this year’s semifinals — the Sugar and Rose bowls — are, unsurprisingly, far better than the long-term norms for those games. But the next several tiers of bowls also contain a large number of better-than-usual matchups, including Ohio State and USC in the Cotton Bowl and the Heart of Dallas Bowl with West Virginia and Utah. Of this year’s 25 best bowls (according to Elo), 18 are higher-quality games than the historical average for those bowls:

The 2017-18 season’s top bowls are better than usual

Elo game quality for the 25 best bowls of 2017, compared with each bowl’s overall 1988-2017 average

Game Quality
Date Bowl Favorite Underdog 2017 Game Avg. Since 1988 Diff
1/1 Sugar Clemson Alabama 2282 2080 +202
1/1 Rose Georgia Oklahoma 2267 2066 201
12/29 Cotton Ohio St. USC 2161 1960 201
1/1 Peach Auburn UCF 2041 1862 179
12/30 Orange Wisconsin Miami 2035 2049 -14
12/30 Fiesta Washington Penn St. 2018 2042 -24
1/1 Citrus Louisiana St. Notre Dame 1989 1947 42
12/28 Alamo Stanford TCU 1951 1812 139
12/30 TaxSlayer Mississippi St. Louisville 1864 1814 50
12/28 Camping World Oklahoma St. Virginia Tech 1855 1772 83
12/28 Holiday Washington St. Michigan St. 1825 1847 -22
1/1 Outback S. Carolina Michigan 1821 1851 -30
12/29 Sun NC State Arizona St. 1788 1759 29
12/29 Music City Northwestern Kentucky 1772 1704 68
12/27 Pinstripe Boston College Iowa 1762 1629 133
12/27 Texas Missouri Texas 1750 1676 74
12/16 Las Vegas Boise St. Oregon 1746 1649 97
12/29 Belk Texas A&M Wake Forest 1742 1747 -5
12/26 Cactus Kansas St. UCLA 1734 1680 54
12/30 Liberty Memphis Iowa St. 1702 1695 7
12/26 Heart of Dallas West Virginia Utah 1687 1565 122
12/27 Foster Farms Purdue Arizona 1661 1689 -28
12/23 Birmingham South Florida Texas Tech 1635 1665 -30
12/23 Armed Forces San Diego St. Army 1581 1520 61
12/24 Hawaii Fresno St. Houston 1580 1523 57

Game quality is determined by taking the harmonic mean of Elo ratings for the two teams in a game. Historical averages for bowls include previous iterations of each bowl under different names and sponsors.

Of course, the dregs of this year’s bowls are truly awful, even by the standards of games named after dollar stores, potatoes and a regional pirate festival. Aside from the guilty pleasure of binge-watching football, those exhibitions might end up making the best case yet for contracting the bottom third of all current bowls outright. But the worst of the worst mostly clear out by Dec. 22, after which we should be treated to an eminently watchable set of matchups.

In a season flush with good college football teams, the same mechanism that produced drama down the stretch of the playoff selection process will also provide some measure of redemption for the bowls. As long as there are 40 of the things, we’ll never be able to go back to a bowl season exclusively reserved for top programs — but for this year at least, we’re trending back in that direction some, once we make it past the first week or so.