Writers of Pro Football Prospectus 2008

Most Recent FO Features

WentzCar16-1.jpg

» Week 11 DVOA Ratings

DVOA has finally climbed on board the Wentz Wagon! The Eagles move into the No. 1 spot, but they aren't the only strong, well-balanced team in the NFL this year. New Orleans, Pittsburgh, and the Los Angeles Rams make this one of the best seasons ever for multiple teams over 30% in DVOA, and Minnesota isn't far behind.

22 Apr 2016

The East Coast Scheduling Bias

Guest column by Brian Stonelake

Are you an NFL fan? Of course you are. OK, think of your favorite team. Other than your quarterback, name the player who most influences your team's success. What if I told you that the NFL created a rule mandating four arbitrary road-game suspensions for this player? Houston fans: no J.J. Watt. Patriots fans: prepare to be Gronkless. Worse yet, this rule only applies to a handful of teams.

Recently the league released its 2016 schedule and (of course) they implemented no such rule. Unfortunately, they did quantifiably worse.

To understand the NFL's blatant error, allow me to provide a crash course in the pertinent area of chronobiology. To oversimplify: many aspects of exercise performance suffer in the morning. I don't mean the morning after a night with a teething toddler, or following a Super Bowl party with too much whiskey, I mean when fully awake and well rested. Numerous biological functions relevant to sport follow the core body temperature cycle, which is at its lowest early in the morning, and highest in the early evening.

This circadian rhythm does not adjust quickly, so while Week 14's 1 p.m. (EST) kickoff will feel like 1 p.m. for the host Panthers, it will feel much earlier for the visiting Chargers, whose body clocks are still synchronized with West Coast time. During such games, dubbed "morning body clock games," the West Coast visitors' circadian rhythms present a chronobiological disadvantage. Do not confuse this with being "tired;" this is a physiologically determined handicap directly affecting many aspects of exercise performance. For example, every 0.1℉ decrease in core body temperature results in a nerve conduction velocity decrease of 0.432 m/sec, directly reducing reaction time. These players aren't sluggish and tired; they're biologically mandated to react more slowly.

One might assume that at worst the physiological handicap provides a marginal advantage for the East Coast home team, an impact that would have little practical implication. But the impact is actually quite significant.

From 2001 to 2015, there were 374 NFL games where the road team faced a "morning body clock" start time. All 374 involved a Central or East Coast team hosting a team from the West or Mountain time zones (hereafter aggregated as "West" due to the scarcity of Mountain teams). The home team won 64.4 percent of these games; far in excess of the 57.2 percent win rate for home teams in general.

This statistically significant difference in win percentage (P = 0.002) did not happen by accident. You can cut the data however you want, controlling for any factor that you wish, and still come to the same indisputable conclusion: teams playing a morning body clock start time face a significant competitive disadvantage.

This shocking result makes more sense when you consider the numerous aspects of performance that oscillate daily with core body temperature. An incomplete list includes respiratory efficiency, blood flow, joint flexibility, reaction time, and muscle strength. Further, at suboptimal body clock times, athletes record lower mean and peak power outputs, lesser peaks of lactate production, decreased high intensity work rates, and lessened stamina. Conversely in the late afternoon/early evening hours, not only are athletes' self-chosen work rates higher, but these efforts are not accompanied by an increase in perceived exertion.

Given the above, it's not surprising that nearly every Olympic world record was set in the late afternoon or early evening. While scheduling considerations may explain some of this, tests isolating body clock time reveal up to a 10 percent decrease in aspects of athletic performance in the morning. For comparison, a 10 percent decrement in performance is also seen when consuming the legal limit of alcohol, or after three consecutive nights of only three hours of sleep. Are the Niners playing like they're horribly sleep-deprived? Do the Chargers look drunk out there? Essentially, they are.

With a better understanding of the chronobiological detriments faced by morning body clock teams, we can look for the expected results in specific areas of play. Playing a morning body clock start increases the frequency of penalties accepted by the opponent (henceforth "penalties"), turnovers per game, and dropped passes. Importantly, from the league's perspective, this serves not only as justification for the competitive imbalance, but also as evidence that these games do not represent the best product that the NFL can put forward.

Both penalty and turnover prevalence are greater in general for road teams than for home teams. Controlling for this fact by comparing morning body clock teams' penalties and turnovers to those of other road teams, we see the physiologically expected increases. From 2003 to 2014 (the available data) visiting teams averaged 6.49 penalties and 1.71 turnovers per game. Teams playing a morning body clock game amassed 7.16 penalties and turned the ball over on average 1.83 times per game; increases of 10.3 and 7.1 percent respectively. These increases are both statistically significant. Due in part to a smaller (relative) variance in penalties, the increase in penalties is statistically significant at a higher level (P < 0.0001) than for turnovers (P = 0.0435).

Further supporting the narrative of sloppy play is the physiologically expected increase in pass drop rate for morning body clock teams. To avoid the possible bias of trailing teams throwing more frequently, pass drop rates are examined in lieu of pass drop totals. (This data was only available from 2006 to 2014.)

We again see the expected increase in pass drop rate among teams playing morning body clock games (5.41 percent compared to the 4.97 percent rate of all other visiting teams), but this time not conclusively (P = 0.0719). The relatively small sample size and factors such as prevent defenses employed late in games by teams with large leads may have masked this correlation.

Looking at every game for the past 15 years, we see that on average scheduling a morning body clock game costs the visitor slightly more than 3 points. Not surprisingly, these results are anticipated by bookmakers who show similar shifts in point spreads for such games. When the league decided to schedule the kickoff for the Rams' Week 13 trip to New England at 1:05 EST instead of, say, 4:05 EST, they handed the Patriots 3 points in addition to the normal home-field advantage. The magnitude of this advantage is hard to overstate.

In August 2013, R.J. Bell wrote an informative piece for Grantland where he found a Las Vegas consensus for the value of NFL players, as measured by a hypothetical point spread change if they were hurt. The highest non-quarterback was Adrian Peterson, who was coming off a 2,097-yard MVP season where he averaged an astounding 6.0 yards per carry. He was estimated to be worth 2.5 points. In other words, Las Vegas measures the competitive disadvantage faced by morning body clock teams to be slightly greater than that associated with losing one of the most dominant offensive players in NFL history. Welcome back to L.A., Rams. We regret to inform you that you will now face an unavoidable handicap far more important than a hypothetical injury to Todd Gurley in half of your 2016 road games.

The next highest valued non-QB in Bell's piece was Calvin Johnson, who led the league in receptions and yards the year prior, and was valued at 1.5 points. The only defensive player worth more than half a point was J.J. Watt, estimated to be worth 1 point. Quarterbacks were worth as much as 7 points (Peyton Manning), with Russell Wilson, Andrew Luck, Tony Romo, and Sam Bradford all being worth 3 points. In other words, when the Seahawks played at Indianapolis in Week 5 that year (a game in which Seattle suffered its first loss), they should have been indifferent between playing the game at 1 p.m. EST as scheduled, and agreeing to play Tarvaris Jackson instead of Russell Wilson behind center in exchange for moving the game to 4 p.m. EST. With all due respect to the former Alabama State quarterback, that is astounding.

Conveniently, last year's playoffs provide a wonderful application of Bell's piece. Two days prior to Seattle's wild-card playoff game in Minnesota, the Seahawks announced that star tailback Marshawn Lynch, who was deemed by Coach Carroll to be "ready to rock" on Monday and a "full go" at practice Wednesday, would not play Sunday. This news sent shockwaves around the football universe and was a top story on many sports news outlets.

The unexpected nature of the Lynch announcement made it easy to quantify, as the betting line quickly adjusted 1.5 points. Losing their pro bowl tailback, in frigid conditions where a running game is paramount, with only a practice squad player as a backup (Thomas Rawls had already been ruled out with a broken ankle), was roughly half as important as the largely overlooked morning body clock start time.

But wait, you counter, didn't Seattle win that game? Yes, they did. The chronobiological disadvantage, like any disadvantage, is not impossible to overcome. Seattle was a far superior team, having the best DVOA in football (Minnesota was 11th), and at 6:1 the best Super Bowl odds of all teams playing wild-card weekend (Minnesota was 35:1). That, and a little bit of luck (like Blair Walsh missing a 27-yard field goal to take the lead with less than a minute to go) can overcome most any disadvantage. But that does not change the fact that Seattle faced a completely unfair disadvantage in the first place.

Well OK, you admit, but maybe Seattle should have gotten a higher seed to avoid playoff road games all together. Yes, a higher seed would have prevented the unfair situation from occurring in this instance, but why should certain teams have to overcome a systematic disadvantage in the regular season just to avoid possibly facing that same disadvantage in the playoffs? Why should Seattle face a disadvantage that East Coast teams with a No. 5 seed would not face?

The very next week Seattle was scheduled to play another morning body clock game, this time in Carolina. In the first half of this game, when the Seahawks were at their largest chronobiological disadvantage, they were outscored 31-0. They eventually lost 31-24.

The staggering effect of the morning body clock disadvantage is nothing new to Seattle. In the Pete Carroll era, they have been outscored 75-0 in the first half of morning body clock playoff games. Dating back to 2003, they have faced six such games in the postseason, their lone win coming in the aforementioned case of Blair Walsh's missed chip-shot field goal.

Fine, you concede, you win; It's a disadvantage. But are these teams any worse off than East Coast teams that have to play West Coast prime-time games ("night body clock")? In a word, yes. Look back to the core body temperature cycle chart. The difference in body temperature between 10 a.m. and 1 p.m. is far greater than the difference between, say 8 p.m. and 11 p.m. Also, body clocks adjust more quickly to westward travel (92 min/day) than they do to eastward travel (57 min/day). And while there is an increase in home team win percentage (60.7) in night body clock games compared to normal home field advantage (57.5), it is not of the same magnitude as morning body clock games (64.4). And even if this disadvantage was the same (it's not), it is far less common; there are roughly five times as many morning body clock games as night. Put differently, a West Coast team faces the larger morning clock handicap roughly 20 times for every one time the East Coast team faces the lesser night body clock handicap.

Inexplicably, despite being made aware of this competitive disadvantage, the NFL still schedules morning body clock games with high frequency. In 2008, East Coast teams claiming a competitive disadvantage associated with travel distance successfully lobbied the league to make schedule changes, yet years of discussion on the morning body clock issue from West Coast teams have seemingly fallen upon deaf ears. I am shocked that West Coast fans, especially those as obnoxiously loud as Seattle's, allow their teams to be so severely handicapped. Some foolishly hide behind anecdotal evidence pointing to games where morning body clock teams win. Yeah, the claim is not that playing in the morning is an insurmountable detriment, but rather that it is a significant detriment. I've seen teams block punts with only ten men on the field. Would that make a new rule limiting West Coast teams to ten payers in punt return situations any less fair?

While total eradication of morning body clock games would be difficult given the complexity of the NFL schedule, it seems as though such occurrences should be greatly reduced (and playoff occurrences eliminated). The most logical fix is to shift the majority of these starts from 1 p.m. EST to 4 p.m. EST. While this creates unwanted competition with the NFL's national game, it seems to be a small price to pay for the integrity of the sport.

It is worth noting that for the non-prime time games of the past 15 years, there is a 68-32 percent split between the first and second TV time slots (games starting at approximately 1 p.m. and 4 p.m. EST respectively). Had all 374 morning body clock games been moved to the second TV time slot, in addition to righting one of the worst wrongs in professional sports, it would have amounted to a more equitable 56 to 44 percent split of games: seemingly advantageous to the fan who prefers to watch as much football as possible. The rise in popularity of fantasy sports, for example, would seem to suggest a preference toward a more equal split of games among these time periods.

This topic was brought to the attention of Commissioner Roger Goodell prior to the 2009 season, at which point he was quoted as saying that he "had not seen specific information suggesting early starts [for West Coast teams traveling east] could create a competitive disadvantage for visiting teams."

To Commissioner Goodell's credit, he revisited the subject in October 2012 with a much more sympathetic view of the situation, saying that "Several of our teams on the West Coast have raised that [issue of morning body clock games] and we have been studying it. We have tried to put as many of those games on the East Coast at 4 p.m. You can imagine the thousands of different issues you have to put into the schedule. But the 10 o'clock starts are pretty tough."

Despite this statement, little has actually been done in terms of scheduling. Since 2001, there have been anywhere from 16 to 31 morning body clock games, with an average of 25 per year. This accounts for more than half of all road games for the affected West Coast teams. This year, after months of consideration and millions of dollars dedicated to creating the schedule, the NFL announced 27 more.

Hardest hit are the Raiders, who face five such games, all in the first half of the season. Should San Diego fans find their team in contention in November, they would be wise to temper their postseason expectations as their last three road games are all morning body clock starts. Seahawks fans still reeling from their unfair playoff matchups can take some solace in knowing that they only face two morning clock games this year; however an educated fan would quickly point out that this is two too many.

A combination of public misunderstanding and the league's denial has allowed this travesty to go on for far too long. The NFL has gone to great lengths to maintain a level playing field, yet allows this audacious injustice to continue. It is the duty of every fan of the NFL, whether your team is adversely affected by this competitive disadvantage or not, to educate yourself and your fellow fans on this matter. This great game deserves better.

Note: A more detailed examination of this matter can be found in "Managing the Body Clock," published in the February 2016 issue of the American Society of Exercise Physiologists' Journal of Exercise Physiology.

Brian Stonelake teaches math and statistics at Southern Oregon University; you can contact him at stonelakb-at-sou.edu.

Posted by: Guest on 22 Apr 2016

69 comments, Last at 12 Oct 2017, 6:29am by GreggCrype

Comments

1
by TGT :: Fri, 04/22/2016 - 2:48pm

1) The first chart: Where is it from? No citation is given. Are there really only 7 data points? Who was studied and when? Why does the best fit line (or what I assume is the best fit line) not go overnight? It's a cycle, so it should be continued.

2) Reaction time calculations: You state that "every 0.1℉ decrease in core body temperature results in a nerve conduction velocity decrease of 0.432 m/sec, directly reducing reaction time." This is a linear relationship? How was this determined? What are the usual rates? How do we know if .432 m/sec is significant?

3) Where are the stats for east coast early games vs east coast late games? By the underlying information we have, east coast teams should have just as much of an advantage in late games as they do in early games. Eyeballing the chart (because there aren't markings for 10:00, 13:00, and 16:00, as would be usefull) suggests the body temp increase is about the same from 1000 to 1300 as from 1300 to 1600.

4) Why did you create such horrible graphs? The Penalties per game / Turnovers per game graph is particularly horrendous. Two y axes with completely different scales? Not starting at 0 to exaggerate differences? Statistics professor, I give you an F.

5) I couldn't actually get through this whole post. It's that horrible.

2
by Kaelik :: Fri, 04/22/2016 - 2:59pm

All the graphs and their associated citations can be found in the link at the bottom, I'm not sure if that will answer all your questions, but that would be the place to start.

7
by brianstonelake :: Fri, 04/22/2016 - 3:37pm

Here are some (possibly horrible) answers:

1) As Kaelik pointed out, that table is referenced in the JEP paper referenced at the end of the article (the transitive property of reference?). I've seen that table referenced in several articles. Rather than recreate the wheel, I reached out to Gregg Atkinson who graciously allowed me to use his graph from his article in Sports medicine.

2) Yes, the research that I reference shows a linear relationship within reasonable bounds. "Significant" is a hard question to answer. I cite it more as a possible explanation for the drastic difference in win percentage.

3) I'm not sure which breakdown you're looking for here, but am happy to provide it. Could you clarify a little, please?

4) I appreciate your implication that the non penalties/turnovers graph are less horrendous by comparison - nice of you to shower me with praise! As to that specific graph, when it was broken up into separate graphs they all started to look the same; I thought this would be more interesting to the reader.

To combine the information into one graph I had to either cite relative differences or have different axes. I thought the latter was less confusing. As to the scaled axes, I used the excel default. Having intercepts at 0 made the information hard to read. It was not my intention to trick anyone. As for your grade of F, it's not the first I've gotten. If you want something more statistical, perhaps you'll read the linked JEP paper. I think you'll agree that it is at least D- quality.

5) Thanks for making it through as much of the horribleness as you did!

8
by Guest789 :: Fri, 04/22/2016 - 3:44pm

Good response to a needlessly antagonistic comment. +1

14
by Cruseydr :: Fri, 04/22/2016 - 5:51pm

I was just about to comment on the graphs with a non-zero axis, as the penalty graph in particular greatly exaggerates the difference. At a glance, it makes it appear that there is a 1000% increase in penalties from home to early body clock away games, when in reality it's around a 20% increase.

Of course this doesn't mean that I disbelieve your findings at all, just that it's always better to have a clear graph.

19
by Dan :: Fri, 04/22/2016 - 7:04pm

I think that a nonzero axis is often fine, and it's fine here. If someone makes a graph of how the world record 100m time has dropped from 10.06 seconds to 9.58 seconds over the past 50 years, it's not necessary to have the axis go all the way to 0.00 seconds. The 0.48 second change in the record is small compared to the entire scale, but it's large in the context of how fast sprinters actually run. And a graph where all the bars from Bob Hayes to Usain Bolt all look roughly the same size is not doing a good job of conveying the relevant information.

It's similar with the penalty graph, which I think is better in its current format than it would be if the axis went all the way to 0. It's probably possible to make it even better (maybe by including whiskers to show the standard deviation in number of penalties in a game), but the current version gave me more information at a glance than I would've gotten from the graph that went to zero (and I did not think, even for a moment, that the effect size was a 1000% increase in penalties).

20
by Cuenca Guy :: Fri, 04/22/2016 - 7:46pm

Your example of world record times would be better demonstrated with a line graph or dot graph instead of a bar graph. Most say that a bar graph should never have a non-zero axis. I would disagree with that point and suggest that there are times that it can be helpful.

In this particular case, simply mentioning that the y-axis isn't zero in order to more easily see the changes would probably have been a good idea. When we're talking about 7-10% changes, those can be significant but hard to see on a "normally" scaled graph. I also did not think that the increase in penalties was to the scale of the graph and doubt that anyone who has read the story thought that way either.

54
by Mo S :: Mon, 04/25/2016 - 11:50am

At the same time we're talking about body temperature. Normal is 98.6 +/1 1 degree and dangerous levels are below 95 and above 106. A zero axis for temperature is just as misleading. "Oh, his temperature only went down 4 degrees, that's not a huge deal, it's less than 5%"

58
by brianstonelake :: Mon, 04/25/2016 - 3:52pm

Ha! This is a great point. Shouldn't the axis-gate folks be just as mad that the temperature graph doesn't have a 0 axis? I'll take it a step further and suggest that 0 degrees F is an arbitrary 0, and might be misleading. Perhaps I should recreate that graph in Kelvins? Funny stuff.

16
by carljm :: Fri, 04/22/2016 - 6:22pm

I don't think this was a horrible article at all; it was interesting to read and (from the source-checking I did) appears well-supported.

That said, the non-0-based graphs are really unforgivable on a statistically-oriented site. Questions of "intent to deceive" aside, such graphs _are_ extremely visually misleading, and "I used the Excel default" is not a reasonable excuse for them. Neither is "having intercepts at 0 made the information hard to read", since in this case "hard to read" is equivalent to "accurately visually conveyed the magnitude of the effect" and "easy to read" is another way of saying "dramatically overstates the apparent magnitude of the effect in a way that supports my argument."

18
by brianstonelake :: Fri, 04/22/2016 - 6:57pm

Fair enough. I'm not sure I agree, but I definitely see your point.

I think readers of FO are sophisticated enough to interpret graphs as their axes present them. Can anyone honestly say that they were confused by the graphs? Is the goal to prevent the confusion of a nonexistent reader?

Below is the graph formatted as you wish - although now I'm sure someone will take umbrage with the maximal values. "I saw a game with 5 turnovers, your graph implies that more than 2 are impossible."

At the risk of sounding defensive, I did note the percentage increases in both turnovers and penalties in the text. At any rate, thanks for the feedback.

http://webpages.sou.edu/~stonelakb/math/pdf/pt.png

22
by doktarr :: Fri, 04/22/2016 - 9:05pm

I disagree. I think it's important to note when you're doing it, but in a case like this where the differences are statistically significant, making it easily visible is not so bad.

It's unsurprising that the differences are small compared to the totals; after all; we're talking about effects that, in total, change the likelihood of one team winning by 8% or so. That's obviously a significant amount but it's small compared to the whole.

3
by Kaelik :: Fri, 04/22/2016 - 2:59pm

Is there any reason that West Coast teams can't just wake up earlier/begin practices earlier/ect. all season round to completely negate this?

5
by brian30tw :: Fri, 04/22/2016 - 3:21pm

That's like wondering why west coast teams don't just add more talent to their rosters to overcome the disadvantage. It misses the point. There should not be a disadvantage to overcome in the first place!

11
by Kaelik :: Fri, 04/22/2016 - 4:15pm

No it is literally not at all like that. It's like if the Seahawks insisted that they don't need to practice, or that their players don't need to exercise. Lots of people all over the world have lots of different sleep schedules, and man, a sure lot of them do wake up really early in the morning to go to work. So if waking up at 6 instead of 8 made a football team win two more games a year, you'd think they'd just do that.

It is certainly possible to claim for example, that waking up earlier wouldn't help, because the beneficial faster reaction times are actually tied to sunlight, not what your body is used to (in which case, FUCKING SOUTH COAST BIAS!) or something else.

But if the actual thing to do to "fix" the "problem" is so easy to do that even people who don't have access to entire support staffs can do it, then yeah, asking them to fix the problem in a way that requires literally zero skill and no salary cap space is a pretty reasonable request.

13
by bluereloaded :: Fri, 04/22/2016 - 5:49pm

That is actually one thing that Pete Carroll has done compared to Mike Holmgren (we'll ignore Mora's singular season). Holmgren was known to hold his practices in the early afternoon, while Carroll has held is practices at 10 AM Pacific since his tenure began.

It still hasn't seemed to help.

51
by Pen :: Mon, 04/25/2016 - 5:55am

It's not a simple fix to just get up earlier. Their biological clocks don't care. Getting up earlier just means they're awake at less than peak performance for a longer period of the day.
For early morning games, these guys have to get up at 4:30.

52
by Kaelik :: Mon, 04/25/2016 - 8:54am

Uh... No? That's completely wrong. If you get up earlier every single day, your rhythms actually change. That's why people in the east and west coast have different rhythms in the first place. That's why people who work night shifts wake up naturally at night and go through the same rising and falling of attention and reaction times at night.

If they woke up at 6 o'clock every single day of the year, they would just be a central time team for the purposes of "unfair rhythm games."

57
by dbostedo :: Mon, 04/25/2016 - 1:01pm

I don't know much about this, but there's a limit to that, right? Doesn't your body also try to sync to daylight? I think people that work night shifts, even after a long time, don't entirely sync. I don't think it's purely about when you wake up and go to bed.

59
by brianstonelake :: Mon, 04/25/2016 - 4:03pm

This is not my area of expertise, but in researching the article I read that there are endogenous and exogenous factors controlling circadian rhythms. To translate this into english, you can reset you own rhythms by changing your sleep schedule, but your efforts will be offset by factors like daylight, etc.

Perhaps more importantly, attempts at changing the sleep schedule of your athletes to one that does not synchronize with their geography is likely not feasible. You'd probably just trade one disadvantage (body clock) for another (sleep deprivation). Teams have been aware of this for a long time, and if there were an easy solution, they would have done it.

But perhaps we are asking the wrong question. Why should teams have to try any of these methods in the first place?

60
by Kaelik :: Mon, 04/25/2016 - 9:47pm

1) SOUTH COAST BIAS. POOR NORTH TEAMS WITH THEIR NO SUNLIGHT IN CRITICAL DECEMBER TO JANUARY. UNFAIR TO MY FAVORITE TEAM NOW!

2) Why should teams have to practice? Why do they have to overcome the disadvantage of being a team that doesn't practice? You certainly haven't made a convincing case that West Coast teams are actually disadvantaged by the schedule more than is reasonable (as was pointed out, the alleged reaction advantage is the same at 4 as it would be at 1). So if it's basically impossible to fix this "disadvantage" without completely revamping the concept of the NFL schedule by scheduling games at 6:00, asking people to wake up two hours earlier doesn't sound particularly unreasonable.

In summary, it sure sounds like a quest for a problem, not an actual verifiable problem looking for a solution. Even your proposed solution has, according to your own graphs, literally no beneficial effect. I wish my team was scheduled at 4;00 more too. But I'm willing to admit that it's just because I want to watch my team at 4:00 when it competes with fewer games, rather than because it would be literally impossible to fix this problem by waking up two hours earlier.

61
by LionInAZ :: Mon, 04/25/2016 - 11:47pm

This is of course nonsense for pro football teams, but it is definitely a bias for springtime college sports. Why should college baseball teams from the South and Southwest get to play longer seasons, which provides more practice and more attention?
Do you think this has nothing to do with why teams from the south and California fare better in poll rankings?

4
by dmstorm22 :: Fri, 04/22/2016 - 3:14pm

So how do you fix this knowing that the reason there are so few 4:00 games is that the networks want to have a 'game of the week' type game to sell at 4:30, with as little competition against that game.

This has been happening for the last 3-4 years now, with fewer and fewer games in the 4:05 window to go up against the 4:25 game.

I guess we can argue it shouldn't happen in the playoffs, but the NFL's hands are again tied by the networks, and ratings.

This really isn't going to change. West Coast teams are better off investing and figuring out ways to make up for this detriment.

6
by Cuenca Guy :: Fri, 04/22/2016 - 3:30pm

So the NFL's hands are tied by the networks? I don't see it changing but not because it can't but because the NFL is far more interested in money than in competitive balance. Not only are they possibly able to negotiate better contracts biven the population distribution of the US, it makes financial sense to give an advantage to East Coast teams.

West Coast teams should do everything they can to try to minimize this competitive disadvantage, but it also stands to reason that they (and any football fan who's not simply looking for the best situation for their team) should speak out against it with statistical evidence like this.

9
by brianstonelake :: Fri, 04/22/2016 - 3:45pm

Well said, Cuenca. Viva Ecuador!

It seems to me (admittedly naive) that the NFL has the power here, not the networks. Yes, forcing eyeballs to their national game might make economic sense, but it seems like that is not in the best interest of the fans, both from an entertainment POV and the aforementioned competitive disadvantage.

10
by Kaelik :: Fri, 04/22/2016 - 4:14pm

Yeah I don't buy that. The NFL schedules games that way because it cares more about money than countervailing concerns. Personally, as a human being NFL fan, I want more 4 o'clock games because right now it looks like one thursday, one monday, one sunday, one 4:25, two 4s, and then all the rest 1 o'clock games (so like fucking 10 on non bye weeks). That's stupid, and it means I get to watch fewer games I actually prefer, because they are all competing at the same slot.

So as a fan, I'm all for pretending that West coast teams have an insurmountable disadvantage that forces them to play later, so that I can have better game choice. But it seems clear that if the NFL fucking wanted, they could easily schedule more games at 4, they just don't want to because they care more about money. If they thought people really believed the fairness of the game is in question, they would probably change that.

12
by dbostedo :: Fri, 04/22/2016 - 5:11pm

Agreed on more 4 o'clock games... I'd be fine with moving any Sunday game where a team travels east 2 or more time zones to 4:00.

62
by LionInAZ :: Mon, 04/25/2016 - 11:55pm

The reality is that the DC- NE corridor gets first preference for TV, because that's where the money is. The league would rather schedule a matchup between 5-6 NFCE teams for primetime than two 6-5 AFCW teams.

15
by Hang50 :: Fri, 04/22/2016 - 6:03pm

SOU faculty on FO! I feel much better about the tuition payments I've been sending down the Ashland for the past three years.

17
by Dan :: Fri, 04/22/2016 - 6:44pm

Nice post! The evidence here is stronger than I expected to find.

21
by JustAnotherFalc... :: Fri, 04/22/2016 - 8:00pm

So, who benefits the most from the scheduling bias in 2016? It's not my Falcons. Despite playing both the NFC and AFC West this season, we get 0 1 PM starts against West Coast teams. We do get 1 PM starts against the Chiefs and Cardinals, but since they are in the Central and Mountain Time Zones respectively, the time effect will be reduced, or perhaps even minimal for the Chiefs.

For the true Pacific Time Zone teams, I count the following 3-hour time shift beneficiaries for 2016:

Lions 1
Jets 2
Patriots 1
Panthers 2
Bills 1
Dolphins 1
Ravens 1
Buccaneers 1
Jaguars 1
Browns 1

Is Jerry Richardson on the Scheduling Oversight Committee?

25
by brianstonelake :: Fri, 04/22/2016 - 11:01pm

Great angle (that I had not considered)!

Here are the results from 2001-2014 (I have 2015 in a different spreadsheet and am too busy/lazy watching playoffs to compile).

1st column is home team
2nd column counts number of times visitor faced a morning start
3rd column is home team win pct in these morning home games
4th column is home team win pct in non-morning home games, for comparison
5th column is difference in 3rd and 4th columns

KC 40 58% 51% 6%
SL 34 59% 44% 15%
CAR 16 50% 50% 0%
MIA 16 81% 48% 33%
ATL 14 71% 58% 14%
MIN 14 86% 61% 25%
NYJ 14 64% 52% 12%
BLT 13 85% 75% 10%
BUF 13 54% 46% 8%
WAS 13 77% 41% 36%
CIN 12 42% 58% -16%
CHI 11 64% 60% 3%
DET 11 55% 41% 14%
GB 11 100% 70% 30%
HST 11 55% 49% 5%
IND 11 82% 71% 11%
JAX 11 64% 47% 16%
NO 11 64% 56% 8%
PHI 11 55% 59% -5%
CLV 10 70% 38% 32%
DAL 10 60% 58% 2%
NYG 10 60% 53% 7%
PIT 9 78% 73% 5%
TEN 9 56% 51% 4%
NE 7 71% 85% -13%
TB 7 43% 49% -6%

Total 349 65% 56% -9%

Not surprisingly, Chiefs and Rams have been the biggest beneficiary of this by virtue of their divisions.

In my very non-scientific polling, rams fans don't seem as concerned as they should be of of the fact that they are switching from a team that benefits from this to one that it hurts.

Your falcons have actually faired pretty well, both in terms of count and results (although with such a small sample there's no real conclusion).

23
by Chip :: Fri, 04/22/2016 - 9:09pm

Isn't a major limitation of this study that the west coast teams were generally terrible to below avg during the study time period (01-15). The NFC West had an atrocious 3-4 stretch with at most one competive team in the entire division along with the Rairsrs who haven't made the playoffs for a decade+.

24
by brianstonelake :: Fri, 04/22/2016 - 10:31pm

Great question!
Fortunately, the answer is no, but it's definitely worth addressing.

From 2001-2015, the 6 "west" teams have a win rate (home and road) of 50.1 percent.
The rest of the league has a 49.7 percent win rate.

By taking 15 years of data and aggregating teams, this (valid) concern is handled.

Perhaps surprisingly, if you remove the 10am starts, "west" teams actually have a slightly higher road win pct than central and east coast teams (see the win percentage graph).

26
by ammek :: Sat, 04/23/2016 - 9:06am

Because I started paying attention to football statistics in the era when the Raiders and San Francisco had the best road win percentages of any franchise – notably during the 49ers' record win streak in the late 1980s – I still have a hard time giving this assertion as much credence as the evidence in this article suggests I ought to.

So I had a very quick look at a different 15-year period: from 1980 to 1994. This was the heyday of the 49ers' dynasty, postdating the passing rule changes, and covering the years when the Rams were still in LA. Because of time limits I only looked at the five teams on the west coast, and simply compared their winning percentages in early games in the eastern and central time zones with their overall road winning percentages. Here are the data:

Win percentage, early games in E&C time zones: 50.8%
Win percentage, other road games: 48.9%

Essentially there is no difference. But when I separated games played on the east coast from those in the central time zone, the result was surprising:

Win percentage, early games in Eastern time zone: 55.5%
Win percentage, early games in Central time zone: 42.9%
Win percentage, other road games: 48.9%

Notably, all five west coast teams had a better win percentage in early games in the Eastern time zone than in early games in the Central time zones. Four of the five had a better win percentage in early games in the eastern time zone than they did in non-early road games. (The Raiders were the exception.)

So I'm not absolutely convinced that a 15-year period gives us enough data to prove Brian's point. Is there a difference in the 2001-15 data between early games in eastern versus central time zones?

27
by ammek :: Sat, 04/23/2016 - 9:18am

Data dump to back up my previous post:

West coast teams, 1980-94, win-loss record & win percentage:

A - In early games in ETZ: 90-72-1, 55.5%
B - In early games in CTZ: 42-56, 42.9%
C - In early games in E&CTZ: 132-128-1, 50.8%
D - In other road games: 156-163, 48.9%

Rams
A - 17-21, 44.7
B - 11-17, 39.3
C - 28-38, 42.4
D - 21-29, 42.0

Seahawks
A - 17-15, 53.1
B - 7-14, 33.3
C - 24-29, 45.3
D - 20-42, 32.3

Niners
A - 31-7-1, 80.8
B - 15-6, 71.4
C - 46-13-1, 77.5
D - 37-19, 66.1

Chargers
A - 15-14, 51.7
B - 4-11, 26.7
C - 19-25, 43.2
D - 31-42, 42.5

Raiders
A - 10-15, 40.0
B - 5-8, 38.5
C - 15-23, 39.5
D - 47-31, 60.3

29
by Bright Blue Shorts :: Sat, 04/23/2016 - 9:40am

Have to agree that my first thought was that the AFC & NFC West were just generally horrible division during the 2000s. The Raiders and 49ers particularly. But the NFC version was probably the worst division ever in football circa 2007-10.

30
by ammek :: Sat, 04/23/2016 - 10:08am

I've had an even quicker look for the remaining Mountain-Pacific teams. Between 1980 and 1994, both Denver and Phoenix* had a better W-L percentage in early road games than they did in overall road games; Houston's record was very similar.

Broncos early road games: 20-21-1, 48.8%
Broncos other road games: 29-44, 39.7%

Oilers early road games: 23-50, 31.5%
Oilers other road games: 14-29, 32.6%

Cards* early road games: 11-24, 31.4%
Cards* other road games: 6-15, 28.6%

All West/Mountain teams in early road games: 186-223-2, 45.5%
All West/Mountain teams in other road games: 205-251, 45.0%

Even so, I still agree with Brian that the NFL shouldn't schedule so many early games for west coast teams.

*Data for Phoenix from 1988-94 only

31
by brianstonelake :: Sat, 04/23/2016 - 12:43pm

Very interesting, thanks for sharing! I wonder how west coast teams did in non morning starts on the east coast? Could the body clock effect be masked by superior west coast teams in this era (harder to beat these better teams at their west coast home than to beat the inferior east coast teams in their home)? Would you mind sharing your data with me? If you're willing, my email is on the article. Thanks!

32
by Kaelik :: Sat, 04/23/2016 - 1:09pm

If you are accepting that overall team/division skill over a 15 year period creates a significant difference, then you are basically telling us that you have no evidence at all, since your entire point was to explain a win differential over 15 years.

33
by brianstonelake :: Sat, 04/23/2016 - 1:21pm

Sorry, I must not have communicated that well. I meant to say that IF a region is far superior it could mask the effect, and needs to be controlled for. In the 15 years I studied this was not the case, as the records indicate (see: comment #24).

35
by ammek :: Sat, 04/23/2016 - 4:09pm

I just used the team game finder facility on pro-football-reference.com to get the data, I haven't double-checked my counting or cut-and-pasting so I wouldn't stand by my numbers 100%. I think they're close enough to make it worth your while checking them out, though, if you want to take your argument further.

I don't have time just now to look at non-morning starts on the east coast. However, the overall W-L record for road teams during the period was 1376-1862-10 (42.5%). That means west/mountain zone teams (391-474-2; 45.2%) did noticeably better on the road than the rest of the NFL (985-1388-8; 41.5%), suggesting that they may have been stronger overall. How much stronger? Hard to say.

I would add, though, that the AFC East had an unusually large share of the terrible teams of the 1980-94 period: for instance, looking at team seasons below .320 (ie, 5-11 or worse) during that time, the AFC East had 19 such seasons, compared with only 7 by teams from the AFC West. Given that more than half of the teams in the dataset played in the AFC West, that might be significant.

36
by ammek :: Sat, 04/23/2016 - 5:45pm

To further test this, I had a quick look at DVOA for the opponents of the four west coast teams (Oak, SD, Sea, SF) during the period 2001-15.

DVOA of opponents in early games:
Avg: -0.50%
Sum: -130.9 percentage points (in 264 games)
W-L: 94-170 (.356)

DVOA of opponents in other road games:
Avg: 0.08%
Sum: +18 percentage points (in 216 games)
W-L: 95-121 (.440)

I can never quite work out how much a percentage point of DVOA is worth, in terms of points, wins, etc. But I reckon this difference in opponent quality between early games and road games is significant, and suggests that west coast teams ought to have won more early games, rather than fewer.

However, the four west coast teams averaged -1.52% DVOA during the period 2001-15, which is still quite a bit worse than their opponents, even the ones in early games. I don't think you can rule out the possibility that west coast teams were terrible enough, on average, to distort the data from 2001-15. (I haven't been able to look at the mountain time zone teams, but I doubt the DVOA for the Cards and Texans is much better.)

38
by ammek :: Sat, 04/23/2016 - 6:37pm

Apologies for replying to myself again, but there's a twist. When I go through these west coast teams' data game-by-game, rather than using 15-year averages, the split in performance between early games and other road games isn't so clear-cut.

I remember reading somewhere that, on average, home-field advantage was worth 17 percentage points of DVOA. So, for each game, I used the teams' annual DVOA and subtracted 17 percentage points from the team on the road. I then picked an expected winner based on which team had the higher location-adjusted DVOA. For west coast teams, the imbalance between early games and other road games was much less marked using this formula.

West coast teams on the road, expected wins-losses versus actual wins-losses:

Early games:
Expected 63-201 (.239)
Actual 94-170 (.356)
Difference (actual minus expected): .117

Other road games:
Expected 70-146 (.324)
Actual 95-121 (.440)
Difference: .116

Using this wins-versus-expected formula, our west coast teams performed identically in early games and in other road games. But homefield advantage was clearly not worth 17 percentage points of DVOA in this dataset, so I reduced it to 10 percentage points:

Early games:
Expected 82-182 (.311)
Actual 94-170 (.356)
Difference: .045

Other road games:
Expected 90-126 (.417)
Actual 95-121 (.440)
Difference: .023

Now there is a small gap suggesting that west coast teams did marginally worse than expected in early games, compared with other road games.

But what is more striking is this. By overall DVOA, west coast teams' opponents in early games were feebler than their other road opponents. Yet, when looking at the head-to-heads, west coast teams were expected to win fewer of these early games against weaker teams. Presumably the 'means' are misleading. I think that's something which needs following up.

40
by brianstonelake :: Sat, 04/23/2016 - 11:40pm

I have little to add, but wanted to reply to show that I appreciate all your analysis!

I'm a little skeptical that the quality of the east coast opponent is drastically different in morning and non morning games. Perhaps pm opponents get a minuscule bump from the fact that "better" games are typically given prime-time status, but I can't imagine it's material.

However, if I understand right, this would mean that the disadvantage is even greater than what is stated in my paper? Not only are west coast teams losing more 10am road games, they're doing so to worse teams?

42
by ammek :: Sun, 04/24/2016 - 5:00am

Perhaps this will cure your skepticism!

West coast teams (Oak, SD, Sea, SF), opponents' average DVOA, 2001-15:

A - Oppt from East/Central Time Zone, early games: -0.50%
B - Oppt from East/Central Time Zone, afternoon/evening games: +3.97%
C - Oppt from Mountain/Pacific Time Zone, afternoon/evening games: -1.19

I wouldn't characterize the prime-time "bump" as minuscule. West coast teams have played New England as many times in the later slots as they did the rest of the AFC East; they have played as many late games against Green Bay as against the rest of the NFC North.

But the bump doesn't just concern opponents. West coast teams themselves are much stronger when playing in the late slots. Here's the average DVOA for west coast teams in the various slots:

A - early games vs teams from east/central time zones: -3.59%
B - late games vs teams from east/central time zones: +6.26%
C - late games vs other teams from mountain/pacific time zones: -1.31%

The DVOA bump is even more significant for west coast teams than for their opponents. It also suggests a more complicated answer to this question:

Not only are west coast teams losing more 10am road games, they're doing so to worse teams?

That was also the conclusion I drew at first (post 36 on this thread). However, if you look closer, using a Predicted Wins formula like the (hastily conceived) one I introduced in post 38, you might begin to question that conclusion.

I definitely think you need to introduce a DVOA-based adjustment in order to test your argument. But the problem with just averaging the DVOA of west coast teams and their opponents (split into 'oppts in early games' and 'oppts in other road games') is that west coast teams simply haven't been average very often! They have taken up more than their share of the very best and very worst seasons of DVOA, and that's why I think a head-to-head, game-by-game, actual wins versus predicted wins formula is more informative.

Here's what I have at the moment, subtracting 10 percentage points of DVOA for each road team in order to determine the predicted winner:

West coast teams on the road, 2001-15:

Early games in east/central time zones:
West coast teams' average DVOA: -3.59%
Opponents' average DVOA: -0.50%
Predicted W-L: 82-182 (.311)
Actual W-L: 94-170 (.356)
Difference (actual minus predicted): .045

Afternoon/evening games in east/central time zones:
West coast teams' average DVOA: +6.26%
Opponents' average DVOA: +3.97%
Predicted W-L: 24-43 (.358)
Actual W-L: 24-43 (.358)
Difference: .000

Afternoon/evening games in mountain/pacific time zones:
West coast teams' average DVOA: -1.31%
Opponents' average DVOA: -1.19%
Predicted W-L: 61-88 (.409)
Actual W-L: 65-84 (.436)
Difference: .027

Using this formula, my conclusion is that west coast teams have performed marginally better in early games than they have in afternoon and evening games, when adjusted for the quality of both teams. Together with the evidence from the 1980-94 period, I'm now strongly disinclined to agree that there's a clear slump in performance for west coast teams playing in early games.

I still think the NFL should do away with as many of those games as possible, however.

45
by brianstonelake :: Sun, 04/24/2016 - 12:13pm

Very interesting! I promise to look into this more, but it might be a few days (I'm elbow deep in mortar, putting in new flooring at the moment).

One question, is your measure of DVOA calculated prior to each week's game? If it's a season end value, for example, are you worried that west coast teams would look weaker than they are precisely because they've faced this a.m. disadvantage 4 times?

One thing I know for sure is that west coast teams think it's a major disadvantage. I have a contact in the 49ers front office who has let me know some of the measures they've taken to combat it, and the actions they've tried to prevent the games from happening in the first place.

46
by ammek :: Sun, 04/24/2016 - 1:05pm

No, it's a season-end value, and you're right that multiple east-coast games could affect the overall DVOA number. Ideally you'd want to use game-by-game DVOA to see whether west coast teams had worse DVOA in early starts than in the rest of their road games. Unfortunately only FO insiders have that information!

One thing I did notice from looking at the data is that teams from the west coast have to be significantly better than average in order to earn multiple afternoon or evening slots for their road games. That seems to me to be unjustifiable. The main cause might be the tv networks' crush on the NFC East. When afternoon/late games featured a west coast team on the road in the east or central time zones, the host team in those games belonged to the following divisions:

NFC East: 18 (Dallas 5, Philly 5, New York 5, Washington 3)
NFC North: 8 (GB 4, Chi 3, Det 1, Min 0)
NFC South: 9 (NO 5, TB 3, Car 1, Atl 0)
NFC West: 4 (StL 4)

AFC East: 6 (NE 3, Buf 1, Mia 1, NYJ 1)
AFC North: 9 (Pit 4, Cle 3, Bal 1, Cin 1)
AFC South: 7 (Ten 3, Ind 2, Jax 2)
AFC West: 5 (KC 5)

As you can see, only five eastern/central teams have played five home games in the late slots against west coast teams, and three of them are from the NFC East.

64
by brianstonelake :: Tue, 04/26/2016 - 2:37pm

Sorry for taking so long to respond, Ammek. I've been trying to keep up with this on various teams sites and several reddit threads. The internet can be exhausting!

First off, I find your analysis fascinating and really appreciate you looking into this. If you'd like to work collaboratively on this or a related project, email me and we can see what we can come up with.

Some general comments - Knowing how morning body clock teams fare relative to expectations is the holy grail. It's awesome that you are looking into this. I'm not sure that YE DVOA is the way to go. I really don't like using a metric that is based on these results to retroactively predict what should have been the result. Of course it's going to say that teams performed (close to) what was expected, because its expectations are based on what did happen.

Unfortunately I don't know how to measure expectations. Pregame power rankings? Too subjective and noisy. Las vegas pre-season win totals? Body clock effects are included in these lines.

My thought is that we don't need to do this (although it would be cool if we somehow could). There's no systematic scheduling bias correlating the difference in quality of the teams playing with the time of day. As long as the regions are comparable in terms of talent (which they are in the 15 years that I examined), we only need to look at the aggregate win rates.

This brings me to your idea of increasing the sample size. At first glance, I like it. Why not include more data? However, I'm concerned by the fact that the 15 year period you chose (if I understand correctly) was chosen because it was an era where some west coast teams were known to perform very well in this situation. I know Scott Kacsmar mentioned in an article that in the niners heyday in the late 80's they won 10 straight a.m. road games (excerpt copied below). Doesn't this violate random sampling?

Also, as you mentioned, there seemed to be less parity then, possibly due to the lack of a salary cap. This might also mask the effect. I imagine we'd still see a difference in point differential, but perhaps less so in win rates. And even the point differential may be hidden due to garbage time in blowout games. I'll still likely look into the data at some point, because I'm curious like that, but I'm worried about any conclusion that can be drawn from it.

At any rate, I don't mean to be argumentative or skeptical of your findings. I'm just worried that they might not properly represent the physiological disadvantages that are known to exist.

Do you really think that the 65% win rate for home teams hosting a.m. body clock teams in the past 15 years is coincidence? That's not a rhetorical question, I'm genuinely interested in your opinion as you have clearly thought carefully about this.

I think my big picture take it this: Physiology literature unequivocally suggests a disadvantage associated with competing in the a.m. It is absolutely astounding that such a disadvantage could manifest itself in game results, given the numerous factors contributing to the result of a given game, yet examining the results from the past 15 season, we see a large difference in win percentage.

NFL teams on the west coast want this change. Some are obsessive about it. I understand those discounting my opinion, but theirs? The league has made similar changes at the behest of east coast teams averse to too much travel. It seems to me that the main factor preventing this change is the league's desire to have minimal competition to their national game. I don't think most fans realize that the league is choosing ratings over the integrity of the competition.

My goal with all of this is just to raise awareness. Everyone should (and will) draw their own conclusion about the magnitude of the disadvantage. Many will disagree with me, which is fine. I'd much rather have someone disagree with me than to be ignorant of the situation in the first place.

HERE IS THE EXCERPT FROM SCOTT'S FO ARTICLE:

Still a little skeptical over the data for Western teams? We understand. This is based on five years; not 30. Had we extended our research back to the 1980’s, we may have finished with completely different results when the West was run by the likes of Joe Montana’s 49ers, San Diego’s Air Coryell, Seattle’s Ground Chuck, the two-time champion Raiders, swap the (St. Louis) Cardinals for the Los Angeles Rams, and John Elway’s Broncos.

No matter what the playing conditions are, we know great teams can overcome such things. There’s no better example than the team that wrote the record book on winning away from home. They just so happened to be an early-rising Pacific team from San Francisco.

From 1988 to 1990, the San Francisco 49ers set a NFL record with 18 consecutive road wins. A whopping 10 of those wins came in games where the 49ers had a body clock of 10 a.m. PST at kickoff.

Not only did they win 18 straight road games in the regular season, but they also won the 1988 NFC championship in Chicago, and of course two Super Bowls were won on neutral fields. That is 21 consecutive wins away from Candlestick – a record that will not likely fall any time soon, and especially not by a team that traveled this much.

67
by ammek :: Tue, 05/03/2016 - 10:07am

Sorry I took even longer to reply.

I'm flattered by your offer of collaboration and will contact you if I have anything new to suggest. But I have no background in statistics or anything other than buzzing around the internet with an eyebrow raised quizzically.

I completely agree with you that the NFL and the networks are putting their interests first. For me, the evidence that 10 am games affect performance doesn't have to be conclusive; so long as there is a suspicion that they do, I think the NFL should reduce the number of early road games for west coast teams. I mean, they're also a pain in the ass for tv viewers on the west coast, whom the NFL seems eager to take for granted.

I think the arguments in your article are generally strong, and that 65% win percentage figure is certainly something that demands to be investigated. My skepticism really boils down to this one assertion – "As long as the regions are comparable in terms of talent (which they are in the 15 years that I examined), we only need to look at the aggregate win rates" – which I don't believe you have proven convincingly yet. It would definitely strengthen your case if data from previous eras fit with your findings from 2001 to 2015 on this question (or if you could look more deeply into the reasons why they might not fit).

The 49ers' extraordinary road success was one of the factors that prompted me to look at the period from 1980 to 1994. But Seattle also had above average success in early road games during that period; the Raiders had below average success; and the other Mountain-Pacific teams all did about average. It's definitely worth investigating. (As is 1995-00; I left that period out because more than one-third of the sample relocated during those years, but it was the heyday of 'parity'.)

28
by Bright Blue Shorts :: Sat, 04/23/2016 - 9:36am

General feedback ... I struggled with the readability of the article ... too many multisyllable words crammed into sentences.

The final paragraph suggesting "The NFL has gone to great lengths to maintain a level playing field" is so not true. If the NFL wanted a level playing field they wouldn't have Thursday night games, they wouldn't make some teams play in England, the Giants wouldn't have won the Super Bowl since Parcells left.

But even so ... isn't this what "home field advantage" is?

34
by brianstonelake :: Sat, 04/23/2016 - 4:05pm

Fair critique. Hemingway I certainly am not! I appreciate the feedback.

It's easy to vilify the NFL, but I don't think they are ambivalent to fairness. They've made scheduling changes in the past to correct for similar unfair situations.

The situations you reference seem like challenges for teams, but I'm not sure they are "unfair" unless they systematically happen to specific teams more frequently than others.

Finally, I disagree that this can just be categorized as home field advantage. The Jets have a much larger advantage playing a visiting 49ers team at 1pm EST than they do if the visitor is the dolphins. It's not home field advantage, it's road game disadvantage that is specific to certain teams, and it is much larger in magnitude than I ever imagined (on par with losing your best non-QB?!?!)

37
by Kaelik :: Sat, 04/23/2016 - 6:04pm

What you are really saying though, is that East Coast teams have a better home field advantage in some of their games (against West Coast teams).

But since Seattle has a better homefield advantage, in 100% of their home games, clearly having a better home field advantage isn't something that the NFL actually cares about, because if it was, they'd through a shit about teams building stadiums to fuck visitors.

So as long as the NFL doesn't do anything about "unfair" home field advantage through stadium design, why would they care about "unfair" home field advantage created by people refusing to wake up earlier?

I mean, I guess you could say they do care, hence crowd noise piping punishments, and they are just really dumb and haven't yet realized that Seattle has a homefield advantage?

But either that incompetent, or not caring, either way, it seems really unlikely that the NFL would care about the alleged fairness concern.

48
by brianstonelake :: Sun, 04/24/2016 - 2:05pm

Apples and Oranges. And the apples are way bigger than the oranges.

The league did not design the stadiums in Seattle, New Orleans, etc nor did they prevent other clubs from designing stadiums similarly. And while the stadium design does contribute to home field advantage (HFA) none of it is possible without raucous fans. bengal fans could be louder, Chargers cant do anything about playing at 10am.

Also, I think I've read that the best HFA are an extra point over normal HFA. The body clock issue is 3 times that.

I see your point, don't get me wrong, I just don't see it as a reason for the league to not change the 10am starts. As I mentioned in the paper, just a few years ago they fundamentally changed the schedule to prevent east coast teams from facing the competitive disadvantage associated with flying to the west coast too frequently. The same should be done here.

55
by Aaron Brooks Go... :: Mon, 04/25/2016 - 12:27pm

"The Jets have a much larger advantage playing a visiting 49ers team at 1pm EST than they do if the visitor is the dolphins."

If you're going to make statements like that, you should really normalize for distance. Geographically Confused East (AFC) is 3000 miles from Geographically Confused West, but 1300 miles from Miami.

A better comparison would be the Bills hosting the Dolphins vs hosting the Broncos, as those two are roughly equidistant, but one is N-S and the other is E-W.

39
by jeff1rey :: Sat, 04/23/2016 - 9:22pm

As a statistician you should be presenting the readers with p values etc. None of your stats as presented would be accepted in a scientific journal. They are interesting but need much further analysis to be meaningful. If this is true then West coast teams coming east for a 1pm game should plan on coming 3-4 days early to adjust or insist on playing at 4 or 8 pm. A big factor is the teams that go to Denver and play at 5000 feet. They definitely should be coming west by Wed for a Sunday game. My son is a big Hawks fan but then so are you. This colors your thinking.

41
by brianstonelake :: Sun, 04/24/2016 - 12:04am

I appreciate your post, but have to correct a few things:

- I did include p-values in the paper

- This article was my attempt at a general audience friendly version of a paper that was already published in a scientific journal (The Journal of Exercise Physiology). That paper is linked at the bottom of the article.

- Trying to manage the body clock as you suggested is a good idea, and many teams try that. SF stayed back east a couple times in the past few years in attempts to regulate circadian rhythms. However the process of managing this disadvantage is itself a disadvantage. This issue is not new to teams. They have been trying to negate the disadvantage for at least 20 years, and as the win percentages show, nothing has worked.

- Most importantly, I am certainly not a Seahawks fan! I was born and raised in the east bay and love the 49ers! (The Raiders were in LA when I was growing up.) While I'm sure I'd love him if he suited up for us, Richard Sherman is probably my least favorite player in the league. I can say unequivocally that my feelings for the 'hawks did not color my analysis.

43
by Bright Blue Shorts :: Sun, 04/24/2016 - 5:41am

Any thoughts about this year's London games? All games are the equivalent of 9:30 ET.

For the LA Rams that will be playing at the equivalent of 6:30am vs the Giants.

The other two games (Colts-Jags, Redskins-Bengals) but obviously the teams involved have similar body clocks.

44
by brianstonelake :: Sun, 04/24/2016 - 12:10pm

Good question. I haven't thought much about it, to tell the truth. I excluded past London games from the study, because "home team" didn't correspond with "no chronobiological disadvantage". To your point, I guess it'll be a much worse time for the Rams than the Giants (but non ideal for either). Any thoughts?

47
by Bright Blue Shorts :: Sun, 04/24/2016 - 1:24pm

I believe teams have started leaving it until the last possible moment (e.g. Friday) before flying to England.

But I think the Rams have to come over after their game in Detroit the previous Sunday. Take the hit of jet lag for 3 or 4 days and be adjusted to GMT/BST by the Sunday so that it becomes an afternoon game for them. Of course the Giants could also do the same.

49
by Bjorn_ :: Mon, 04/25/2016 - 4:38am

Well, from a comercial standpoint it would seem that the main "sell" of the London games for the NFL and the big Money reason for a potential London Franchise is the possibility to sell an additional "game slot" to the TV networks for a Sunday Morning game at 9:30 Eastern.

I know that it was already heavily discussed back in the Days of the Start of the WLAF (that later became NFL:Europe) what a "Jet Lag advantage" the european teams had.

Quality of play for many of the London games so far has been less than great, this factor could certanly have something to do with that.

50
by Bjorn_ :: Mon, 04/25/2016 - 5:07am

While the potential addition of London based team would to some extent mitigate this discrepancy between East and West coast teams (assuming that team would be put in one of the East divisions) since now also east coast teams would have a bunch of disadvantaged games.

But this one team in London would then have a pretty sweet competive advantage:
A bunch of games 14:30 for them, 09:30 for opponent (west coast team 06:30) -> Great advantage
A bunch of games 18:00 for them, 13:00 for opponent (west coast team 10:00) -> Great advantage
A few games 21:00 for them 16:00 for opponent (West coast teams 13:00)-> more or less a wash
The occasional prime time game 01:00 for them, opponent at 20:00 -> significant disadvantage (but with travel east to west being possible to manage)

AND, if the team starts to be succesful marketwise, you will start to have whoever has the european TV deals start to lobby for them having less late and prime time games!

56
by Aaron Brooks Go... :: Mon, 04/25/2016 - 12:35pm

Only if you discount travel times (and remember, all road games are international travel; unlike UEFA, this isn't inter-EU). That London team is going to wrack up a ton of extra road miles and air hours.

Ask the Red Wings what the Western Conference playoffs were like in the NHL when you were the only team in your time zone, and one of five east of (or adjacent to) the Mississippi. A single trip to Vancouver was more travel than NJ had in the entire playoffs.

53
by Kevin from Philly :: Mon, 04/25/2016 - 9:24am

More importantly, how does this phenomenon impact RaiderJoe's spelling?

63
by LionInAZ :: Tue, 04/26/2016 - 12:06am

I wonder if byes should also be accounted for. I recall one season in which the Chargers had to travel more than once to plasy AFCE teams that were coming off the bye. Talk about unfair advantages!.

65
by BretU :: Wed, 04/27/2016 - 7:38am

You're obnoxiously loud. Now apologize for saying Russell Wilson is only incrementally better than Tavaris Jackson.

66
by brianstonelake :: Thu, 04/28/2016 - 10:36am

It's Tarvaris. Who knew? (Not me)

I bet you could win a bar bet or two with that knowledge. Does that make up for the bad "obnoxiously loud" joke?

68
by BrianWhomy :: Tue, 10/03/2017 - 1:37pm

[url=http://www.globalsolutions.in/CssGlobal/fliure.php?q=24]Levitra Generic[/url]
We all want to obtain sparkling, healthful-looking your hair. The simplest way to achieve the look you want is to achieve a comprehensive knowing of the most effective good hair care methods and routines. By using the minds in this particular piece to heart, you will soon become a locks expert and become the envy of your respective good friends.
[url=http://www.universalgolffoundation.org/Scripts/cache.php?t=59-Lovegra-Price-Buy-Viagra-Australia-Online-Cheap-Viagra-Melbourne]Lovegra Price[/url]
When you are a youthful man who is starting to shed his your hair you don't must seem like your passionate and sex lives have ended. There are several individuals who definitely are actually drawn to males with less locks plus an even more quantity of people who are just indifferent into it.
[url=http://www.tobyaboutiquehotel.com/editorfiles/file/park.php?PageId=129-Viagra-I-Turkiet-Lovegra-Gel-Viagra-Online-Sweden]Viagra I Turkiet[/url]
An incredible hint which will help for you continue to keep your stress threshold lower would be to go on a warm bath tub. The drinking water forces you to sense peaceful and you'll overlook all of your issues. Getting a very hot bathtub might help keep the tension under control and it's some thing you could do every day.
[url=http://www.ita-eg.com/event/form.php?productID=167-Viagra-Hamburg-Viagra-Für-Frauen-Test-Viagra-Ohne-Rezept-In-Holland-Kaufen]Viagra Hamburg[/url]

69
by GreggCrype :: Thu, 10/12/2017 - 6:29am

ryofkxa

http://www.geaweb.it/726-air-max-command-rosse.php
http://www.bjdn.co.uk/dolce-gabbana-sneakers-for-sale-009.asp
http://www.crescentmoonmusic.co.uk/630-mbt-black-womens.htm
http://www.150store.it/adidas-tubular-prezzo-035.htm
http://www.mhcreativemedia.co.uk/fila-shoes-womens-white-674.htm

[url=http://www.zapatillasmizunomujer.es/zapatos-salomon-mujer-gore-tex-852.html]Zapatos Salomon Mujer Gore Tex[/url]
[url=http://www.geaweb.it/402-air-max-90-green.php]Air Max 90 Green[/url]
[url=http://www.timberlandsejlersko.dk/776-billige-timberland-støvler-til-damer.htm]Billige Timberland Støvler Til Damer[/url]
[url=http://www.mujerzapatosaltos.es/643-manolo-blahnik-descuentos.asp]Manolo Blahnik Descuentos[/url]
[url=http://www.pro-forma.co.uk/309-asics-tiger-gel-lyte-iii-spectra-green.shtml]Asics Tiger Gel Lyte Iii Spectra Green[/url]