Archives for category: Uncategorized

Do the Saints avoid drafting LSU players or is it just coincidence?  During Sean Payton’s tenure the Saints have only drafted 1 LSU player (All Woods, 2010).  Is it because they purposely avoid LSU players or is it just the way the boards have fallen over the years?  Is there a statistical way to answer that question, and if so, what do the numbers say?

Since 2006, (not counting the currently still on-going, at the time of this writing, 2018 draft) there have been 77 LSU players drafted.  The Saints have only drafted 1 of those 77.  By the numbers, with 32 teams, the average team drafted 77 / 32 = 2.4 per team.  That stat doesn’t really tell us much, so let’s take a look from another angle.

If we took the same group of players drafted each year into the NFL since 2011, and then randomly distributed them among the NFL teams, what is the probability the Saints would have 0 LSU players?  I calculate that number at about 19.6%.  In other words, there would be a better-than-80% chance (80.4%) the Saints would have drafted at least one LSU player since 2011 if the drafted players were randomly distributed.

year number of LSU players drafted that year number of LSU players drafted by the saints probability of getting an LSU player that year probability of not getting an LSU player that year cumulative probability of not getting an LSU player over the years
2011 6 0 18.75% 81.25% 81.25%
2012 5 0 15.63% 84.38% 68.55%
2013 9 0 28.13% 71.88% 49.27%
2014 9 0 28.13% 71.88% 35.42%
2015 4 0 12.50% 87.50% 30.99%
2016 5 0 15.63% 84.38% 26.15%
2017 8 0 25.00% 75.00% 19.61%

The above table (via research done at PFREF.COM) shows the year, the number of LSU players drafted that year, the number the Saints took that year, and then 3 probabilities: probability of getting at least one LSU player that year in a blind randomly distributed draft, probability of *not* getting at least one LSU player that year, and in the final column a cumulative probability of not getting at least one LSU player in any of those years.

In 2011, there were 6 LSU players drafted.  With 32 teams, if these 6 players are randomly distributed among the teams, the percentage probability of the Saints getting at least one of them would be 6 / 32 = 18.75%.  That’s obviously not a very high probability, and so shouldn’t raise any eyebrows when it happens exactly that way.  Saying there is an 18.75% chance of getting at least one LSU player in the draft is the same thing as saying there is a 100% – 18.75% = 81.25% chance of *not* getting an LSU player in a blind, randomly distributed draft.  In other words, I’m equating blind, randomly distributed with meaning there is no bias either in favor of LSU players or against LSU players by the Saints.

Okay, so with an 81.25% chance of *not* getting an LSU player, we can’t draw any conclusion from the fact the Saints did not get an LSU player in 2011.  But let’s continue on to 2012.  5 LSU players are drafted that year, and doing the same math we get the number at 84.38% that none of the 5 would land with the Saints.  So we have 2 probabilities to work with: 81.25% and 84.38%.  If we multiply those together we get 81.25% * 84.38% = 68.55%.  68.55% is the probability percentage that the Saints would not get *any* LSU players in *either* year.  It’s still over 50%, so it shouldn’t be a shocker that the Saints didn’t end up with any LSU players in 2011 through 2012.  Let’s continue on for the following years, where the numbers continue dropping to 49.27% in 2013, 35.42% in 2014, 30.99% in 2015, 26.15% in 2016, and now 19.61% as of 2017.

Already in the first 2 rounds there have been 4 LSU players drafted in the 2018 draft.  A reasonable expectation is there will be 4 or 5 more LSU players drafted, so call it 8 LSU players in total that get drafted in 2018.  If the Saints don’t get any of them, the number would be the same as it was in 2017 when also 8 LSU players were drafted.  This means we would need to multiply 19.61% by 75%, which gives us a 14.71% that the Saints would have skipped out on all of those LSU players just by random coincidence.  Personally, I’m not buying it.

I’m not a conspiracy theorist, not by a long shot, but something is fishy here.  You can’t tell me something with a 14.71% chance of happening has happened all just by random coincidence.

On the other hand, the Saints *did* draft Al Woods back in 2010.  Also, we’re looking at overall numbers and assigning equal probability to all 32 teams when, in fact, the Saints have very happily traded away 42 picks (in exchange for 30 picks) over this time frame, not to mention *never* getting a single compensatory pick due to losing free agents and not replacing them in free agency.  In short, the Saints have had the fewest draft picks of any team, and by a considerable margin, over the years.  Having fewer picks (about 2 fewer per year than the average team) means the Saints are the least likely team to have ended up with an LSU player, assuming blind random distributions.

Final conclusion?  I think there’s a bias at work here against drafting LSU players.  Not sure why, could be the Saints just don’t think LSU does a good job coaching those players up (contrary evidently to how the rest of the league views the situation).  Maybe it’s just a location bias, in that the Saints want to try to cast a wide net rather than just focusing on the local players, and hence are bending over backwards too far.


Saints just drafted a player, Defensive End Marcus Davenport, who weighs in at 264 pounds.  Is weight an indicator of how successful a pass rusher might be in the NFL?  If so, how does Davenport’s weight figure into the numbers?

Using the play index at PFREF.COM (and doing the addition in my head) I broke it down in ranges of 10 pounds, e.g. 240-249, 250-259, etc.

Stats compiled from 2000-2017
Weight 10+ Sack Seasons
200-229 0 (0 players)
230-239 23 (8 players)
240-249 19 (7 players)
250-259 67 (29 players)
260-269 79 (30 players)
270-279 50 (25 players)
280-289 46 (21 players)
290-299 17 (8 players)
300 or more 10 (7 players)

Bear in mind we’re using the same weight value for each player for his entire career, and obviously players (like the rest of us) tend to put on weight as the years go by, and weights will also tend to fluctuate throughout the year and even within the same game.  So, take these numbers with a grain of salt and with the realization they’re not necessarily telling the complete story.  It’s what we have to work with.

The thing in these stats that really stands out to me is the sweet spot is fairly broad (from 250 to 290), with 260-269 being the peak of the bell curve.  Bottom line, at least according to these numbers, if you want a 10+ sack player, try to draft one in that 250-290 range, preferable between 260-269.

Another interesting thing about Davenport is his speed in the 40 at his weight.  According to my research using the Combine search tool at the same site as mentioned above, Davenport is the only player in the last 3 Combines (not since Bud Dupree did it in 2015) to run a 4.58 or faster while weighing in at 264 pounds or heavier.  In other words, this guy is freakishly fast considering his weight.

Does that mean this guy is a lock to have his bust in Canton some day?  No, of course not, statistically speaking he is probably more likely to be a bust of a different sort.  But one thing is very clear, the Saints believe he will be a big time player because there’s no way they would have spent 2 1st rounders and a 5th rounder to get him otherwise.  Personally, I would not have made that deal just based on what *I* know about the player (which is next to nothing aside from what I’ve already detailed herein).  The Saints made the deal based on what *they* know (or think they know) about the player, and presumably, they’ve done the necessary homework in the scouting department.

By the way, kudos goes out to Larry Holder of the TP for calling this one about 2 hours before the draft.

Happy Pi Day (March 14th).  Today I thought I’d put together a little something unrelated to sports (although I suppose with more forethought I could have tied it into sports somehow), but instead related to math, and the funnest of fun numbers: Pi.

As you all well remember, I’m sure, from grade school, Pi (3.14…) is the ratio of a circle’s circumference to its diameter, such that given a perfectly round lake one mile across at its center, one would need to walk Pi miles completely around the lake, but only swim 1 mile across the lake.  With that said, on to our story of genies, kings, pies, and Pi.

One Pi Day the king of the genies said to his math genie, ‘Math, bring to me Pi pies that I may feast upon them this Pi Day.’  To which Math replied, ‘Would that I may, your majesty, but that would be more than 4 times thine allowed daily caloric intake, and should her majesty the queen of genies learn of it…’

‘Very well, very well’, said the king, ‘then bring me a 4th of Pi pies that I may feast upon it this Pi Day.’

‘Very well,’ said Math.  ‘How precise an approximation would suffice, your majesty?’

‘Approximation abomination!’ replied the king.  ‘I want *exactly* a 4th of Pi pies, no more and no less.’  Whilst his majesty might have to settle for a fraction of a Pi pies, by no means would he settle for an approximation of that fraction.

‘This shall require an infinite series, but it can be done in finite time.  We need merely produce and consume each subsequent piece with twice the rapidity of the first, and since the subsequent pieces will be ever smaller, this will be QED, quite easily done.’

With that, Math produced a pie, and placed it on the king’s table, but before allowing the king to eat it, he cut away 1/3 of it.

‘You must partake of this first piece in 6 minutes exactly, your majesty.’

After the king eating the first piece, Math produced 1/5th of a pie and placed it on the king’s table, but before allowing the king to eat it, he cut away 1/7th of a pie from the 1/5th of a pie, which isn’t very much pie, and the king of genies was not particularly happy about it, but ate it nonetheless in the prescribed 3 minutes allotted.  Recall, each subsequent piece of pie had to be consumed in half the time of the previous.

The next piece of pie was 1/9 – 1/11, and had to be consumed in 1 1/2 (1.5) minutes.  Subsequent pieces were 1/13 – 1/15, 1/17 – 1/19, 1/21 – 1/23, and so on.  Summary:

1/1 – 1/3 + 1/5 – 1/7 + 1/9 – 1/11 + 1/13 – 1/15 + 1/17 – 1/19 + 1/21 – 1/23… = Pi/4.

My question to you, should you wish to take it upon yourself, is how long did it take the king of genies to eat a 4th of Pi pies, divided into infinite pieces, and assuming the first piece took 6 minutes, and each subsequent piece 1/2 the time of the previous, and further assuming no time was wasted in between each piece?

As a corollary to this, there was another genie, who was fascinated with birds.  They called him Birdie.  One day, Birdie found 2 trees exactly 1 mile apart.  He transfigured himself into a bird, a magical bird, unbound by the laws of physics, mind you, and began flying from one tree to the other and back again.  He made the first trip at a leisurely 10 miles-per-hour, but then doubled his speed to 20 mph for the next trip, re-doubled to 40 mph for the next, 80 mph for the next, and so on and so forth.  The question (which is the same as the question of the king eating his 4th of Pi pies) is how long did it take before our magical bird was in *both* trees at *the same time*, presuming he wasted no time between trips?

[spoiler follows]

[spoiler follows]

[spoiler follows]

[spoiler follows]

[spoiler follows]

[spoiler follows]

Okay, so the answer is twelve minutes.  If we take 6 + 6/2 + 6/4 + 6/8 + 6/16… we get what is called an infinite series, which does, in fact, converge to 12.  There is a great site called where you can get the answer to that question, provided you can formulate your query.  Go there and enter the following:

Sum [6 / 2^k, {k, 0, Infinity}]

The answer you will get is this:


As to whether this was exactly a 4th of Pi pies, we can get this similarly, though not quite as neatly:

Sum[((-1)^(k + 1))/(2*k – 1), {k, 1, Infinity}]


This is another sum of an infinite series.  Recall, we are trying to get 1/1 – 1/3 + 1/5 – 1/7 + 1/9 – 1/11… = Pi/4

We’re adding and subtracting each subsequent term of 1/k where k takes the range of 1 through Infinity for all odd numbers: 1, 3, 5, 7, 9, 11….  We can get there by using a power of -1, which works because -1 * -1 = +1, but -1 * -1 * -1 = -1, and -1 * -1 * -1 * -1 = +1 and so on.  In other words (-1) ^ k is positive if k is even, but negative if k is odd.  The bottom number (2*k – 1) is *always* odd because it’s always 1 less than some even number (2*k).  Let’s consider the first few terms.  The first one (k=1) is:

((-1)^(k+1)) / (2*k-1) = ((-1)^(1+1)) / (2*1-1) = ((-1)^(2)) / (2-1) = (1) / (1) = 1

Now, (k=2):

((-1)^(k+1)) / (2*k-1) = ((-1)^(1+2)) / (2*2-1) = ((-1)^(3)) / (4-1) = (-1) / (3) = -1/3


((-1)^(k+1)) / (2*k-1) = ((-1)^(1+3)) / (2*3-1) = ((1)^(4)) / (6-1) = (1) / (5) = 1/5

Similarly, (k=4) results in -1/7, (k=5) in +1/9, and so on down the line, alternating basically 1/k as +1/1 – 1/3 + 1/5 – 1/7 + 1/9 – 1/11… all of which adds up to Pi/4.

There are lots of other similar means to get Pi from various infinite series.  For example, had the king wanted 1/6 of Pi^2 pies:

(Pi^2)/6 =  Sum[1/k^2, {k, 1, Infinity}]


In other words, Pi^2 / 6 = 1/1 + 1/4 + 1/9 + 1/16 + 1/25…

Stated another way, Pi^2 / 6 = 1/1^2 + 1/2^2 + 1/3^2 + 1/4^2 + 1/5^2…

We can also say Pi/6 = the square root of (1/1 + 1/4 + 1/9 + 1/16 + 1/25…)

I’ll leave you (not that any of you are still reading) with the following nugget about the interesting number Pi.  The winding river takes a curved path that is Pi times the length of the direct path to its destination.  Yes, friends, the offseason is loooong.

PS: Congrats to LSU track star, Aleia Hobbs, who took 1st place and a national championship in the NCAA indoor meet recently.  Aleia ran a blazing 7.07 seconds in the 60 meter sprint.  We’re more familiar, as football fans, of the 40-yard dash run at the Combine.  If Aleia ran the 40-yard dash at the same *average* speed with which she ran the 60-meter sprint, what would her 40 time have been at the Combine, running with the fastest male NFL prospects?

Wolframalpha to the rescue again.  Go there and enter:

Solve 60 meters in 7.07 seconds = 40 yards in n seconds

to get the answer (n = 4.310 seconds).  Aleia’s time would have been the fastest at the Combine (4.32 by LSU’s Donte Jackson, Tulane’s Parry Nickerson, and Ohio State’s Denzel Ward) this year!  Now, consider also the 40 is much shorter than the 60-meter, so she had to pace herself a bit over 60 meters compared to perhaps being able to go full out in the shorter 40, so it’s possible she could have shaved a few more fractions of a second off that average time.  (It’s also possible, I suppose, the longer run gave her more time to offset any disadvantage she might have had, if any, at getting up to full speed at the start.)

In some ways the Pelicans resemble the Saints from 2012-2016, good offense, bad defense, but let’s not oversimplify the issues.

Shots per game

Obviously, you can win if you make more shots than the other guys, but in order to make more shots you either need to take more shots or hit a higher percentage of the shots you take or some combination thereof.  One area where the Pelicans are really struggling at is in the shots taken per game stat.  Pels are currently attempting 84.9 field goals per game, which is 20th in the league.  League average is 85.4, so they are 0.5 shots per game below the league average.  That’s not so bad, but their opponents are attempting 90.2 shots per game, 2nd most in the league.  Pels are getting out shot by 90.2 – 84.9 = 5.3 shots per game.  That, my friends, is dead last in the NBA.  But Golden State is 29th and Cleveland is 28th, so maybe this is not the be-all end-all of stats to be concerned about.

Why would a team be -5.3 in shots per game?  Turnovers can certainly be part of it.  Rebounding could be another part.  Reluctance to take the end-of-the-quarter percentage busting heave shots could be another.  You will sometimes see clever veterans dribble towards the other end, wait for the buzzer to sound, and then heave the shot, knowing it won’t count against their stats.  But let’s not get sidetracked.


In terms of field goal % per 100 possessions, the Pelicans are very good, 2nd best in the league at 48.8%.  They’re 3rd best in 2 point % and 9th best in 3 point % on their way to 2nd best overall.  Defensively, however, not nearly so good.  Pelicans are tied for 15th best in FG% allowed at 46.1% (per 100 possessions).  They are 15th best in 3 point % allowed and 16th in 2 point % allowed (per 100 possessions).  So, field goal percentage is really not the issue.  They’re excellent offensively and average defensively.  Sure, they could be better defensively, but it’s middle-of-the-pack in terms of field goal percentage allowed per 100 possessions.

In terms of net FG% (your FG% – your opponents’ FG%) the Pelicans are +0.027, which is 2nd best in the league.  The problem isn’t so much the defensive FG% allowed, but rather it’s the total number of shots being allowed.


In terms of rebounds per 100 possessions, the Pelicans are 24th with only 42.7 total rebounds per 100 possessions.  Opponents are getting 43.3 rebounds per 100 possessions.  Thus, the net rebounding per 100 possessions is -0.6, which is 18th in the league, so another middle-of-the-pack ranking.  Rebounding, while it could be better, is not the big issue.

A note on rebounding.  I think too much is made of offensive rebounding versus defensive rebounding because every defensive rebound you get is an offensive rebound the opponent doesn’t get, and vice-versa.  Let’s say we change the rules such that every offensive rebound gets you 10 points and defensive rebounds count for 0 points.  You’d want to get more offensive rebounds, right?  Wrong.  It would still be just as important to get the defensive rebound because that’s an offensive rebound (and 10 points) the opponent is not getting.  Total rebounds is really all that matters, and it doesn’t matter whether they’re offensive or defensive.


Pelicans are averaging 15.8 turnovers per 100 possessions, which is 6th worst in the league.  Opponents are averaging 14.4 turnovers per 100 possessions, which is 21st in the league.  Thus, the Pelicans have a net of -1.4 turnovers per 100 possessions, which is 25th worst.

The Pelicans are getting out shot (in terms of field goal attempts) by 5.3 per 100 possessions.  Turnover differential makes up for 1.4 of those lost shots, while rebounding differential makes up for 0.6.  Thus, turnovers and rebounds account for 1.4 + 0.6 = 2.0 lost shots per 100 possessions.



Biggest issue facing the Pelicans is the net shots per 100 possessions of -5.3.  If you’re shooting 84.9 shots per your opponents’ 90.2 shots, that means you’re taking 84.9 / 90.2 = 94.12% as many shots as they are.  Just to break even you would need to shoot 1.06243 times better than your opponents are shooting.  Pels are shooting 48.8%, their opponents are shooting 46.1%, so they’re outshooting them by 48.8 / 46.1 = 1.05857, which is short of where they would need to be to break even (1.06243) and make up for the lost shot opportunities.

Pelicans are shooting it about as well as they can, so there isn’t much room for improvement there.  There is room for improvement in FG% defense, but it’s middle-of-the-pack.  Biggest room for improvement is in turnover %.  Cutting down on the turnovers should be priority #1 going forward.  Better team rebounding should be priority #2.

Time to pay up, Saints fans.  Time to pay our membership dues.  You see, it’s not free being a member of the Who Dat Nation.  The cost is an emotional one.  Another way to look at it, it’s the less-than-anticipated return on an emotional investment.  Either way it sucks, but it has to be paid.

Scott Prather, @Scott_1420 over at ESPN1420 in Lafayette, wrote a nice piece on this subject.  Scott found inspiration in the great Alfred Lord Tennyson’s quote about it being better to have loved and lost than to never have loved at all.  That’s a good take on it, but mine is slightly different.

I once read a book, but I can’t remember the name of it or the author.  It wasn’t a classic, but it made an interesting point.  The story (and this has been many years ago that I read it) was about a sort of Sherlock Holmes type detective who did yoga.  He was an interesting character in that he didn’t try to lessen the blow when life punched him in the gut (as life is wont to do from time to time).  Instead, he relished the lows as much as the highs.  He said it was akin to riding a roller coaster.  You had to experience the highs and the lows to get the full effect.  Else it wasn’t a roller coaster.  It was something else.

Part of being a fan is experiencing the lows of losing as well as the highs of winning.  Without those lows the highs just don’t mean as much.  Just as a roller coaster with nothing but highs would be boring, a team that did nothing but win would be…  Okay, maybe the analogy falls short here.  I wouldn’t know what it was like to root for a team that *always* wins, but I think it really would get boring after a time.  Even Alabama fans get the occasional loss once in a while to keep things interesting.  It’s the threat of losing that makes the winning fun.

I’ll leave you with one more story.  There was a fisherman who died.  He knew he was dead.  He appeared in the afterlife in a boat with a rod and reel in his hand.  Bass fishing had always been a great love of his.  He eased up to a beautiful cypress tree and cast his lure right on side of it.  It’s one of those Boy Howdy lures, the kind that floats and has the little propellers on it.  Soon as it hit the water.  Pop!  He set the hook.  After fighting the fish for a while he reeled him in and it was big lunker bass, maybe a state record.  I’m in heaven!, he shouts.  The boat, of it’s own volition, eases him up to another tree, big beautiful cypress.  He casts his lure again.  Soon as it lands on the water, Pop!  He reels in another one, just as big as the first.  He grins from ear to ear.  This continues for a while.  Tree after tree.  Fish after fish.  Hour after hour.  Always a fish with every cast.  His smile fades as revelation dawns.  This isn’t heaven after all.


There is an old adage, “It’s hard to beat a good team 3 times.”  True or false?  The answer is yes.  The answer is no.  It’s yes and no.  It’s both.  It depends on *when* the question is asked.

If we ask the question a priori, before the season begins: “Will Team A beat Team B 3 times?”  The answer will be almost certainly no.  First of all, it’s rarely the case Team A will even get the opportunity because for it happen A has to beat B 2 times *and* A and B have to both make the playoffs *and* A and B have to meet in the playoffs.

But if we ask the question *after* A has already swept B in the regular season and after both teams make the playoffs and after the teams are scheduled to play each other a 3rd time, then the answer to the question becomes, “Yes, Team A probably will beat Team B 3 times.”

Conditional probabilities can be tricky.  Consider this conditional probability: Your friend has 3 children, and you know they are a mixture of boys and girls.  You don’t know how many boys, could be 1 or 2, but you know there’s at least 1 boy and 1 girl.  Just knowing this information (and assuming a 50/50 split of males and females in the general population ) what is the probability that the oldest child is a boy?

If you answered 50% you are correct.  Now, let’s add some more information (the conditional part).  Let’s say your friend shows you a picture of one of his children, a little girl, and he says she is his youngest child.  Now, armed with the knowledge of the gender of the youngest child being a girl, does this change the answer to the question of the gender of his oldest child?  Is it still 50/50 or has it changed with the new information?

You might say, no, it’s still 50/50, but you would be incorrect.  The probability (based on what we know at this point) is actually 2/3 or 66.7% that the oldest child is a boy.  Let’s consider the possibilities:

  1. Oldest is Boy, Middle is Boy.
  2. Oldest is Boy, Middle is Girl.
  3. Oldest is Girl, Middle is Boy.
  4. Oldest is Girl, Middle is Girl.

Well, 4 is invalid, so we need to strike that one out.  This is because we know he has at least one boy and at least one girl.  They can’t all 3 be girls.  So, the probability the oldest child is a boy is 2/3, or 66.7%.  If this seems counter-intuitive it’s because conditional probabilities can be tricky.

Consider the so-called Monte Hall puzzle.  Monte Hall was the host of Let’s Make A Deal back in the day.  Let’s suppose you are a contestant and you are offered the choice of 3 doors.  Behind one of the doors is a brand new car or some such great prize and behind the other 2 doors is a zonk, which amounts to no prize at all.  You select Door #1.  Monte says, well, it’s a good thing you didn’t take Door #2 because he reveals it to be a zonk.  Then he asks you, do you want to keep Door #1 or do you want to switch to Door #3?  What should you do?  Does it matter?

If you said it doesn’t matter, the odds are 50/50 whether you switch or keep Door #1, you are wrong.  The odds will favor switching to Door #3.  You will have a 2/3 chance of winning by switching and a 1/3 chance of winning by staying with Door #1.  Like I said, conditional probabilities are tricky.

The thing about this puzzle is we know he’s going to show us a zonk, and then ask us to switch doors, no matter what.  If you pick Door #1 and the car is behind Door #3 he’ll show you Door #2.  If the car is behind Door #2 he’ll show you Door #3 and ask you to switch.  If it’s behind Door #1 he’ll show you either Door #2 or Door #3, doesn’t matter, and then ask if you want to switch.  He’s not trying to trick you, it’s just how the game works.

The car is randomly behind one of the Doors.  Your odds of winning are 1 in 3 when you select Door #1.  If you stick with Door #1 your odds remain 1/3 because he can (and will) always be able to show you a loser door.  When he showed you Door #2 was a zonk it gave you more information.  You’re still at 1/3 if you stay, but can improve to 2/3 by switching.

Best way to wrap your head around this puzzle is imagine there are 10 doors.  You select one at random, say Door #1 again.  Your odds are 1/10 of winning the car (assuming all the other doors are all zonks).  Monte then proceeds to reveal 8 of the other doors, and then asks if you want to switch doors.  (All of this presumes the game is on the up-and-up.)  Should you switch?  Your odds were 1/10 before, and if you keep Door #1 they remain 1/10, but by switching you improve your odds of winning to whatever the opposite of 1/10 would be (9/10).  Imagine there is a thousand doors, you pick Door #362, and then he reveals 998 doors, leaving #362 and #787 unrevealed.  Should you switch to #787 or stick with #362?

The answer, which I hope you will agree upon is, you should switch to #787, the only other unrevealed door.  If you keep #362, your odds would be 1/1000 of winning the car, but by switching your odds of winning become 999/1000.  Similarly, with the 3 door puzzle you should also switch to Door #3 to improve your odds from 1/3 to 2/3.

So, what does all this have to do with a team completing a clean sweep in the NFL?  The difference is the new information we have once we learn Team A has already defeated Team B 2 times.  Without any other information (which team is better, or that both teams are playoff caliber) the preseason prediction would be it’s very unlikely to happen that A would beat B 3 times that year.  But once we learn A and B are playoff teams and A will play B a 3rd time and A already beat B twice, it changes the probabilities in favor of Team A completing the clean sweep.  History tells us it will happen about 2/3 of the time when A has that opportunity to do it.




These are all regular season numbers from  The number in each cell is the team’s ranking (1-32) in that stat.  The right column (NO NET ADV.) gives the overall net advantage for the Saints in the rankings matchup for this stat (negative number implies a disadvantage).  The Vikings are a very, very solid team with few weaknesses.  They are top 10 in 9 / 10 of these statistical categories on offense and 6 / 10 on defense.  But Saints are also top 10 in 9 / 10 on offense.  It’s on defense where the Saints struggle in the comparisons here, top 10 in only 3 / 10 categories.

EXP 9 17 2 3 -7
Scoring % 8 17 2 3 -8
Turn over % 3 9 8 22 8
points /drive 8 15 2 2 -7
passer rating 4 7 1 3 -1
Sack % 8 8 2 17 15
rush yds/att 22 27 1 5 -1
rush yds/gm 7 16 5 2 -12
punt yds/ret 8 27 25 22 -22
kick yds/ret 3 29 6 25 -7

The chart below makes it easy to see where the advantages are for the Vikings.  Passer ratings are very close as are rushing yards / attempt.  The 2 areas where the Saints have the edge are in sack % and turnover %.

nfl no vs min graph1

In the chart below the shorter bars mean the better (lower number) ranking.  Blue and Yellow are for the offenses, red and green for the defenses.  The long red bars are areas of relative weakness for the Saints defense.  Even though the defense is vastly improved from recent seasons it’s still just basically an average defense whereas the Vikings have one of the best defenses in the league.

nfl no vs min graph2

Based on this analysis the Vikings should win this game, especially since they’re at home.  The two areas where the Saints have advantages are in turnovers and sacks.  Saints absolutely need to win the turnover battle, which is, fortunately, one of their advantages.  They’re also likely going to need Drew Brees to be Drew Brees again, like he was against the Panthers.  Luckily, the Vikings foolishly built an indoor stadium, so the sub-zero weather conditions won’t be a factor.  Saints will need to win a shootout in this one.

Using‘s draft index we can compare drafts using their Career AV stat.  AV is a number PFR gives every player as a means of comparing production among different players at different positions, a tricky proposition to say the least.  Who had the better year, Ryan Ramczyk (OT) or Juju Smith-Schuster (WR)?  That’s what the AV number attempts to answer.  (Answer is: according to the AV stat, both scored 10.)

These were top scores in AV stat for 2017 rookies:

Player Pos CarAV
Alvin Kamara RB 16
Kareem Hunt RB 15
Ryan Ramczyk T 10
JuJu Smith-Schuster WR 10
Budda Baker S 9
Christian McCaffrey RB 9
Cooper Kupp WR 9

Here are the team sums:

NOR 46
SFO 28
JAX 23
LAR 23
CLE 22
DET 22
HOU 22
CHI 21
MIN 21
CIN 20
PIT 20
BUF 19
KAN 19
CAR 18
NYG 18
NYJ 18
WAS 18
SEA 16
TAM 16
ARI 15
GNB 15
TEN 14
DAL 12
IND 12
LAC 12
MIA 12
OAK 12
PHI 11
BAL 10

Here’s a handy dandy chart:

nfl 2017 draft avs

Saints, 49ers, Jaguars, Rams, and Browns led the way with getting value out of the 2017 draft.  Saints, Jaguars, and Rams all had big improvements, and the 49ers had a late-season surge.  Browns, Lions, and Texans did well in the list, but for various reasons it didn’t translate into improved performance in the win/loss columns.


There were three injustices perpetrated upon college football fandom this season: 1) Big 10 champion did not make the playoffs, 2) Pac-12 champion did not make the playoffs, and 3) an undefeated team (UCF) did not make the playoffs.

Don’t misunderstand me.  I’m not saying it was an injustice that Alabama got it, I’m saying it was an injustice the other 3 teams got left out.  Let’s go to 8 teams so we can get all power 5 champs in.  Pac-12 and Big 10 fans should already be with me, but so should SEC, ACC, and Big 12 fans, because next year it might be your champion left behind.  No champions left behind.

UCF didn’t belong?  They beat Auburn.  Look, I’ll go along with they didn’t belong in a 4-team field because of their schedule, but they would belong in an 8-team field.  What if they really are the best team?  What if 20 years from now there are 5 players from this team in the pro football hall of fame?  Never know.  Could happen.  Who knew Tom Brady was gonna be that good?  He lasted until the 6th round.

Some arguments, and my counters and suggestions for workarounds:

Too many games

They already play 12 games, plus conference championship games, plus with 8 teams in the field, that would make (for some teams) 3 more games.  Georgia will have played 12 regular season + 1 conference championship, plus now 2 playoff games, for a total of 15 games.  That’s almost an NFL 16-game schedule for college kids that need to at least pretend to be getting an education.  If we go to 8 teams you will have some teams playing 16 games.  That is too many.  I agree.  (But consider this: not all 130 teams are affected, only a handful of teams will actually play that many games.)

Here are a few ideas for how to fix the too-many-games problem: 1) go back to playing 11 games, 2) eliminate conference championship games, 3) use flexible scheduling to eliminate 1 game from the schedule for teams that are in the conference championship games.  Can’t go back to 11, too much lost revenue.  Can’t eliminate conference championship games for the same reason.  Enter my idea: flexible scheduling.

Flexible scheduling

All of the power 5 teams (far as I know) schedule these so-called rent-a-win games.  These are the games where a big, rich school pays some small, poor school a million smackers to come play.  The big school (95+% of the time) gets a win, and the small school gets to fund its athletic programs (including the money losers, which most college sports are) for another year.  The small schools *need* those games just to make ends meet.  And the big school, of course, gets to buy itself a win (but caveat, see Troy vs LSU, App State vs Michigan, UL vs Texas A&M, etc.).

My proposal is, let’s schedule these rent-a-wins during conference championship week.  If you make the conference championship game, the conference pays off the rent-a-win schools (so they each get their million smackers) and they play each other.  Everybody wins.  Too-many-games issue resolved.  As an example, this year Georgia played App State and Samford,  while Auburn played Mercer, Georgia Southern, and UL Monroe, all clearly rent-a-win games.  Could have scheduled, say Mercer and Samford, during conference championship week.  The games would be canceled, but the teams would get paid and one of them gets a bonus home game.  Everybody wins.  Smiles, smiles, everyone!  I feel like Ricardo Montalban of Fantasy Island.  Where’s Tattoo, and where’s that damned plane?  Should have been here by now.

Bowl tradition

One of the things lost with a playoff system (even the BS BCS) was the traditional matchups had to be tossed.  We could get back to having the Pac-12 champion play the Big 10 champion in the Rose Bowl, the SEC champion in the Sugar Bowl, the Big 12 champion in the Orange Bowl.  As they should all be.  The Fiesta Bowl can be the other 2 teams.  These could be your first round games.  Alternatively, bring in 3 more bowls and let the major bowls alternate the championship and semi-finals, or just play those games the following weeks at those same sites.

There will still be arguments anyway about who was left out

There will always be arguments about whether this team or that team got snubbed, but this time the arguments are about the 8th and 9th teams rather than about the 4th and 5th teams.  The there-will-still-be-arguments argument is without merit.

Only the best should be in, those other teams don’t belong

Do we really know who the best teams are until they play?  Did you know UCF was going to beat Auburn?  If you did you should have flown to Las Vegas because Auburn was a 15-point favorite.  Would UCF have a shot against Clemson, Oklahoma, or Alabama?  My guess they’d be 15-point underdogs again.  My point is we don’t know and cannot know who the best teams are, but what we can know is which teams are most deserving, and by putting the 8 most deserving teams in we have a much better shot at crowning the best team champion.  And, look, crowning a champion is not about finding the best team.  It’s about determining a champion.

If you give these people 8 teams next they’ll want 16

I don’t want 16.  That’s too many.  Maybe I’ll change my mind, but as of now (at least for me) this is not a stepping stone to eventually expand even further.  Let’s go to 8 teams.  I’ll settle for 6, but I think 8 would be better.  With 6 teams you could still get the 5 power champs in, plus 1 at-large, and with 6 teams you’d need to have 1st round byes, which would help against the too-many-games argument.  Either way, let’s expand.  At the risk of two too many classic tv references, eight is enough, but four is not.

It might seem like something akin to blasphemy to even suggest pulling for the hated rival Falcons, but bear with me.  The only chance for the Saints to host 2 games in the playoffs is if both the Saints and the Falcons advance to the NFC Championship Game.  But it’s not as cut and dry as that because (in my view) the most difficult leg on the journey to the Superbowl is if the Saints have to go to MIN, which would be required with the ATL-beats-LAR scenario.

Potential paths to the Superbowl (presuming a win over CAR):

@MIN → (home to) ATL

Which is the easiest path? Here are the win probabilities per matchup:

@LAR 0.4
@MIN 0.35
@PHI 0.5
(home to) ATL 0.6

The above win probabilities are just numbers I pulled out of my, er, hat.  You might disagree on them.  Just as a brief explanation,  I think the Saints have a 40% chance of going to LAR and coming out with a win, 35% chance @MIN, 50% @PHI, and a 60% chance of beating ATL in the Superdome.  Again, those are just my numbers I came up with on the spur of the moment.

Using the above numbers I get the following result:

(Presuming win over Panthers)
Individual matchups for Saints
Game Win probability
@LAR 0.4
@MIN 0.35
@PHI 0.5
(home to) ATL 0.6
Paths to get to Superbowl
Path Win Probability
@PHI → @MIN 0.175
@PHI → @LAR 0.2
@MIN → @PHI 0.175
@MIN → ATL 0.21

Unfortunately, wordpress doesn’t allow javascript on this site or else I could put together a little tool for adjusting those individual matchup numbers and have it automatically update the results.  If you want the spreadsheet file, here’s a link to it in Open Document Spreadsheet (.ods) format.  It’s for Open Office, but Excel might be able to open it.  If you can’t open it, leave a comment and I’ll send it to you in Excel format.

To come up with your own path probabilities you would need to download the spreadsheet and change the win probability numbers for the individual matchups at the top of the spreadsheet, whereupon the path probabilities at the bottom would get updated.  As it stands using my numbers above the @MIN -> (home to) ATL path is slightly easier than the @PHI -> @LAR path.

%d bloggers like this: