Preseason Projections 2019

Updated 14-Jul-2019; 16:00 ET

It's finally that time of year, Browns fans. Football is so close we can practically taste the nachos, buffalo dip and beverages of choice that welcome us warmly to Sunday afternoons in the Fall.

We will line up down the block under the merciless July sun to watch our Browns practice during training camp, which - having been packed with fans even during the lean years - will be one of the hottest tickets in Cleveland later this month. Football TV like "All or Nothing", "Hard Knocks" and "Last Chance U" will release and whet our appetites even more for football season. EA Sports newest rendition of Madden will release in August and I know I'm not alone in saying that I can't wait to see what this Browns team looks like in that game when it's finally available to the public. Through all of these little appetizers to football season, we'll all be left to wonder just what this team can be in 2019 until that first preseason game kicks off on August 8th, when we'll finally get to see some semblance of our beloved Browns take the field against Washington.

Expectations for the Browns this year are high and I'll go on record now to say that this team (on paper and in my daydreams) can absolutely contend for a division championship, host at least one playoff game and win the AFC. But what level of expectations are realistic for the Browns this year? To answer that question, I'll walk you through my attempt to project the 2019 NFL season by using a combination of 2018 statistics with adjustments made to account for season over season player additions and departures for each NFL team.

Previous Preseason Projection Results

The last time I did preseason projections was ahead of the 2017 season (my wife and I had our first son at this time last year, precluding me from completing a 2018 version). I wish I could claim to have predicted the Browns' 0-16 campaign that year, but a projection of 3-13 for that season gave me a good idea that I wasn't in for a fun season watching Browns football. This hunch was confirmed when our very first offensive series ended with a blocked punt for TD given up to Pittsburgh in week 1.

There are some hits and some big misses in those projections for sure, but preseason projections will always be challenging for a few very practical reasons like injuries and late-summer suspensions (Zeke Elliot missing 10 games in 2017 didn't help my SB pick of Dallas very much) not to mention the fact that they're made without any actual data for how teams play in their updated configurations. Being off by around +/- 3 wins per team on average and correctly identifying about 58% of the playoff teams in that year is a fairly decent jumping -off point in my personal opinion.

Preseason Projections 2019: Methodology

For the most part, I used the same methodology for 2019's projections to produce a mathematical estimate of how many wins each team will end up with as was done in 2017. Essentially, the approach was to identify statistics that correlate closely to predicting NFL wins, to project those statistics into 2019 from 2018 data, and then to plug those projected values into a statistical model whose output is ultimately a game-by-game win probability for each team given their own projected 2019 stats and accounting for their opponents' similarly projected statistics. For this season's effort, however, I wanted to account for the fact that these year-over-year statistical predictions can change substantially given a key player's addition or subtraction from a given roster.

To accomplish this, I used overall ratings from EA Sports's Madden gaming franchise to approximate player value and referenced ourlads.com's 2019 depth chart projections to build a somewhat realistic view of what each NFL team's 2019 roster will look like at each position according to Madden 19 ratings. I then compared these depth charts versus each team's previous depth chart from the previous season to identify where overall Madden rating points were either added or subtracted from a given team at any given position. Using models created from correlating overall Madden ratings to real-world NFL statistics, I was able to calculate the value added to projected 2019 NFL statistics by a particular team's adding of, for instance, an elite WR to an an already strong group returning from the 2018 season. After adjusting for player movement between teams (and approximating ratings for rookie additions), I ran these team-level statistics through my win-probability model to come up with the results that follow below.

Preseason Projections 2019: Results

Projected AFC Standings:

According to the brownalytics model, much like in 2007 Cleveland puzzles at the idea that a strong 10-6 record finds them squarely on the playoff bubble. Paired with projected losses to both Pittsburgh and Tennessee, the Browns hang on to the 2nd Wild Card slot, but just barely. A wild card weekend match up with New England is all the more challenging to stomach given the knowledge that finding just one more win would have resulted in a bye and home field advantage throughout the playoffs if the standings were to play out as our model suggests.

Projected NFC Standings:

In the NFC, familiar faces Philadelphia, Chicago and Seattle are back in the playoffs according to the brownalytics model. Carolina rebounds from a couple frustrating seasons in a big way, easily winning the NFC South. As a side note, in the several iterations of these predictions that I produced throughout my development process, one team was consistently predicted to have a strong year: Carolina, based on their having a solid defense and an offense that moves the chains well without relying too heavily on the run. New Orleans is projected to regress after a very strong 2018, but ultimately hangs on to a playoff spot by eking out a tiebreaker for the second wild card spot over a Green Bay team that continues its recent run of mediocrity.

Projected Wild Card Bracket:

Our Browns, having not played a playoff game since the 2002-2003 season, head to New England to attempt to dethrone the defending champs. Winning in New England in January would be a tall order for any NFL team, and the young, relatively untested Browns come away empty handed in what ends up being the second-closest result of the weekend. Brownalytics's statistical projections envision a game where Cleveland struggles to to keep up with a steady Patriots offense in a hostile environment with most of this young team's roster getting their first taste of NFL playoff action. Key turnovers exacerbate the situation for Cleveland as the Browns are unable to get their defense off the field efficiently enough throughout the day and the offense is unable to capitalize when they do. Ultimately, it's a disappointing early end to a promising season where big additions had everyone around the team frothing at the mouth for a deep playoff run as far back as March 2019. But it takes time to turn talent into a true contender in the NFL, and these projected Browns would leave Foxboro having been competitive with a team whose success at staying competitive for more than a decade they'll hope to emulate with the young players who form the foundation of their 2019 roster.

Projected Playoff Bracket:

All that's left to root for in a Browns-less playoff bracket is for the Steelers to stumble. Hang on tight Browns fans, this is the part of the projections where things get bleak.

Brownalytics's projections see the Steelers beating Tennessee in the Divisional round prior to winning a very close conference championship game against Houston enroute to a berth in the annual "Supper Bowl". Of course, this means that we'll all be subjected to hearing all about Big Ben "riding off into the sunset" after potentially winning another ring for the two agonizing weeks that we're forced to wait for the big game after the conference champions are decided. We'll see all-time hated Steelers like Jerome Bettis and Hines Ward yukking it up with hosts on the myriad of big pregame hypecasts, telling fluffy stories about how they "...knew all along that Ben would be a special quarterback, but that's because he's an even better man". Peee-yuke!

But after all the hoopla, Cleveland-bred Mitchell Trubisky and his Chicago Bears come to Browns' fans' rescue when we all need it most. Chicago upends Pittsburgh's run in the big game (hopefully, poetically, on a last second field goal). They silence the narrative of Big Ben potentially heading into retirement like Elway and Manning did, as champions, and focus us all on the question of whether this is the last deep run that Pittsburgh's current roster has in them.

Projected Draft Order:

In the scenario described above, the Browns would pick 24th in the 2020 NFL Draft, and would likely look to bolster the Offensive Line with one of several premier starting prospects that are expected to be available at that point of the first round in way-too-early mocks published through July of 2019.

Preseason Projections 2019: Assumptions and Outstanding Issues

One of my favorite quotes is often brought up in an effort to forecast a future scenario; the quote goes as follows:

"Essentially all models are wrong, but some are useful."

- George E.P. Box, British Statistician

What is important about this quote is that we recognize, as Box would, that no prediction is likely to be perfect. As such, the brownalytics methodology and the results described above are based on many assumptions and do raise eyebrows with a few of the projections presented. In this section, I'll attempt to address a few of those.

1. "Browns not 16-0! What gives?"

The 2019 Browns are set to field their most talented roster since 1999 and, possibly, their most talented roster since the 1960's. With that said, it's wise to remember Freddie Kitchens's gold plated quote about expectations for the Browns going into the season: "Right now, we are just a bunch of good individual players. Yeah, our roster looks great on paper -- whoopty-hell, alright?". So much of this season's promise rests on Baker Mayfield's continued development into year two and this is an input that is not distinctly factored-in to any of brownalytics's projections. Likewise, year-over-year projections for Cleveland (as a team) in our models are weighed down by the ineffectiveness of Hue Jackson and Todd Hailey's offense in the first half of 2018.

Baker's Areas for Improvement in 2019 by EPA

Comparing Hue's and Freddie's offenses in 2018

On that note, coaching changes are also not distinctly factored into our mathematical projections for the Browns, so the addition of "pass-happy" Todd Monken's offensive tendencies and the departure of the "bend but don't break", c-gap hunting, defense of Gregg Williams are not accounted for. It's for a combination of these reasons that the brownalytics model is lower on the team's offensive outlook than most fans and the media are. Defensively, our projections have the Browns being very strong at limiting points, which evidences a projection that is heavily influenced by the aforementioned "bend-don't-break" style of the previous regime. Again, subjectively speaking as a fan, I see some issues with these numbers and as an analytics practitioner I see room for my own improvement of the models that generated them, but it's worth mentioning that even given these lower-than-expected results, the Browns are projected to net ten wins in the alternate reality described above. Ten wins and a playoff berth, no matter how we get there, would be a massive step in the right direction for this team, franchise and fan base.

2. "Pittsburgh to win the AFCCG? LMAO!"

Brownalytics's year-over-year statistical projections are very high on the Steelers' offense in 2019. Subjectively speaking, this seems a bit optimistic to me, but the Steelers have been one of the most consistently competitive AFC teams for what seems like forever and everyone who has predicted the demise of Big Ben has been wrong so far. Arguably, were it not for the Belicheck and Brady connection being so prolific in New England, we Browns fans would probably be subjected to a lot more grief from Steeler fans than we are today. The loss of Antonio Brown will certainly affect them, but if any team has shown the ability to 'reload' and compete after the loss of star players, it's Pittsburgh. The Steelers are a great NFL franchise and a very worthy rival. Is my model a bit too high on them in 2019? I think it is. But anyone who believes that this team is going to roll over and relinquish the division to Cleveland may have a surprise coming in 2019.

3. "What happened to Kansas City, New Orleans and the LA Rams?"

Brownalytics's modeling projects overall record regression due to porous defenses in KC, NO and LA despite these teams having strong offensive outlooks going into 2019. By almost any measure, KC, NO and LA lit up the league with their extremely potent offenses in 2018; so much so that regression from last year's heights to some extent is to be expected. But, somewhat surprisingly, our models remain down on these teams' defensive moves made over the course of the 2019 offseason and this achilles heel is what mostly drives predictions around middling records for each of these three teams in the upcoming season.

Conclusion:

It was probably my late Grandpa who told me at one point of my (very opinionated) youth that "Opinions are like [butt] holes, everybody has one" in the midst of one of our discussions.

Essentially all that is captured in the projections above is an opinion about what might happen in the 2019 NFL season. Yes, it's based on mathematical projections, some very complicated logic, and reams of data, but those facts don't mean that there isn't right and wrong in this work. In this way, these projections are just like in the other projections that are sure to flood our timelines over the next few weeks as training camps open and "Football SZN" starts to pick up steam. Preseason projections, as touched on above, are a challenging thing to pin down with any real confidence for a number of very practical reasons. Since 2018, teams have changed coaches, they have signed and released players, they have installed entirely new playbooks and have had time in detailed film study to digest the strategies that either made themselves or their opponents successful last year. Players have gotten stronger (and weaker), they have worked on weaknesses and studied their own tendencies in an effort to improve their own individual games and some, doubted their whole lives, are ready to let that chip on their shoulder lead them to proving the whole league wrong about what their "ceiling" is. Projecting wins and losses is fun to do and to read about in the summer, but the game is ever-changing because the people who play and coach it are constantly trying to be better than they were before; to break the scouting reports and defy the analytics that seek to define their potential. Until we get some real data from the actual 2019 season that directly describes how a player and/or team has performed relative to their competition, it's all just an estimate built on a series of other estimates; an opinion, and everybody's got one.