What is EP in Football? An Expected Points 101 Crash Course for Rookie Fans

Sup football geeks! Terry from 33rdsquare here. Today I‘m breaking down one of my favorite next-gen stats that‘s revolutionizing how we analyze the game – Expected Points.

Even if you‘re a total rookie when it comes to analytics, keep reading and soon you‘ll be an EPA expert equipped to settle any analytics argument at your next watch party. Let‘s dive in!

EP and EPA – The Next Big Stats on the Block

Imagine it‘s 4th quarter of the Super Bowl, down 3 with a minute left. Your team just ripped off a 40 yard pass down to the 30 yard line. How much did that improve your chances of scoring the game tying touchdown?

Your gut says a lot. But ESPN‘s fancy new expected points model can quantify it.

Expected Points (EP) uses historical data to estimate how many points an offense is likely to score from a given field position, down and distance. Before that 40 yard pass, from your own 30 yard line, you had an EP of around 0.6. After the pass, now setup at the opponent‘s 30, your EP jumps to around 3.5.

That single play increased your expected points by nearly 3 – meaning it tripled your odds of tying the game! Pretty clutch if you ask me.

Expected Points Added (EPA) measures this difference in EP between each play. It captures how plays directly swing scoring chances in a way that vanilla stats like yards and touchdowns don‘t.

So next time your buddy yells "What a play!" you can quantify its greatness by shouting the EPA. He‘ll be so impressed with your analytics skills that you‘ll forget you‘re both actually accountants.

From Obscure Academics to Football Mainstream

Back in my college days writing papers on optimal 4th down strategy, Expected Points was mostly an obscure stat you‘d find in engineering academia.

But over the past decade, the NFL analytics revolution has transformed EP and EPA from niche metrics into essential mainstream tools. Teams are using EPA more in their internal scouting and decision making. ESPN prominently features EPA stats in broadcasts. And fans now have apps that calculate EPA on every play.

The rise of Expected Points mirrors how analytics has permeated the game. We‘ve come a long way from football being jarheaded coaches spitting dip and bashing heads. Well actually many coaches still do look like that. But you get the point – football strategy today is more scientific than ever.

Yet hardcore analytics still butt heads with old school skeptics. So for EPA to go truly mainstream, it‘s on stat geeks like me to educate fans on what advanced metrics reveal that box scores don‘t.

Once fans grasp how EPA captures the context and difficulty of each play in a way that summary stats like yards and TDs cannot, they gain a whole new understanding of the game.

So let‘s dive into the nitty gritty details of how expected points work. Grab a pen and paper if you want to follow along, or just keep sipping that brew – whatever helps this click!

Expected Points Step-by-Step – From Raw Data to EPA

Alright here‘s a quick 101 on how the sausage of Expected Points is mathematically made:

Step 1) Record the down, distance and field position of EVERY play over 10+ seasons of play-by-play data. We‘re talking hundreds of thousands of data points here people.

Step 2) Tally up the ACTUAL points scored on the ensuing drives starting in those field positions. Calculate the league average points per drive for each down, yard line and distance.

For example, over recent seasons, NFL teams scored on average 2.3 points on drives that started with a 1st & 10 from their own 25 yard line.

That 2.3 points becomes the Expected Points for that situation. Do this calculation for every conceivable down, distance and field position on the 100 yard field.

Step 3) Using this EP model, you can measure EPA on any play.

Start with the Expected Points for the offense at the beginning of the play based on the down, distance and field position.

Then plug the ending field position into the model to get the Updated Expected Points.

EPA = Ending EP – Starting EP

That‘s it! EPA quantifies exactly how much each play impacted scoring chances. Now rinse and repeat for every play, every game, every season.

Why EPA is Superior to Traditional Stats

Alright, let‘s look at why Expected Points is such a valuable addition to the stat nerd tool belt:

Accounts for situational context: EPA gives you credit based on the difficulty of the scenario. A 5-yard scramble on 3rd & 15 is huge. On 1st down? Meh. Traditional stats ignore this crucial context.

Better correlates with scoring: Studies have found EPA has a stronger correlation to points and wins than stats like total yards. Because that‘s the name of the game baby!

Isolates individual impact: EPA distills a QB‘s direct effect on EP rather than being influenced by external factors like receiver YAC. No padding your stats (looking at you Kirk Cousins!)

Apples-to-apples comparison: EPA provides a common scale to judge any type of play. We can directly compare a 60-yard bomb to a 3rd down conversion since they‘ll have similar EPA.

Matches the eye test: The QBs with the gaudiest EPA totals usually match our subjective sense of who the best performers are. HELLO PATRICK MAHOMES!

For these reasons, EPA gives us football analysis in the crisp 1080p quality the stat geek desires compared to the fuzzy VHS tape of conventional stats.

EPA in Action – 2022‘s Top QBs

Let‘s see EPA in action by looking at 2022‘s leading MVP candidates:

Quarterback	Passer Rating	Yards/Game	EPA/Play
Patrick Mahomes	105.2	317	0.293
Jalen Hurts	101.5	265	0.242
Josh Allen	96.2	297	0.230

By traditional stats, Mahomes looks dominant while Hurts doesn‘t stand out as much. But their EPA reveals Jalen also had massive impact given his style of play and system. Hurts‘ dual threat skills made him almost as efficient per play as Mahomes in reality.

And Josh Allen compiled shiny yardage partly thanks to an ultra-aggressive passing approach. His EPA advantage over Hurts was much smaller than raw totals suggest.

This pass the eye test too. Ask any football fan and they‘ll say Mahomes and Hurts were the game‘s best QBs with Allen a small step behind. EPA aligns with these real perceptions.

Does EPA Correlate with Winning?

But does EPA translate to the one stat we actually care about – wins? Short answer: yes!

In 2021, the top 10 QBs in EPA/play included 9 playoff teams. Conversely, QBs with awful EPA like Zach Wilson anchored bottom feeder squads.

More rigourous analysis also shows EPA‘s strong predictive power. This study found EPA per drive was more correlated with future wins than points per drive. Moving the chains matters folks!

EPA‘s impact is clearest on pivotal plays. Patrick Mahomes generates the highest EPA per attempt on 3rd or 4th downs over the past 3 seasons. No wonder he‘s so clutch!

Advanced stats back the eye test. The battle tested quarterbacks with the "it" factor shine brightest in EPA in crunch time.

EPA vs. Other Advanced Stats

Expected Points isn‘t the only nerd stat out there. Let‘s see how it stacks up against some other heavy hitters.

PFF Grade – Measures efficiency on a per play basis like EPA. However, PFF remains proprietary and subjective. EPA provides more transparent quantification.

DYAR – Football Outsider‘s DYAR adjusts for situation like EPA but uses cumulative totals rather than per play. Also includes opponent adjustments.

QBR – ESPN‘s Total QBR factors in EPA and other stats like dividends, sacks and clutchness. More of an aggregate stat compared to EPA‘s micro-level focus.

Each metric has pros and cons. Combining EPA with PFF grades, QBR and DYAR gives the most complete performance profile. EPA alone does not tell the whole story – but it reveals key insights other stats miss.

EPA Caveats – No Stat is Perfect

EPA isn‘t a magic bullet though. Here are some factors it does not account for:

• Strength of opponent

• Supporting cast

• Garbage time

• Luck/randomness

EPA also requires larger sample sizes than traditional stats to stabilize – around 100 plays per player.

For these reasons, the wise football scholar sprinkles EPA into their analysis like a secret sauce instead of drowning food in it. EPA enhances traditional scouting – but should not fully replace what coaches and scouts see with their own eyes.

Wrapping Up Expected Points 101

And there you have it – a crash course on the game changing world of expected points! We covered:

• How EPA quantifies a play‘s impact on scoring chances

• Why EPA is a huge leap forward from box score stats

• How the EPA sausage is mathematically made

• Examples of EPA in action and its correlation with winning

• How EPA stacks up against other next-gen stats

• Caveats and limitations to keep in mind

Got an EPA question I didn‘t cover? Hit me up on Twitter @33rdsquare and I‘ll get you straight, young protégé!

Now if you‘ll excuse me, I have to prep for my fantasy draft by running EPA regression models to find next season‘s overlooked studs. Fantasy championships wait for no nerd!