The Challenges of Pigskin Simulation in a Bull Market

With Indianapolis squeaking out a 45-44 victory over Kansas City last weekend, it’s hard not to notice all the offense in the NFL these days. The two teams came within a touchdown of setting the record for combined points scored in the playoffs. And they set a combined record for yardage.

Overall this season, teams are combining for a shade under 47 points per game. That’s a full point above last year’s NFL record. We’re now almost a touchdown over the 41-42 averages we saw when I first started work on Front Office Football. That’s 10-15 percent more offense.

If you track points scored going back to the beginning of when football started to resemble the modern game (I’d call it 1943, when helmets became mandatory, and you started to see free substitution), the NFL has always regulated to scoring in the 40-43 point range. If scoring drops, you see “rules emphasis” on contact from defenders. If scoring increases, the referees allow a little more.

The last four years have been the exception. In 2010, we hit 44 points for the first time since Bryan Bartlett Starr won back-to-back Super Bowl MVPs. Now we’re at 47 – by far the most in NFL history.

This is a tremendous challenge for those of us who like to simulate sports. If I increase offense to reflect this explosion, records are going to fall like dominoes. Early testing on the league-normalized Front Office Football Seven saw a world where you really wouldn’t recognize offensive records. Particularly with the running game, because variance with yards per carry is much higher these days. A workhorse, in theory, could gain 2,500 yards these days with the right offense.

So I decided to move offense back 5-7%. I did this mainly because most casual players of Front Office Football would be put off if they see someone throw for 6,000 yards. This could easily happen in the NFL with what we’re seeing today. For instance, nine of the 17 quarterbacks who have thrown for more than 36 touchdowns in a season in NFL history did this in the last four years. I’m operating under the assumption that someone in NFL headquarters is troubled by records losing their meaning. That when we reach a point where scoring is so easy, the game is out of balance.

If this continues, of course, I will have to make adjustments and hope for the best. I think we need that 2,500-yard runner in real life to get the casual gamer on board with this, however. This is certainly my biggest challenge when it comes to keeping the Front Office Football simulation engine both relevant and accurate.

10 Responses to “The Challenges of Pigskin Simulation in a Bull Market”

  1. Alan Jobb Says:

    In the real NFL I think we will see a 6,000 yard QB long before we see a 2,500 yard back. The 6,000 yards can be spread over lots of receivers, but for one back to take 2,500 yards worth of punishment and stay healthy seems unlikely to me.

    As for the game, I will say I am perfectly fine with the way the game is modeled now, but love the idea of it being more dynamic.

    IMO, the best case scenario is OOTP style options to adjust the way the game is played during certain eras. However I understand that might be a lot easier to model in baseball then it is in football.

    The next best case scenario is for the baseline to fluctuate with perhaps clues such as an email from the league “The NFL rules committee this off-season implemented rules that are expected to reduce the effectiveness of offense” We don’t need further detail, just the general idea.

  2. Jim Gindin Says:

    I’ve thought about adding that type of feature to the game. I don’t think the casual gamer appreciates just how different passing is these days. There’s a general understanding that teams pass more and more effectively than they did in the 1970s. But the changes in the last four years, which hit running almost as much as they do passing, would cause many to dismiss the game. I saw plenty of that even with moving toward the 2009 model with FOF7. If I offered a “2013″ model and people selected it and saw the occasional 6,000 or higher season, they’d stop taking the game seriously.

    Durability in running backs is what prevents 2,500 today, but I wouldn’t dismiss it entirely. The numbers I’m seeing suggest that some of what Baylor/Oregon are doing is somewhat possible at the next level.

  3. Alan Jobb Says:

    But is your game one that appeals to “the casual gamer” in the first place?

    my teenage son plays a billion hours of Madden, but looks at me like I am crazy when I try to show him FOF or OOTP. Even worse when I show him the strat o matic cards from my youth.

    I believe your demographic is far more in tune with football statistics in general and the changes of the game across the years. We are an NFL nerdy bunch.

    just something to think about.

  4. Jim Gindin Says:

    I’ve thought about it a lot. I don’t think anything generates as angry an email as I receive when customers perceive that records are being broken too frequently. They post these records online whenever someone discusses the game. It really upsets them. Maybe it’s a minority of my customer base, but it has a large effect on perceptions.

    We’ve had about 40 years of modern NFL, with a passing game that we’d recognize as a passing game. That’s tiny when you look at the vast statistical universe we can create through simulation. From a business perspective, and perhaps from a game-play perspective as well, I think it’s best to treat the last four years – especially 2013 – as an aberration that the NFL hopefully will take steps to fix.

    Or perhaps it’s close to the top of what the NFL can produce with its rule set. Rather like this stretch of record-breaking snowfall followed by record cold is as close to the most extreme weather lower Michigan can produce.

    Either way, I have given it a lot of thought and think it would be a terrible decision to normalize FOF around 2013 numbers.

  5. Chris Cockerille Says:

    I have thought about this also. The nfl is becoming more unrealistic and less entertaining because of this imbalance. The fact is that with a simulation you can keep things balanced which is more important than matching a game where defense doesn’t matter any more. I agree with you Jim the average points and yards should be kept pre 2010. I would rather play fof 2007 than have todays nfl numbers in FoF7.

  6. Ace Says:

    When I play a sports simulation, I want it to simulate the real-life product. Frankly, I was going to upgrade to FOF7, but when I read this admission that it does not accurately portray today’s NFL offensive numbers, I passed.

    An easy solution to this is to implement era modifiers like OOTP does. If this happens, then you will have my money :)

  7. Neel Says:

    As you said Jim “Either way, I have given it a lot of thought and think it would be a terrible decision to normalize FOF around 2013 numbers”…

    I agree. To me the playoff game between the Colts and Chiefs was a bit boring because you just had to wait and see how quickly someone would score. No defense.

    Yet the 49′ers vs Packers game was great. Every first down mattered. But I grew up watching 1970′s era football and was a CB, so maybe that clouds my vision.

  8. Jim Gindin Says:

    I’d argue, Ace, that it is accurate. In the past, I had centered numbers on the past season. Since the NFL was relatively consistent, this changed a little over time.

    But the last four years, particularly 2013, have been completely out of line with any curve depicting NFL statistics over time. I can treat that in one of two ways.

    1) I can react and recenter on 2013.

    2) I can accept (and there’s reams of data on all sports suggesting this is a better approach) that 2010-13 are part of a temporary aberration and will be assimilated into the long-term NFL statistical model.

    In other words, it’s like looking at a 95-degree heat wave in July, seeing the average high for July is 85, and 1) reacting by saying the average high temperature for July is 95 or 2) reacting by turning on the air conditioner and saying, “it’ll probably be cooler next week.”

    I think it would be a poor decision to go with model 1 because then the heat wave in July would be 108-110.

    It’s also difficult to compare baseball simulation to football simulation. What might be easy in one is far from easy in another. Thousands of constants control simulation. Era modifiers might be possible, but they would be incredibly difficult.

    Plus, the default setting would logically be 2013. Which produces those 108-110 degree days that would turn off 100 potential buyers for everyone who is turned off by my centering on 2009.

  9. Ace Says:

    That’s a fair point, Jim. I guess I firmly believe that today’s high outputs on offense are here to stay because of the rule changes to the game that favor offense and appear to be permanent.

    But who knows? Perhaps we’re due for some great defensive minds to step in and slow down the wave of yardage and scoring.

    I don’t want to pick at FOF7 because, like I said, I haven’t tried it yet. You know you’re customers better than I do, but I can’t imagine I’m the only one who read your post about dialing back offense in the game and decided not to buy.

  10. Jim Gindin Says:

    My hope is that for everyone who is turned off by that post, many more appreciate my explanation and agree that it’s better not to set the center at the absolute historic peak of NFL offense.

    I do know for certain that even at 2009 levels, higher offense is possible.

Bad Behavior has blocked 75 access attempts in the last 7 days.