NFL Preseason -- Does It Matter?

Pre-season games, BAH! Who cares right?

Well USA Today posted an article titled "The NFL Preseason Matters More Than You Think." In this era of Big Data and Advanced Analytics the data is out there for any and all to see and use. The question mark comes from the "Analytics" part. What data is relevant and how do you use it?

The USA Today article title is misleading. In my opinion, the data shows that preseason records are totally irrelevant.  At first they had me going with a chart that showed that 75% of the teams that finish with 2 or more wins made the playoffs. That's a pretty significant percentage. Sure there's that 1 in 4 team that makes the playoffs with 1 win or less, but 3 out of 4 is a lot better odds.

Then they showed the 2nd Chart. Basically it shows that 75% of the teams (data collected over 5 years) finished with 2 or more wins. Of those teams, 40% made the playoffs. Of the 0 win preseason teams, you guessed it, 40% of the teams made the playoffs. There were just a lot less of them, which accounts for the 25/75 split in the first chart. The 1-win teams made it only 28% of the time, but even the article admits that it's probably just an outlier and it would probably move more towards 40% with a bigger sampling size.

Really, the preseason records don't matter. What a Shocker huh?

Just before the NFL draft I read an article about analytics in regards to drafting players (I would link it if I could find it). They interviewed several organizations and basically what it came down to was using analytics to back up your findings. It's when the numbers don't match what your findings tell you that you have to reevaluate. They gave an example of how every successful edge rusher in the NFL has a 10-time (the time a player takes to go 10 yards from a three-point stance) under a given time that I can't remember. They zeroed in on two unnamed players. One had the quickest 10-time ever recorded. He had top 5 totals in sacks and pressures in college, but scouts questioned his strength and technique. The other player also had top 5 totals in sacks and pressures. Great work ethic, great technique, overpowering at times...but a 10 time that was slower than any successful edge rusher in the NFL today.

Both were drafted (I don't remember how long ago). Where are they today? Both are out of the league. Neither made it. The Analytics lesson? If there's a data point that matters don't ignore it. But it's also not the only thing. One rep for an organization that wasn't named even said that they've been doing Advance Analytics for decades, they just didn't call it that and that the data available now is so much more that you need people just to make sense of it all.

So back to preseason games. Players are playing that won't be around when the season starts. Established starters may play 3-4 quarters out of the 4 games...some not at all. Most coaches are running vanilla game plans. They are not trying to create mismatches, on the contrary, they are trying to put players in certain match ups so they can evaluate them. There are several players on the field still learning the offense/defense so the pace is slower. Put those findings altogether and reasonable thinking is that the preseason records don't matter....and the data backs it up.