Win. Your. League.

Receive 3 Free Downloads More Details

The Problem with Expert Mock Drafts

Expert Mocks, or Industry Mocks, can be fun and insightful tools... if you're careful not to fall for the wrong takeaways.

You've got a draft coming up, and you want to see what some of your most trusted fantasy football follows are doing this year, so you check out a quick mock draft. It's understandable, and also totally normal.

Every site does mock drafts, and this one here is no exception. In fact, our staff is doing 15 mocks this year for a variety of different league setups. They're some of our most popular features.

And our mocks don't just include picks, they include post-draft analysis so you can see not just who someone liked, but why. You can get a glimpse at what goes on in the heads of the Footballguys staff, what players they like, what strategies they employ.

Mock drafts are awesome. They're fun to do. They're fun to read about. They're useful for holding people accountable and putting their beliefs to the test. They're great.

They're also kind of problematic.

Industry Mocks Aren't Templates

This article is inspired by something I've known a while, but which I've seen brought up independently by three different long-time Footballguys staff members in the last week and a half. It seems like this idea is having a bit of a moment right now, and I think that's a good thing.

Here's Jason Wood raising the issue last week on Twitter.

Here are Matt Waldman and Clayton Gray bringing it up in our staff chat.

Quarterbacks go late in industry mocks. Really, really late. Here, let me give you a "for instance". In one of the mock drafts I participated in I grabbed Tom Brady as the second quarterback off the board. One of my post-draft questions was "Explain why taking a quarterback early on is a wise decision in a best ball league".

This is a good and fair question. Except the problem here is that I took Tom Brady with the 48th pick of the draft. His consensus ADP has him going with the 24th pick of the draft, on average. I got him a full two rounds later than usual.

So my answer reflected this. "And here I thought 4.12 was pretty late to still be able to nab a top-tier quarterback."

Now, I do not mean to call out the original question. It's a good question! For some reason, I was more willing than the other staff members to commit to a quarterback at the top of the draft. The questioner was trying to get me to expound on that reason.

The purpose of these questions is to get inside our heads, and this particular inquiry allowed me to talk about how frequently scores get counted in best ball leagues and the danger of positional runs when drafting from the turn.

Because at the end of the day, a draft is all relative. If everyone is grabbing quarterbacks in the 2nd and 3rd, then the 4th round is late. If everyone is grabbing quarterbacks in the 6th and 7th, then the 4th round is early.

Anyway, the problem with industry mocks isn't quarterbacks, per se, they're simply the most visible manifestation. Here's Ryan Hester hitting on another example.

Ryan's last comment is right. In a typical league, you're usually not going to see favorite sleepers going several rounds before their ADP. But this happens all the time in industry drafts.

Someone like Matt Waldman watches a lot of film. He has a very firm opinion on who he likes and who he doesn't like, and that opinion isn't influenced very much by outside sources, so sometimes that opinion is very different from consensus.

In fact, Matt Waldman has even written an excellent article this year devoted to blowing up ADP and taking guys he loves way before he's "supposed to". For instance, he advocates taking Marvin Jones Jr, whose consensus ADP puts him somewhere in the 12th round, as early as the 8th round.

These aren't just words on a page, either. Here's an industry draft Waldman participated in where he took Marvin Jones Jr with pick 8.03. While we're at it, look at the ADP of the four receivers taken from pick 8.01 to pick 9.01: Desean Jackson (ADP of pick 86, taken with pick 86), Marvin Jones Jr (ADP of pick 139, taken with pick 87), Rishard Matthews (ADP of pick 125, taken with pick 93), and Eric Decker (ADP of pick 98, taken with pick 97).

In a regular draft, Jackson and Matthews would be viewed as horrific reaches. In an industry draft, though, that kind of thing is just business as usual.

It's not just the "film guys" who have strong opinions, either. Someone like Danny Tuccitto is to numbers what Matt Waldman is to film. He calculates, for example, how many carries a player has to get before his sample size is large enough that his yards per carry represents 50% skill and 50% luck, and then he uses these calculations to find a player's "true" yards per carry.

Danny is likewise going to have players that he's way outside the consensus on. I've seen Danny take players dozens of picks early because he liked their "true yards per route run" or "true yards per carry" values.

Most people are not watching that kind of film or doing these kinds of analytics. This group includes nearly everyone outside of industry drafts, and even a fair few people inside industry drafts. (For instance, I very much count myself among this group.)

If you don't grade players yourself or build predictive models, then valuing players relies more on consensus building. You listen to the opinions of lots of other people that you trust, and then you merge them all. This results in a set of preferences that are far less extreme than what you'd see from a Matt Waldman or a Sigmund Bloom or a Danny Tuccitto.

The end result is that industry drafts have 3+ round reaches as a matter of course, while in more traditional drafts they're as rare as double rainbows.

The Underlying Problem with Industry Drafts

Again, falling quarterbacks and routine reaches aren't the "problem" with industry drafts, per se. They're just ways that the problem manifests itself. The underlying problem is one of expectations.

If you put an industry insider in a draft with 11 casual players, the industry insider will draft differently than he will in a draft with 11 other industry insiders. Matt Waldman might want Marvin Jones Jr, but the reason he's willing to reach is that he knows anyone else in the draft might also want Jones and be willing to reach for him.

If you're playing with guys who can go "off script" in a heartbeat, you have to be much more aggressive to get the guys you want. But in a more typical league, a 1-2 round reach is plenty to ensure you still get your player of choice.

Similarly, I'm willing to wait until the 4th or 5th round to grab Tom Brady in an industry draft. But part of the reason I'm willing to do that is that I know there's a good chance Brady falls that far in an industry draft. And I know that if I miss him, there'll be tons of great quarterback value available later as a fallback.

In a more typical league, I'm not going to wait that long at quarterback, because I know I can't wait that long at quarterback. If I want Brady, I have to grab him in the 3rd at the latest, because he's not making it back to me in the 4th.

This is what I was talking about earlier, how drafting is relative. Who I draft is based on who I prefer... but it's also based on who everyone else prefers. And industry mocks will tell you an awful lot about who guys in the industry prefer, but they will tell you very little about who your leaguemates are going to prefer and where you should be looking to draft specific players.

This isn't to say that industry drafts are useless for casual players. Again, they're an opportunity to see what kind of strategies we employ and players we like. The key is simply to remember that they don't serve as a template for what your draft is going to look like this year.

In applied mathematics, a distinction is made between "locally optimal" solutions and "globally optimal" solutions. (Trust me, I'm going somewhere with this.) Locally optimal solutions are the best solution out of all the neighboring points. Globally optimal solutions are the best solutions out of the entire set.

Consider this graph, for instance:

If your goal is to find the lowest point, then the trough on the right is a locally optimal solution. There is no point nearby that is lower. But it's not globally optimal because there's another trough a ways to the left that gets lower still.

The fantasy football advice industry is all about globally optimal solutions. Who is the best running back this year? Which players will score the most points this week? In an ideal world, how long should you wait to draft a quarterback?

Industry mocks are the another piece in that tradition. They ask "what's the optimal way to draft if everyone else in your league is also drafting optimally?" But fantasy football is interesting in that what you're truly after are locally optimal solutions, instead. After all, in most leagues, the question you want answered is "what's the optimal way to draft if everyone else in your league is not drafting optimally?"

This emphasis on locally optimal solutions is one of the reasons I'm such an evangelist for Footballguys' unique tools, like MyFBG, the Draft Dominator, and the new Lineup Dominator and Waiver & Trade Dominator. Because for the longest time, globally-optimal solutions were the best we could do in the industry.

But advances in technology have allowed us to start seeking out locally optimal solutions, instead. The Draft Dominator, for instance, can now attempt to estimate where your leaguemates will actually draft players, and it can use this "Estimated Draft Position", (a local value), in place of "Average Draft Position", (a global value), to give you better and more specific advice.

Or the Lineup Dominator can take a look at your team, take a look at your opponent, and tailor specific advice to your specific matchup. If you're an underdog, it can say "hey, you need a breakout game here, so don't start the guy projected to score the most points; start the guy with the highest upside, instead!" That kind of granularity wasn't possible five years ago.

Every league is different, and those differences mean every league has a different recipe for success. Industry drafts are a cool look into the thought processes of people who spend massive amounts of time thinking about fantasy football, but they're never going to give you a personalized path to victory.

This doesn't mean they're useless. It's just important to keep in mind what they are, and more importantly, what they aren't.