There is very little difference between the #1 and the #10 kicker and predicting future performance from past performance is a crapshot. Plus if you draft a kicker early, you will be exposed as a fraud :) Every year there are no-name players that have breakout seasons, so use your late round picks trying to hit the lottery.
* Try out Value Based Drafting
This is a more advanced metric for optimizing your drafting strategy. The basic idea is to not pick players based on the raw number of points they are projected to get, but rather how much better than a "baseline" replacement. It's very nerdy but something that the traditional "office jock" isn't going to bother with. If you want to dig a little deeper, read this: http://sports.espn.go.com/fantasy/football/ffl/story?page=nf...
* Set aside time to manage your roster every week
You only need 20 min to do the minimum each week: set your lineup based on expert rankings and add waivers/free agents to replace injured/non-performing/bye-week players. If you do this every week, I can almost guarantee you won't come in the last place (the ultimate objective!). Pick a single source of rankings (ESPN is fine) and don't worry about reading 50 different blogs.
* Remember to have fun and don't talk shit
If you don't follow sports at all, now you can get a taste. It's pretty fun! If you normally follow the hometown team, congrats - you will now find yourself caring about 8-12 other NFL teams every week - if only for your one fantasy player. And don't be an asshole. No one likes playing fantasy with jerks.
PS Be careful, you've just stepped into the most addictive hobby I've experienced since World of Warcraft :)
This is all good except the part about not talking shit.
Talking shit makes fantasy sports even better. There is, however, one rule to this: only do it in leagues with your friends. Yes, you're an asshole if you're in a league with 11 other random internet people. With your friends it's just fun.
The authors' (completely algorithmically managed) team managed to get a rank of ~25000 when competing against 2.5Million human players. Pretty damn good if you ask me.
This paper is about Fantasy Soccer though. However the same rules apply to Fantasy Football too.
The biggest aspect of Fantasy Football is the draft - where you select your players before the start of the season. So the area where you could apply some actual hacking is the projection/simulation of player performance to come up with your own rankings. Hopefully, this could allow you to pick undervalued players - so similar to predicting stocks. Basically you can treat it as a statistical modeling exercise and go from there.
There is no realtime aspect to fantasy sports, you are usually setting a roster weekly, waiting for the games to be played for the week, and then seeing the results. So something like high frequency trading doesn't really apply.
This is the first year that I've tried Value Based Drafting. It's also the worst team I've had in years. I struggle to get 90 points in a standard league. Part of it is because Spiller has been a big bust this year, but in general, this team relies on a lot of potential instead of safe bets. Bad luck for me, as few of my guys have shown that potential. :/
Just saying that YMMV. It's fantasy football, nothing is guaranteed.
I actually like the idea behind Value Based Drafting, and will probably still use it the next few years, though, not for my primary league anymore.
* The first rule of Fantasy Football is: You do not bitch about Fantasy Football
Literally, no one cares about your team. Maybe the person you're playing against might have an iota but chances are, he's in other leagues and has his own FF problems.
I have actually gone the opposite when it comes to Fantasy Football. I am a huge football and basketball fan. I am also a very competitive person. So when I played Fantasy sports I found myself too often rooting against my team when it came to getting fantasy points. It started to kill all the fun from watching the game.
So last year I stepped back and didn't play FFB. First season I was able to enjoy without the extra stress of competing in 5 years. This year I didn't play and once again I am enjoying the game more. So I don't think I am going back. I will save my competitive juices for either the basketball court or playing a game online.
The condescending tone of this article makes it almost unreadable.
Play FF because you enjoy it.
Watch football because you enjoy it.
Don't secretly deride your coworkers for enjoying something you don't. That makes you a shitty person to be around. And chances are you enjoy doing something that is equally irrational (like playing video games or solving Rubik's cubes or playing Settlers of Catan).
So the TLDR is if you invest a lot of effort, you could win a game that doesn't matter, that you don't care about, all to impress people you don't care about?
Its a good article in that if you've already decided to do something like it, FF is a sort of interesting, unusual choice.
One minor failing of the article is failing to deal with the issue of FF being extremely slow paced. Like playing pac man by submitting punch card jobs on an ancient mainframe.
By far the most accurate description of fantasy sports I have heard is that they are "Dungeons & Dragons for jocks." Because they are about sports, they are basically the only socially-accepted way for jocks to live in a fantasy world without being labeled "nerds."
(Yes, I realize non-jocks play fantasy sports too, but they are definitely not the main target audience.)
Presumably the fun part comes from the Final Fantasy aspects of it, and not from the football aspects of it. Min-maxing can be very fun by itself, regardless of the subject at hand.
For this post, I'm going to make a couple of simple, reasonable assumptions.
1) there exists only two types of people in this world: jocks and nerds
2) you're the latter
Assumption 1) is not reasonable, these two groups are not mutually exclusive. I am definitely in both groups, although I might use different terminology.
I'm a Fantasy Football (not egghand) player myself and a big football fan in general.
Funny enough, Fantasy Football is what taught me how to code, I wanted to enhance the crappy official website ( http://fantasy.premierleague.com/ ) with nice tools so I went from knowing html to deploying my own rails apps ( http://www.insidefpl.com ) in a year. I'm still learning everyday and new tools ideas are what make me push myself to learn more.
I admire the enthusiasm of the author, and I do encourage others to try fantasy football (or other fantasy sports), as it's really fun and I enjoy it a lot.
I just want to point out that he's in an 8 man league, which is on the smaller side. That's why his team is so stacked. Everyone in his league has a stacked team. Just saying, so anyone who is new to fantasy doesn't have false expectations.
(A 12 man league is considered best, at least IMO, because everyone will have a few good players, but they will have to make some smart trades and waiver wire pickups to actually win.)
The way this is written makes it sound like in American fantasy football it's not tied to the real games, is that right?
I don't follow football (the British kind) but for the first time this year I'm playing fantasy football. It started as a very clever marketing hack my business partner came up with. See, in ff there are league tables so you can see how you're doing against other people you know. The industry we're breaking into all play this game so they see our names and our company name every week while checking the scores. It also provides an instant talking point - "bit of banter".
Aside from that I've been enjoying playing. It's a game of chance and there is advantage to be had from knowing about the teams / players. There's stack loads of data though. More then you can handle. Everything about every player is tracked. The performance of each player in real life each week determines the points they win you.
There's a whole other aspect too - a market for players. Based on demand the cost of a player changes each day. So you can play the markets to build team value so you can afford more expensive players.
There's a lot going on and even those of us who have no interest in football itself (I'm yet to watch a game) can find plenty to enjoy in the fantasy side of the game.
I don't know the ins and outs completely (tried to be up front about that), but I do think it's very possible to win a week in FF and have all your players be on losing teams in real life.
Depending on the league and how points are your specific rules, your players get points for things like passing yards, touchdowns, interceptions, rushing yards, etc.
These things are all well and good, but having one good player that does a lot of them doesn't mean that their whole NFL team is going to put those things together and beat their real world opponents. And vice versa, you can have several players on a winning team that each get 0 points because they happened to be injured or were benched.
Cost is a completely different can of worms.... sounds more interesting, but everything I've done is first come first serve. You draft in order and you can pick up as many free agents as you have space for.
The fantasy points are tied to individual stats, that are reported in close to real-time from the NFL (or stats provider). The league begins with a draft, much like real football. Random positions are assigned, and each player takes turns picking from the full roster of available players until each player has filled up his team, with a couple of spots left over for backups.
Like on a real team, your fantasy team consists of a quarterback, a couple of wide receivers, a couple of running backs, a tight end, a defense, etc. Different leagues have different roster spots -- some have tight end as a position, some don't; some allow you to pick individual defenders to comprise your defensive team, some just let you pick defensive units (e.g., "Kansas City Chiefs Defense") as a whole. Generally though, they're pretty similar.
The way that it works, in a nutshell, is that you have fantasy points that attach to real-life events. E.g., a touchdown is worth x points for the player that scores it. As such, because you're playing individual players, instead of just playing a team, you can mix and match to get great results.
A great example is Adrian Peterson, who is the best player on a pretty middling team. Adrian Peterson is the running back, which basically means that his primary job is to be handed the ball and run. His quarterback is not very good (comparatively, all the players in the NFL are good), so they over-rely on the running game to advance the ball. This means that, in real life, the team is very unlikely to win a game, but in fantasy-land, Adrian Peterson is likely to get a lot of points... also, he's an exceptionally good running back, and was likely picked #1 in pretty much every league this year.
The stats get a lot more complicated than that though. Quarterbacks get points for throwing the ball, but at the same time, receivers get points for catching the balls thrown. These points are scored differently. The result, basically, is that if you have a team with a mediocre quarterback, but that also has only one good wide receiver, that wide receiver might get more fantasy points than a receiver on a much better team, with a much better quarterback, because in the former, the one good receiver is likely to get a disproportionate amount of receptions (because there are fewer targets for the Quarterback), while the better team may have more recipients for the quarterback to throw to.
It is a lot more complicated than just picking 'good' players, because too many scenarios arise where good players don't play well in a given game, or act as decoys to allow lesser players to do more of the work. A great example here is Calvin Johnson, who is one of the best receivers in the game. Because he is so good, he often has games where he scores very little in fantasy points, because his role in those games is to draw double-teams from defenders. Those defender, now paying disproportionate attention to Calvin Johnson, leaves holes in the defense for lesser receivers. That's just an example though, because in practice, Calvin Johnson usually draws double-teams, and is still usually able to make lots of plays because he's insanely talented.
Each week, you select the players on your roster you want to start, taking into consideration things like whether your player is playing that week (football has rotating bye weeks, players may be out on injury), how each player is likely to do (e.g., a quarterback playing a very tough defense may be expected to play more poorly than a worse quarterback playing against a looser defense), etc.
Your roster is pitted against another player in your league, and whomever's roster combines for the most fantasy points wins the week, and gets a new opponent next week. Your total team's points are compared against your opponent's, and whoever has more 'wins' the week. That win gets tacked onto your record -- e.g., if you win the first week, you're 1 and 0. Lose next week, you're 1 and 1. Competition happens tournament style.
As a fan, fantasy football makes things... complicated. My fantasy quarterback is Aaron Rodgers, and the team I root for in real life is the Baltimore Ravens. It leaves for very conflicted feelings when Aaron Rodgers is playing an amazing game against the Baltimore Ravens, as happened two weeks ago.
Or, your opponent may have the quarterback that throws to your wide receiver, which means you end up rooting for ridiculous scenarios like "Adam Smith needs to throw a very short pass to Dwayne Bowe, who then runs for a very long touchdown" to minimize the QB points and maximize the WR points.
It is fun though, and it appeals to me both as a football fan and a geek, and I'm honestly surprised that more geeks don't play. Leagues can be very competitive, and people can spend as much time scouting stats and players week to week as top tier WOW players can spend prepping for raids... okay, not that bad.
Offtopic, there's a reference that I can't remember the attribution for, but seems appropriate for this thread.
"Fantasy Football is a glorified D&D for the jocks who used to beat us nerds up for playing D&D." -- is not too far off the mark.
Why should you learn about sports just so you can participate a conversation of content won't matter in a week? I recommend a different approach: Have a conversation about something constructive that your peers might not know about. "Did you read gwern's article about spaced repetition?"[1] "I played around with WebRTC last week and got video chat working. I found a lot of bugs in Chrome and Firefox..." "Have you ever wondered how async.js avoids exploding the stack if you give it synchronous functions?"[2] Etc.
This can enrich your peers as well as yourself. Heck, they might even follow your lead and do more interesting things in their spare time.
Not everyone works in a San Francisco-based start up that employees 90% technical folks. Perhaps people are just interested in a low-effort conversation item so they can interact with co-workers, neighbors, and a different social circle? Some engineers like the fact that they can enjoy fantasy sports for the nerdy, statistics aspect and still be involved with something enjoyed by "normal" people.
Not everything in life has to be about programming. And caring about sports does not preclude you from having technical discussions either.
I gave programming examples because of the HN crowd here. In my own life, it varies quite a bit. In the past week, topics have included:
* Fluorescence. (At a cocktail bar full of blacklights. Any cocktail containing tonic glowed due to the quinine.)
* Comparing the stories of Alistair Reynolds and Lawrence Watt-Evans. Their fans seem to overlap quite a bit, even though their genres are different.
* Beard growth. Why most militaries don't let soldiers grow beards. (Beards prevent gas masks from sealing well.) How fast beards grow. (About 5 nanometers per second, hence the beard-second as a unit of measure[1].)
* Noticing that dentists can replace teeth, but doctors can do little for chronic back pain.
Aside from some topic selection issues, you might want to consider passive vs active topics. Can't do any activity while awake that's more passive than watching TV, almost anything else is more active.
I'm not too interested that a coworker watched someone else exercise on TV. On the other hand, start talking about experiences on local medium distance hiking trails, and I start listening...
You may already be aware of this, but if you're not, it's something you eventually need to hear, so I'm going to tell you now. Not everyone wants to talk about tech stuff all the time. As someone who does, I feel your impulse, but it is not going to be good for your reputation as a conversation partner if you do this.
This is a great article, and I was considering starting Fantasy Football this season for the same effect. I will make one finer-grained distinction, however, that the ability to shoot the shit about football is a useful one depending on what you're trying to do with your life.
The trade-off is that everything takes time and attention. Lots of different priorities compete for our attention, and while registering in your office's fantasy league might be a small time investment, it might be a disproportionately larger attention investment. You will start noticing games, players, conversations about football, etc., with much more frequency than you previously had.
In my personal life, I've sat through many sporting events and conversations with the attention of a dullard with ADD. I generally tune out all sports-related things and devote my extra attention at these events to thinking about other things which are more interesting to me, or ruminating on a problem I am trying to solve. These are potentially useful things I would give up if I am more actively focused on sports-related things as a by-product of simply choosing some big names in a Fantasy draft.
At the end of the day, I still find it worth the trade-off, but only because I am optimizing for a certain level of "ubituitous social lubrication" at the expense of, say, skipping the office happy hour and watching a Coursera course on machine learning (a contrived example, but you get my point). You may not want to make this trade-off if you are not trying to be a "general people-person", or if you can obtain this same social lubrication through other avenues (IE, politics).
tldr; This article is very important for startup founders, perhaps not so much for systems engineers / data scientists.
If you're in the target audience for this blog, I'll add that football itself is very enjoyable. The strategic elements of it really are game playing at its finest. The basic goal is for the offense to advance 10 yards down the field within 4 plays (inclusive). Here are videos covering formations, which are core to understanding the game:
Also, if you have a chance, I recommend watching any football games which don't have commercials (including not going to live games which are televised) because the commercial breaks slow down the game unnaturally by about an hour. Arena football is especially fun to watch (basically a minor league and semi-pro type team with ex college-level stars playing under modified arena rules... google to find a local team) and high school football is also quite fun to watch.
In my nerd-office we've started getting serious about https://www.fanduel.com/ - think normal fantasy football, but each league only lasts a week long. It means you can put much less effort into a long draft, and you can quickly iterate & recover strategies.
Another fun angle to look at fantasy football is to apply financial concepts to it.
For example, I have Tony Romo and Dez Bryant on my team, who are the QB and WR for the Cowboys and if one puts up a bunch of points, the other will as well, which means I take on a lot of variance for an expected high return. Or I could take both Jimmy Graham and Pierre Thomas, who play TE and RB for the Saints; since the Saints have a fixed (well, sorta) amount of offensive opportunities, you're giving up a certain amount of possible return in order to have a guarantee that, between those two players, you'll have a decent number of points -- even though every rep Graham gets is a potential rep taken away from Thomas and vice-versa.
(The way this differs from traditional financial planning is that success is unary; you win or you don't. So you get a great balance of trying to only take on as much risk as you need to win.)
"With very little effort (and absolutely zero time spent watching grown men throw around a dead animal), you'll be able to disguise yourself as a true fan."
Is this not deception? And condescension? I'm not saying everyone needs to be into sports but mocking and trivializing things you personally aren't into is neckbeardy and gross.
As a fantasy football player turned programmer.. it's great to see posts like this and I completely agree with the OP. It's fun, surprisingly logical (thus easier when playing against emotional people), and great conversation starter (co-workers, family, fathers-in-law).
My friends and I actually created what has effectively become "fantasy football for sales organizations" where managers can choose/weight metrics for employee-formed teams to compete over across seasons. It's early but so far the results have been impressive. The demo is here if anybody wants to check it out: http://tryambition.com
My family did a league this year...to be honest I've always felt left out whenever Football season came around, I just couldn't get into it because I didn't have "My team" like everyone else in my family.
I think it's brought me closer to my family, since I always felt like an outsider whenever I did watch a football game, I had no clue who the good players were and what teams were doing.
Now not only do I enjoy watching the games but I went from an 0-3 record to 4-3 (soon to be 5-3, fingers crossed, 2nd place).
It definitely makes the games more interesting to watch and has a healthy side effect of building better relationships with those you care about.
If you're looking at fantasy sports as a way for "geeks" to play a part in the social interactions which revolve around sports, I think baseball is the one to focus on. We're talking about a game that is seemingly designed for stats nerds and which understanding of basic statistics concepts actually gives you a huge competitive edge.
The rise of sabermetrics (defined as "empirical analysis of baseball" by wikipedia) has been pretty amazing given that its just math, and it has gotten to the point where TV/Radio commentators talk about this stuff every game.
Seems like the short version is "apply principles of 'Moneyball' to American fantasy football, and win".
American football still holds no interest to me, and I'm still the guy at the office who would rather discuss F1, rugby, and proper football with Brits and Aussies in the building.
As an aside - if anyone has a good "cricket for dummies" recommendation, I'm open to suggestions. As much as I try, the rules still don't really make sense to me. Perhaps I'm permanently damaged having grown up on baseball.
Yea... I started hearing "doe" a lot recently and got pretty irritated in a "get off my lawn" kind of way.
Then I realized there's a 1993 Ice Cube song called "Really Doe" and I can't get mad about "kids these days" when it's the title of a 20 year old song. http://youtu.be/07vkbKhS6CY
This is dead on. Another developer and my self reluctantly joined our office's league this year, despite knowing nothing about football. We're both dominating. He's 7-1 and I'm 6-2. I'd be 7-1 too if I'd have known what a 'bye' week was and paid closer attention to my team.
I've had this exact thought. I could create an api for ESPN with Nokogiri or Capybara or whatever and get the relevant stats for everyone. If it's really just points, then it would be an incredibly simple thing to write.
Will probably do exactly this next year, just for fun and the opportunity to wreck my coworkers.
Check out Daily Fantasy sports... you get to draft a new team each week for football. With other sports it really is Daily, but football is based on weeks.
And at least you were able to get to the playoffs... better than I ever did.
there's a nice chapter on breaking down Fantasy Football performance between the two skills of drafting and and line up setting in Kaiser Fung's book Numbersense
* Don't draft a kicker before the 16th round
There is very little difference between the #1 and the #10 kicker and predicting future performance from past performance is a crapshot. Plus if you draft a kicker early, you will be exposed as a fraud :) Every year there are no-name players that have breakout seasons, so use your late round picks trying to hit the lottery.
* Try out Value Based Drafting
This is a more advanced metric for optimizing your drafting strategy. The basic idea is to not pick players based on the raw number of points they are projected to get, but rather how much better than a "baseline" replacement. It's very nerdy but something that the traditional "office jock" isn't going to bother with. If you want to dig a little deeper, read this: http://sports.espn.go.com/fantasy/football/ffl/story?page=nf...
* Set aside time to manage your roster every week
You only need 20 min to do the minimum each week: set your lineup based on expert rankings and add waivers/free agents to replace injured/non-performing/bye-week players. If you do this every week, I can almost guarantee you won't come in the last place (the ultimate objective!). Pick a single source of rankings (ESPN is fine) and don't worry about reading 50 different blogs.
* Remember to have fun and don't talk shit
If you don't follow sports at all, now you can get a taste. It's pretty fun! If you normally follow the hometown team, congrats - you will now find yourself caring about 8-12 other NFL teams every week - if only for your one fantasy player. And don't be an asshole. No one likes playing fantasy with jerks.
PS Be careful, you've just stepped into the most addictive hobby I've experienced since World of Warcraft :)