login
register
drafts
tools
games
forum
GOAheadScore.com» FF Articles » Math and Theories Behind Snake Fantasy Drafting
Share This
Why do we draft fantasy football drafts the way we do? One question that springs to mind straight away is why so many running backs go in the first round, and why people wait on by far the highest scoring position - quarterback - until often as late as the 5th or 6th rounds.
Of course, we could just accept it and just draft the way that convention dictates. If everyone else does it, it must be right. Right?
But if you like fantasy football there's a little problem with that attitude. Because on some level - like it or not - you're a stat geek, and stat geeks should need to understand everything. Because who knows how we should change conventional drafting tactics according to the roster lineups and scoring systems of a specific league?
nb: For convenience, the drafting situations discussed below assume players will exactly replicate their 2010 output. Exactly the same theories apply no matter what you think players will score though!
All the theory below is implemented for you on our draft advisor app - check it out!
I'll just take the dude who will score the most points!
We've seen it all before, and most leagues have one. The guy who this year actually hasn't turned auto-draft on, but turned up with too many beers (c'mon man) and then pulls the #1 pick out of the hat.
He stands up in a manner akin to George Bush stepping onto the tee, and takes great pleasure in announcing to the room he's drafting Peyton Manning.
Really George? A first overall pick on me?
But is he that stupid? After all, quarterbacks score more than other positions in fantasy, and as quarterbacks go it's hard to argue Peyton's not near the top. He goes on to tell you that Manning will score more points than your first round pick by a mile.
And he's probably right there. Last year, Manning scored 285pts in a standard scoring system. QBs aside, he was only outscored by Arian Foster.
We all know that he's an idiot though. Because when he has to take a running back, he'll be dumped with trash. Let's say he goes down to the 4th round, and 15 running backs have been drafted. Last year, the 16th best back was Cedric Benson (167pts over the season).
In the same round as Benson, the guy who took Adrian Peterson in the first round (240pts), will draft Josh Freeman (256pts).
Manning + Benson = 452pts
Peterson + Freeman = 496pts
Idiot loses.
You can't just take raw point projections in a fantasy draft. A player's value is instead first dictated by how much better they are than the worst player at their position who you could possibly have to field. Because even if the expert in the example above hadn't taken Freeman, by the time he finishes his starting lineup around round 8, the 12th best QB should probably still be on the board.
Applying this to a Non-Standard League
Last year, my main league went with 2 spots at quarterback. 8 out of the 10 teams blindly followed the "running back first" convention. A buddy and I spent a little more time thinking about it. He ended up with Rodgers and Manning, myself with Brees and Rivers. In the end, he finished third and I won the league (the runner up was the guy who got Vick in free agency to go with Big Ben).
The other guys lost the whole league at the draft - it was painfully that simple. They might as well not have bothered playing the season out because the guy with Chris Johnson first overall didn't go QB until round 4, where he was lumbered with Favre and Cutler.
The "worst starters" in the 2QB, 2RB, 2WR format were Bradford (191pts), Tolbert (155pts) and Welker (133pts). Note of course that in a 1QB format, the worst starter was Joe Flacco (231pts). So we can simulate what the top ten might have looked like with more intelligent drafting. For each player, we'll take their score and subtract from it the worst starter. This gives us the value of the player over the worst starter - who you should be able to draft with your last pick before the bench spots start to get filled.
1. (326/171) Foster
2. (308/117) Vick
3. (304/113) Rodgers
4. (299/108) Brady
5. (286/95) P. Manning
6. (279/88) Rivers
7. (240/85) Peterson
8. (237/82) Charles
9. (234/79) Hillis
10. (269/78) Brees
(total points scored/value over worst starter)
By this theory, 6 quarterbacks should have gone in the top 10. 5 of the top 6 spots should have been quarterbacks.
But where Joe Flacco is the "worst starter" at quarterback in an otherwise identical 1 QB league, this is what we get:
1. (326/171) Foster
2. (240/85) Peterson
3. (237/82) Charles
4. (234/79) Hillis
5. (308/77) Vick
6. (209/76) Lloyd
7. (229/74) Chris Johnson
8. (207/74) Bowe
9. (304/73) Rodgers
10. (299/68) Brady
3 quarterbacks, and just Vick sneaking into the top 6.
So we've managed to find out why running backs get drafted first - and understanding why people started doing this allows us to know when to follow the rule and when to ignore it, and how it applies to other positions.
"Tiers" of Players Complicates Matters Further...
When we're selecting players in earlier rounds, especially for the WR1 and RB1 slots, we're not so concerned with worst starters - they're only really going to come into play for the guys in the flex or WR2/RB2 slots.
I touched on this earlier with the Ced Benson talk - in every draft, each team is going to have at the very least 1 running back and 1 wide receiver on their roster in the first 4 rounds.
Looking at tight ends going into 2011 is a good example here - while there's a lot of talk about Jimmy Graham showing the potential to join the elite group, let's just assume there's a group of 5 guys way above the rest (Gates, Witten, Finley, Clark and Davis).
Let's also assume that we're picking 10th overall in the 3rd round, and all of the tight ends are still on the board - so the draft order looks like this:
3.10 My Fantasy Team (current selection)
3.11 Team 11
3.12 Team 12
4.01 Team 12
4.02 Team 11
4.03 My Fantasy Team
It may be at this point that one of those guys, say Antonio Gates, has the best "worst starter" rating. But looking at this in more detail, at this point it makes no sense to draft him - teams 11 and 12 will at worst take 1 tight end each, leaving you with your choice of one of the 3, 4 or 5 remaining elite players at 4.03. Your "loss" then over the worst starter at tight end can still be very well mitigated.
Imagine however being at the other end of the 4th round, if we'd drafted 3rd overall instead of 10th:
4.10 My Fantasy Team (current selection)
4.11 Team 2
4.12 Team 1
5.01 Team 1
5.02 Team 2
5.03 My Fantasy Team
The difference here, being later in the draft, may be that only Jason Witten is still on the board. Now the argument becomes a lot stronger to draft him - especially if teams 2 and 1 did not go tight end at the 2/3 turn (unlikely) - as with 4 picks in there, there's a high chance that at 5.03 all the elite guys are gone. You'd feel it's too early for Graham and later in the draft you end up with someone a lot closer to the "worst starter" at tight end.
Many people may go into the draft with a strategy to pick up tight ends later this year, and that's absolutely fine - what's important here is the theory of trying to see what will be left on the board when you next make a selection, and realizing that needs to have an impact on your decision with the current selection.
Being able to project the mean player left at each position when you have your next selection allows you to use them as a benchmark for rating higher players on your board, instead of the worst starter, and until the starting rosters are about to be filled, is therefore a superior strategy.
How you go about making these projections is difficult. Our FF Draft Advisor app looks at each team's rosters, their position needs, and uses its locally stored projections to guess the likelihood of each player being picked at each selection.
In your own drafts, you may have some more information - for instance, the Colts fan is likely to go Manning in the first 2 rounds, or you may know some players who certain owners will refuse to draft - this information is all useful and certainly worth making notes of before you draft.
Leaving Yourself Short at a Position
Many people speak about balanced drafts, and making sure you are strong at every position. Really it's difficult to access this strategy because it depends on your league too much - in a trade happy league value is value, especially if you're confident of taking advantage of other owners - and if there's a WR1 still on the board when you've already drafted your top guy there, it makes complete sense to stack up on receivers and look elsewhere for a back.
While you still have roster spots open at a position, mathematically they should be considered just as strongly as any other position. As long as week to week they are going to have an opportunity to start, they'll still contribute to your points total just the same as anyone else.
Of course, in conservative leagues or leagues where trading and/or waivers are restricted, look to balance out your roster more.
One thing to note is that the franchise guys in fantasy leagues are very difficult to acquire in trades. Many owners become emotionally attached to the players they draft early and will live or die by those picks. Yet another reason to make sure that within your first 3 picks you get at least 1 stud WR and RB apiece.
Avoid the worst case disaster - don't go for the home run
So when we've taken a look at the theory and math, and I'm not going to apologize for a lack of equations, everything points to chucking the following statement out of the window:
"With every pick, look for that guy who will put up enough points to win you matchups",
and replace it with this:
"With every pick, do whatever you can to mitigate the worst case on your roster".
You can't win a standard fantasy league in the draft, but you can certainly lose it.
What About Player Upside?
Go back and take a look at your leagues last year - make a note of your roster at the end of the season and your roster just after the draft.
How many players are on both? For my main league with a 16 man roster, the number is 6. Well, 7, if you include Cedric Benson, who was ceremoniously dropped during the playoffs when free agency and waivers were closed. Frustration had to be vented somehow....
There comes a point in every draft where you're selecting wideouts and backs who could be fairly easily replaced during the season. Think about the free agency crop last year led by Hillis and Lloyd - they would not only have replaced a WR/RB2 but a first string guy as well.
That's why we draft upside and why there's no point going low risk later in the draft. If you can place a ceiling on a guy, let's say Eddie Royal as a fantasy WR3 - why bother drafting him? Within the first 4 weeks of the season it is a guarantee you will be able to find an undrafted player off waivers or free agency to fill that role, so why not roll the dice on some upside?
In the draft of my main league last year, I took Steve Breaston, Mike Wallace and Bernard Berrian, amongst other receivers. All guys I hoped had some upside. One of them worked out and his value increased by the end of the season to the extent where I could package a deal together to replace Benson. Two very quickly found themselves dumped to waivers - and replaced by Mike Williams and Mark Clayton - both guys who, at the time, were establishing themselves at the top of their depth charts.
It's good and sensible draft strategy to make sure your early picks will be earning you points season long, and your later picks have the potential to produce big weeks or increase in value.
So Am I Expected to Use a Calculator During Drafts Now?
No. Not all all!
Thankfully we've got a ton of resources on the website that will do all this work for you.
Members (joining is free) can setup their scoring systems on the website to get worst starter draft orders from our pre-draft rankings (note these rankings assume a standard roster).
Cheat sheets can only take you so far though, so enter the new era with our FF Draft Advisor app. All the theory above is implemented along with all our projections, MyFantasyLeague integration, strength of schedule information and extensive 2010 reviews and 2011 outlooks for all the top fantasy players.
It's available on the Apple app store and will be coming to an Android or Blackberry store near you soon.
10490 points
Pro Bowler
This is just like horseracing some horses have all the numbers and still lose,what you need most of all is lady luck.
10490 points
Pro Bowler
I got to get off the Practice Squad,how many points do you need for a pair of tube socks
33985 points
Franchise Player
Of course you need luck, but some players have more risk attached to them than others - so my tendency is to play it safe early because I think it's very tough to win a league without your round 1 and 2 picks playing in the playoffs, then shoot for those risky high ceiling guys later on.
Think 500 will get you up off the practice squad or buy a custom title in the GOAheadScore shop! It costs 20 points a day so right now you could buy about 18 days worth or so...
72730 points
NFL Draft Guru
I wrote a similar article (not as in depth,) last year. Always enjoy a mathematics look at a math game such as fantasy sports, thus why I like Mike Clay's work and articles like this. Thanks man! I'll def. be using some of the resources you guys have for my drafts.
17180 points
Pro Bowler