Now that I've set some ground rules in part
one of this article, let's look at the gory details of last
year's draft. Given that we have some ideas about how many players
are going to be taken at each position and how valuable each pick
throughout the draft is, now we can turn to a more detailed analysis
of last year's draft.
The predictions mentioned earlier, especially the curve fit for
draft pick value, assume that everyone who takes a player takes
the best available person with each pick. That will not happen
unless everyone has the same draft prioritized draft sheet. Since
that will never happen, let's look at some of the variations that
tend to occur from the predictions.
First off, some definitions of how I assessed the picks made
in last year's draft:
Steal = picked more than two rounds below final ranking
Good = picked within one round of final ranking
Early = picked between one and three rounds above final ranking
Bust = picked more than three rounds above final ranking
I weighted each of these types of picks to generate a combined
assessment of each person's draft. The formula behind this rating
weights Steal and Good picks positively, Early and Bust picks negatively,
and adjusts the resulting values to a percentage. A rating of 0%
would indicate a draft completely filled with busts, while a rating
of 100% would indicate a draft with two Good picks and the rest
Steals (by definition, Steals can't occur in the first two rounds,
therefore the best you can do there are Good picks). The figure
below compares how well each owner drafted (Rating) with how their
team performed for the year (Score). For clarity, I've ordered these
ranking from one to twelve, with one representing the best draft
(and the best scoring team).
There are a couple of observations to be made from this figure (aside
from the colorful nicknames in the league). The teams with the top
three drafts (Mooch, Frac, and Rusty) all finished in the top three.
Additionally, of the top six drafts, all but one finished in the
top six (green box); for the bottom six drafts, all but one finished
in the bottom six (red box). The two exceptions were Hank (red arrow)
and Flash (green arrow). Hank started off the season well with a
good draft, but then went in the tank because he only made one transaction
during the season. Flash scrambled during the season to recover
from a poorly rated draft and had the highest number of transactions
in the league.
Hmmmm ... maybe there's something to this preparation thing...
Overall Analysis of Picks
The first thing immediately apparent when looking at the breakdown
of pick types by round (see figure below) is that there were a
LOT of busts. In fact, outside of three rounds (1, 11, 12) every
round had more busts than any other type of pick. Yikes! How did
anyone actually even score in this league? Are they all a bunch
of freaking idiots?
To be fair, I must rise to the defense of my league. Over 46%
of the draft picks were of Good or Steal level excellence. That's
not too bad, especially considering how many big name players
went belly-up last year. However - fully 47% of the picks in last
year's draft were Busts. The good news is that's the lowest percentage
of Busts in the eleven year history of the league. The bad news
is that it could have (should have!) been better.
How could the draft have been better? Examining the distribution
of pick types throughout the draft will shed some light. Aside from
the seeming preponderance of Busts, there are several interesting
trends in this chart. Let's look at each of the pick types to see
things more clearly.
Obviously, high draft picks that turn out to be busts really
hurt. These were the guys counted on as starters, and when they
bomb it really leaves a hole in a roster. Conversely, busts in
the later rounds don't really hurt at all. At that point, you're
looking to add depth only; maybe you're taking some fliers on
the off chance that you'll get lucky. The high number of busts
in the final round is expected, as that's the proper time to be
making those serious sleeper picks. But when it comes down to
it, there's almost no way to avoid at least some busts every year.
People get hurt, especially QBs.
Let's look at the fourth round. Seven of the picks in that round
were busts. Not good. By looking back up at the table for the
distribution of positions taken in each round, we can see that
the fourth round was where there was a decided emphasis on WRs
after everyone salted away their starting RBs (in fact, 75% of
the picks were WRs). Jeremy Shockey being taken in that round
as the 17th receiver was a bad pick. In a league that requires
playing a tight end this might have been a good pick, but in this
league, where WR and TE are equivalent, this was much too early.
Why? Well, only two TEs ended up ranking in the top 20 for receivers
(Tony Gonzalez at #12, and Shannon Sharp at #20) and the next
closest TE was #37. This was a flier pick, made much too early,
especially since it was a rookie. I'm not going to name any names,
but the offending owner's initials are Fungspo, and they bought
into the New York City hype machine. The safer pick would have
been to take any one of a number of WR picks (Chad Johnson, for
one, was still available at that point) and wait to take a TE
until later (or take an established TE like Gonzalez).
Another point of interest is the Good picks and their locations.
There were high numbers in the first two rounds, and then very
low numbers throughout the rest of the draft. This makes sense,
as in the first couple of rounds most of the studs are being picked,
and for the most part they're easy to predict. Being able to make
a good pick later in the draft really requires a lot of work (and,
admittedly, some luck).
There were very few Early picks (only seven), and they were scattered
evenly throughout the draft. What that means is that when people
missed on a pick, they tended to miss big (as evidenced by the
number of Busts).
As expected, the number of steals went up as the draft progressed,
peaking with the next-to-last round of the draft. Most of the
owners did their due diligence and had identified high value picks
later in the draft.
The most important item to take away from this discussion is
that early round Busts can kill your draft. There are two ways
to avoid this for the first half of the draft: be conservative
and avoid taking fliers, and draft the best available players.
Wait until the second half of the draft to take chances on sleepers,
rookies, and the can't-miss media darlings.
Fine. Right. Easier said than done. Yep, I hear what you're all
saying. Let's take a look at specific examples of the different
pick types to get a better idea how to do this.
Top Twelve Steals of 2003
The table below lists the best steals from last year's draft,
as indicated by the difference between where they were picked
and where they ended up ranking in the VPR assessment of player
value. One observation of note is that all of these highly rated
steals occurred at the end of the draft (round 9 and later).
Top
Twelve Steals of 2003
Round
Owner
Player
Degree
12
Fric
Jeff Wilkins
11
11
Unk
Mike Vanderjagt
10
12
Ox
Keenan McCardell
9
12
Mooch
Ryan Longwell
7
10
Unk
Peter Warrick
7
9
Fungspo
Santana Moss
7
13
Neck
Ricky Proehl
6
12
Flash
Eddie Kennison
6
11
Flash
Jason Elam
6
11
Marko
Ike Hilliard
6
10
Rusty
T.J. Duckett
6
12
Unk
Corey Bradford
5
Something that jumps out immediately upon examining this list is
that there are four kickers on it, three of whom are in the top
four. How can a kicker be a steal, you ask? Well, what this indicates
is that there is great value to be had by waiting until the end
of the draft to acquire a kicker. With the exception of T.J. Duckett,
who benefited from Warrick Dunn's fragility, all seven of the remaining
steals on this list were middle-of-the-road WRs. This shows (at
least in this league) that there is value to be had even with the
second and third receivers on an NFL team.
But how to predict guys breaking out like this? That's not too
easy to do. Certain things to look for include: playing in the
final year of a contract (playing for $$), WRs in their second
to fourth year with a team who have steadily increased their stats
but not yet exploded, backups for fragile starters, and situational
players who are favorites of their coach. Of course, there's also
a great deal of luck involved in snagging the steals in the draft.
But there are things for which to look.
Top Ten Busts of 2003
As you can see in the table here, the worst busts of 2003 where
predominately QBs. For the most part, presence on this list was
predicated on a player either getting hurt badly enough to miss
most of the season or sucking enough so that they lost their starting
position to someone else who didn't suck. How many of the QBs
on this list were hurt last year? They're certainly fragile, aren't
they?
Top
Twelve Busts of 2003
Round
Owner
Player
Degree
4
Marko
Kurt Warner
-26
3
Fric
Rich Gannon
-26
7
Fungspo
Drew Bledsoe
-21
2
Mooch
William Green
-21
2
Flash
Corey Dillon
-21
2
Ox
Trung Canidate
-21
9
Flash
Michael Vick
-20
7
Hair
Chad Pennington
-20
2
Fungspo
Donovan McNabb
-20
12
Frac
Joe Nedney
-19
6
Frac
Amos Zereoue
-19
10
Fric
Kelly Holcomb
-18
Looking at this list of the worst busts with hindsight, it certainly
isn't difficult to see why they had such problems. Injuries, as
I've mentioned above, certainly played a major factor for the QBs
and for Joe Nedney. Drew Bledsoe came back down to earth after playing
out of his mind for the first half of 2002. Of course, it didn't
help that his supporting cast kinda went belly-up as well. William
"420" Green is doing his best Lawrence Phillips impersonation
in Cleveland, and performed accordingly. Corey Dillon's attitude
finally became an immovable object, and now he's moved to the north-east.
Trung Canidate was never really given a chance, but who really thought
he had a chance with wacky Stevie as his coach? Not-so-famous Zereoue
just didn't really do a whole lot of anything, even given tons of
opportunity.
20/20 Hindsight Picks
There's one final class of picks that I haven't addressed yet:
those players who weren't drafted at all, but based on their performance
throughout the year, probably should have been. Should any of
them really been drafted, or is this just hyperbole? Let's look.
Tweleve
Players That "Should Have" Been Drafted
Round
Pick
Pos/Rank
Player
2
20
WR10
Anquan Boldin
3
29
RB13
Brian Westbrook
3
30
WR14
Steve Smith
3
31
WR15
Javon Walker
3
32
RB14
Rudi Johnson
4
37
PK3
Matt Stover
4
38
TE2
Shannon Sharpe
4
41
WR20
Justin McCareins
4
45
WR21
Marcus Robinson
4
46
RB17
Domanick Davis
4
47
PK4
John Kasay
4
48
WR22
Bobby Engram
Let's get the out-of-left-field players out of the way right
away. There's no way anyone could have known that Anquan Boldin,
Rudi Johnson, and Domanick Davis would emerge from the depths
of the rosters to enjoy the success they had. How about the other
RB on the list: Brian Westbrook? Yeah, he probably should have
been drafted, especially considering the number of rookie RBs
that were drafted instead of him. Shannon Sharpe? His return to
Denver obviously reinvigorated him, and knowing his history with
Shanahan, he should have definitely been drafted. The two kickers?
Eh, who can ever tell with kickers? It's all pretty much a crap-shoot.
That leaves the five WRs. Steve Smith and Javon Walker most definitely
should have been drafted, as they personified the definition of
sleepers. Justin McCareins? Nope - there was no real indication
that he'd make the leap last year. Marcus Robinson and Bobby Engram?
They pretty much emerged from several years of obscurity and therefore
couldn't have been predicted to succeed.
So that means that four of the surprises from last year should
have been drafted. If the guys in the league had done their homework
a little more diligently, they would have been able to snap up
these players at the draft instead of having to compete for the
regular season addition.
Wrapping This Up...
So, what does this all mean? First off, if you lasted all the
way to this summary, you're a true fantasy football freak. Welcome
to the club!
There really is a method to the madness; it works if you know
what you're doing. Sure, you can rely on luck, but the odds are
that you're going to come out of the season poorer than you went
into it.
So, if you have indeed read all of the preceding analysis, perhaps
you've already made the intuitive jump to what I'm going to say
here. For the rest of the class, let me be explicitly clear: the
most important part of the season is the draft. Therefore, it
is something for which you must prepare thoroughly. Sure, you
might be able to recover from a poor draft through a series of
roster moves throughout the season, but more often than not if
you mess up your draft, you've messed up your season. I'll be
even more explicit:
If you draft well and don't fall asleep at the switch during
the season, you're going to do well.
Even if you draft well but then don't pay attention, you're
not going to do well at all.
If you don't draft well, you still have an outside chance at success
during the season, but you're going to have to churn your roster
like crazy to have any shot.
Look, it really is quite simple. Let me put it in terms that might
be more accessible: think of an analogy to home brewing. Sure, you
can make home-brew with unfiltered tap water, that stale yeast that's
been sitting in your cupboard for three years, and in a dusty basement
next to the cat litter, but no matter what you do after that your
beer is going to blow chunks. Same thing with the draft: you've
got to be prepared by knowing the players and the situations in
which they will be performing this year, and then know when to draft
them.
So let's recap the take-home points:
Minimize the number of bust picks you make, especially early.
Be conservative as long as you possibly can, only taking flier
picks in the late rounds of the draft.
Let the other guys make the risky picks.
Don't get suckered into runs on a particular position: always
draft for value.
Wait as long as possible to draft a QB, preferably until halfway
through the draft.
(Bonus corollary: Don't draft rookie QBs.)
Take your PK in the last two rounds, depending upon availability.
Avoid players ripe for a fall.
Do your homework to find those sleepers ready to explode.
If you've got a decision to make between loading up on WRs/TEs
or RBs late in the draft, take WRs. More of them score points.
Develop your own method for assessing the relative worth of
players and use that to project how well you think each player
is going to do this year. Then, use that list to draft players
according to their relative value to other players in the draft.
I won't lie to you: it's not a trivial task to prepare for the
season like this. But it can add to your enjoyment of fantasy football,
especially considering that more often than not there's a reward
at the end of the year.