8/27/09
This article is a continuation of the last two articles where we
investigated historical QB and
WR performance. Recall that
in those articles we dug into the distribution of each player’s
historical scores. We noted that if two players both score a lot
of points, that in some cases we might prefer the player to be consistent
than to not. And we isolated some specific players that stood out
as being more or less consistent than their peers. We concluded
with practical advice that we hope might be useful to some in upcoming
fantasy drafts. This week, we apply the same idea to RB’s.
Approach
We focus on only the last four seasons. Our intention is to consider
only games in which a fantasy football manager would have considered
starting the player. Prior articles (on
QB and WR) go into our subjective method of choosing the games to
include in some more detail. This time around we take a similar
approach and apply it to RB’s.
The table below shows the sample size we get for each player.
Note that since we look at 4 years of games in Weeks 1 through 16
the maximum possible total is 60 (4 years of 15 non-bye weeks).
LaDainian Tomlinson is the only running back with a sample size
of the full 60 games. Thomas Jones and Jamal Lewis are not too far
behind with 59 and 58 respectively. Pierre Thomas has the smallest
sample size. Sometimes having more information is useful, and we
should keep this small sample size for Pierre Thomas in mind when
interpreting results.
Scoring
We focus again on a PPR scoring system, consistent with last week’s
article on WR’s.
PPR Scoring |
FPts Per Yard -
Rushing/Receiving |
FPts TD -
Rushing /Receiving |
PPR |
Fumbles Lost |
0.1 |
6 |
.5 |
-2 |
|
We award points for both rushing and receiving, and we give no credit
for special teams or passing.
Totals
Let’s start by looking at the average fantasy points scored
per game under this set of assumptions. We look at the points per
game rather than the total points, since we don’t want to
imply a player has been more valuable just because they have been
around for more seasons.
LaDainian Tomlinson is at the top, which probably isn’t going
to shock anyone. Pierre Thomas is next, which is probably less expected.
Keep in mind, we’re only considering 7 of his games, the last
7 games of last season. So you’ll have to interpret for yourself
how meaningful those are. DeAngelo Williams stands out at being
close to the bottom, highlighting the fact that last season’s
breakout was not a standard performance for him. His value is obviously
sensitive to the week-to-week carry split with Stewart, and we assume
that a fantasy owner can’t tell in advance who will get more
carries, which is probably consistent with reality for the most
part.
This gives us some perspective on the average fantasy points per
game the top RB’s have been scoring over the past few years.
We continue our analysis by next investigating the distribution
of their scores per game.
Volatility (and Coefficient of Variation)
Volatility is a measure which quantifies how widely a data set varies
from its mean. If a player scores about the same amount of points
almost every game, his scores will tend to have low volatility.
If a player is just as likely to score 40 points in a week as 0,
then his scores will tend to exhibit higher volatility. Let’s
look at the volatility of the RB scores.
Note the high scoring RB’s tend to have higher volatility
and the low scoring RB’s tend to have lower volatility. (two
notable exceptions are Matt Forte and Pierre Thomas). Part of this
might be due to the higher scoring RB’s being more volatile,
but another piece is due to the fact that just looking at the volatility
without scaling it for their average scores will tend to effectively
overstate the volatility for the high players and understate it
for the lower players.
Let’s scale their volatility by their average score. In other
words, we’ll look at their coefficient of variation (CV).
This will allow us to compare RB’s who average low scores
against those who average high scores a little bit better.
- Matt Forte and Pierre Thomas score a lot of points, and they
do it with a low coefficient of variation. That is, in their
sample, they not only score a lot but they do it consistently.
Matt Forte especially stands out here in that he has a bigger
sample and his CV is much lower than anyone else. It’s
worth noting that both of these players have a small sample
size.
- Tim Hightower and Kevin Smith don’t score many points,
and their CV is low. This means that they do not score many
points, and they are pretty consistent at doing it.
- LenDale White does not score a lot of points on average,
but he has a high relative CV.
So does this tell us we should take Pierre Thomas #2 in our fantasy
drafts? Not exactly. It does tell us that Matt Forte and Pierre
Thomas have scored a lot of points historically and they have
done it consistently. Those are favorable characteristics to have
and they compare favorably to some of their peers in that regard.
Their sample (especially Pierre Thomas’s) is fairly small.
When choosing a running back in the later rounds of a draft, the
available options probably will not be scoring a lot of points
historically on average. If they were, then they probably would
tend to have been picked earlier (Larry Johnson is one exception
this year, as views of his future performance tend to be a lot
worse than those of his historical performance). When choosing
a RB like that, a high coefficient of variation tends to be more
attractive than a low one. In other words, if a RB doesn’t
score many points on average, then it’s preferable for him
to be inconsistent than consistently bad. In some ways when people
talk about a player with a low average score with upside, they
are indirectly referring to a high coefficient of variation.
Distribution of Scores
Next we investigate the distribution of scoring per game of each
running back. The following table shows the maximum, minimum,
and percentiles of scores for each player. 90th percentile indicates
90% of the time the player scores less than that score. Median
indicates 50% of the time he scores more, 50% of the time he scores
less. The table is sorted by the median score.
So 90% of the time Brian Westbrook scores less than 35.7 fantasy
points, but 10% of the time he scores more.
The least points Matt Forte has scored is 11. That’s a lot
of points for a worst day. Note that it is higher than the median
performance of players including Ryan Grant, LenDale White, DeAngelo
Williams, and Cedric Benson.
The following table shows the same ideas expressed as a ranking
rather than raw scores (again sorted by the median).
The table is sorted by the ranking of their median performance from
best to worst. The 2 for Brian Westbrook under 0.9 indicates that
the 90th percentile score of Brian Westbrook is the 2nd highest
90th percentile score of all the players considered.
The 1 for Adrian
Peterson under Max indicates that his highest score is the highest
high score of all the players considered. The 22 under Min indicates
that his lowest score is lower than all but 6 wide receivers.
Let’s step back and see if anything stands out.
- Pierre Thomas – there he is again. He has the highest
median score. And he scores in the top 6 at all percentiles.
This means his good days are better than most other’s
good days. And his bad days are better than their bad days too.
This is a good combination. Keep in mind (especially in this
stat) that his sample size is small. It remains to be seen if
he can keep this up, but this is certainly a great start for
him.
- Joseph Addai is in the top 5 at the high end, indicating
his good days have been better than most. But they drop quickly
and at the middle and bottom his scores range from mediocre
to bad, indicating his average days and bad days are worse than
most.
- Adrian
Peterson scores well at most percentiles, but lower than
some might expect.
- Maurice
Jones-Drew scores in the middle of the pack to above average
for the most part, and has no glaring bad points. It will be
interesting to see if he is able to break out now that he is
moving more towards a primary back role than his traditional
shared role with Fred Taylor.
- Even in this PPR format, Reggie Bush has a profile that is
not all that attractive. Note that his 75th percentile good
day is just barely ahead of Cedric Benson and Tim Hightower.
For a RB known for his “upside” this is a little
disappointing to see.
- Larry Johnson’s profile is better than you might expect,
even after last season’s disappointment. He scores in
the top 6 of all RB’s at most percentiles.
- LaDainian Tomlinson has the most attractive profile, scoring
in the top 4 at every percentile. If you think about this, this
is actually kind of amazing. This means that his good days are
better than almost everyone, his average days are better than
almost everyone, and even his bad days are better than almost
anyone He also happens to have the highest average score too.
And on top of that, he’s played in every game in the 4-year
sample. It’s no surprise that LaDanian Tomlinson shines
through positively every way you look at him historically.
There’s not a lot of shocking
stuff in the analysis we presented. The big name RB’s shine
through positively for the most part, roughly in line with (at
least my) expectations at the start.
Conclusion
We highlighted some wide receivers that have scored well against
their peers anyway you look at it. Most of the “top-tier” RB’s
score well across the board, but LaDainan
Tomlinson really stands out from the crowd. Matt
Forte is another RB who stands out favorably among the top
RB’s. Pierre
Thomas stands out as the biggest surprise in just how favorable
he looks, although it’s important to keep in mind he has a small
sample. LenDale White stands out favorably against his peers at
the bottom end of the top RB’s.
Sometimes having good perspective about the performance of available
players historically is useful. This article aims to help build
some of that perspective, and I hope you find it useful.
Obviously you would rather have your players score a lot of points
and be a little inconsistent than score few and be consistent.
When comparing peers who score a similar amount of points, sometimes
you are looking for someone a little more consistent, and sometimes
you are looking for someone with a little more upside (i.e. volatility).
Hopefully this article provides some numbers that help you to
do some comparisons like that in practice.
Next Steps
Next week will dig into the difference of PPR and non-PPR leagues,
and try to isolate some players who have stood out in the past
and might stand out in the future. As always, feel free to contact
me with any questions or suggestions. See you next week!
|