Early Returns on the Warriors-Pacers Trade

January 27, 2007
Al Harrington

Al Harrington is looking good in Oakland.

It’s been a little over a week since the Warriors and Pacers swapped quartets of players. Let’s take a quick look at the fantasy impact so far.

To date, the biggest beneficiary (by far) has been Al Harrington, who’s been playing some strong ball since arriving in Oakland, and is the 14th-ranked player over the past two weeks. Stephen Jackson has had one great game and two bad ones (not to mention his off-court distractions). On the Indy side of things, Troy Murphy has seen a bump up in value, playing at about a top-80 level since the trade, and Mike Dunleavy has improved a little as well, playing near a top-100 level.

Overall, from a fantasy perspective, this has been one of those trades that truly helps both sides (at least so far).

Advertisements

Correlating the Categories

January 18, 2007

The other day, Patrick suggested something (in response to the post comparing category distributions) I’ve been thinking about doing for a while too: Generating correlations between all the roto stat categories. This way, you could see which categories “go well” together, for the purposes of fine-tuning your team.

Without any further ado, here it is (based, of course, on the RotoPoll data):

FG

FT

Points

Reb

Ast

Threes

Steals

Blocks

Overall
FG


FT

0.18

Points

0.35

-0.26

Rebounds

0.86

0.15

0.35

Assists

0.34

0.53

-0.34

0.20

Threes

0.49

-0.30

0.25

0.37

-0.29

Steals

0.31

-0.45

0.35

0.33

-0.42

0.27

Blocks

0.50

-0.26

0.15

0.30

-0.13

0.52

0.20

Overall

0.18

0.41

-0.43

-0.01

0.67

-0.33

-0.43

-0.18


For those not familiar with the concept, a correlation is a statistical measure of how well two sets of data match up. In the case of basketball stats, for example, two categories will have a high correlation if players who do well in one of them also tend to do well in the other. A negative correlation suggests that players doing well in one category tend to do poorly in the other. A correlation around zero suggests there’s no relationship at all.

Ok, so what does this mean? To help demonstrate, I’ve highlighted the best (FG and rebounds) and worst (FT and steals) correlations. If your team needs to improve in both FG% and rebounds, the good news is you’ll be able to find a lot of players who can help you in both. If you need FT% and steals, however, you’re going to have a much harder time finding someone.

This data may also be useful if you’re considering tanking (giving up in) a category. For example, if you intend to tank FG%, it’ll be hard to do it without taking a hit in rebounds as well.

Another interesting thing to observe is the correlation between cateogories and overall value. Two cateogries, FT and assists, stand out as being the most correlated with overall value. I’m not sure, but this seems to suggest that those categories are more valuable in some way. For one thing, it means that it’s harder to find low-value players who can help you in those categories, as opposed to, say, steals or points, where the correlation with overall value is actually negative. So hold on to Steve Nash.

And of course, if anyone knows stats well enough to suggest or criticize something here, by all means, go ahead.


Comparing the Categories

January 14, 2007

In fantasy, each category is a little different. For example, players who get a lot of blocks are rare, rebounds aren’t so rare, and a single free throw albatross (such as Shaq for the past 10 years) can sink your team. Unfortunately, however, most of this type of knowledge comes from intuition, observation, editorial opinion, or other informal sources. Let’s look at some hard data that will hopefully help you evaluate the differences across the categories, and make decisions to help your teams.

The following is the distribution of scores, in each category, for the top 180 players in the RotoPoll rankings.

Category Maximum Median Minimum
Blocks 3.22 -0.33 -0.82
Assists 3.08 -0.62 -1.58
Points 2.60 -0.93 -2.32
Rebounds 2.28 -0.44 -1.42
Steals 2.11 -0.40 -1.69
Threes 2.02 -0.36 -1.14
FT 2.01 -0.00 -2.57
FG 1.78 -0.10 -1.40

And here’s a distribution of each category (click to see a larger version):

category distributions

So what can we learn from this data? Here are some initial thoughts.

  • The category with the highest average is free throws, and if you look at the chart, you can see that it drops off precipitously at the end (in fact, it has a noticeably different shape than all other categories, which tail off gradually), meaning that there are just a few guys who are really dangerous. It seems you want to avoid those guys (Ben Wallace, Tyson Chandler, etc.), and if you need to bump up your FT% a little, there are a lot of guys out there who can help you.
  • By contrast, the category with the lowest line on the graph is points. This means that it’s harder to find someone who can help you in the points category than any other category. This might seem counterintuitive, since almost everyone scores some points, and it’s not that hard to find a waiver pickup who can give you 10-12 per game, but it’s harder to find someone who can actually help you than in other categories.
  • Blocks is the most top-heavy category on the graph. It has the highest minimum and maximum, which suggests that production is concentrated in fewer players than in other categories. Conversely, a player who gets no blocks doesn’t hurt you as much as a player with a zero in another category (or poor percentages).

And of course, these observations apply to each category in greater or lesser degrees.

There’s certainly a lot more to take away from this data, but that’s a start.


Buy Low and Sell High Opportunities

December 30, 2006

This idea first appeared in a comment on the post about who’s over- and underperforming, but I think it’s good enough that it deserves a quick post.

Basically, the idea is that the list could be a good way to find guys to buy low or sell high.

While this might not be all that helpful in trying to predict a player’s performance going forward, it could be very useful in helping pinpoint which players to buy low/sell high on. Take someone like Kirilenko, who shows a -3.06 score differential. This is a quick and easy way to see that Kirilenko’s owners are probably pretty frustrated with him right about now (I know, I’m one). If you think he might somehow turn things around in the last 2/3 of the season, you can probably pick him up on the cheap. Conversely, Gilbert Arenas’ owners who don’t believe he’ll sustain his otherworldly level of play much longer can probably swap him for a KG/Marion/LeBron, straight up.

I suggest using the RotoPoll list in conjunction with “real” lists (created by humans) like the one over at GMTR. There’s a lot of bargain hunting you can do out there.


Who’s Outperforming, and Who’s Disappointing

December 29, 2006
Kevin Martin

Kevin Martin is flying high.

Ok, this should be fun… Today we’ll take a look at which fantasy players are outperforming their preseason expectations (that is, who was underrated coming into the season), and who’s been underperforming (the biggest busts).

The obvious way to do that would be to compare their preseason ranks or draft position (I’ll use ESPN’s) against where they are now. Doing a simple subtraction yields the following list:

Player Rank diff
J.R. Smith 122
Josh Childress 99
Al Jefferson 94
Kevin Martin 84
Sean May 74

However, I don’t think that gives us the full story. Clearly, players who were ranked very low have an advantage in this measurement, and a guy like Gilbert Arenas, in the midst of a monster year, has no chance of showing up. So another idea would be to use the ratio of the ranks:

Player Rank ratio
Gilbert Arenas 8.00
Kevin Martin 4.11
Caron Butler 3.62
Carmelo Anthony 3.38
J.R. Smith 3.03

Aha, there’s Gil. This list looks a little better to me, and notice J.R. Smith is still up there. But actually, I think the best way to meausure this is to measure the player’s current RotoPoll score against the score he would have, if he still occupied the same position as the preseason rankings. That is, since Gilbert Arenas was 8th on the preseason list, we measure his current score against that of the player currently ranked 8th (Carmelo Anthony). This should give us the truest measure of how much more a player is helping his fantasy teams than we expected him to.

So, without any further ado, here’s the full list:

Player Score diff
Kevin Martin 4.90
J.R. Smith 4.15
Caron Butler 3.44
Josh Childress 3.42
Mo Williams 3.10
Josh Howard 3.02
Al Jefferson 2.86
Andre Iguodala 2.79
Andris Biedrins 2.66
Baron Davis 2.59
Jarrett Jack 2.54
Leandro Barbosa 2.48
Sean May 2.37
Carlos Boozer 2.37
Zach Randolph 2.36
Mike Miller 2.35
Zaza Pachulia 2.14
Gilbert Arenas 2.11
Carmelo Anthony 2.00
Udonis Haslem 1.95
Jamal Crawford 1.93
Rashard Lewis 1.89
Antawn Jamison 1.69
Brevin Knight 1.68
Danny Granger 1.53
Andres Nocioni 1.51
Luke Ridnour 1.46
Chris Duhon 1.41
Deron Williams 1.32
Rafer Alston 1.32
Shane Battier 1.30
Richard Hamilton 1.29
Manu Ginobili 1.23
Alonzo Mourning 1.23
Yao Ming 1.22
Luol Deng 1.16
Kwame Brown 1.13
Luke Walton 1.13
T.J. Ford 1.03
Dwyane Wade 1.01
Jamaal Tinsley 1.00
Jorge Garbajosa 0.98
Travis Outlaw 0.95
Mike Dunleavy 0.87
Tayshaun Prince 0.87
Grant Hill 0.83
Devin Harris 0.83
Allen Iverson 0.74
Ryan Gomes 0.73
Steve Nash 0.73
Andre Miller 0.65
Kenny Thomas 0.63
Emeka Okafor 0.61
Marcus Camby 0.59
Ben Gordon 0.58
Joe Johnson 0.54
Andrew Bogut 0.47
Josh Smith 0.40
Samuel Dalembert 0.37
Erick Dampier 0.31
Michael Redd 0.25
Brendan Haywood 0.24
Wally Szczerbiak 0.23
Derek Fisher 0.19
Jermaine O’Neal 0.13
Nene Hilario 0.11
Tony Parker 0.10
Kurt Thomas 0.10
Jason Williams 0.09
Troy Murphy 0.09
Jason Terry 0.09
Nenad Krstic 0.07
Shaun Livingston 0.05
Shawn Marion 0.00
Raja Bell 0.00
Rasheed Wallace 0.00
Vince Carter 0.00
Paul Pierce -0.00
Hakim Warrick -0.01
Kevin Garnett -0.02
Kyle Korver -0.02
Nazr Mohammed -0.02
Kendrick Perkins -0.06
Chauncey Billups -0.16
Cuttino Mobley -0.17
Shareef Abdur-Rahim -0.18
Jason Kidd -0.23
Eddy Curry -0.24
Quentin Richardson -0.24
Donyell Marshall -0.25
Shelden Williams -0.27
Stephen Jackson -0.28
Tim Thomas -0.33
Marko Jaric -0.33
Lamar Odom -0.34
Juan Dixon -0.36
Earl Boykins -0.36
Pau Gasol -0.36
Drew Gooden -0.40
Marvin Williams -0.46
Jerry Stackhouse -0.47
Smush Parker -0.52
Dwight Howard -0.57
Paul Millsap -0.68
Ray Allen -0.70
Chris Wilcox -0.72
Ron Artest -0.79
Tyson Chandler -0.81
Matt Harpring -0.83
Charlie Villanueva -0.86
Ricky Davis -0.93
Raymond Felton -0.96
Mehmet Okur -1.08
Joel Przybilla -1.09
Vladimir Radmanovic -1.10
Martell Webster -1.13
Al Harrington -1.13
Dirk Nowitzki -1.20
David West -1.21
Sebastian Telfair -1.22
Ben Wallace -1.29
Andrew Bynum -1.29
Steve Francis -1.36
Amare Stoudemire -1.37
Theo Ratliff -1.37
Sarunas Jasikevicius -1.41
Antonio Mcdyess -1.41
Chris Paul -1.45
Kobe Bryant -1.46
Delonte West -1.48
Andrea Bargnani -1.51
Tim Duncan -1.53
Ike Diogu -1.60
LaMarcus Aldridge -1.63
Elton Brand -1.64
Hedo Turkoglu -1.67
Larry Hughes -1.67
Jameer Nelson -1.75
Chris Webber -1.81
Mike James -1.82
Morris Peterson -1.82
Antoine Walker -1.84
Peja Stojakovic -1.85
Kirk Hinrich -1.90
Carlos Arroyo -1.90
Zydrunas Ilgauskas -2.00
Boris Diaw -2.09
Desagana Diop -2.09
Jordan Farmar -2.15
Rudy Gay -2.18
Corey Maggette -2.23
Speedy Claxton -2.25
LeBron James -2.27
Darko Milicic -2.34
Chris Bosh -2.45
Marquis Daniels -2.48
Sam Cassell -2.54
Channing Frye -2.55
Primoz Brezec -2.61
Jamaal Magloire -2.71
Chris Kaman -2.79
Juwan Howard -2.85
Rajon Rondo -2.94
Marcus Williams -3.02
Mike Bibby -3.02
Andrei Kirilenko -3.06
Tracy McGrady -3.13
Michael Finley -3.14
Eddie Jones -3.27
Brandon Roy -3.30
Gerald Wallace -3.33
Randy Foye -3.55
Stephon Marbury -3.81
Adam Morrison -3.86
Richard Jefferson -3.97
Tyrus Thomas -4.02
Jalen Rose -4.36
Bonzi Wells -4.42
Brad Miller -5.29
Shaquille O’Neal -5.38
Jason Richardson -5.39
J.J. Redick -5.77
Kenyon Martin -5.98

Congrats if you drafted guys at the top of this list.


The Cost of a Missed Putback

December 27, 2006
Josh Howard

Please don’t miss, Josh.

Just a quick little tidbit for you. Have you ever been watching a game when one of your guys tries to tip in a rebound and misses? You probably thought to yourself, “Oh well, he missed the shot, but at least it’s a rebound.” But is the rebound worth more than the missed shot cost you? I think the answer is pretty surprising.

Using the RotoPoll Rater Creator, we can see the relative effect of the rebound and the missed shot. Based on the numbers for the season to date, a rebound adds 0.36 (roughly) to a player’s overall RotoPoll score. Incredibly, missing a single field goal subtracts 0.59 from a player’s score. That’s right: Missing a shot costs you almost twice as much as a rebound helps you.

For those of you suspiciously thinking, “Wait a second, that doesn’t seem right. Maybe your scoring system is screwed up,” I can only offer to let you take a look at how the scores are calculated and offer your opinions. I stand firmly behind these scores and the methodology for calculating them (and I think this is the same way most sites do it).

Sample data:
Before the rebound and missed putback
After the rebound and missed putback
For a net result of -0.23. The difference holds up, no matter what stats you use as a baseline.

Just another thing to think about while you’re watching your fantasy teams in action.


Who is the “most average” NBA player?

December 26, 2006
Monta Ellis

Monta Ellis: Mr. Average.

Most average sounds like an oxymoron. Let me explain.

The RotoPoll player rankings are based on the statistical distributions, in each stat category, of the top 180 NBA players. This means that the “average” player, at least for the purposes of roto leagues, would have a score of 0.0 in each category. To put it another way, the “most average” player is the one closest to the league average in all categories.

Obviously, nobody has perfect zeroes across the board, but who is the closest? Let’s take a look. So far this season, the Top 10 Most Average players are:

Monta Ellis
Lamar Odom
Tayshaun Prince
Troy Murphy
Luke Walton
Mike Dunleavy
Mo Williams
Andre Iguodala
Jameer Nelson
Luke Ridnour

(These rankings are created by measuring the absolute value of the standard deviation from the mean, in each category, and summing them for each player.)

I guess a more positive (and flattering to the fine men listed above) way to describe average would be well-rounded, although I’m kind of partial to the way The OCC put it, fantasy-neutral.

Why is a score of zero good for 40th in the rankings, not 90th?

As of today, a player with a score of 0.0 would be ranked 40th, right between Deron Williams and Amare Stoudemire. Clearly, those guys are not “average” players in any sense of the word. So why aren’t their scores higher? Why is zero such a good score?

This is an interesting question. Since the rankings are based on the top 180, you’d expect the 90th-ranked player (or someone thereabouts) to have an overall score of zero. It turns out that, as with most statistical populations, the stat categories are skewed toward the top of the distribution. That is, the top few guys are way out ahead of the pack, and it levels off after that. When you add up all the categories and their skews, you get a top-heavy list. This doesn’t mean the results are inaccurate. Rather, it highlights the overwhelming value of the elite players. Experienced fantasy owners already know this, and therefore try to be on the slim end of unbalanced (2-for-1, 3-for-2, etc.) trades.

I wonder how Amare would react if I told him he was an average player. Mabye you should tell him instead.