SaberCards: A (concept of a) Baseball Card for Stat Geeks
by adarowski on Jun 2, 2010 11:00 AM EDT
in Visuals Above Replacement
Email
PrintGraph of the Day
Like many readers of this site (I assume), I used to collect baseball cards as a kid. I loved reading the backs. Now that we're older and less tolerant of the stats available during Baseball 1.0 (you know, before sabermetrics), what would our ideal baseball card look like? Today I decided to take a first stab. I bring you: SaberCards.
This is merely Concept #1—I wouldn't mind continuing to work on this. Feedback is welcome and appreciated.
The RAA categories:
- Offense: Batting, Reach on Error, and avoiding the Double Play
- Defense: Range (Total Zone), Infield Double Play, Outfield Arm, Catching
- Baserunning
- Position
Data comes from Rally's WAR database. The excellent photo of Mr. Jones was posted to Flickr by Keith Allison with Creative Commons rights.
67 comments
|
7 recs |
Do you like this story?
Comments
Ok, this is by far the coolest thing I've seen in a while
Adam wins Graph of the Year as far as I’m concerned.
Graph of the Forever.
See Data Differently. Beyond The Boxscore. | Follow me @justinbopp
Two Out Rally, the new BASEBALL MMORPG! | Facebook | Twitter
Cool idea and great execution!
Jonathan Sanchez: Often maddening to watch, but capable of perfection on a moment's notice---just like his adoptive father.
The main thing I’d like to see added is a cumulative career bar/line.
Jonathan Sanchez: Often maddening to watch, but capable of perfection on a moment's notice---just like his adoptive father.
It might be cool to make an online version of these
with several different views of the data on the back. That would resolve a lot of these requests.
See Data Differently. Beyond The Boxscore. | Follow me @justinbopp
Two Out Rally, the new BASEBALL MMORPG! | Facebook | Twitter
Maybe, but the appeal to me is having an actual card to hold
"There's never enough time to do all the nothing you want" -Bill Watterson
That's beautiful.
Great work, Adam.
My first thought is how will it scale for a truly huge season (say, 90-100 RAA)? And is there any way to include individual season WAR on the bar chart? Obviously 513 RAA in Chipper’s career means 76.7 WAR, but how does that play out per year? Is the conversion consistent for each year, or are the 90s different from the 00s and so on?
Again, though, it looks fantastic.
==
Check out Wezen-Ball.com
I love the concept
I do feel like it should have a little more, though; at the very least, it should include the team(s) played for each season (obviously not an issue for Chipper, but…)
Also, I feel like some AVG/OBP/SLG data should be included; off the top of my head, I feel like either listing career values in the top portion, or listing seasonal values in faded text over the RAA bars (but that could easily get too busy, so maybe it’s best not to).
Another thing I think is important to keep in mind is what scale you’d use (as lar alludes to above); on the one hand, it makes sense to scale it to whatever the individual player needs; on the other hand, it would probably be ideal to have one scale used for everyone, because then you can more easily compare them visually.
Anyway, though, all in all this is seriously awesome, and don’t take my quibbles as trying to deny that in any way.
Brewers Baseball and other assorted nonsense (mostly the assorted nonsense) at my blog, What's a Tararrel?
Great feedback
The scaling is interesting. I remember Donruss and Upper Deck would just show the past 5 years for a while there while Fleer and Topps would just squish the numbers until you could barely read them on the back of a Dave Concepcion card. I’d have to weight whether or not using the same scale for Barry Bonds and Bill Bergen would be worth it. Some cards would look mighty empty.
Of course, this assumes these are actual cards that would be created for every player. Perhaps it’s just for the WARlords out there.
Single-season WAR is something I tried to get in there but left out for now. I’d like to rework the back a bit to get more season-by-season stats in there. Perhaps an OPS+ would help add a little context as well. Team name is something I hadn’t even thought of.
On Twitter: @baseballtwit
I like the "most recent X years" thought.
As far as a “comparable” scale…
What about some type of diminishing returns type scale.
Like the first 1/3 of the scale is to 10 RAA, the 2nd 1/3 goes to 40 RAA, and the final 1/3 goes to 100 RAA.
You’d be able to line up cards to see if player X is better than player Y visually — maybe not on an actual “this line is twice that line basis, so that guy is 2x better!”, but still see that Bonds was clearly superior to McGuire for instance.
It would seem to make those really great years seem much less impressive, but it would put the cards on a standard that could be applied to any player and be graphically representative.
You know, maybe the last 5 isn't a bad idea.
Maybe version 2 can give that a whirl.
On Twitter: @baseballtwit
McGwire
I agree with the fixed scaling point. The whole point of a set of cards like this, for me, would be to compare players, not to compare the different seasons of a player. The diminishing scale idea is interesting, but the question is whether it is necessary The only way to know would be to run cards for players of various skill levels to see how it looks.
"There's never enough time to do all the nothing you want" -Bill Watterson
another idea would be to include a "percentiles" bar at the bottom
so if the scale shrinks or expands, you can still have a frame of reference.
Would you like to follow me on Twitter, Facebook, or my blog...well you can't.
by SagehenMacGyver47 on Jun 2, 2010 2:37 PM EDT up reply actions
Although you are missing interesting trivia about the player
Like “Chipper named his son Shea after the rival Mets’ stadium”
BAD ASS
you win this round, Adam!
See Data Differently. Beyond The Boxscore. | Follow me @justinbopp
Two Out Rally, the new BASEBALL MMORPG! | Facebook | Twitter
Awesome.
Yeah, that looks much better Adam – draws more attention to the information that you are trying to get across. Although I love the earlier idea of adding the “Did you know?” line that is always on sports cards too, and general biographical info…
Yes, thanks for the help!
Did You Know? stuff is always cool. I loved the early Score cards because of the additional biographical info. I’ll have to watch out before these things turn into Collect-a-Books! (remember those!)
On Twitter: @baseballtwit
Anytime!
That collect-a-book is awesome, though.
Jose always reminds kids not to drink and drive.
Awesome. :-)
Anyway, think about adding biographical information and some more stats … but of course that has to be balanced with the fact that it will become a mess if you put too much on it.
Talkin' Baseball!
See Data Differently. Beyond The Boxscore. | Follow me @justinbopp
Two Out Rally, the new BASEBALL MMORPG! | Facebook | Twitter
Well
Story should go with player…
"There's never enough time to do all the nothing you want" -Bill Watterson
I'd prefer a quote or an anecdote about the player.
Something non-sabery. Maybe a little Joe-Morgany? :)
See Data Differently. Beyond The Boxscore. | Follow me @justinbopp
Two Out Rally, the new BASEBALL MMORPG! | Facebook | Twitter
and what was the story for Darryl going to be...?
I don’t blame them on this one.
Would you like to follow me on Twitter, Facebook, or my blog...well you can't.
by SagehenMacGyver47 on Jun 2, 2010 2:38 PM EDT up reply actions
Impressive!
That is a fantastic idea and excellent execution!
MileHighReport.com member since 02/06/07, promoted to "Position Coach" (i.e. new staff writer) on 02/16/10!
A few suggestions
First, I don’t think you need to use as much real estate on the back with the picture and heading from the front of the card. Might be the traditionalist in me, but I like the photo-less card back.
I would also like to see some more stats so we know what kind of player we’re dealing with. DiamondView is the most obvious example of a clever way to do that in limited space.
I also think that whether or not you do the full career, it’s worth providing some additional info about the previous season (whether it’s traditional counting stats, salary data, other saber stats).
I look forward to your next version, and would absolutely buy a set if this really happens.
"There's never enough time to do all the nothing you want" -Bill Watterson
Diamondview is not the way to go, if that's what you're suggesting
And personally I wish people would stop using it on this website. It gives an inaccurate glance at a player’s skill due to its inherent nonlinearity.
I'd be completely open to a more accurate glance, if there's one out there
I’d just like to know more than that Chipper is a good hitter in the aggregated area. I’d like to know why he’s a good hitter. I think that’s diamondview’s goal (whether perfectly realized or not), so absent a better way I’d vote for it. If there are alternatives I’d be open to them too.
"There's never enough time to do all the nothing you want" -Bill Watterson
There are ways
that it could be used appropriately with a static data set.
See Data Differently. Beyond The Boxscore. | Follow me @justinbopp
Two Out Rally, the new BASEBALL MMORPG! | Facebook | Twitter
Uh...
Not to get too far off topic, but how is it an inaccurate glance due to its inherent nonlinearity?
See Data Differently. Beyond The Boxscore. | Follow me @justinbopp
Two Out Rally, the new BASEBALL MMORPG! | Facebook | Twitter
I didn't make the comment, but...
…two issues I have with it are:
1. Visually our brians determine magnitude based on area, and the area colored for each category is not only nonlinear (a 10% improvement in a category does not result in a 10% increase in the area shaded that color) but also is dependent on unrelated categories (as adjacent categories increase, the area for a category increases even if that category’s value remains constant).
2. Sort of related to #1, when you have a negative value for a category it detracts from the area on the opposite category, again limiting visually how we assess the magnitude. I don’t have one open now so I might have my order wrong, but for example a negative fielding value might shrink the area shaded for hitting and make it appear the player is a worse hitter.
Of course this is totally off-topic and hopefully doesn’t detract from our fantasies of saber-cards leading the revival of the baseball card!
I can understand the problem of assuming that area = greatness
1) That makes sense, but the shape gives you an idea of a player’s “dimensionality”. A 1 dimensional player will be a very thin triangle into the area that he’s good at. A two dimensional player will have two strengths that form some type of shape. An all-around player / master of none, will be a nice little square.
2) I think the most iteration did away with this piece as it was scaled as a percentile of the population. Therefore, a player with a negative -500 UZR would rank as “zero” while a playe rwith a positive 1000 UZR would rank as 100. Yes, those are bullshit and unreachable numbers just to point out how the extremes would work.
3) Another problem that I’m surprised wasn’t mentioned is that not all categories are created equal. A 100 in batting is much much better than a 100 in baserunning. But, the point is to compare how a player ranks in a stat, and it does well on that. It doesn’t give you overall value, so as a tool on a baseball card it might be nice as a “summary” as the background picture to show how the career has been compared to the population?
by Trickman on Jun 2, 2010 6:21 PM EDT up reply actions 2 recs
Agreed.
It shows a thorough understanding of not just the concept and intention, but the flaws and how they could be mitigated.
See Data Differently. Beyond The Boxscore. | Follow me @justinbopp
Two Out Rally, the new BASEBALL MMORPG! | Facebook | Twitter
I wonder if bending the lines on the box wouldn't mitigate the area issue
Something like the p=4 star here.
It wouldn’t solve the issue completely, but then solving the issue completely would kill the area effect, which is neat.
"There's never enough time to do all the nothing you want" -Bill Watterson
That'd be interesting to see
Sorry for dragging this off topic. Maybe a fanpost on diamondview’s pros and cons is in order. I agree the shape is the most interesting part, and cool to look at.
You know...
What if you took this RAA concept and instead of having a “picture” behind it, had the diamondview for the guy’s career numbers prorated to a seasonal basis on it.
So that you’d have the dark lines indicating his last 5 years of actual value, and then a diamond view behind it showing his “relative” value?
Looks great
Love the idea and the execution. Personally I like the picture on the back.
Small typography note, you are going to run into trouble when the player’s first name includes a descender, as it could overlap with the team name or position.
Yes, that's why I'm looking forward to generalizing this...
That is definitely an issue I’m aware I’m going to run into. The font choice doesn’t have that many descenders (since it is all caps), but they do exist and need to be dealt with.
On Twitter: @baseballtwit
I have a special love
for typography. There are numerous solutions for this issue, though, including moving the team/position pretty much anywhere on the card.
See Data Differently. Beyond The Boxscore. | Follow me @justinbopp
Two Out Rally, the new BASEBALL MMORPG! | Facebook | Twitter
Yup.
Goes to my “no need to duplicate header” point above.
"There's never enough time to do all the nothing you want" -Bill Watterson
But I REALLY
like the picture. Maybe a smaller typeface for the name on the back?
See Data Differently. Beyond The Boxscore. | Follow me @justinbopp
Two Out Rally, the new BASEBALL MMORPG! | Facebook | Twitter
Well, at minimum it should be a different picture
"There's never enough time to do all the nothing you want" -Bill Watterson
Unless
it requires additional investment/royalties. ;)
See Data Differently. Beyond The Boxscore. | Follow me @justinbopp
Two Out Rally, the new BASEBALL MMORPG! | Facebook | Twitter
Pictures are going to be an issue anyway, but I'd rather have a more expensive/better set
"There's never enough time to do all the nothing you want" -Bill Watterson
Also, fwiw, the Chipper Jones picture is NOT issued under a cc non-commercial license
It does, however, require that our use be under a CC license (absent author’s permission)
The bigger issue is going to be the fact that the pictures are of baseball players in their uniforms.
"There's never enough time to do all the nothing you want" -Bill Watterson
I have no idea what it would take to get a license from the MLBPA
But I’m pretty sure that’s what you’d need to do.
"There's never enough time to do all the nothing you want" -Bill Watterson
I admit...
Front was an afterthought. About 99.5% of the focus was on the back while the front was created from a “Save As”. So VERY good feedback that I defnitely plan to incorporate.
On Twitter: @baseballtwit
Hey, fair enough
I’m (obviously) emerging from BtB lurkerdom because this is a sweet idea. No first draft is ever perfect, and I’m sure some of my suggestions are comfortably within that category.
"There's never enough time to do all the nothing you want" -Bill Watterson
I just wish there were more stats
WAR is nice and all, but what about rate-based stuff? What if I want to know a player’s wRC+, wOBA, UZR/150, etc.? I feel these should be reflected on the card. Great idea, though!
I completely agree that the card needs a line of numbers
And they don’t all need to be 2009 numbers either. UZR/150, for example might be better as a multi-year weighted average.
"There's never enough time to do all the nothing you want" -Bill Watterson
I like this.
I was thinking of going with a simple OPS+, but adding wOBA or wRC+ along with UZR/150 sounds awesome. Although, I’m pretty much married to Total Zone because of my ridiculous fanboy attachment to Rally WAR…
On Twitter: @baseballtwit
wRC+ > OPS+, if you're going to use just one of them
Great job, though. I love the concept and a lot of the suggestions are great, too.
while I think those are helpful too, the nice thing about AVG/OBP/SLUG
(despite their deficiencies) is that they give you a look at the hitter’s profile (power, light hitter, etc.)
Would you like to follow me on Twitter, Facebook, or my blog...well you can't.
by SagehenMacGyver47 on Jun 2, 2010 2:42 PM EDT up reply actions
For a geek stats card...
maybe ISO to get the hitter profile?
Just checking back in
Looks like getting bumped down the front page has finished the discussion.
Figured I should offer to help/stay in the loop if there are tasks that break off easily. My screen name at gmail.
"There's never enough time to do all the nothing you want" -Bill Watterson
I'm still up for further discussion if you are.
See Data Differently. Beyond The Boxscore. | Follow me @justinbopp
Two Out Rally, the new BASEBALL MMORPG! | Facebook | Twitter
I'm not sure how much it'll help, but...
I encourage people to check out this page more often: http://www.beyondtheboxscore.com/posts/archive
It’s simply a list of the title of every post going backwards in time, with the number of comments you haven’t read listed, too. Simply, effective.
Agreed
I find on SBN blogs, the discussion on “main” posts is nearly always shorter than on fanposts since there’s an out-of-sight/out-of-mind issue.
It’s one of the larger flaws with the software platform.
"There's never enough time to do all the nothing you want" -Bill Watterson
I'll be sure to post updates, too.
I’m thrilled with the response. I can’t wait to work on version 2, which will make it’s way to a post in one form or another. Thanks!
For v2, I’m thinking:
- photo, name, team, position stay on the front
- show more stats, like wRC+/wOBA or UZR/TZ
- only show last 5 seasons
- perhaps a short bio
Sound good?
On Twitter: @baseballtwit
Sounds much better to me
I look forward to seeing what you come up with.
"There's never enough time to do all the nothing you want" -Bill Watterson
Comments For This Post Are Closed
Secondary Sidebar
User Tools
FanPosts
Community blog posts and discussion.
Recent FanPosts
Follow us on Twitter!
BTB: @BtBScore
Justin: @justinbopp
Satchel: @SatchelPrice
J-Doug: @RationalPastime
Josh: @garik16
Julian: @giantsnirvana
Bill: @BillPetti
Lucas: @DBITLefty
Jeff: @jeffwzimmerman
Lewis: @LewsOnFirst
