Why I don’t create new stats

I’ve been writing about baseball online for four years, and early on, I learned I needed to work hard to create anything meaningful. A few years ago, I killed a weekend creating a stat I dubbed, “Runs Accounted For.” It looked at a player’s run and RBI totals compared to his team’s run total and, as I later learned, was more or less a simplified version of Bill James’ work, Runs Created. I didn’t know this when I posted my piece (I hadn’t read a James book up to this time), and proudly, naively, I submitted a link to Baseball Think Factory expecting to be applauded.

The response I got is fairly typical for anyone who creates a new baseball metric and is one reason I don’t devote much time inventing stats.

Before I go further, let me be clear: I applaud everyone I know or know of in the baseball writing community willing to do this kind of research. I can’t imagine analyzing baseball history, at least analyzing it well, without utilizing so many of the new metrics that have popped up in the last 40 years, stats like WAR, OPS+ or ERA+, among many others. Sabermetrics offer a better, more-comprehensive look at the goings-on in baseball. They cut through hyperbole.

In the competitive world of online baseball writing, more stats are coming out all the time, such as Hall Rating from my friend and HHS colleague, Adam Darowski. To this I say: great. Keep the stats coming. More research will unearth more useful advanced stats and help refine the ones that exist. It’s needed work. It just isn’t for me.

I learned this reading BBTF forum members mock my work a few years ago. (On a side note: I finally quit reading BBTF comments on my writing a year ago. It’s been liberating.) People repeatedly pointed out the obvious, that I had essentially invented a stat that was already long since in existence. There was lots of snark and not much positive or useful feedback. It paralleled the reception I received a couple months before that when I killed another weekend looking at how many more votes non-enshrined Hall of Fame candidates got than players who were already in. Statistical analysis is, admittedly, not my strongest suit in baseball blogging. I’m better at writing, journalistic reporting and historical research.

What’s funny is that the reception I got was actually far milder than what other stat inventors have received. Just ask anyone who’s contributed to WAR. Voros McCracken fought an uphill battle in forums about a decade ago when he created the idea of defensive independent fielding stats. Adam had his handiwork, which he’s spent the couple years I’ve known him tinkering with, dubbed an “utter load of garbage.” Countless other baseball writers, I imagine, have faced similar criticism.

I don’t know what motivates readers to level vitriolic complaints. Maybe they feel threatened or that their criticisms will stymie work they don’t agree with. It certainly helped end my brief foray into stat creation. I just don’t see the point in an endeavor with heavy labor requirements and a better chance of backlash than reward. These days, I try to stick to the things in baseball writing I’m good at while letting people like Adam fight the good fight.

0 0 votes
Article Rating
Subscribe
Notify of
guest

20 Comments
Oldest
Newest Most Voted
Inline Feedbacks
View all comments
Steve Helsing
11 years ago

Great article, Graham. Not only would I never try to come up with a new stat, I also will seldom, if ever again post a comment on a sabermetrically inclined web site, at least not if my comment were to in any way, shape or form have to do with stats. The trash talk if you misunderstand a stat can be brutal!

Hartvig
Hartvig
11 years ago
Reply to  Steve Helsing

To be fair, while I’ve seen some people make some reasonably forceful arguments for or against many statements made on High Heat Stats I don’t recall anyone being personally eviscerated anywhere near to the extent you find on so many sports sites comments sections. In fact it’s been my experience that if you do say something that is an obvious misunderstanding of a stat- something I personally have been guilty of on more than one occasion on this site- people on HHS are pretty civil when pointing out your brain fart or ignorance of the topic at hand. The rare… Read more »

no statistician but
no statistician but
11 years ago
Reply to  Steve Helsing

SH:

The fact that I’ve come to be tolerated here at HHS speaks volumes about the mannerly give and take of the site’s stats fanatics, and it isn’t just about stats anyway. Stick around and say your piece when you’ve got a point you think needs making.

Ed
Ed
11 years ago

This is definitely a very unique site. Not sure why or how but I’m glad it is.

Chuck
Chuck
11 years ago

“It looked at a player’s run and RBI totals compared to his team’s run total..”

I did the same thing. In 1991.

I sent it to SABR thinking I had re-invented the wheel, only to receive a Dear John letter claiming it of no value because Thomas Boswell’s “Total Average” was the same thing. (It wasn’t).

A year later, imagine my surprise when my work appeared, with a few tweeks but clearly identifiable, in a future publication under someone else’s name.

Lesson learned.

Copyright.

Tim Pea
Tim Pea
11 years ago

Speaking of dumb questions, how come if you add dWAR + oWAR the number is not equal to WAR? What am I missing?

Insert Name Here
Insert Name Here
11 years ago
Reply to  Tim Pea

@Tim Pea: I believe the positional adjustment to WAR is counted in both oWAR and dWAR, for some reason. So WAR = oWAR + dWAR – Positional Adjustment

RJ
RJ
11 years ago
Reply to  Tim Pea

Because oWAR and dWAR both contain the positional adjustment; adding them together would double count this. Someone like bstar will be able to explain it better, but I think that’s all there is to it.

Tim Pea
Tim Pea
11 years ago
Reply to  RJ

OK, I’ll do a little more research.

Ed
Ed
11 years ago
Reply to  Tim Pea

What RJ and INH said is correct. oWAR is basically for people who don’t trust the defensive stats. It assumes that everyone is average defensively. dWAR…well I kind of understand what is represents but wouldn’t be able to explain it to someone else.

Artie Z.
Artie Z.
11 years ago

But you shouldn’t let the comments keep you from thinking about creating new stats and discussing what you are thinking about. You could possibly write a blog post like “Hey, I was thinking about this type of stat …” and if it was something that had already been discussed someone could point that out in a few minutes. Think of it this way: If you are on a deserted island and have no knowledge of the wheel, it’s a big discovery for you even if everyone on the mainland knows what a wheel is. If you think up something that… Read more »

John Autin
Editor
11 years ago

Never doubt that a small group of snarkless netizens can change the blogosphere…. 🙂

Don’t be discouraged, Graham.

Adam Darowski
Admin
11 years ago

Great post, Graham. I finally let a BBTF post get to me recently. It was when we announced our franchise pages over at Hall of Stats. It was a fun tool that I released with a ton of caveats, which were of course ignored. One commenter went on and on and eventually said this that stuck out at me: how is that not worthy of some ridicule? All I could do was come back with: I dunno, just not the way I operate. If I see a person make an attempt at this, point out where the calculations have flaws,… Read more »

Paul McCord
11 years ago
Reply to  Adam Darowski

Smack ’em with the “nice” stick. I like it. In fact, I like pretty much everything you wrote just there. Well done.

Bryan O'Connor
Editor
11 years ago

Graham, here’s my advice, based on personal experience. Develop a new stat and introduce it on a blog that gets about 15 hits a day. When you realize a year later that the stat doesn’t accomplish what you wanted it to, refine it and reintroduce it on the same blog. Then, when you realize that you can’t figure out how to park-adjust it to give it any real value, stop writing about it and stick to your day job. At that day job, of course, you should spend way too much of your time at this site, arguing about the… Read more »

Chris
Chris
11 years ago

Wow. Some of those people are just mean. I think they are jealous of you. I too have created new stats and have been met with similar responses. I would talk about stuff like this with fellow SABR members and it is amazing how crude some can be. You should be lauded for devoting your time (weekend) into helping baseball progress. (You did not regress). There are not enough people like us who develop our own stats. As I have found sometimes in baseball research there is new ways of explaining old concepts but what you did was a novel… Read more »

Thomas Court
Thomas Court
11 years ago

We all have moments where we think we are onto something, but it turns out that someone thought of it first.

Jeff
Jeff
11 years ago

I don’t think the problem was the stat itself. Its just that its not an evaluative stat per se. It seems that the commenters won’t accept it if it doesn’t evaluate the player in question.

The stat to me seems to be a descriptive stat. It tells you what they did, relative to the team they did it on. And what’s wrong with that?

The commenters kept referencing Bill James. One of my favorite James stats is Power/Speed number. James himself said its not a useful tool. But its a descriptive stat, not an evaluative stat.

Paul E
Paul E
11 years ago

While noticing that Bryce Harper and Mike Trout had the highest all-time single-season totals of WAR for their respective ages last season, I couldn’t help but notice what a relatively strong season, as per WAR, for a 22 year old, Hank Blalock had in 2003. However, Blalock’s total was garnered in somewhat puzzling fashion when considering his 3.0 dWar “earned” at 3B: INN CH PO A E F% DP RF/9 dWAR 1167 363 110 238 15 .959 32 2.68 3.0 Hank Blalock 1333 482 125 343 14 .971 33 3.16 1.0 Eric Chavez 1184 399 91 289 19 .952 29… Read more »