THE BOOK--Playing The Percentages In Baseball

When you say “video games,” are you talking about text-based simulations like OOTP, or console games like MLB: The Show? I guess LI could be used to drive manager strategy, at least in part, for either kind of game.

MLB 2K8, by 2K Sports, has WPA in this year’s game, I was shocked to see it and frankly amazed that it was.  In each box score it lists the players WPA, it shows the game graph, much like what fangraphs does, and it shows the WPA season’s leaders.  They seem like they are using new information for good not evil, hopefully others would do the same with LI.

Can you post some of the numbers when you get the chance, in various game states (3 or 4 of them)?  I know I would direct these games to other sites when I would reject them.  I just want to make sure they didn’t get their hands on my stuff.  Thanks…

Someone offlist suggested a referral system, whereby I get a chunk of any leads that end up in sales.  Seems like a good idea.

Also sounds like a good idea if B-r.com wanted LI through its Play Index premium feature (or BJ online wanted LI on its website).

This really depends on how well it can be tracked.  Amazon likely has a great tracking system, so it’d be a challenge to come up with one.

An option is to have “promotional codes”, whereby I’d get you guys a 10% discount if you mention the code TANGO, and it would come out of my share.  Everyone wins.

You could simply insist that they include wOBA - along with an explanation - in their in-game stat reports. Would make a hell of a teaching opportunity if nothing else.

Is “Leverage Index” a copyrighted name? If it’s in the public domain, why do they need to offer to pay you for it?  Answer (I think):  you created it, and they want to keep other game manufacturers or any other commercial venture from using the term.  Presumably that would also keep MLB from adopting the term as an official statistic as well, unless they paid a royalty to whoever owned the name.

(I assume they’re not going to call it “Tango’s Leverage Index” and therefor want to use your name as well.)

They can create their own version, but it is time-intensive, especially to get it right by run environment.  I don’t think the name matters.  They can call it “Stress” or something else really.  They are compensating me for all the time I spent on it, and that they get to use it in a commercial product.

If you flipped on a game tomorrow night and ESPN had a “pressure index” on the screen, and it said “151” where you calculated an LI of 1.49, or theirs said “82” when your LI was 0.79, would you have any ground to challenge them? Unless you have a patent, I don’t know if you would. They wouldn’t even have to divulge their formula to you. 

Because it’s relatively easy to “roll your own” leverage index, I’m not sure if it really has much value, because why pay for something you can do yourself in an evening? Nobody would pay for the formula for Total Average, for example, or even for the right to use it. Just add a sac fly here or a GIDP there, slap a new name on it, and you’ve got your own metric. And that’s exactly what happened to Barry Codell’s idea—popularized by Tom Boswell and all the others that followed who reinvented it without acknowledging their predecessors.

On the other hand, pitchFX and Retrosheet data have value, because you can’t just supply that information yourself. It would be harder to come up with your own pitchFX database than your own LI, because with pitchFX you need a bit of programming savvy.

So to answer your question, I don’t think LI, or any other statistic, really has much commercial value at all. If it were me, I’d let the James guys have it for free provided they mentioned my name/book/website in the glossary.

Dackle, I think it has some commercial value, the concept anyway, as Fangraphs is showing.  Also, Leverage Index has to be tweaked by run environment, which is not an easy thing to do.

As I said, I don’t care much about exposure, as Fangraphs and B-r.com already give LI that exposure.

You are correct that I’d have no grounds against someone else to rolling out their own version of LI.

I wonder if they are interested in LI for simulating clutch. If they were (making some players more ‘clutch’ than others), would you approve or disapprove Tango?

They could use it for more, like Fangraphs is using.  Imagine that they would supply, in real-time, WE, WPA, and LI to the teams (i.e., what Fangraphs gives us), and they charge them an extra 5K for it.

In an open source world, I don’t mind giving my stuff away, as long as that chain keeps going.

In any case, the parameters here, for now, is only for their book, and nothing else.

I think the referral idea makes the most sense at this point.

Rather than paying $5K, don’t you think that one of the guys already on the MLB payroll could come up with his own WPA and LI table in an afternoon? I’m sure there are plenty of employees who keep up with the current research and have decent computer skills.

MLB has range factor and “defensive efficiency rating” in its fielding stats. No mention of James anywhere, not even in the popups.  I wonder if he gets a royalty? But isn’t it better for range factor and DER to be out in the public domain (ie helping casual fans learn that there is more to fielding than errors), than for James to refuse to allow MLB to use his formulas unless he’s paid X dollars?

I never said to pay 5K. I said imagine a team would pay BIS 5K extra for the WE/LI data and analysis.

***

In an afternoon?  It depends how good you want to make it.  You want to create a very basic WE chart that is based only on a 5.0 runs per game environment, fine.  And an LI chart based off that, fine.  Very doable.  (But you don’t even have to since the WE is already in The Book, and the LI is on my site.)

What if you want one for a 4.3 rpg park, or one for a 5.8 rpg park?  Or anything in between?  Book a week.

Dackle, I don’t understand your point.  I *have* given away LI already.  To Fangraphs and Baseball-Reference.com.  And, not just the basic LI (and WE), but for all run environments from 3.0 runs per game to 6.5 runs per game.  And, I’d be happy to provide them for runs per game even outside that range if they need it.  (And they might, because of the Astrodome and some Coors years.)

The issue is bundling my work in a consumer-paid product.  This bolded part is the key.  It basically knocks out your objections.  I don’t see why I need to give away the fruits of my labor under those conditions.  And I don’t understand your strong position on the matter of my donating my work to a business that wishes to use it as a selling point.

If BillJamesOnline.net were to be free to the public, then I would gladly give them the LI and WE data.  And I would to MLB.com and ESPN.com and BP.com if they were to only use the LI so that Joe Fan gets it for free.

You seem almost indignant that I would want a consumer-driven business to pay me for my work.  Am I reading you right?

Here are the terms of my licence, of which Fangraphs and B-R.com have not paid me for:

Terms of Licence
================
You are hereby granted one (a) limited-commercial, (b) non-exclusive, (c) non-transferable, (d) perpetual-use licence, to the Win Expectancy and Leverage Index file. 

These terms are defined as follows:
(a) limited-commercial: you can use this file as a secondary, virtually valueless, feature to an already-developed advertising-driven, non-consumer-paying product.

(b) non-exclusive: you are not the sole recipient of this licence

(c) non-transferable: you cannot transfer this licence to anyone

(d) perpetual-use: you can use this file for as long as you wish

Limitations of Use
==================
1. You cannot reproduce, retransmit or otherwise copy the file. 

2. You cannot sell the file, any parts of the file, or any derivative of the file.

3. Any use of this file must be accompanied with any one of the following notices on any page in which this data or its explanation is used:
A. The Win Expectancy and Leverage Index data is licenced from http://www.InsideTheBook.com
B. WE, LI data licenced from http://www.InsideTheBook.com

Don’t you find that very reasonable?

If “leverage index” is a spreadsheet you’ve created, and a commercial outlet approached you and said they didn’t want to do the work themselves, so could they buy the spreadsheet from you for X dollars, then I can see that “leverage index” has value.

But if “leverage index” is a general concept—divide potential change in WPA by average change in WPA, then ... I don’t know if there’s much value there. I don’t think you can require people to ask your permission to make that calculation.

Dackle/14:
- paragraph 1, yes, that’s exactly it.  This is what I’m talking about.

- paragraph 2, yes, I agree completely.  My permission is not required if BP or ESPN or anyone wants to create a similar type metric, nor do I want anyone to seek my permission.  Just don’t give it the same name.

Glad we got that straightened out!

Well, it seems that their requirement is far more modest than I expected: they just need the overall LI for all relievers, and only for this year (no historical data).  So, I’m going to work with David to provide that for them as they need it, and I told them that all they need to do is provide the attribution and nothing else.  I also offered to write a mini-article if they need it.

Thanks guys.  Will be cool to see LI in print though.  I can’t wait for Mariano Rivera to say “Tangotiger.net”!

I received confirmation that LI will appear in BJ09 Handbook.

I’d love to be in the first arbitration hearing that will actually quote LI. 

***

Time to get on BP’s case now.  Don’t know why they’d be so resistant.  I thought they were avant-garde.  They’re acting more like garde-derriere.

Re-reading, it’s hard to tell if I’m being chevalier or derriere-de-cheval.

I’d like to point out, that creating your own version of WPA and Leverage Index is going to take well more than an afternoon and $5k to baseball teams or major corporations is really just a drop in the bucket.

If I were an MLB team, or ESPN, or MLB.com, why would I waste my time developing my own system when I can just buy this one for 5k.  I bet it would take these guys well over 100k in development, QA, etc.. to just get it up and running and it’s possible they’d need a new hire to work on it.

Even at FanGraphs, if there’s work that’s already been done for me somewhere and it’s affordable and it’s of high quality, why should I waste a month or more trying to come up with something that will inevitably be worse, when this off the shelf product has been slowly perfected for years. It just doesn’t make business sense.  It’d be like, why use MySQL or Oracle when I can code my own database (a bit extreme, but you get the idea).  I’ve seen stuff like this happen at companies where they do decide to code their own database (metaphorically) and it kills them in the long run.

Will be cool to see LI in print though.

I just want to point out that we’ve had LI in the THT Annual for the past two years.  First!

Tango/#17,

I don’t know about you, but it seems to me that BP, while at the forefront of sabermetrics, prefers to have things developed by writers at their employ. Thus, my guess is they’d rather have Silver or Davenport create something than buy yours. Or they’d have to hire you to feel like they could accept it, which I’m guessing you’re would not be too keen on either.

BP may have been at the forefront of sabermetrics at one point, but they certainly are not now.  Hardball Times has easily outpaced them.

If you are asking if I would do a work-for-hire for the exclusive use of BP, you are right, I would not do that.  Not unless they pay at corporate America industry rates.  Any discount of my pay is basically a donation of my work and time to somebody.  And that donation of my efforts will be made to you guys, and not for a commercial entity.

They know how to reach me otherwise…