Questions & Answers with Mitchel Lichtman

The place where GRB members interview baseball authorities.
Locked
mgl
AA Minor League Player
Posts: 35
Joined: May 10 06, 7:44 pm

Post by mgl »

vance wrote:Thanks for taking the time to answer our questions.

1. Based on the work you have done, could you rank the top five hitters of all time? How about top five pitchers?
I don't really delve into that sort of stuff (historical research). All of my work is pragmatic, utilitarian, and relates to modern baseball.

Ranking top hitters and pitchers, especially going back many years, is problematic because the context from era to era and sometimes even from year to year can change so dramatically. Unless one can account for context, those "all-time lists" based on any kind of statistics, even saber ones, are going to be failry meaningless.
2. I know when the SABR revolution first began, very few of the researchers, writers and thinkers were a part of the baseball establishment. Because of that most all of the cutting edge work was published and available to the masses. That leads me to this question: How much of your research, if any, are you unable to publish because it falls under the domain of your work with the Cardinals?
Good question. First of all, there were teams that used sabermetricians (or whatever they were called back then), like Chadwick and Roth, way back when. Even 20 years ago, a few teams experimented with using analysts and sabermetricians, like Mike Gimbel and Craig Wright.

You are right in that sabermetrics is a field where most of the work was and is available to everyone, mostly via the internet, and to some extent via printed material (books, journal articles, etc.). Teams could improve dramatically without ever hiring an analyst or sabermetrician, just by scouring the internet for sabermetric material, especially the many good player projection databases available for free or for a small fee.

I don't keep secret a whole lot of stuff. Most of my research is freely available or is in our book, The Book. Sure, my projection databases and UZR numbers are largely propietary, but the methodology is freely available and other people do essentially the same work.

mgl
AA Minor League Player
Posts: 35
Joined: May 10 06, 7:44 pm

Post by mgl »

onlyfacts wrote:Where do you stand on the latest theories of pitchers not having as much control over batted balls as they were always assumed to have before?
I don't know that it was "always assumed to have before." I think it was more that no one really thought too much about it, or couched it in quite the way that Voros did.

For every event and every type of player (position player, pitcher, offense, defense, etc.) there is a random element and a skill element to the event. You can estimate the skill element for just about any event, using fairly rudimentary statistical techniques, like a simple linear regression (and then looking at the correlation coefficient) over some period of time (like a year-to-year regression or correlation).

It just so happens that in the major leagues, there is a small spread of talent among pitchers in turning batted balls into outs. I suppose that is a little surprising. A small spread of talent essentially means that pitchers (at least major league ones) do not have a whole lot of "control" over their opponents' BABIP (batting average on nonHR balls in play). Most of the differences you see in BABIP, especially over the short run, is due to random fluctuation. Pitcher talent mostly lies in K, BB, and HR rates. I think we have known that for a long time. It is just that no one ever expressed it in the way that Voros did.

For my pitcher projections, I have always used individual regressions for each of the offensive events. I have always used much smaller regressions for HR, BB, and K rates, then non-HR hit rates. That is essentially the same thing as saying that "pitchers have little control over their BABIP."

For a nice follow-up discussion and analysis to Voros' work, read my "DIPS Revisited" in the BTF archives. Other people have done some nice work on DIPS as well.
Specifically the idea of giving pitchers control over all the HRs they yield seems wrong to me; after all I've seen Vlad hit one off the ground that certainly can't be placed under the pithcer's control but rather belogs to Vlad.
It is not an either/or thing. Every event has a luck and skill element for both the batter and pitcher. As I said, there is a random element to HR rate as well (for the pitcher), just not as large as non-HR hits. As well, we are not talking about the dichotomy of the batter/pitcher matchup when we are talking about whether and by how much pitchers have control over a certain outcome. We are talking about the "spread of talent among pitchers."

Whether there is a little or a lot of spread of talent among pitchers with respect to a certain event, the batter still has around half of the control over the outcome of a PA (certain events batters have more than 50% "control" and vice versa). So in the example you are talking about, one, the batter has some control over the outcome of the PA and two, luck plays a role as well.

mgl
AA Minor League Player
Posts: 35
Joined: May 10 06, 7:44 pm

Post by mgl »

ghostrunner wrote:Hi Mitchel,

I remember seeing your posts on Baseball Think Factory about the Spivey signing, and you considered it a steal at the time. I had similar feelings, but of course we know what's happened since then.

To the extent you're allowed to comment, what do you think is the cause of Spivey's apparent decline? Is it a physical/health issue, did he play above his head for a while, or is the jury still out?

Thanks.
First of all, we never know a player's true talent level, even if he is healthy and has always been healthy. We can only estimate it using his historical numbers. There is always a finite chance that we are "wrong" in our projection in any direction and at any magnitude.

I don't know what happened to Spivey or even that he is indeed substantially worse than we thought he was initially. Players can have fluke poor periods of time whereby teams give up on them (thus making a mistake). Plus poor (or good) performance over a short period of time does not signficiantly change our projecion for that player, especially when we have lots of historical data on him.

Of course, that is where scouts can come into play - identifying whether poor performance may be a fluke or a reflection of poor talent, for whatever reason. I did hear that Spivey had some physical problems, but I don't know what they were.

mgl
AA Minor League Player
Posts: 35
Joined: May 10 06, 7:44 pm

Post by mgl »

thrill wrote:Thanks for stopping by and bringing your expertise to this site!

My question is two-fold:
How directly do you work with the Cardinals front office/scouting with evaluating players, and how much stock does the Cardinals front office place in SABR analysis when evaluating free agent and trade options?
They rely on saber analysis a lot I think when it comes to FA and trade options. Obviously there are other considerations, some of which are probably a good thing and some of which are a bad thing, IMO.

I don't work with Walt or Mo (or Tony) directly. I occasionally meet with Mo about various things. I mostly send in my projection databases which pretty much tells the whole story, again, IMO. Sometimes (fairly often actually) they ask me specifically about a particular personnel move they are considering or have been approached with. My "opinion" on those moves or potential moves are based largely on those projections.

Knot Hole Gang Vet
All-Star
Posts: 1464
Joined: April 21 06, 4:15 pm
Location: Monterey, California

Post by Knot Hole Gang Vet »

This is a really great thread, thanks to all for cranking it up.

I was asked to pose a question, and having started as a Card fan in 1941, I guess I should pose something from years long gone.

I know the game is always evolving, and I do not want to take on the mantle of being against anything new. That said, I have a problem with the relief pitching handling. Your work may not have anything to do with this aspect, but the constant rightie/ leftie matchup, and a new guy to pitch to one batter drives me nuts. In the dark ages, if a pitcher gave up a few hits and runs, say 4 or 5 runs, it would look like it was time to try someone else before the game was beyond winning. That fellow would go till it looked like he was getting pounded as well. Sometimes he did so well, everyone would wonder why he did not start the game in the first place. It is a bit of the luck of the draw. But, if someone comes in and retires the side and seems in command, why pull him in favor of a guy who might not be able to find home plate. If he cannot pitch to right and left handed batters, he does not belong in the majors, in my view. Also, how is he to learn, if he never gets a chance?

Also, I have stated on here, that I feel that pitching is watered down due to there being twice as many teams as there were with the old 8 team leagues. That may account for all the heavy mashing prevailing today. Are batters fattening up on lots of second rate pitchers?

mgl
AA Minor League Player
Posts: 35
Joined: May 10 06, 7:44 pm

Post by mgl »

Reverend Redbird wrote:MGL,

Thanks for answering our questions and all your work you do!

I'm curious on your thoughts on the 2006 Cardinals.

One concern I have is about the pitching staff. They are the best at inducing ground-balls, but one of the worst when it comes to getting the strikeout. Does power pitching really win in the playoffs, or is that just a myth? How does this lack of having a dominant ace (outside of Carpenter, of course) bode for the Cardinals?

I've noticed just from the naked eye it seems that Hector Luna has amazing range at second base. What are your thoughts on Luna?

Is it too soon to begin to assess if the new Busch is a hitter's park, or a pitcher's park?
Good questions. I made some comments about the 06 Cards in a previous post. Yes, the team is not as good as it has been over the last couple of years.

No, power pitching does not win any more or less in the playoffs. As I have said many times, pitching talent, like anything else, is the sum total of a pitcher's true ability to keep runners from scoring (duh!). That is generally reflected in HR, K, and BB rates. If you are not good in one area, you can make up for it in another. If you are good in all 3 areas, you are generally a very good or great pitcher. And it does not matter what kind of team or hitter you face, in general. If you are a good pitcher, you depress offense in all hittters by around the same amount. Ditto for bad pitchers in the other direction. Good pitching does not "beat" good hitting and good hitters do not "feast" on poor pitching. Etc.

As I said in a previous post, in the playoffs, it is nice to have 2 or 3 very good starters (and 1 or 2 crappy ones that you don't have to use in your rotation). But having one great one, like Carp, who is one of the better pitchers in baseball) is a nice advanatage in the playoffs. Mulder is not so bad for a #2 starter either. I certainly wish we had a good #3 starter, but we don't (Suppan is OK and Marquis and Ponson are not good, IMO).

I have not seen Luna enough to have an "opinion" on his infield defense. And of course he plays the OF as well at times.

I don't have enough data on him to confidently say anything about his defense, objectively speaking (UZR). My guess is that he is not considered a particularly good or great defensive player by the Cards, but I don't really know. He also looks a little heavy for a great infielder. Then again, I think he came up as a SS, did he not? A decent SS is usually a very good 2B. (And a bad 2B is usually a terrible SS, like Michael Young!)

Yes, it is too early to say anything with confidence about whether Busch II is a pitcher's or hitter's (or neutral) park. Considering that it is in the same area (same altitude and weather) as Busch II, we can make some pretty good assumptions based on the dimensions of the new park as compared to the old one.

AFAIK, the foul territory is about the same size, the alleys are a little deeper and the lines are a tad shallower (I think). Therefore, I expect that the new park is a little more of a pitcher's park than the old one (which was around neutral). I could be wrong as the wind currents and hitting background have some affect on offense as well.

I'll be back later or tommorow folks! Again, thanks for the great questions. I am going to see Jerry Seinfeld tonight!

mgl
AA Minor League Player
Posts: 35
Joined: May 10 06, 7:44 pm

Post by mgl »

CardFanInChi wrote:Mitchel,

I have heard that the Red Sox use the computer simulation game Diamond Mind Baseball as a tool in doing analysis. What kind of tools are used with the Cardinals, and have you considered using something like Diamond Mind?
I did not know that about the Red Sox, but it makes sense. A good "Monte Carlo" type baseball sim can be a valuable tool for all kinds of analyses, although most of what a sim can do, a Markov sim can do as well, although once you start throwing in all kinds of variables, the line between a true Monte Carlo baseball sim and a Markov sim gets blurred.

I have a baseball sim that is as good or better than DM. We use that for all kinds of analyses as well. For example, it is an easy way to analyze and compare different lineups, as well as certain in-game strategies, like the IBB.

A sim can also be used (although I use a more blunt "instrument") to simulate a season of course. Now that we know more about the $ value of marginal wins to a team (in terms of how those marginal wins affect their chances of making the playoffs), a good estimate of a team's chances of making the post-season before the season starts and all the way through the trading deadline is essential information. For example, if a team is in the "sweet spot" in terms of their chances of making the post-season (not too little and not too great), based on their pre and in-season talent, a marginal win may be worth several million dollars or more. If that same team is not in the sweet spot (either they have little chance of making the playoffs or they have a great chance - like 80 or 90% - for example, us last year), then a marginal win might only be worth several hundred thousand dollars. The $ value of a marginal win, BTW, is how much in extra revenue an extra win "buys" a team over the next 10 years or so.

If you want to know more about how a team's chances of making the post-season dramatically affects the $ value of a marginal win, read BP's book, Baseball By the Numbers (BBBTN), which I touted in a previous post. It is an excellent book. IMO, if you buy 2 books this year (3 including The Fielding Bible), buy The Book and BBBTN!

mgl
AA Minor League Player
Posts: 35
Joined: May 10 06, 7:44 pm

Post by mgl »

RC21 wrote:Mitchel-

After Tango posted the link and the recommendation, I read the DIPS Revisited article. What, if any, headway has been made by individual teams to marrying traditional scouting info and sabermetric conclusions?
I really don't know what other teams are doing and I can't speculate on something as specific as that. I think that the goal of even the most saber-oriented team is to create a "marriage made in heaven." That does not mean that that is the optimal strategy for running a team, however. It has a lot to do with the personnel in the front office and in player development and the minor leagues. Even on saber-oriented teams, there is a lot of pressure to do things the "traditional" way. There is also pressure from the media and the fans, although different teams respond to that kind of pressure in different ways.

Bill James recently told me something I thought was interesting. I am paraphrasing him (probably butchering what he said). He said that Francona (the Red Sox manager) will sometimes ask him about a certain lineup (whether he thought it was optimal or not). James said he basically told him that within reason most lineups are worth around the same (which is true, more or less) and that optimizing a lineup is probably not worth the heat he might receive from players, coaches, media, fans, etc., if that optimal lineup is non-traditional. I don't disagree with that.
Specifically, I wonder if there are any common traits among pitchers who tend to yield fewer hits on outfield line drives or who yield more infield popups.
I think there is some good research along those lines somewhere on the web and/or in one of the recent publications. The Hardball Times does a lot of work with respect to batted ball research. I highly recommend their annual, BTW. I am not personally able to answer that question. We do know, for example, that knuckle ball pitchers have a true BABIP less than that of the average non-knuckler. So when you compute projections for knuckle ball pitchers, you have to use a (slightly) different regression formula when "crunching" their historical data.

mgl
AA Minor League Player
Posts: 35
Joined: May 10 06, 7:44 pm

Post by mgl »

stewie13 wrote:It seems to me that the biggest thing limiting the next level of sabermetrics is technology.

Just in the past few years there has been so much more data readily available for people to mine with more and more people getting their hands on batted ball data. Obviously the data has it limits as seen by some of the inconsistencies in BIS line drive data people noticed when Pinto put out his latest PMR but clearly it is good data eventhough the methods are pretty simplistic.

How much longer will it be till teams or companies are tracking the actual speed, distance, trajectory, path, etc. of every batted ball? It seems once you have that data then it be a lot easier to truely measure range/defense/etc.

I'm sure the Stats. Inc and BIS batted ball data is all very good but how hard are people pushing for even better data?

Any chance of you releasing any new super lwts data? There really isn't a lot of total player value type stats out there right now beside Win Shares.

Thanks
Good comments and questions! The Fielding Bible has pretty much the same "conclusory data" as UZR, as the methodologies are very similar. One merely has to translate their "plus minus" ratings into runs saved or cost, which is trivial (I don't know why they did not do that in the book - they should have). They also should have done some park adjustments, which they didn't, I don't think.

Also, if one wants to turn their 3-year data into a "projection" (or true talent estimate), one has to weight the yearly numbers before combining them (to get a weighted average) and then one has to add in a "regression factor."

Maybe sometime soon (in the next year or two) I'll start publishing my UZR data. I have actually done a lot with the methodology such that it is a lot better than when I used to publish the whole database.

You are right about the technology. There will soon be a lot more ganular data available for defensive research (as well as pitching and batting). Whether that comes from pressure from saber teams or simply a normal evolution in data technology, I don't know. I suspect the latter. The Cards have asked me this year what extra data we would like to see from companies like BIS and STATS.

Even when that kind of data becomes available (somewhere), there is still going to be quite a bit of lag time until good use is made of it.

What is nice now is that data that was heretefore only available to teams and private people with lots of money to spend is becoming more and more readily available to the masses for little or no money. When that happens, we see a flurry of good research that uses that data. There is still a long way to go, however, in that regard.

mgl
AA Minor League Player
Posts: 35
Joined: May 10 06, 7:44 pm

Post by mgl »

CardsnOK wrote:Thanks so much for answering our questions and being so detailed.
Thank you!
Chatting with someone like you always reminds me that there's far more to baseball than meets the eye. There's always more to learn and analyze.

Sorry if this or something like it has been asked already...

First, when Jocketty or the front office are looking into acquiring a free agent or trading for a player, what part do you and sabermetrics play in the process?
Yes, I think I discussed this already. We definitely look at my Superlwts (offense, defense, baserunning, etc.) and pitcher projections, which "tells" us one, what is this player worth in marginal runs/wins, as compared to a replacement player, and two, what is this player worth, again, in marginal runs/wins, as compared to an alternative player. To me, for a reasonably established player, there is no substitute for a good all-around projection. There really isn't. One can then look at a player's "garbage stats," his most recent stats (both of which are going to be over-emphaisezed by most teams), and a player's general reputation and scouting reports, to see which players are likely to be overvalued or undervalued. This is especially true of pitchers, where the disconnect between garbage stats and reputation and actual true talent (as reflected in a sabermetric projection) can be enormous. Sometimes that is the case with position players, but not nearly as much these days as in the past.
Second (and you knew this was coming), if Jocketty asked you to target one available player for a deadline trade who would you target and why?
I have not really thought about this, and who knows who is going to be available. And it entirely depends on where we stand in the standings at the trade deadline. Obviously we can be upgraded at 2B, the corner OF, and starting pitching (another ace releiver would be nice too). As far as specific players, I have no idea off the top off my head. Throughout the season, I will send memos to the front office when some player (almost always a pitcher) pops into my head as I am doing some research, that appears to me to be a "sleeper."

There are many pitchers (starters and relievers) that have little or no name recognition and yet are quite good.

There are also many pitchers who should not be anywhere near a major league field, but who somehow manage to do so. I won't mention any names (ahem...Lima..ahem MccLung...ahem Mays) though.

It is getting harder and harder to get good value in the FA market and even in trades. That is why the draft and player development is so important. Remember also that it is human nature to be short-sighted, which is why it is even more difficult for teams to focus on improving in those areas.

Locked