Baseball Prospectus - Chat with Keith Woolner

Chat: Keith Woolner

Welcome to Baseball Prospectus' Tuesday March 21, 2006 7:00 PM ET chat session with Keith Woolner.

Keith Woolner is an author of Baseball Prospectus.

Keith Woolner: Good evening everyone. Before we get to the questions, I wanted to make a brief personal appeal. As you may have noticed in the "AL Kings" Scoresheet league article today, I am active in raising money for multiple sclerosis, a disease my wife has dealt with for a decade. The MS Walk, taking place in RTP, North Carolina on April 1st, is an event to raise funds for MS research. My wife has organized a team of at least 25 people who will be walking, and we would very much appreciate your support. For more information, or to make a donation, please visit http://tinyurl.com/9f9as Thank you, and let's get chatting!

chunkstyle (Toronto): BP2006 has not shipped in Canada. I saw BP2006 in bookstores on my Spring Break trip to Florida last week, but did not buy one because I had preordered one, and figured it would HAVE to be here by the time I got back. It STILL isn't here, and that was over a week ago. What's going on? Needless to say, I am about as disappointed as I could possibly be.

Keith Woolner: I checked with BP customer service about this and here's what I learned:

All Canadian BP customers should know that Amazon.ca has been unreliable, and that everyone should order from http://chapters.indigo.ca instead.

On behalf of the BP team, my apologies for the shipping delays.

Wily Mo Pena (Boston, MA): Do you share PECOTA's excitement about me?

Keith Woolner: Lots of Wily Mo questions in the queue today...

It's difficult for me to separate my Red Sox fandom from the careful analyst, but I'm pretty optimistic. Pena has already shown that he's capable of hitting in the majors (albeit with less plate discipline that I'd like to see), and is young enough to still take several steps forward. When your upside is Sammy Sosa, and your downside is Joe Carter, you've got a pretty good future ahead of you.

DavidCrowe (Canada): CC Sabiatha - is he still on the verge of a breakdown? or has he proven that he's one of the special few with a rubber arm and his best days are still to come?

Keith Woolner: Sabathia's been around so long that people forget just how young he still is. His baseball age this year is 25 -- he turns 26 in late July.

When people your age are being considered for prospect lists, I don't think you can anything about proving longevity.

Anthony (Long Island): After reading Baseball Between the Numbers, I have a question about the situational leverage chart in the chapter on closers. You wrote that teams should bring in their closer in the eighth inning or earlier when leverage exceeds 2.32, and in the ninth when leverage exceeds 1.66. The chart lists the leverage for ninth inning, no outs, tied game, bases loaded as 1.04. If that's true, then that seemingly critical situation is of only average importance, and the Yankee (for example) would be wiser to bring in Ron Villone and save Mariano Rivera for another day. Really? In fact, the chart consistently indicates that when a team is up by one, tied or losing, leverage goes down as more baserunners reach. Can you explain this finding?

Keith Woolner: Awesome question, and one that gets to the heart of how "leverage" is defined, and how to use it.

In a bases-loaded, no out, tie game situation, the expected number of runs that will typically score (given an average pitcher) are figured into the win probability. Thorn & Palmer's matrix (which I showed how to approximate in your head here: http://www.baseballprospectus.com/article.php?articleid=4842)
indicates 2.25 expected runs in that situation (and that's based on data from a lower scoring era). From a leverage standpoint, its almost as if those expected runs have scored. How much does one run cost you if you're down two runs with bases empty? Not nearly as much, so the leverage is lower.

But you rightly point out that a perfect performance -- getting out of the jam allowing no runs, dramatically increases the win probability. And such performances are more likely with a Mariano Rivera than a Ron Villone. But Rivera isn't perfect (however nearly so he seems sometimes), and from a leverage standpoint, you have to ask yourself what the chances that another, higher leverage, more winnable situation would happen tomorrow (presuming that Rivera would be unavailable if you used him today).

Your example suggests another way relief appearances could be rated: the swing in win probability if the inning is finished with no more runs scoring. That would capture the "saving our bacon" aspect of getting out of deep trouble that you mention.

Ineftin23 (Baltimore): Hey Keith, do you think Melvin Mora can regain his pre-2005 form this year?

Keith Woolner: What are the odds that a 34 year old, lifetime .281 hitter can hit .340 again? Orioles fans should be happy if he hits like 2003 again, and leave it at that.

Evan (Vancouver, BC): For the record, shipping to Canada using Amazon.com works well, and there's only about a 50% chance the government will remember to charge taxes and handling fees and the like when your package crosses the border. Now, given the general strength of the Mariners bullpen and the weakness of their rotation (save Felix), should they be working at turning Soriano into a starter or a closer?

Keith Woolner: Thanks for the ordering tip Evan.

As for Soriano, I'm a pretty firm believer that unless there's reason to believe a pitcher can't succeed as a starter (lack of endurance, limited repertoire), you should always plan to make him a starter.

The short-term and long-term value of throwing 200 innings exceeds the situational importance of a closer.

A bullpen can't preserve a lead that the starters don't hand them.

ssimon (Pelham, NY): Keith, could you describe a scenario where the Devil Rays win the 2006 World Series? Is that scenario any more than a pipe dream?

Keith Woolner: Well, I think that any team that can get to the playoffs has a nontrivial chance at winning the World Series, so the question then is whether the D'Rays can either win the AL East or sneak in as the wildcard.

The rosy scenario? Baldelli comes back healthy and plays CF, and they trade Gathright for an arm. BJ Upton comes up to play 3B, shifting Huff to 1B. Delmon Young comes up shifting Gomes to DH. That would give you a decent offense.

The rotation is still a big question mark after Kazmir. If Seth McClung can harness his velocity, Edwin Jackson returns to the form that made him a top prospect, and they find another passable starter among Fossum, Waechter, Niemann, Hammel, and probably a couple of other guys I'm forgetting, then they could win 90 games.

Is it a pipe dream? Yes.

Race Bannon (Montevideo): Thanks for the chat, Keith! Why are AL Central prognosticators so dismissive towards the Twins? They won the division in '02, '03, and '04, and they have several young breakout candidates (Mauer, Morneau, Kubel, Liriano, maybe Cuddyer and even Lohse.) Plus a full year of Torii, and a massive upgrade at 2nd base! Looks like 90-94 wins to me....

Keith Woolner: This isn't 1884, and Johan Santana can't start 162 games.

Bob R. (Clearwater): Three questions: 1. If you could change 3 scoring rules in baseball, what would they be? (Would adding a "team error" be one of them?) 2. If you could change 3 rules of how the game is played, what would they be? 3. If you could change three awards, or the criteria for them, what would they be? (For example, two current glamour awards are BA & Saves. Or, while not official, usually the criteria for Cy Young includes wins while for MVP often includes the traditional triple crown categories. Would you alter any of that?)

Keith Woolner: How about I take one example from each category.

1) Scoring rule: eliminate sacrifice flys, and treat them as at-bats

2) A relief pitcher who enters mid-inning must either face three batters, or finish the inning.

3) Replace the MVP with the "Best player award" to eliminate the weaselly word-games voters play with the definition of "valuable" to avoid voting for someone they don't like.

Larry Bowa (FL): Will I get a shot at the Yankees job someday?

Keith Woolner: As a Red Sox fan, I sure hope so.

jacksonreams (Washington, DC): It seems to me that PECOTA has difficulty predicting significant advances in pitchers' K/BB rates. Is there some other way to see Sheets-style jumps coming? For instance, some at BP are high on Brandon McCarthy, but Pecota seems to indicate "league average" for the forseeable future. How is this disconnect explained?

Keith Woolner: PECOTA is, by design, pretty conservative. The truth is that the average performance of a large group of players doesn't change tremendously from year to year -- the fluke seasons and true breakouts are balanced by the slumps and injury decimations. Add to that the injury risk factor built into PECOTA, and you see why from an "actuarial" standpoint, PECOTA is warranted in being conservative.

One tip for how I personallyuse the PECOTA cards (e.g. http://www.baseballprospectus.com/pecota/mccarbr01.php)
I find that there are too many percentile lines to really assess the range, so I break it mentally down into three categories:

High (25%) - 90th percentile
Medium (50%) - 50th percentile
Low (25%) - 10th percentile

When you look at McCarthy's upside as a 1-in-4 chance of around a 3.32 ERA in 173 inning, you start to see why the optimistic views exist.

Benton Quest (Summerville, Ma): When you say W.M. Pena's upside is Sammy Sosa, do you mean the 40-HR, .888 OPS, 27-year-old Sosa of 1996...or the, um, "enhanced" version of five years later (1.174 OPS)? Thanks, Keith.

Keith Woolner: Obviously, predicing multiple 60 HR seasons for anyone is foolhardy, chemically-assisted or not.

How about the we split the difference and use the 33-year old Sammy (.288/.399/.594) -- powerful, still striking out prodigiously, but also learning to take a walk?

Johan Santana (Minny, Minn.): Hey, Keith! Dontcha know I'm not the only arm up here in the frozen North; our rotation was 2nd in A.L innings last year, and 4th in ERA--and all we've lost is Joe Mays and his cruddy 5.65 ERA over 156 IP. (I know you don't need to be lectured about our bullpen.) So here's a second chance to hop on the Twins' bandwagon!

Keith Woolner: Carlos Silva's GB rate's aren't low enough to compensate for his strikeout rate, and can we agree that he's likely to walk more than *9* people in 2006? Radke similarly has excellent control, but still allows fairly many HR and strikes out too few. If he slips even a little, it would be ugly. And for reasons I admit aren't totally rational, I'm not a Kyle Lohse believer.

If the Twins put Francisco Liriano and Scott Baker into the rotation, then you might have something.

rgndvo (Phoenix, AZ): I picked up Between the Numbers and reading here and I'm amazed by the level of sophistication. However, looking over win/run expectancy tables it seems that you still have to know the makeup of the team or game to know the "true" values. If there is only one quality hitter on the team, taking a walk could kill a rally. If the same situation comes up when a guy is in a potent lineup he may be best off taking a walk or choking up even if hes a power hitter. It also seems to ignore if youre in a 13-12 game as being different than being in a 1-0 game.

Keith Woolner: The distribution of talent throughout a lineup will have an effect on specific situations. If your big bats are far away, the probability of getting a key hit is lessened.

This is always the tradeoff between controlling for every variable and minimizing model complexity. The more variables we try to include -- opposing pitcher, platoon splits, day/night, grass/turf, streakiness, potential managerial strategies, player's mood, etc. the harder it is to really understand what factors are key drivers in any situation. And you run the risk of invalidating the whole model if you make a bad decision on how to implement the impact of one particular factor.

That's not to say that my way is the right way, and others are wrong. Just that where I chose to draw the line (base-out situation, run differential, strength of offense, and park) seemed specific enough to be helpful in a variety of situations.

Dick Versace (Peoria, Il.): Duke is to the Yankees, as Bradley is to...?

Keith Woolner: Well let's see. Zach Duke pitches for the Pirates, and the Yankees are in the same division, opposite league, so Milton Bradley, being on the A's, flip the league... carry the one... apply the hyperbolic cosequent transform....

The Padres?

(What do you mean there's some other sport going on right now? Don't you know it's spring training!)

Tom (River): How does one decide its a good decision to trade your young OF with HOF potential, for a FB pitcher with low K numbers when you play in a park that turns flyballs into HR's, and have Griffey in CF to make room for Hatteberg or Womack?

Keith Woolner: When it's tossup as to whether the person who leads the team in homers is your left fielder, center fielder, or #3 starter, you need more pitching.

I like Arroyo more than most (probably the Bosox factor again), but from Cincy's point of view you can't ignore the fact that Bronson is signed to a 3 year extension at below market rates (which kind of stinks for him, since he did so because he wanted to stay in Boston, from what I understand). His willingness to take less money to stay in the town where he earned his success made him that much more desirable as trade bait.

Something to think about when the "players don't have any loyalty to the team" crowd gets going again.

Jonathan (LA): The Pirates are in the central!! For that, answer a question for me. What are the Dodgers' chances of avoiding injury and winning the NL West?

Keith Woolner: GACK! Color me embarrased. Or, uh, I was answering like someone would from 1991's perspective. What's Zach Duke would be in grade school? Ugh, maybe no one else will notice...

Dodgers chance of avoiding injury?

Nomar...
Kent...
Mueller...
Lofton...
Drew...
Furcal...
Odalis Perez...
Brad Penny...

Need I continue?

Brian Cashman (NY): Keith, the hyperbolic cosecant is a trigonometric function, not a transform! You could apply a Legendre or a Fourier Transform to Bradley, but then you'd get... Jeff Kent?

Keith Woolner: If you read closely, you'll notice I said the "hyperbolic cosequent transform", not "cosecant". The HCT is a new way of measuring hustle I introduced in Baseball Between The Numbers. It's right there on page e^(pi)*i

tschiera (Brooklyn): Keith, I am a new subscriber, and not exactly sure if PECOTA is your department, but I guess I am looking for some reassurance as to its reliability. I just finished my fantasy league draft using BP's PFM (I know PFM could be mistaken based on the info I gave it, but it seemed to me to be rendering legitemate results), however, I did not pick players I normally would pick (I love flashy talent and young guys with big upside, I used PFM -- which uses PECOTA -- to keep me "grounded"). I feel "uneasy" taking Andy Pettitte over Felix the King, Rich Harden, and Josh Beckett in a league that counts IP, W, SV, ERA, SO, WHIP -- and taking Josh Willingham over Brian Giles in a leage that counts R, AVG, HR, RBI, SB, OPS. Are these quams founded (PECOTA or PFM was off) or am I just feeling my emotional attachment to my go to fantasy names being severed by the clinical reasoning of BP?

Keith Woolner: Hello and welcome to the BP family...

I explained a bit about PECOTA's conservativeness, and how it bases projections on populations of comparable players. One thing it doesn't specifically take into account is the manager the player plays for. And in particular, whether a young player who starts off slow is at risk of losing playing time or being demoted. Under the right circumstances, the concerns about Willingham over Giles might be well-founded. Ditto for Felix Hernandez -- historically, how many 20 year olds have a lock on a spot in the rotation? While I think PECOTA is excellent for overall projections of rates of production, there's always room to add specific knowledge about a player and his relationship with his team or manager.

Mary S. (ShowMeState): Thank you for the chat, Keith. Bobby Cox has lost in the first round of the playoffs 5 of the last 6 years, but Tony LaRussa gets criticized by the national press (and many Cardinal fans) for his playoff record, while I have yet to hear or read one remotely negative comment about Atlanta's postseason flops. Should Cox be analyzed/criticized more, or LaRussa less?

Keith Woolner: That's an interesting perspective because my gut reaction (just off the top of my head) is that Cox has gotten a fair share of criticism for the Braves' lack of post-season success. Maybe not as much as LaRussa, perhaps because a bigger "star" among managers.

Cubs' PR Director (Wrigglyville): Hello Keith. If you could pick a team of one sabermetrician and one traditional baseball guy to run a team, who would they be?

Keith Woolner: I'd pick a baseball guy who knows a lot about the traditional side of the game, but is open to new ideas and experimentation. Earl Weaver, Davey Johnson, maybe someone Larry Dierker.

On the sabermetric side... is it cheating to selfishly pick myself? As long as we're fantasizing that had that kind of decision-making authority in the first place... :-)

Brian Cashman (NY): Shamefully, I haven't picked up BBtN yest...I would have if I knew it began with page -1! Hey what are your thoughts on Chien Ming Wang? Great groundball tendencies, low K rate last year, but his K rate in the minors was much better. If Burnett is a bargain at $11 mil...

Keith Woolner: If I recall, Wang had arm problems, and I think his strikeout rate is more related to how well he recovers from that than his pre-injury minor league record.

As for his GB%, it is astoundingly good, and frankly, I don't have an intuitive feel for how much to weight GB% compared to strikeout rate and walk rate. For no research-based reason, I tend to mentally associate ground-ball pitchers with high-control pitchers, and if that has any basis in reality, teasing apart the impact between BB rates and GB rates should be an interesting topic to study.

Mark Prior (Naperville): Keith, could you please explain to me why the Cubs or any other team would work so hard to hide an injury like Prior's, or lie about it? I don't understand what theyre hoping to accomplish. Its not as if Priors shoulder will get better by denial, nor will opposing teams try harder or less hard if someone's hurt. Is it season ticket sales in this case, and if so, isnt the extra few sales outweighed by the increase in fan distrust of the organization? Why do teams lie?

Keith Woolner: Let me shock the audience by saying something in the Cubs' defense...

Suppose that Cubs knew about Prior's injury well in advance. Maybe they even know better than is public even now how severe it is. If they think they're going to be without Prior for an extended period of time, they very well might turn to the trade market to pick up some starting pitching. Would you want your potential trading partners to know how desperate you are? Seems like leverage that could be used against them in a negotiation.

Bob R. (Clearwater): Do you think that given their lack of an established closer, and their limited resources, that the Rays are in a good position to try to use their bullpen differently from current practice, bringing their best reliever in when the game is on the line rather than waiting for the ninth inning save? Wouldn't it be smart of them to avoid designating one pitcher as "the closer"?

Keith Woolner: I think the Rays may not have enough 9th inning leads to worry about it.

But as we've said many times over the years, and I repeated in my reliever chapter in Baseball Between The Numbers (I'm legally obligated to plug it every 5th chat response) we've gone backward in maximizing relief aces since the 1970's. Recognizing high leverage situations and using relievers where they will have the most impact (perhaps allowing for the "value of a perfect appearance" factor mentioned at the start of this chat) would be a good risk for any team.

Dan (PA): How do you see the Phillies doing this year? Is there any chance at a playoff or is the rest of the division too good?

Keith Woolner: You gotta love Ryan Howard and Chase Utley on the right side of the infield. I'm not impressed with Nunez at 3B, and you have to wonder how long Mike Lieberthal will be an acceptable catcher, but the rest of the lineup doesn't have any gaping holes.

There's not a true ace among the starting pitchers (Myers could get there if he cuts down on the homers), but there are some possibilities (I like Madson moving to the rotation, FWIW). I could see them hanging around the outskirts of contention, and if things all break right, winning the division outright. But they will need help.

Keith Woolner: That's all I've got time for tonight, but I hope you enjoyed it. Thanks very much to those of you who already made a donation to our MS Walk team -- your support is very much appreciated. If you haven't donated yet, but are willing to help sponsor me and my team, that URL is: http://tinyurl.com/9f9as Thanks for the great questions!