Hindsight Values - Pitching Edition (Part II)

On Thursday, I wrote about how hitter values would change if, instead of using the draft population like Alex Patton recommended, we used the best 168 hitters at the end of the season instead.

The results were less than overwhelming. Obviously, values changed. However, these changes were less than revolutionary. No hitter's value changed by more than $3. Predictably, the top hitters lost more value than the bottom hitters, but this exercise didn't do much across the board.

Let's do the same thing with the pitchers. Once again, I'll start with the Top 10 American League pitchers in 2006:

Top 10 American League Pitchers, Draft Value vs. "Season" Value

Draft Rank	Player	Draft $	"Season" $	"Season" Rank
1	Johan Santana	$45	$43	3
2	Joe Nathan	$44	$45	1
3	Jon Papelbon	$43	$44	2
4	Francisco Rodriguez	$43	$41	4
5	B.J. Ryan	$42	$42	5
6	J.J. Putz	$40	$38	6
7	Mariano Rivera	$39	$38	7
8	Roy Halladay	$34	$29	12
9	Huston Street	$33	$30	8
10	Chris Ray	$32	$30	9

To the untrained eye, this is a pretty boring list. However, if you've been looking at values for long enough, this is radical.

Unlike the A.L. Top 10 hitter list, where each hitter lost $2-3, there is what appears to be some fluctuation here. Joe Nathan and Jon Papelbon gain a $1. This sounds like nothing, until you consider that the goal is to make sure the top 108 pitchers are worth $936.

The fall-off begins in the second half of the top 10. The clue as to where the fall off is coming from is wrapped up in Roy Halladay's $5 dip, and the second grouping of pitchers provides even more insight.

11-20 American League Pitchers, Draft Value vs. "Season" Value

Draft Rank	Player	Draft $	"Season" $	"Season" Rank
11	Akinori Otsuka	$31	$29	11
12	Francisco Liriano	$30	$30	10
13	Bobby Jenks	$28	$23	14
14	Todd Jones	$27	$23	15
15	Jeff Weaver	$26	$25	13
16	Mike Mussina	$26	$21	16
17	C.C. Sabathia	$25	$19	17
18	Chien-Ming Wang	$24	$17	18
19	Kenny Rogers	$22	$14	22
20	Curt Schilling	$21	$13	24

Now this is radical!

Using "season" dollar values, the fall off in the next best 10 pitchers in the American League is dramatic. Six of this grouping of 10 pitchers loses $5 or more in value. Starters, in particular, take it right on the chin. Rogers and Schilling lose $8. Wang loses $7. Sabathia loses $6. On the other hand, it's not every starting pitcher. Weaver only loses $1 and Liriano holds his ground.

Using the column to the far right, we can also see that, up until Wang, that the entire pitching population is losing value. Wang loses $7 but his ranking doesn't change. There is some jostling amongst the first 18 pitchers, but these are the same 18 pitchers using the draft values and the season values.

You'll see why this is important in a moment.

21-30 American League Pitchers, Draft Value vs. "Season" Value

Draft Rank	Player	Draft $	"Season" $	"Season" Rank
21	John Lackey	$21	$13	25
22	Justin Verlander	$20	$13	23
23	Dan Haren	$20	$11	30
24	Ervin Santana	$19	$11	31
25	Scot Shields	$17	$16	21
26	Kelvim Escobar	$17	$10	33
27	Joel Zumaya	$17	$16	20
28	Scott Kazmir	$17	$12	29
29	Erik Bedard	$17	$9	36
30	Nate Robertson	$16	$7	41

Now the fluctuation in player rankings using the two separate populations is impossible not to see. And look at who "gains" when the full A.L. pitcher pool is used to calculate value. Scot Shields loses a $1 but jumps four spots to #21. Joel Zumaya does even better. He loses $1 as well but jumps seven slots to #20.

So what's happening here?

First, here's a look at the baselines using Alex Patton's draft values. Each win is worth $0.83. Each save is worth $0.62. Why wins are worth more than saves is tilled ground that has been covered exhaustively by Alex in his books.

The baselines for ERA and WHIP are 4.31 and 1.356. A pitcher who puts up these ratios, regardless of how many innings he does it in, will neither earn nor lose money in ERA and WHIP. Every point below these qualitatives will earn the pitcher money, whereas every point above these numbers will lose the pitcher money. The baseline for innings is 1,200. To simplify things, I'll start calling these neutral ERA and neutral WHIP.

As I keep mentioning, these baselines are derived from the typical rosters drafted in late March or early April by your average A.L.-only Rotisserie League. So what happens to these baselines when you simply factor in the best 108 pitchers? More importantly, why does it happen?

First of all, the quantitative categories. Wins are now worth $1.03 apiece. Saves barely move, dropping to $0.60 a save.

So why are starting pitchers dropping so much in value? An 18-game winner is worth another $3.60. This isn't a lot, but given that saves aren't changing in value at all, shouldn't starters move up?

Ah. The ratios.

Neutral ERA drops from 4.31 to 3.62. Neutral WHIP plummets from 1.356 to 1.239. And now the innings factor drops from 1,200 to 910.

Experienced users of Patton values will quickly understand what's happening to dollar values. I'll explain the effect to everyone else.

Before the season starts, we mostly draft pitchers for their roles and not for their ability. A significant number of starting pitchers are purchased who will eventually post bad ERAs/WHIPs and lose their jobs. The same is true, however, for relievers. We not only purchase every known closer, we also draft two or three pitchers from those bullpens who will either pick up saves should the established get hurt or get bad, but also vultures who might pick up wins.

As we all know, this is impossible to predict. Pat Neshek will come up from the minors out of nowhere and earn a lot of money. Dennys Reyes and his 50 IP wound up being worth a lot of money in 2006, but not too many leagues owned him at the beginning of the year because specialists generally aren't worth that much.

As a result, the ERA/WHIP baseline drops. Once it drops, pitchers with ERAs in the mid 4s and WHIPs in the 1.4 range start go from losing a little money in ERA/WHIP to losing a lot of money. They actually lose enough money so that their wins don't even keep them in the black.

As a result, some pitchers who barely pitched last year start making the top 108. Pitchers like Devern Hansack and Agustin Montero suddenly appear in the top 108. Even Zack Greinke, whose 4.50 ERA and 1.67 WHIP are negative, pops into the top 108.

This is because the more innings a pitcher throws up above a 3.62 ERA and a 1.239 WHIP, the more money that pitcher loses in the qualitatives. And those numbers are extremely difficult to come by. Only 63 pitchers equaled a 3.62 ERA or did better. Only 54 pitchers equaled that WHIP or did better.

So another result is massive losses when the entire American League pool of pitchers is now brought back into the fold. The entire American League pitcher pool loses $405 in 2006 using the top 108 pitchers as a baseline.

What does all of this mean?

Well, it tells us what we all kind of knew intuitively. Pitching hurts us, and the top pitchers are extremely valuable. The fact that Johan Santana is still worth $43 even after we lower the bar for ERA and WHIP should tell us something.

Problem is that we need innings. A staff with Devern Hansack, Jon Huber and Mike MacDougal looks pretty, but we know we won't make innings picking up pitchers who put up 10-30 innings apiece.

We also can't identify who these guys will be in April. Middle relievers are as unpredictable as starting pitchers.

So this exercise is interesting from a theoretical standpoint. But it doesn't help us much in trying to put together the best Rotisserie team we can. It's interesting to know that Jon Huber was one of the best 108 pitchers last year in the context of the known pitching universe. But it still doesn't do me a lot of good when I'm trying to put together a staff in April.