Baseball for the Thinking Fan

You are here > Home > Baseball Newsstand > Baseball Primer Newsblog > Discussion
Baseball Primer Newsblog
— The Best News Links from the Baseball Newsstand

## Saturday, December 20, 2008

#### Rob Neyer: Explaining some new math

As you know, I’ve been wrestling with this notion of positional adjustments, which on Thursday led to the following e-mail exchange with FanGraphs’ Eric Seidman …

Eric: Rob, I thought I would try to see if I could clarify the adjustments for you. Basically, it is all due to defense. Tom Tango likes to refer to the adjustments as “Runs Over Willie” (in honor of Willie Bloomquist, who can play everywhere). Essentially, the idea is that if you took Willie Bloomquist and put him anywhere on the field, what would an average fielder produce, runs-wise, compared to his production?

So, if you put Willie at 1B, the average 1B would cost his team 12.5 runs more than Willie. If you put Willie at shortstop, the average SS would save 7.5 runs more than Willie.

It’s really just a quantifiable way of showing which positions are the toughest to play. Catcher gets +12.5 runs because not everyone can play there. Shortstop gets +7.5 runs because it is the toughest non-catching position. Then comes 2B/3B/CF, at +2.5 runs apiece. LF/RF are docked -7.5 runs, and 1B docked -12.5 runs. Using these adjustments allows us to compare Carl Crawford in LF to Chase Utley at 2B.

If Crawford is +15 runs via UZR and Utley is +19 runs, it really isn’t as close as it seems, given that LF are docked -7.5 runs and 2B gain +2.5 runs. Before even factoring in offensive contribution or adding two wins (20 runs) to be above replacement level, Utley would be a +21.5 run defender, Crawford a +7.5 run defender.

Hope that makes some more sense. It is a very confusing concept, but basically it just allows us to make cross-positional comparisons so someone like Crawford doesn’t have an overstated defensive value.

Me: Is this a new thing? I don’t recall seeing any discussion of positional adjustments before the last couple of weeks. I’m just wondering if everything we thought we knew about player valuations have been wrong. Is Dave Concepcion worth more than we thought? Tony Perez less?

Eric: I wouldn’t necessarily call it two weeks new, but definitely new in the last year or two, as far as I know. Granted, I didn’t really “come onto the scene” until May 2007, but I can recall as far back as June 2007 reading Tango’s positional adjustment work.

I don’t think it means that everything we have already done is wrong, per se, because defensive stats weren’t really ever taken into account outside of Fielding Percentage for the longest time.

Huh, I didn’t realize that positional adjustments is considered so ‘new’.

Tripon Posted: December 20, 2008 at 05:01 AM | 301 comment(s) Login to Bookmark
Tags: general

Go to end of page

Statements posted here are those of our readers and do not represent the BaseballThinkFactory. Names are provided by the poster and are not verified. We ask that posters follow our submission policy. Please report any inappropriate comments.

Page 2 of 4 pages  < 1 2 3 4 >
101. Chris Dial Posted: December 21, 2008 at 01:15 AM (#3034902)
I understand that, if your goal is to account for what happened last year, you can compare only like positions to like positions and still have it all "add up".
This sentence refutes this statement: "That is why you would want to consider the shortstops' abilities when you consider the second basemens'."
But 1) that becomes of little use once you move away from the "accounting for last year" scenario
Why would you think that?
(What are our team's strengths and weaknesses? Who is more valuable to our franchise?
I disagree that OPD doesn't answer that 100% as well as anything else.
If we're "timelining" and thus not assuming that every year of baseball history is zero-sum, who is greater, Player X from 1950 or Player Y from 2000?, etc. -- all of these questions are better answered by the more abstract method than by the "accounting" method),
I disagree with that assertion (although my defensive evaluation only goes back 20 years)
and 2) although you can make things add up that way, I don't see why it's actually preferable to do so.
Because it doesn't use a static value for positional adjustment, which, in Tango's case, is clearly wrong ith regard to shifting baselines. I have to make no "other calculation" that any metric wouldn't when a player changes positions.
102. The District Attorney Posted: December 21, 2008 at 01:23 AM (#3034906)
(What are our team's strengths and weaknesses? Who is more valuable to our franchise?
I disagree that OPD doesn't answer that 100% as well as anything else.
I don't see how, when you just admitted that it wouldn't tell you that in the hypothetical all-2B-are-scrubs league, calling up a replacement-level SS and moving Escobar to 2B would really help the team. It'd inform you of that after you actually did it and played the season that way. Great?

Look, just on the most basic level, if you can replace a 2B with a SS, then how are the shortstops' abilities not relevant to the value of the second basemen? I don't see either "you can make it add up anyway" or "but he was a second baseman!" as a real answer to that.
103. Chris Dial Posted: December 21, 2008 at 01:35 AM (#3034917)
I don't see how, when you just admitted that it wouldn't tell you that in the hypothetical all-2B-are-scrubs league,
No such thing happened.
104. Chris Dial Posted: December 21, 2008 at 01:36 AM (#3034918)
Look, just on the most basic level, if you can replace a 2B with a SS, then how are the shortstops' abilities not relevant to the value of the second basemen?
Are you talking about ability or value?
105. Tango Posted: December 21, 2008 at 01:36 AM (#3034919)
But where do the numbers come from?

Comparing players who played multiple positions, using UZR. If Alou is +2 as a LF and +2 as a RF, then we conclude that the average LF = average RF. If Ichiro is +10 as RF and +1 as CF, then the CF baseline is 9 runs higher than RF. This is the short answer. This is just an illustration.
106. SkyKing162 Posted: December 21, 2008 at 01:37 AM (#3034920)
Absolutely. His value to his team is increased. His value is increased. He is more valuable.

He's in the hitting lineup either way. He's playing two similarly difficult defensive positions with similar skill. He's not providing significantly different production, he's just being accounted differently. To me, and I might be explaining myself poorly, that doesn't make him awesomely more valuable.

I think the high school question is valid. How would we determine HS value? If we're going to make the assumption that MLB teams get it close enough for a fudge factor, we no longer disagree on anything, but we've changed the discussion from which side is theoretically better to which side is closer and calling it a tie.

*** ***

On the DH issue, do you compare DH's to league-average DH offense like other positions? So they're held to a lower level of production than 1Bs and corner OFs (potentially, not sure they're always below LF/RF.)
107. Mefisto Posted: December 21, 2008 at 01:40 AM (#3034923)
THat's because they are. I don't know why you can't accept it. Their contributions on the field outpace their counterparts equally. They *are* equal with respect to what we are measuring, and that is how much they help their team win on the field. I think you may be trying to measure something else.

I think this overlooks the fact that the CF can play RF, but not vice versa. There would be a cost to moving the RF to CF. Because we don't have to pay that cost with a CF, we want to recognize that value he adds.
108. Mefisto Posted: December 21, 2008 at 01:46 AM (#3034927)
Let me phrase it another way. Baseball teams don't have on their roster SS and 2B and CF. They have players -- Barry and Ryne and Paul. The managers have to decide how to get the maximum value out of those players in where they bat them in the line up and where they place them on the field. Measuring the relative value of the different positions is what allows them to do that systematically and efficiently.
109. Chris Dial Posted: December 21, 2008 at 01:48 AM (#3034928)
He's in the hitting lineup either way. He's playing two similarly difficult defensive positions with similar skill. He's not providing significantly different production, he's just being accounted differently. To me, and I might be explaining myself poorly, that doesn't make him awesomely more valuable.
To me, adjusting for position for a player at a position he did not play, IMO, assumes facts not in evidence. I don't really think that Escobar as a "0" at SS is necessarily a +5 at 2B (or whatever the adjustment is). I think the biggest argument against this type of positional adjustment is ARod. ARod is a significantly worse 3B than he ever was SS. Nearly 20 runs worth. I don't like the assumption that makes. That means it is much more important (to me) to calculate realtime.

YEs, there are times when I need to consider (as I was doing in the Mets 2B Options thread) what a SS would do playing 2B. For me, I think the defensive value of the positions is very similar, and I think it is folly to count the chickens of moving a SS over to 2B before they hatch. There's enough difference in the DP to matter. I don't think that moving a SS to 1B increases the defense. It completely depends on how a given player takes to it. I haven't studied it as much as Tango, but I think his present sample sizes have been too small.

BUT, while I "look forward", having Tango's positional *defensive* differences is valuable. Looking at MVP, I think those positional adjustments, as compared to OPD, are a step backwards.
110. DJ Endless Grudge Plays All The Hits (CW) Posted: December 21, 2008 at 01:49 AM (#3034929)
I'm trying to catch up on this thread - lot of good discussion here.

Clearly not, and I should have been more precise. 2B have many more assists than 3B (more PO too). 1B have lots more PO, but many fewer assists.

Second basemen have their assist totals larded by the double plays they turn. (They also lead all other IF positions in what I call "secondary assists," or assists on balls they didn't field. These seem to include DPs, cutting off throws, and fielding balls deflected by other infielders.)
111. Tango Posted: December 21, 2008 at 01:49 AM (#3034930)
I hope everyone is following the discussion between Chris and I, because we are both trying to be pretty clear about what we are doing.

Their contributions on the field outpace their counterparts equally.

Chris is going on the basis that your "counterpart" is the opponent who plays the same fielding position. He also said there's no such thing as a "hitter" position.

As I said, I don't go along with this premise at all. Someone else can come along and say you have to outperform your counterpart based on the batting order.

But, I see it that your "opponent" has nothing to do with the fielding position or batting order or anything. It's just the average guy on the other team.

***

As for the high school issue: we agree that there is a disproportionate number of great fielding and great hitting players in high school who play SS, correct? And, the worst player is probably playing 2B or RF, right?

Whatever it is, it is nowhere close to being balanced like in MLB. In MLB, the average at each position is not so skewed. But in high school, it's enormously skewed.

But, the way Chris is arguing it, the "counterpart" is always the opponent at the same position. The great hitting, great fielding SS is being compared to the great hitting, great fielding SS he's playing against. The crappy hitting, crappy fielding 2B is similarly compared.

The end result is that the average high school SS will rank side-by-side with the average high school 2B. Perhaps this answers Chris' question, but I can't see how this answers the question that readers would be asking. Are they aware that this is the implication of Chris' process?

(Presuming I'm understanding Chris' process well-enough.)
112. Chris Dial Posted: December 21, 2008 at 01:50 AM (#3034931)
I think this overlooks the fact that the CF can play RF, but not vice versa. There would be a cost to moving the RF to CF. Because we don't have to pay that cost with a CF, we want to recognize that value he adds.
I don't think it overlooks anything in their performances. He doesn't "add value" that I can see - he either creates runs or saves them, I think.
113. The District Attorney Posted: December 21, 2008 at 01:50 AM (#3034932)
No such thing happened.
Oy. Whether you "admitted it" or not, what I said was accurate, right?
Are you talking about ability or value?
Are you joking? That's the very distinction I've been trying to make! I said several times that there is a difference between value in the sense of accounting for where the wins came from in a given year, and value in a more abstract sense (i.e., ability). Your response, paraphrasing, has essentially been that we're hardly ever worried about ability -- only if we're considering a move such as converting a SS to 2B would we need to know such an obscure thing, and even then, it'd only be a few runs anyway -- and that "accounting" value is just as good as anything else at determining team strengths and weaknesses and player worth in the general-manager sense.¹ Then you ask me if I'm talking about ability or value. WTF is this?

¹ If these are mischaracterizations, please give me something that advances the conversation more than "I didn't say that."
114. Chris Dial Posted: December 21, 2008 at 01:55 AM (#3034934)
But, the way Chris is arguing it, the "counterpart" is always the opponent at the same position. The great hitting, great fielding SS is being compared to the great hitting, great fielding SS he's playing against. The crappy hitting, crappy fielding 2B is similarly compared.

The end result is that the average high school SS will rank side-by-side with the average high school 2B. Perhaps this answers Chris' question, but I can't see how this answers the question that readers would be asking. Are they aware that this is the implication of Chris' process?.
I don't see what's wrong with that. You are interchanging ability with value, I think. And it isn't as simple as "who he's playing against". There's a large aggregate of players. There would have to be a HUGE glut of graet HS SS, which would mean that to each individual team, that players doesn't provide any advantage (or value) over their opponents. The crappy 2B that outpaces other HS 2Bs would provide more value in terms of runs saved and created than a great SS where every team has a great SS. If everyone is a great player, then no one is.
115. Chris Dial Posted: December 21, 2008 at 02:00 AM (#3034937)
Oy. Whether you "admitted it" or not, what I said was accurate, right?
Truth be told, I couldn't parse that. Bear with me, I have sixteen different arguments going here.

Are you joking? That's the very distinction I've been trying to make! I said several times that there is a difference between value in the sense of accounting for where the wins came from in a given year, and value in a more abstract sense (i.e., ability). Your response, paraphrasing, has essentially been that we're hardly ever worried about ability -- only if we're considering a move such as converting a SS to 2B would we need to know such an obscure thing, and even then, it'd only be a few runs anyway -- and that "accounting" value is just as good as anything else at determining team strengths and weaknesses and player worth in the general-manager sense.¹ Then you ask me if I'm talking about ability or value. WTF is this?
No, I am not joking. I am measuring value. Everyone in this thread was talking about value, as best I could tell. What are you asking? The bolded part is accurate.
116. DJ Endless Grudge Plays All The Hits (CW) Posted: December 21, 2008 at 02:03 AM (#3034939)
If convering a shortstop to a second baseman is an "obscure" thing then can I ask what sport we're talking about?
117. Tango Posted: December 21, 2008 at 02:04 AM (#3034941)
Chris/114 illustrated perfectly why he and I are on opposite sides of this issue.

Just to give one last example and following up on Chris' statement here:
"If everyone is a great player, then no one is."

If in high school, all the SS are guys like Hanley Ramirez or Jimmy Rollins, and most of the 2B are guys like Willie Bloomquist, but there's one 2B that is like Orlando Hudson or Mark Ellis, Chris' list will look like this:
+30 Hudson
+0 Hanley, Rollins, all other SS
-1 Bloomquist and guys like him

In terms of what Chris is trying to do (counterpart by fielding position) it answers his question. But, seeing this list, aggregated this way, is mixing apples and oranges from what I'd be trying to do.

So, the reader needs to decide if this kind of list (though obviously not so greatly exaggerated in MLB) is what they want.

Well, it will be greatly exaggerated a couple of times, notably NL LF from 2001-2004. If Bonds is in the league or not makes a substantial difference in the rankings of NL LF.
118. Chris Dial Posted: December 21, 2008 at 02:08 AM (#3034942)
the reader needs to decide if this kind of list (though obviously not so greatly exaggerated in MLB) is what they want.
Not only "not so greatly exaggerated", but nearly non-existent in MLB. Only with Bonds in recent history does this effect exist, and doing it with "positional adjustment" greatly undervalues Bonds. Even Pujols doesn't see (or create) this effect. It takes Barry Bonds, and Bonds only in those seasons, not over his career, or Babe Ruth to create a HS effect, IMO, rendering this argument fallacious.
119. Chris Dial Posted: December 21, 2008 at 02:08 AM (#3034943)
But, I think Tango's example is a proper illustration of our differences - unfortunately, the HS effect isn't applicable.
120. Chris Dial Posted: December 21, 2008 at 02:10 AM (#3034945)
If Bonds is in the league or not makes a substantial difference in the rankings of NL LF.
Bonds hitting those seasons would skew every "normal" valuation. When I calculated OPD those years, I did it with and without Bonds.
121. DJ Endless Grudge Plays All The Hits (CW) Posted: December 21, 2008 at 02:10 AM (#3034946)
Chris, how do you handle the issue of determining how valuable a player's offense is relative to position for the following season?
122. Chris Dial Posted: December 21, 2008 at 02:12 AM (#3034948)
how do you handle the issue of determining how valuable a player's offense is relative to position for the following season?
First, I ignore salary. Second, I assume he'll perform similarly to Marcel and adjust accordingly. Defensively too.
123. The District Attorney Posted: December 21, 2008 at 02:17 AM (#3034950)
Again, I think this is a semantic discussion, this time about the definition of "value." That'd be fine normally, except you're not only asserting your own definition but claiming that the other definition has little relevance to anything we might ever want to know, and that to my mind has gotta be wrong.

In so many words:

1. Yes, you can account for what happened last year by comparing each team's C to the other C, its 1B to the other 1B, etc. If you do that for each team, it'll add up to what everybody did, so, yeah.

2. Even so, doing that isn't truly measuring players by their replaceability, because some players can be replaced by other players who are currently playing other positions.

3. For instance, in the all-2B-are-scrubs league, the accounting analysis wouldn't alert you that if you have a halfway decent SS to replace him, you would get a huge advantage by moving Yunel to 2B.

4. At the very least, this means that the accounting isn't giving you the best possible sense of team strengths/weaknesses, or of what we might call "worth in a general manager sense."

5. I would also say that it isn't even giving you the best explanation of what happened last year, since you're allowed to make in-season transactions -- the option of, for instance, replacing your 2B with a player currently playing SS is an option that is always available to you.

6. You deny 2 through 5... seemingly by circling back to #1 and reasserting that the accounting adds up to what happened last year and that that is the only relevant definition of value... or if that's not the argument, I'm not sure what it is.
124. DJ Endless Grudge Plays All The Hits (CW) Posted: December 21, 2008 at 02:19 AM (#3034953)
Okay, but here's a question. Let's say you're working for the Brewers, and the team is considering moving Hardy to second base to make room for Alcides Escobar. They ask you how Hardy's value would change going from SS to 2B.

In the 2008 NL, the average batting line at each position:

SS - .276/.334/.404
2B - .271/.338/.408

Here's the issue, though - if you move Hardy to second base, you're replacing him with a worse hitter overall. And you're replaceing a worse hitter at second base. So by moving that one premier hitter, you're making the gap between SS and 2B bigger. How would you account for that change in value?
125. Chris Dial Posted: December 21, 2008 at 02:27 AM (#3034958)
you're not only asserting your own definition but claiming that the other definition has little relevance to anything we might ever want to know, and that to my mind has gotta be wrong.
No, I am not doing that at all.

In so many words:

1. Yes, you can account for what happened last year by comparing each team's C to the other C, its 1B to the other 1B, etc. If you do that for each team, it'll add up to what everybody did, so, yeah.
Check.

2. Even so, doing that isn't truly measuring players by their replaceability, because some players can be replaced by other players who are currently playing other positions.
No system can nor does this.

3. For instance, in the all-2B-are-scrubs league, the accounting analysis wouldn't alert you that if you have a halfway decent SS to replace him, you would get a huge advantage by moving Yunel to 2B.
That's not accurate. It's not as if all the data that goes into my OPD vanishes after those numbers are created. We're still aware of what the raw numbers say.

4. At the very least, this means that the accounting isn't giving you the best possible sense of team strengths/weaknesses, or of what we might call "worth in a general manager sense."
You haven't described to me how it doesn't any differently from any other system.

5. I would also say that it isn't even giving you the best explanation of what happened last year, since you're allowed to make in-season transactions -- the option of, for instance, replacing your 2B with a player currently playing SS is an option that is always available to you.
Again, of course, OPD allows for that.

6. You deny 2 through 5... seemingly by circling back to #1 and reasserting that the accounting adds up to what happened last year and that that is the only relevant definition of value... or if that's not the argument, I'm not sure what it is.
Deny? They are untrue statements (or true about all valuation systems).

Define the "V" in MVP? Are you of the belief that Ryan Howard was arguably the MVP?
126. Mefisto Posted: December 21, 2008 at 02:31 AM (#3034959)
Second basemen have their assist totals larded by the double plays they turn. (They also lead all other IF positions in what I call "secondary assists," or assists on balls they didn't field. These seem to include DPs, cutting off throws, and fielding balls deflected by other infielders.)

Joe and I were discussing deadball 2B, so the double play totals don't amount to much. That said, I'd call turning the DP a skill, so in my view it should be counted as an assist. The cut-off throws a 2B has to make don't require the same skill as fielding a grounder and then throwing, but they have value nonetheless (obviously, they cut down a runner). Fielding deflections can be harder (having to reverse direction) or easier.

The point I was making is that we can't just assume that the more difficult balls a 3B fields make him more valuable than a 2B despite the extra chances handled by the 2B, we need to prove that.
127. Blackadder Posted: December 21, 2008 at 02:34 AM (#3034963)
This is related to a point Tango brings up above, but I think it is worth re-iterating and spelling out further (besides, I thought of this argument independently when reading through the thread, and I wouldn't let a little thing like being late to the party get in the way :) ) Before starting, though, I wanted to commend both the civility and the content of this discussion, both of which have been admirable.

As I understand him, Chris takes, as the starting point for his conception of value, the idea that in order to win a pennant a team must be sufficiently far above average in either their offense or defense. This is, of course, a tautology. At the team level, it is trivial to determine how much above average a team is offensively and defensively (well, maybe not trivial, especially if pitching doesn't count as defense, but you get the point). The trick, then, is to figure out how to divvy up the total runs above average in each team among the team's players.

Chris thinks that the appropriate way to do it is to compare each player to their positional peers. This is at least a consistent way of cutting up the value pie, in that adding up every player's value above average at their position on a team will give you the team's total value above average. The problem, however, is that a player's position, in this context, is really just a label. You are simply picking some group of players on other teams to compare him to; a choice that seems natural, of course, since someone's position is a very salient point about them as baseball players, but still a choice nonetheless. The crux of the matter is that, however, there are lots of other ways of grouping the players in the league that would lead to radically different distributions of value.

One could, to borrow Tango's example, group players by lineup order instead of position. I don't see how this could be any less justified than grouping players by position. Wanting to compare players to their "peers" is not going to block this; after all, in lots of ways the players most "similar" to a given player are those that bat in the same spot, not those that field the same position. It is also true that a team "wins" by being above average at each lineup spot just as much as it "wins" by being above average at each position. It goes without saying, I trust, that comparing players to others in the same batting order spot is absurd. One could go even further; one could, say, rate players by how they compare to other players who had the same rank ordering in their team in overall offense+defense.

The problem with these groupings is that they are manifestly biased. They arbitrarily group better players with better players and worse players with worse, creating the illusion that the bad players are, relatively speaking, much more valuable than they actually are. Grouping players by their position is not more theoretically justified than these groupings, and moreover runs into the same biases. The average shortstop is simply more valuable than the average second baseman, and the average CF in the 1950's was a lot more valuable than the average LF, in the same way that the average quarterback is more valuable the average punter. One needs to consider position, of course, in calculating the total value of any player; but doing so because a team "wins" by being above average at each position is, I think, a poor reason for doing so.
128. Chris Dial Posted: December 21, 2008 at 02:41 AM (#3034966)
Let's say you're working for the Brewers, and the team is considering moving Hardy to second base to make room for Alcides Escobar. They ask you how Hardy's value would change going from SS to 2B.

In the 2008 NL, the average batting line at each position:

SS - .276/.334/.404
2B - .271/.338/.408

Here's the issue, though - if you move Hardy to second base, you're replacing him with a worse hitter overall. And you're replaceing a worse hitter at second base. So by moving that one premier hitter, you're making the gap between SS and 2B bigger. How would you account for that change in value?
This isn't a difficult question at all. Well, what is Escobar going to hit - that's the big question. The comparison isn't Hardy very much - it's wether or not Escobar is better than the 2B I am replacing, because SS and 2B hit so similarly.

Hardy was a +13.3 runs as a SS (86 XRp), in 433 outs. Moving him to 2B would drop him to +10 at 2B. Three runs. He still created 86 runs in 433 outs, but 2B created slightly more in those outs. It's a very small change.
129. Chris Dial Posted: December 21, 2008 at 02:47 AM (#3034968)
You are simply picking some group of players on other teams to compare him to; a choice that seems natural, of course, since someone's position is a very salient point about them as baseball players, but still a choice nonetheless. The crux of the matter is that, however, there are lots of other ways of grouping the players in the league that would lead to radically different distributions of value.

One could, to borrow Tango's example, group players by lineup order instead of position. I don't see how this could be any less justified than grouping players by position.
This isn't true from a practical point of view. I don't think it is just as justified. A fielding position requires a set of skills. A slot in teh batting order does not.

No one plays the field as a #2 hitter. Any position can hit second, but the skills to play 2B differ from those to play C. Either can bat second, but either cannot (at the price of admission to play MLB) play the other's slot in the field.

This isn't a theoretical exercise - players have to meet a certain criteria to play in MLB. That's why teh HS example fails, as well as any justification for using batting order as compared to fielding position.
130. Ivan Grushenko of Hong Kong Posted: December 21, 2008 at 02:49 AM (#3034969)
This isn't a difficult question at all. Well, what is Escobar going to hit - that's the big question. The comparison isn't Hardy very much - it's wether or not Escobar is better than the 2B I am replacing, because SS and 2B hit so similarly.

Hardy was a +13.3 runs as a SS (86 XRp), in 433 outs. Moving him to 2B would drop him to +10 at 2B. Three runs. He still created 86 runs in 433 outs, but 2B created slightly more in those outs. It's a very small change.

But then in determining the defense value of Escobar vs the New SS don't we need to estimate the prospective defensive contribution of Escobar at 2B -- a position he's hardly played? Is Tango's system useful in doing that? Can we use Tango's system to estimate that if Escobar was +5 at SS, then he'd likely be +10 at 2B?
131. Chris Dial Posted: December 21, 2008 at 02:50 AM (#3034970)
As I understand him, Chris takes, as the starting point for his conception of value, the idea that in order to win a pennant a team must be sufficiently far above average in either their offense or defense.
This isn't really accurate. I think everyone understands having better players should mean more wins. How are better players defined? How they perform relative to their peers. So, teams maximize their chance to win a pennant by increase their players performance relative to their peers. Is that not a correct set of assumptions?
132. DJ Endless Grudge Plays All The Hits (CW) Posted: December 21, 2008 at 02:50 AM (#3034971)
Hardy was a +13.3 runs as a SS (86 XRp), in 433 outs. Moving him to 2B would drop him to +10 at 2B. Three runs. He still created 86 runs in 433 outs, but 2B created slightly more in those outs. It's a very small change.

If it's a very small change, then why do more teams not move their 2B to SS? Instead of the Orioles fielding Brian Roberts at 2B and a parade of replacement-level shortstops, why not move Roberts to short and find a replacement-level second baseman? After all, a replacement-level second baseman is typically a better hitter than a replacement level shortstop, and they're both typically +/- 0 relative to position on defense.
133. The District Attorney Posted: December 21, 2008 at 02:50 AM (#3034972)
2. Even so, doing that isn't truly measuring players by their replaceability, because some players can be replaced by other players who are currently playing other positions.
No system can nor does this.
As I see it, if you add in the Tango/Fangraphs positional adjustment, you're at least trying to do it. I do grant your point that it breaks down to individual people and not every individual SS would be a better 2B, not every CF would be a better RF, etc. But I certainly don't think that's a reason not to make any estimate about something that seems quite important.

Are you of the belief that Ryan Howard was arguably the MVP?
No, and I'll tell you why. It's not solely because if the Phils had replaced Howard with someone else who is currently a 1B, it would have cost them more than (to stick with the same team) if they had replaced Utley with someone else who is currently a 2B. It's also because the Phils could have replaced Howard with someone not currently playing 1B -- in fact, pretty much anyone other than David Ortiz could have been used to replace him -- while replacing Utley with anyone besides a SS or 2B would most likely have been catastrophic.
134. Chris Dial Posted: December 21, 2008 at 02:52 AM (#3034974)
But then in determining the defense value of Escobar vs the New SS don't we need to estimate the prospective defensive contribution of Escobar at 2B -- a position he's hardly played? Is Tango's system useful in doing that? Can we use Tango's system to estimate that if Escobar was +5 at SS, then he'd likely be +10 at 2B?
Yes, Tango's system is useful in doing that - when you are working with an unknown, and projecting.

As I said in #109:
"BUT, while I "look forward", having Tango's positional *defensive* differences is valuable. Looking at MVP, I think those positional adjustments, as compared to OPD, are a step backwards."
135. Ivan Grushenko of Hong Kong Posted: December 21, 2008 at 02:55 AM (#3034976)
It's also because the Phils could have replaced Howard with someone not currently playing 1B -- in fact, pretty much anyone other than David Ortiz could have been used to replace him -- while replacing Utley with anyone other than a SS or 2B would most likely have been catastrophic.

How does Tango's system solve this problem though? Does his system then tell us that moving a 3B to 2B would work just as well defensively as using someone who's got the same defense rating relative to average that's been playing 2B? I assume that this doesn't work for moving someone from CF to 2B just because the positional adjustments are the same.
136. Chris Dial Posted: December 21, 2008 at 02:55 AM (#3034977)
If it's a very small change, then why do more teams not move their 2B to SS? Instead of the Orioles fielding Brian Roberts at 2B and a parade of replacement-level shortstops, why not move Roberts to short and find a replacement-level second baseman?
The defensive challenges.
137. Chris Dial Posted: December 21, 2008 at 02:57 AM (#3034979)
As I see it, if you add in the Tango/Fangraphs positional adjustment, you're at least trying to do it. I do grant your point that these are individual people and not every SS would be a better 2B, not every CF would be a better RF, etc. But I certainly don't think that's a reason not to make any estimate about something that seems quite important.
I don't think Derek Jeter is included in anyone's calculation of replacement level at 2B. Perhaps I am mistaken.
138. Chris Dial Posted: December 21, 2008 at 02:59 AM (#3034980)
Does his system then tell us that moving a 3B to 2B would work just as well defensively as using someone who's got the same defense rating relative to average that's been playing 2B? I assume that this doesn't work for moving someone from CF to 2B just because the positional adjustments are the same.
I haven't spent much time memorizing the positional adjustments, but I am supposed to think that David Wright moving to 2B would have the same defensive impact as Carlos Beltran moving to 2B??? EDIT: WRT runs saved at 2B.
139. Ivan Grushenko of Hong Kong Posted: December 21, 2008 at 03:02 AM (#3034983)
I haven't spent much time memorizing the positional adjustments, but I am supposed to think that David Wright moving to 2B would have the same defensive impact as Carlos Beltran moving to 2B???

2B, 3B and CF are all +2.5 runs in the positional adjustment at the top of the page. I'm pretty sure Tango's not saying that Beltran is about equally good at playing CF and 2B. I'm not sure whether he's saying Wright would be equally good playing 2B or 3B.

EDIT: I'm asking whether he thinks that Wright would be equally good at both positions.
140. DJ Endless Grudge Plays All The Hits (CW) Posted: December 21, 2008 at 03:11 AM (#3034990)
If it's a very small change, then why do more teams not move their 2B to SS? Instead of the Orioles fielding Brian Roberts at 2B and a parade of replacement-level shortstops, why not move Roberts to short and find a replacement-level second baseman?

The defensive challenges.

And shouldn't that be something that our value systems capture?
141. Chris Dial Posted: December 21, 2008 at 03:19 AM (#3034997)
And shouldn't that be something that our value systems capture?
To be clear, because the Orioles feel teh defensive challenges are too great - not that they necessarily are for Roberts. If you follow my posts, I routinely argue for teams to move players to the left in the spectrum. I think it is overstated at how hard it is for a nimble athlete to move that direction. And his bat will generally overcome those errors.

Also, the move is specific person dependent, and not as simple as the present numbers given. Orlando Hudson is very athletic, and could probably handle SS. But teams don't usually try players whose "body types" don't fit the idea. So the data of who does change positions is Ryan Freel or Chone Figgins.
142. Chris Dial Posted: December 21, 2008 at 03:20 AM (#3034998)
Oh, and no, the value I am talking about is his performance, not his trade value or his ability - whether or not he "can" play SS isn't relevant to whether or not he's more valuable than another payer that didn't play SS either.
143. Ivan Grushenko of Hong Kong Posted: December 21, 2008 at 03:21 AM (#3035001)
If it's a very small change, then why do more teams not move their 2B to SS? Instead of the Orioles fielding Brian Roberts at 2B and a parade of replacement-level shortstops, why not move Roberts to short and find a replacement-level second baseman?

The defensive challenges.

And shouldn't that be something that our value systems capture?

Do I understand correctly that in Dial's system, this change is not considered a possibility? 2B and SS are two different positions and twain never meet -- ie. Roberts hasn't played SS and therefore his potential defense at SS is unknown. Roberts' defense at 2B tells us nothing about Roberts' prospective defense at SS.

In Tango's system Roberts defense at SS would be his projected defense at 2B less 5 runs if he's never played anything other than 2B.
144. Chris Dial Posted: December 21, 2008 at 03:26 AM (#3035003)
Roberts' defense at 2B tells us nothing about Roberts' prospective defense at SS.
Easy. It suggests. It is just far from a foregone conclusion, and cannot be added in to his value.

In Tango's system Roberts defense at SS would be his defense at 2B less 5 runs if he's never played anything other than 2B.
If you want to say "Next year we're going to try Roberts at SS, so we can (ballpark) expect he'll be -5 there", dig in. But I think it is incorrect to say or even assume that Roberts is a 0 at 2B therefore he *IS* a -5 at SS.

ARod was a +10 at SS. He's a -5 at third base. It is positively incorrect to say that during ARod's last season at SS, his ability at 3B would be a +15 at 3B. That is clearly not what happened.
145. DJ Endless Grudge Plays All The Hits (CW) Posted: December 21, 2008 at 03:31 AM (#3035007)
If you want to say "Next year we're going to try Roberts at SS, so we can (ballpark) expect he'll be -5 there", dig in. But I think it is incorrect to say or even assume that Roberts is a 0 at 2B therefore he *IS* a -5 at SS.

ARod was a +10 at SS. He's a -5 at third base. It is positively incorrect to say that during ARod's last season at SS, his ability at 3B would be a +15 at 3B. That is clearly not what happened.

But - as you keep asserting - we're after value, not ability. We don't particularly care, in a value metric, if a particular fielder maps well to these estimates.

The problem is that the way we measure a player's defensive value, we measure it relative to position. Just because that's how we record the data, doesn't mean that's what we're trying to measure! It may be easier or more convenient to say that because someone plays second base, we'll only compare his fielding performance to other second basemen. It's an accounting trick, not a statement of any real underlying truth.
146. Ivan Grushenko of Hong Kong Posted: December 21, 2008 at 03:32 AM (#3035008)
ARod was a +10 at SS. He's a -5 at third base. It is positively incorrect to say that during ARod's last season at SS, his ability at 3B would be a +15 at 3B. That is clearly not what happened.

So another possible way to estimate A-Rod's ability at 3B based on his play at SS would be to go back to Tango's scouting report and look at A-Rod's grades at those skills that 3B need, and then look at the grades at 3B of real 3B who had similar grades. The trouble then is that those 3B are probably being compared to other 3B in the minds of the fan/scout and A-Rod to other SS.
147. The District Attorney Posted: December 21, 2008 at 03:34 AM (#3035010)
Naturally it's a general estimate that you'll end up applying to a specific player. I think it should be pretty clear that it's two completely different statements to say A) a SS can be responsible for five fewer runs created on offense/runs allowed on defense than a 3B and still be as good a player as the 3B is, vs. B) if you moved David Wright to SS, his hitting would be the same and he'd field five runs worse.
148. Chris Dial Posted: December 21, 2008 at 03:37 AM (#3035012)

But - as you keep asserting - we're after value, not ability. We don't particularly care, in a value metric, if a particular fielder maps well to these estimates.
To be clear, *I* am. I think Tango is after ability.

The problem is that the way we measure a player's defensive value, we measure it relative to position. Just because that's how we record the data, doesn't mean that's what we're trying to measure! It may be easier or more convenient to say that because someone plays second base, we'll only compare his fielding performance to other second basemen. It's an accounting trick, not a statement of any real underlying truth.
I don't see any purpose to that distinction. What I do is grounded is specific reality. DA/DR used to included every BIP for every player, but when you include BIP that are fielded at a zero rate, it's a fruitless activity. That's fine to describe it that way, but it is less applicable to reality, IMO.
149. Chris Dial Posted: December 21, 2008 at 03:38 AM (#3035013)
Naturally it's a general estimate that you'll end up applying to a specific player. I think it should be pretty clear that it's two completely different statements to say A) a SS can be responsible for five fewer runs created on offense/runs allowed on defense than a 3B and still be as good a player as the 3B is, vs. B) if you moved David Wright to SS, his hitting would be the same and he'd field five runs worse.
Unfortunately, the second is how it is used.
150. The District Attorney Posted: December 21, 2008 at 03:39 AM (#3035014)
What does that matter, when you disagree with the first anyway ;)
151. Chris Dial Posted: December 21, 2008 at 03:41 AM (#3035016)
A) a SS can be responsible for five fewer runs created on offense/runs allowed on defense than a 3B and still be as good a player as the 3B is,
Doesn't that make a huge assumption about the average performance? Like in your world of all replacementlevel players at a position?
152. Chris Dial Posted: December 21, 2008 at 03:42 AM (#3035018)
So another possible way to estimate A-Rod's ability at 3B based on his play at SS would be to go back to Tango's scouting report and look at A-Rod's grades at those skills that 3B need, and then look at the grades at 3B of real 3B who had similar grades. The trouble then is that those 3B are probably being compared to other 3B in the minds of the fan/scout and A-Rod to other SS.
I don't think those people would have picked up on whatever his problem is. I don't know what it is.
153. DJ Endless Grudge Plays All The Hits (CW) Posted: December 21, 2008 at 03:46 AM (#3035020)
Maybe it'll be clearer if we discuss outfielders here for a second.

From 2002-2008, average ZR for the three outfield spots:

POS ZR
LF 0.864
RF 0.875
CF 0.879

I think we can all agree that the act of shagging a fly ball is essentially the same regardless of the OF spot involved - a person who can shag flies in CF can do so equally as well in LF or RF. (There are differences, but they are very middling ones, I think.) And we can also agree that balls hit to CF or RF aren't significantly easier than balls hit to LF to field.

And this holds up under scrutiny - players who played all three OF spots tended to have the same ZR regardless of which position they played.

So why the discrepancy between the average ZR at the positions? Because teams tend to put their best fielders where there are more balls to field. In this case, a left fielder who is +5 in 100 BIZ is less valuable than a center fielder who is +5 in 100 BIZ, because the left fielder has made fewer plays.
154. Chris Dial Posted: December 21, 2008 at 03:48 AM (#3035021)
So why the discrepancy between the average ZR at the positions? Because teams tend to put their best fielders where there are more balls to field. In this case, a left fielder who is +5 in 100 BIZ is less valuable than a center fielder who is +5 in 100 BIZ, because the left fielder has made fewer plays.
No, he's equivalent in performance.
155. DJ Endless Grudge Plays All The Hits (CW) Posted: December 21, 2008 at 03:53 AM (#3035026)
No he flipping isn't. If you go out to the course and play Tiger Woods in a friendly game of golf, and you play with a handicap and the two of you shoot the same score, who shot better?
156. The District Attorney Posted: December 21, 2008 at 03:56 AM (#3035027)
Doesn't that make a huge assumption about the average performance? Like in your world of all replacementlevel players at a position?
Sure, and we could definitely argue about whether these are the right numbers for either this moment in time or for all of baseball history... but as I understand it, your objection is not to the exact numbers, it's to the concept of even doing this at all.

Just to clarify, although the data for this inevitably comes from players who switch positions, I don't think it's ultimately primarily about the few players per year at the major league level who actually end up doing that. The point is not that we actually are going to move Pedro Feliz to 1B to replace Ryan Howard. It's that we could do that (or move Chase Utley, or Raul Ibanez, or...), that those are all options available to our team, and that fact is relevant to Howard's... I won't say "value" if you prefer... but it's something that we really need to factor in if we're going to say much of anything meaningful about Ryan Howard.
157. Ivan Grushenko of Hong Kong Posted: December 21, 2008 at 03:56 AM (#3035028)
Naturally it's a general estimate that you'll end up applying to a specific player. I think it should be pretty clear that it's two completely different statements to say A) a SS can be responsible for five fewer runs created on offense/runs allowed on defense than a 3B and still be as good a player as the 3B is, vs. B) if you moved David Wright to SS, his hitting would be the same and he'd field five runs worse.

Unfortunately, the second is how it is used.

I'm trying to figure out whether it is intended to be used in the second way. It would help if I could follow how these positional adjustments were derived.
158. DJ Endless Grudge Plays All The Hits (CW) Posted: December 21, 2008 at 03:58 AM (#3035031)
159. Ivan Grushenko of Hong Kong Posted: December 21, 2008 at 03:59 AM (#3035032)
Just to clarify, although the data for this inevitably comes from players who switch positions, I don't think it's ultimately primarily about the few players per year at the major league level who actually end up doing that. The point is not that we actually are going to move Pedro Feliz to 1B to replace Ryan Howard. It's that we could do that (or move Chase Utley, or Raul Ibanez, or...), that those are all options available to our team, and that fact is relevant to Howard's... I won't say "value" if you prefer... but it's something that we really need to know if we're going to say much of anything meaningful about Ryan Howard.

Is Mark Teixeira vs Matt Holliday a good comparison? Are we trying to assess the additional value Holliday brings by being able to play LF, RF, CF (sort of), DH and maybe 1B, whereas Teixeira can only play 1B and DH?
160. Chris Dial Posted: December 21, 2008 at 04:01 AM (#3035033)
No he flipping isn't. If you go out to the course and play Tiger Woods in a friendly game of golf, and you play with a handicap and the two of you shoot the same score, who shot better?
Well, you are right in that you are framing the question in an ambiguous manner. If Tiger Woods and I shoot the same score, we shoot the same score. That's not particularly debatable. No one "shot better". We shot the same. I either played above my talent level or he played below his, but the fact of the matter is, if we shoot the same score, we shoot the same score. Now, if our scores are the same because I subtracted shots from my score, then we didn't shoot the same score.

Now back to your earlier question :if your comment is meant to be a rate stat, that's different, the CF gets more chances (500 vs 400). Or if you are referencing plays because a play to CF is worth more runs than a play to LF. But, if you are saying over the course of a season, the LF is a +5 DRS and a CF is +5 DRS, then they have contributed equally in turning balls into outs relative to their counterparts.
161. The District Attorney Posted: December 21, 2008 at 04:03 AM (#3035035)
Are we trying to assess the additional value Holliday brings by being able to play LF, RF, CF (sort of), DH and maybe 1B, whereas Teixeira can only play 1B and DH?
IMO, yes.
162. Chris Dial Posted: December 21, 2008 at 04:05 AM (#3035036)
Sure, and we could definitely argue about whether these are the right numbers for either this moment in time or for all of baseball history... but as I understand it, your objection is not to the exact numbers, it's to the concept of even doing this at all.
It's both. I don't personally believe that's the right way to evaluate players, but it is arguable to do it the other way. I dont agree with these adjustments because they are not going to be right more often than not.
163. Chris Dial Posted: December 21, 2008 at 04:07 AM (#3035040)
IMO, yes.
That's a completely different "value", not related to the actual output of that player on the field. Within a team construct, Holliday is worth more money because of that, but his OPD still is only relevant to his performance.
164. DJ Endless Grudge Plays All The Hits (CW) Posted: December 21, 2008 at 04:09 AM (#3035042)
Dial: I'm not much of a golfer, so I'm probably not being clear. You didn't shoot the same score, but had the same score after subtracting the handicap.

To the point: the only reason that the LF is a +5 plays and the CF is a +5 plays is because you're comparing them to different baselines. Our LF made 91 plays and our CFer made 93 plays. (Roughly - it's technically impossible to be +5 exactly at either position in 100 BIZ given the average ZRs I listed.)

If instead I simply use an outfield ZR of .873 (the average ZR of ALL outfielders in that timespan) then our LF is +3.7 and our CF is +5.7, and then our CF is +2 plays above our LFer.

I'm willing to be convinced that this is less correct than the way you do things, but I'd have to see it.
165. Chris Dial Posted: December 21, 2008 at 04:19 AM (#3035050)
To the point: the only reason that the LF is a +5 plays and the CF is a +5 plays is because you're comparing them to different baselines.
They *have* different baselines. For a LF to perform above his counterparts he has to catch 91 FBs. For a CF, he has to catch 93. Either way, the LF prevented as many runs above average at his position as the CF did. There is no "average of all OF positions". That isn't a position.
166. Ivan Grushenko of Hong Kong Posted: December 21, 2008 at 04:23 AM (#3035054)
158 -- Thanks Colin, I've read that before and I'm still not entirely clear, but it did help to re-read it. It looks like Tango's saying there are 3 positions -- C, IF and OF -- and not 8 (not counting pitcher).

Let's say you're considering acquiring a hypothetical SS Bob and moving him to 3B for 2009. Bob has never played 3B. So to project his 3B defense for 2009 you first project his SS defense for 2009 (say +5) based on his SS performance in the past (say 2008, 2007 and 2006) plus the Fan Scouting Report. You then add 5 runs and get +10. Bob then proceeds to field - 5 in 2009 at 3B [EDIT not SS]. Then, after 2009, you want to project his defense for 2010.

In Tango's system you then project this by considering Bob's 2009 defense at 3B, the Fan Scouting Report AND his SS defense in 2008 and 2007 (if you use 3 years in your projections).

In Dial's view you ignore 2008 and 2007 and only use Bob's 2009 3B number and maybe the Fan Scouting Report in your calculations. Dial believes that considering 2008 and 2007 SS performance is more likely [EDIT sp] to do harm than good. Further Dial believes that these defense stats relative to average or replacement are really designed to estimate historical value and not so much to project future performance.

Have I accurately summarized?
167. Chris Dial Posted: December 21, 2008 at 04:28 AM (#3035057)
In Dial's view you ignore 2008 and 2007 and only use Bob's 2009 3B number and maybe the Fan Scouting Report in your calculations. Dial believes that considering 2008 and 2007 SS performance is more likey to do harm than good. Further Dial believes that these defense stats relative to average or replacement are really designed to estimate historical value and not so much to project future performance.

Have I accurately summarized?
No. The bold part, yes.I certainly consider his performance at a tougher position, but it is merely a temperence. I am going to weight his actual play at 3B much heavier than his SS play.
168. DJ Endless Grudge Plays All The Hits (CW) Posted: December 21, 2008 at 04:29 AM (#3035059)
They *have* different baselines. For a LF to perform above his counterparts he has to catch 91 FBs. For a CF, he has to catch 93. Either way, the LF prevented as many runs above average at his position as the CF did. There is no "average of all OF positions". That isn't a position.

You gave them different baselines. That was a conscious decision.

There's two different definitions of value here - value relative to position and value relative to league. You're presupposing - or if you're not, then it's an unintended consequence - that in aggregate, each position is equally valuable.

And for the most part, this is correct, but that assumption will break down whenever players are unevenly distributed among positions. During Barry Bond's best seasons, the average LFer was more valuable than the average player at other positions. And that's an example - it's not the only case.
169. Chris Dial Posted: December 21, 2008 at 04:36 AM (#3035061)
You gave them different baselines. That was a conscious decision.
No, *I* didn't. The positions exist. They were established 150 years ago. Balls put in play in this section of the field are turned into outs at this rate. If LFs converted at a higher rate than CFs one season, my system holds.

There's two different definitions of value here - value relative to position and value relative to league. You're presupposing - or if you're not, then it's an unintended consequence - that in aggregate, each position is equally valuable.
No, I am not. Not even unintentionally. I am saying that each positions runs created or saved above the average at that position is equal. A run saved at SS is no more valuable than a run saved in LF.

And for the most part, this is correct, but that assumption will break down whenever players are unevenly distributed among positions. During Barry Bond's best seasons, the average LFer was more valuable than the average player at other positions. And that's an example - it's not the only case.
170. DJ Endless Grudge Plays All The Hits (CW) Posted: December 21, 2008 at 04:46 AM (#3035066)
1999 AL - hand to God, the average LFer was out-OPS'ed by the average CFer.

CF - .275/.346/.430
LF - .272/.339/.432

I'm reasonably confident that a LWTS-based analysis would widen the gap (probably just a shade) in CF's favor - OBP being worth more than SLG and all. But LFers comfortably outhit CFers in 1998 and 2000; did anything change in 1999? Is anything really different?
171. Mefisto Posted: December 21, 2008 at 04:55 AM (#3035069)
How are better players defined? How they perform relative to their peers. So, teams maximize their chance to win a pennant by increase their players performance relative to their peers. Is that not a correct set of assumptions?

I would say it's not a correct set of assumptions. It leaves out how they get to be "peers". Under your approach, someone simply "is" a CF. But that's not the full story. The full story is that the manager chose to play him in CF. And the reason he did so is that this player, as opposed to the others, could handle the defensive responsibilities of CF. That ability to handle the position is what your approach is missing -- it has value.

Oh, and no, the value I am talking about is his performance, not his trade value or his ability - whether or not he "can" play SS isn't relevant to whether or not he's more valuable than another payer that didn't play SS either.

I think this is a significant reason for the debate. Trade value is critically important -- it's what allows managers to construct teams which (in your words) "increase their players performance relative to their peers". GMs don't always trade like for like (CF for CF). More often, they trade one position to fill another. They need to know the value of what they're giving up when they do that.
172. Chris Dial Posted: December 21, 2008 at 05:01 AM (#3035072)
It leaves out how they get to be "peers". Under your approach, someone simply "is" a CF. But that's not the full story. The full story is that the manager chose to play him in CF. And the reason he did so is that this player, as opposed to the others, could handle the defensive responsibilities of CF. That ability to handle the position is what your approach is missing -- it has value.
*How* they get to peers doesn't matter for the value I speak off. Only performance does.
They need to know the value of what they're giving up when they do that.
My system does that, I think.
173. DJ Endless Grudge Plays All The Hits (CW) Posted: December 21, 2008 at 05:03 AM (#3035074)
*How* they get to peers doesn't matter for the value I speak off. Only performance does.

Only if the process of how they get to be peers equally distributes talent in the timespan in which you are measuring.
174. David Cameron Posted: December 21, 2008 at 05:04 AM (#3035075)
I have to be honest, Chris - your narrow definition of value is nearly worthless on a lot of practical levels. It only works going backwards, assumes a whole bunch of variables are fixed, and cares nothing about the potential value of the things that it isn't actually measuring. If we were having a discussion about who was better in 1947, maybe I'd care about your definition of value, but since I care about who is better in 2008 and how it might affect team's decision making for 2009, I absolutely want to know about ability as well as value.

Ignoring ability, and ignoring the potential for change in variables that you consider fixed (especially as it relates to each player's "peers"), makes your definition of value useless in looking forward. Most of us care about looking forward. If you don't, that's fine, I guess, but I'm not sure why'd want us to think OPD is something we should find useful.
175. DJ Endless Grudge Plays All The Hits (CW) Posted: December 21, 2008 at 05:18 AM (#3035081)
Or let's go back to, say, the 2008 NL. The average shortstop hit about like the average second baseman. Meanwhile, over in the AL, the average shortstop hit worse than the average catcher. We know Dial's OPD seperates the leagues - does it really make any sense to say that the average AL SS was equal to the average NL SS? Or should we think that for some reason, all of the best shortstops ended up in the NL this year, and it's biasing our results?
176. Chris Dial Posted: December 21, 2008 at 05:23 AM (#3035083)
Chris - your narrow definition of value is practically worthless.
Thank you.
It only works going backwards, assumes a whole bunch of variables are fixed, and cares nothing about the potential value of the things that it isn't actually measuring. If we were having a discussion about who was better in 1947, maybe I'd care about your definition of value, but since I care about who is better in 2008
Well, yes, that's all I've claimed. That's not worthless at all.
how it might affect team's decision making for 2009, I absolutely want to know about ability as well as value.
How a player performs in 2007 and 2008, as defined by OPD is a terrific descriptor of how he'll perform in 2009. How is that not useful for looking forward?

Ignoring ability, and ignoring the potential for change in variables that you consider fixed (especially as it relates to each player's "peers"), makes your definition of value useless in looking forward.
This simply isn't true because what you claim I claim is fixed *doesn't* vary very much season to season.
Most of us care about looking forward. If you don't, that's fine, I guess, but I'm not sure why'd want us to think OPD is something we should find useful.
First, please say "I", not "we". And I disagree that "Most of us" is accurate. Some of you. And past performance is a good predictor of future expectations. So OPD, as a good descriptor of past performance is a good predictor of future expectations.

In the very rare situations where a team signs a player to play a position he doesn't have experience at, there is some utility in approximating his future performance there. However, pick a FA - he's very likely to play the position he did in 2008. So, what value isn't being captured? Moreover, what value do you think OPD doesn't capture that something else does? It works fine for looking forward, and I use it for that - as accurately as the next methodology.
177. Mefisto Posted: December 21, 2008 at 05:29 AM (#3035084)
*How* they get to peers doesn't matter for the value I speak off. Only performance does.

Not if I'm a GM looking to trade a SS for a 2B. That new guy I get can't play SS. Even if the two players are both exactly average, I'm giving up something by making that trade. I need to know what that "something" is.
178. Chris Dial Posted: December 21, 2008 at 05:37 AM (#3035089)
Not if I'm a GM looking to trade a SS for a 2B. That new guy I get can't play SS. Even if the two players are both exactly average, I'm giving up something by making that trade. I need to know what that "something" is.
I double dog dare someone to define that accurately.

Okay: the difference is less than the accuracy of the methodologies, isn't it?
179. DJ Endless Grudge Plays All The Hits (CW) Posted: December 21, 2008 at 05:41 AM (#3035091)
Chris - I've now presented three cases in which I feel that the way you handle positional adjustments is inaccurate - Barry Bonds, 1999 AL left fielders, 2008 NL shortstops. Should I keep looking for more? Where's the threshold for number of examples before you become concerned?
180. DJ Endless Grudge Plays All The Hits (CW) Posted: December 21, 2008 at 05:46 AM (#3035092)
2006 AL - League-average catcher is a league-average hitter, largely based upon a career-best .347/.429/.507 season out of Joe Mauer. Greg Zaun also had a career year and guys like Jorge Posada and Victor Martinez had very good seasons as well.

EDIT: Misread sOPS+ versus tOPS+ there; the league average catcher was a 93 OPS+, not quite league average but still a hell of an offensive baseline for a catcher.
181. Chris Dial Posted: December 21, 2008 at 05:50 AM (#3035095)
I've now presented three cases in which I feel that the way you handle positional adjustments is inaccurate - Barry Bonds, 1999 AL left fielders, 2008 NL shortstops. Should I keep looking for more? Where's the threshold for number of examples before you become concerned?
Pardon? How did you present anything? You merely noted that players as a group performed well. Barry Bonds was a problem because he alone moved the number.
The average shortstop hit about like the average second baseman. Meanwhile, over in the AL, the average shortstop hit worse than the average catcher. We know Dial's OPD seperates the leagues - does it really make any sense to say that the average AL SS was equal to the average NL SS? Or should we think that for some reason, all of the best shortstops ended up in the NL this year, and it's biasing our results?
What does that mean? I've stated before that OPD doesn't erase the raw data. I don't see how anything of what you write says anything about my methodology. You'll have to explain better.
182. Chris Dial Posted: December 21, 2008 at 05:52 AM (#3035096)
2006 AL - League-average catcher is a league-average hitter, largely based upon a career-best .347/.429/.507 season out of Joe Mauer. Greg Zaun also had a career year and guys like Jorge Posada and Victor Martinez had very good seasons as well.
What do you think that means? It means that the Twins and teh Yankees and teh Indians got less of an advantage in value from having a good catcher than a more typical year.

If everyone is special, then no one is.
183. DJ Endless Grudge Plays All The Hits (CW) Posted: December 21, 2008 at 06:02 AM (#3035101)
What do you think that means? It means that the Twins and teh Yankees and teh Indians got less of an advantage in value from having a good catcher than a more typical year.

If everyone is special, then no one is.

Yeah, but that doesn't matter if we're comparing players between positions. Does Mauer become more deserving of the 2006 MVP ballot if Greg Zaun took a foul tip to the face in May and was out for the season instead?

Let's look at it from the context of the 2006 Twins for a second. Mauer and Morneau both put up practically identical OPSes; I don't have the defensive metrics in front of me for either but let's presume that they were both even on defense for their position - I don't think that's too far off the point. Does anyone other than the MVP voters that year think that Morneau was even AS valuable as Mauer? Ignore the league for a second - if we put Morneau behind the plate, do we think there's any way he's even adequate back there? If we put Mauer at first base, do we have any reason to believe that he won't be a perfectly acceptable defensive first baseman?
184. Jeff K. Posted: December 21, 2008 at 06:23 AM (#3035105)
He's in the hitting lineup either way. He's playing two similarly difficult defensive positions with similar skill. He's not providing significantly different production, he's just being accounted differently. To me, and I might be explaining myself poorly, that doesn't make him awesomely more valuable.

You use a key word in there: accounted. Just as accounting and finance are different, so are production and value different. Production can be thought of as the raw amount of runs created/saved. Value is just that, the *value* of those runs. This would mean, to me anyway, that in its extreme, a perfect valuation of a season takes in every context of every event, happenstance, and irregularity. That would include lineup position adjustments (mentioned earlier). No matter how small the adjustment, it should be there. Guy A can account for +30 runs AA in RF while Guy B can account for +30 runs at SS. That doesn't mean the value of those runs, in full or partial context, is the same. I realize that this would eventually bring in a host of contextual factors like team success (Is Guy A's +10 run season more valuable than Guy B's +20 if A's team makes the playoffs by one game and B's finishes 30 games back?) Frankly, in the proper discussion, that's a worthy question.

Because of the rules surrounding it, accounting basically tells you only how much money was made in the accounting sense. Finance strips away most of those rules and asks the question "How much is that worth?" Both are valuable; for one thing, it's accounting and its rules that gives the raw numbers for finance to start with. Similarly, the rules of VORP or whatever total player metric tell you a raw number. But you don't have to stop there. Two different raw levels of production that are equivalent can and likely will have different values in the end.
185. Jeff K. Posted: December 21, 2008 at 06:25 AM (#3035106)
Does Mauer become more deserving of the 2006 MVP ballot if Greg Zaun took a foul tip to the face in May and was out for the season instead?

Skipping everything in between real quickly (so sorry if this has been said): Yes. Does the value of your gold bar go up if the world's largest gold mine blows up?
186. DJ Endless Grudge Plays All The Hits (CW) Posted: December 21, 2008 at 06:29 AM (#3035107)
Skipping everything in between real quickly (so sorry if this has been said): Yes. Does the value of your gold bar go up if the world's largest gold mine blows up?

But what if Greg Zaun isn't a gold mine, just a comodity (like oil) whose price is being driven up by a short-term bubble that won't last?
187. DJ Endless Grudge Plays All The Hits (CW) Posted: December 21, 2008 at 06:49 AM (#3035108)
Okay, stepping back for a second.

Dial's main contention is that, since a team has to field both a 2B and a SS, the average SS and the average 2B are equally valuable. The problem is that MLB is not very efficent at distributing talent among teams - it's very possible for one team to have the best AND the second best shortstop in the league at the same time, for example. Thanks to long years of club control for young players, long free-agent contracts, and the relative difficulty of making trades, we actually have a very inefficent process for putting every player at his best position.

This method of value only works if we assume that every player has only one position, or at least that each player is being used optimally - and even then, concepts like "average hitting at LF" or "average fielding at SS" seem to at least bend, if not break, during the peaks of Bonds' or Ozzie Smith's career.
188. Jeff K. Posted: December 21, 2008 at 07:20 AM (#3035117)
To be clear, I'm not defending Dial's assertions. For one, I haven't read them yet. If your statement of his position is accurate, then I disagree with it on face before reading his explanation.

But what if Greg Zaun isn't a gold mine, just a comodity (like oil) whose price is being driven up by a short-term bubble that won't last?

First, you are mixing the terms price and value. That's a little pedantic, but I'm just noting it. Second, you're bringing up an externality that isn't germane to the point of the analogy, that scarcity creates value. Assuming he is over replacement level, the existence of Gregg Zaun drives the value of all other catchers down, and his disappearance drives their value up. Given the independence of pennants, Gregg Zaun missing the season from May to the end makes Joe Mauer's season more valuable than if Zaun had played the whole time.
189. Exploring Leftist Conservatism since 2008 (ark..) Posted: December 21, 2008 at 07:34 AM (#3035121)
This is why, in my HOF analysis of him, I treat Jeter as a slightly subpar defender at 3b.

Why on earth would anyone do this? He's a SS. You can't treat him as a 3B . . . you need to figure out his actual value as a SS, anything else is a wildly inaccurate shortcut, IMO.
Not at all, though I admit that to some degree it's an existential issue. If you accept an analysis of potential HOFers as an exercise in comparison, it doesn't follow that you need to accept that the current roster of HOFers is the most just roster it was possible to compile. If it doesn't follow, then reimagining a player's value is a useful tool, done in order to broaden the range of comparison. It's common to think of Jeter as a ready HOFer because he's such a good hitter "for a shortstop". If he was a LFer I don't imagine most of us would consider Jeter a lock for election. The problem is, he's a faux shortstop. He's probably more suited to 3b (and even there his limited mobility would likely make him a below average defender), and if he played that position, we would consider him a slightly less valuable hitter than Ron Santo, and a somewhat less valuable fielder than Santo, a man who has not yet been able to make the HOF. Add to that the cost to the Yankees of putting Rodriguez at 3b, a decision Jeter surely had control over, and the debit against Jeter only increases. But, as I've said, this is a tool.

"Why on earth" would I do this? Because Jeter is a third basemen pretending to be a shortstop. Why am I obliged to accept without reservation the man's impersonation?
190. CrosbyBird Posted: December 21, 2008 at 07:44 AM (#3035122)
Let's look at it from the context of the 2006 Twins for a second. Mauer and Morneau both put up practically identical OPSes; I don't have the defensive metrics in front of me for either but let's presume that they were both even on defense for their position - I don't think that's too far off the point. Does anyone other than the MVP voters that year think that Morneau was even AS valuable as Mauer?

Using OPD, Mauer would be worth more runs on offense than Morneau if they put up the same OPS, because the average C is a worse hitter than the average 1B, so he would be considered more valuable. Using positional adjustments, Mauer would get extra runs added to his total and Morneau would have runs deducted from his total, so he would be considered more valuable.

If the Twins had a +5 overall catcher as their best replacement for Mauer but a -5 overall 1B as their best choice for replacing Morneau, then Morneau would be the more valuable player as far as the 2006 Twins are concerned.
191. fret Posted: December 21, 2008 at 07:47 AM (#3035123)
This is a great discussion.

My approach is, first figure out what question you want to answer, then design the stat to answer that question.

If I am measuring value, my baseline is going to be replacement, not average. Consider two players, A and B, who play the same position (say 3B) and are both average fielders.

Player A: +20 RAA in 140 games
Player B: +20 RAA in 162 games

Player A missed some games with an injury, during which time 3B was filled by a replacement level player. His team got less overall production from 3B than Player B's team. By my lights, Player B was more valuable than Player A.

Dial may agree or disagree with this, I'm not sure. If he disagrees, then we just have different notions of value. I think mine is more meaningful, but then I would, wouldn't I? Not really much else to say. If he agrees, then OPD needs to be adjusted for playing time. That can be done, but it goes against my approach of question first, stat second.

If I am measuring ability, the choice of baseline doesn't make much difference. But that's not really germane to this discussion.

For value, I begin in the same place as Blackadder in #44:

what ultimately matters for measuring value has to be the quality of the player who could replace for the player in question for essentially no cost.

Take Chase Utley. I don't care how he would have performed as a SS, 3B, CF, etc. I care how much better he was than a replacement 2B. And, the pool of replacement 2B includes players whose primary position is SS or 3B.

Dave Cameron contends (I think?) that there are "freely available" 3B who could convert to 2B and be better than the players currently regarded as replacement 2B. These are big guys whose "natural position" is 3B, which clouds teams' judgment. Assuming for the sake of argument that this is right, there are two options.

Option 1: Follow the teams' lead in deciding which players are qualified for which positions. This is Dan R's choice. It allows us to do positional adjustments based mainly on levels of offense. (Still, I would follow Tango's approach for the OF, and stipulate that shortstops should be able to move to 2B and 3B.)

Option 2: Decide players' qualifications based on our own judgment. This is Tango's choice, and his method is appropriate if we choose this option. As Mefisto points out, it opens up a big can of worms about 2B vs. 3B in the dead ball era.

Under option 1, I think 2B would be about 5 runs higher than 3B on the positional spectrum. Under option 2, they seem to be about equal.

I go back and forth on which option I think is better. Right now I lean towards option 1, but there are certainly arguments on both sides.
192. DJ Endless Grudge Plays All The Hits (CW) Posted: December 21, 2008 at 08:04 AM (#3035127)
One last crack at this and I'm going to bed.

Dial:

A run saved at SS is no more valuable than a run saved in LF.

Sure. Fine.

And a single hit by a SS is no more valuable than a single hit by a LF - this should be even more obvious, because you don't bat as a SS or a LF, you bat as a batter. You can be a PH or DH and not even bat as a fielder at all!

We adjust for position because of the difference in defensive skill required - and because that's the difference between positions that's where we should make our adjustments.

The other objection Dial has is that he doesn't trust that Tango's positional adjustments are based on a large enough sample. But shouldn't Dial want a smaller sample? After all, he considers a one-year variation in difference between offensive positions to be meaningful information.
193. fret Posted: December 21, 2008 at 08:12 AM (#3035129)
Jeff K, you wrote in #9:

Just like the difference between a 200 OPS+ and a 190 OPS+ is more valuable than the difference between 110 and 100, the claim here is that being 7.5 runs over average defensively at short is more valuable than 7.5 runs over average at first. It's not that you're comparing him to second basemen now. It's that the 7.5 run difference is harder t accomplish at short than 1b, and hence, more valuable. You can't straightline compare the two as we've been doing lo these many years.

The difference between 200 and 190 is the same as the difference between 110 and 100 in terms of runs. The reason you might prefer the difference between 200 and 190 is that the conversion from WAR to salary is nonlinear, with superstars deserving more \$/WAR. (In practice the superstars get the same \$/WAR but longer contracts, which comes to the same thing. I know Tango has a different view, but I don't quite buy it.)

The only reason a 7.5 run gap at SS would translate into a bigger salary difference than a 7.5 run gap at 1B would be if the shortstops are better players overall than the first basemen. In order to figure out if that's true, you need a position adjustment in terms of runs. And judging from this thread, there's a long way to go before we can all agree on that!
194. fret Posted: December 21, 2008 at 08:18 AM (#3035131)
Clarifications to #191:

Dave Cameron contends (I think?)

Not in this thread, but elsewhere.

Under option 1, I think 2B would be about 5 runs higher than 3B on the positional spectrum. Under option 2, they seem to be about equal.

This is right now, not in the dead ball era.
195. Exploring Leftist Conservatism since 2008 (ark..) Posted: December 21, 2008 at 08:24 AM (#3035132)
A run saved at SS is no more valuable than a run saved in LF.

Sure. Fine.
Except that it's harder to save a run in LF than it is as SS.
196. Erik, Pinch-Commenter Posted: December 21, 2008 at 10:22 AM (#3035141)
Ok I haven't read all of the second page but I think I have a good grasp on what the real issue is here.

The problem is that there are at least 3 specific ideas and corresponding valuations being mixed and matched together. There is the old 'skill' vs 'value' issue going on but in addition to that there is an 'on the field value' and 'market value' issue within 'value'. Market value is likely a more specific way of quantifying skill in dollars, or wins. Let me explain each further...

ON THE FIELD VALUE:

In my opinion a lot of the positional adjustments being done are things that belong in 'market value' ratings. When you break it down from runs scored vs. runs allowed each player adds real actual runs when they bat, while they all save runs when they go into the field. Over time all positions should be equal to each other because defensive first positions will be saving runs at the same rate offense first positions score them. If not then teams will be more inclined to add better offensive players who are worse defenders to these spots.

To measure true on the field value of a player you simply calculate their total offensive contribution from a 0 baseline [plus base running], because we are trying to find the exact number of runs a player contributes. Then, instead of a positional constant like Tango uses [depending on how exactly he uses it this may be the same], there needs to be a positional multiplier that represents the average difficulty of the plays at each position. This would reflect how first base has the easiest plays while third base is more difficult. Then multiply it by the number of plays made to get a fairly accurate run value for defensive contributions. The offensive and defensive values added together would give you a true on the field value for the player. If you do this right then when you add all the players on a team together you should get a very similar expected runs allowed vs runs scored total to the actual numbers. Positional scarcity is a non-issue here. This is simply what you did on the field and I think this area is far too often ignored in these discussions and in 'all inclusive' statistics.

MARKET VALUE:

The other aspect of value was the market value of a player, which should be what teams calculate before deciding on their free agent targets and their respective dollar values. This is where theoretical defensive values come into play and positional scarcity is an issue. When figuring out market value one needs to know a players defensive value at virtually every position. The players 'on the field' defensive values need to be projected out to each position. Through studies, like Tango has done, we can find the average drop off or increase in defensive value over changes in various positions. A player will have various values for each position. A player like Carlos Beltran will have a good but not great offensive value overall with a high defensive rating in all three outfield spots, the highest defensive value being in centerfield where he has more, tougher plays that he can make. Then obviously compare the offensive plus defensive value of the player to all theoretical candidates at a position [so a first basemen would be judged against just about all players, while a SS would be judged against just SS's and those rare other players who could play there]. This is the market value of the player. Of course this is complicated by varying defensive values for each team [the value of defense on one team will be less than on other teams due to the high number of strikeouts their pitchers get].

So in conclusion, what we really need are for stats to be clear in expressing on the field values or market values of players.
197. Blackadder Posted: December 21, 2008 at 10:42 AM (#3035142)
A small point, fret: Dan's system actually sees 2B and 3B as almost identical today, just as Tango does. The biggest discrepancies, as I mentioned earlier, are in CF and SS.
198. Jeff K. Posted: December 21, 2008 at 10:57 AM (#3035145)
Jeff K, you wrote in #9:
(snip)
The difference between 200 and 190 is the same as the difference between 110 and 100 in terms of runs.

The 200/190 and 110/100 were stated as OPS+ numbers, just to note. But yes, I know this. That was my whole point.

The reason you might prefer the difference between 200 and 190 is that the conversion from WAR to salary is nonlinear, with superstars deserving more \$/WAR. (In practice the superstars get the same \$/WAR but longer contracts, which comes to the same thing.

Well, no, not really. The reason you prefer the jump from 200 to 190 is that it's a harder jump to make. It's more of an outlier, it's more standard deviations away from the median, however you want to put it. If we assume a normal distribution of performance, 200 to 190 is a more valuable difference than 110 to 100. That's why teams pay more for it, which is reflected in \$/Wins Added, which is not the same causality model. Your progression is akin to paying for the cart with the horse.

(In practice the superstars get the same \$/WAR but longer contracts, which comes to the same thing. I know Tango has a different view, but I don't quite buy it.)

I haven't seen this discussion, but on face, I'd have to agree with Tango, assuming the disagreement is whether higher \$/WA is equivalent to equal \$/WA for more years. For one, even if you purposefully machined the contract to do this, discount rates will vary in the real world, and you'd never hit it on the nose.

(EDIT)
Is fret someone I should know? I'm having a hard time deciding. Regardless, no offense intended in my second answer. You never know with unfamiliar handles around here how versed that person is in stats. Or anything, really. :) I'm certainly no maven, myself.
199. Jeff K. Posted: December 21, 2008 at 11:12 AM (#3035146)
Over time all positions should be equal to each other because defensive first positions will be saving runs at the same rate offense first positions score them. If not then teams will be more inclined to add better offensive players who are worse defenders to these spots.

Whoa, wait a minute. This is making an enormous assumption that I don't necessarily buy. It could well be that there's something physiological that causes more or less variance in the defensive play at a single position, or that causes one position to be harder (this is basically true prima facie.) You're assuming that the entire equation is in the control of the GM and manager.
200. Tango Posted: December 21, 2008 at 03:00 PM (#3035188)
Page 2 of 4 pages  < 1 2 3 4 >

You must be Registered and Logged In to post comments.

<< Back to main

### News

Old-School Newsstand

### Support BBTF

Thanks to
greenback fixes the cable
for his generous support.

### Bookmarks

You must be logged in to view your Bookmarks.

### Hot Topics

NewsblogOT - March 2017 NBA thread
(7405 - 8:20am, Jun 28)
Last: STIGGLES don't want to talk about cracker barrel

Sox TherapyRed Sox Retire Rich Garces' Number
(17 - 8:10am, Jun 28)
Last: Jose is El Absurd Bronson Y Pollo

NewsblogCENTRAL PARK - JULY 1, 2017 10 AM - BBTF ANNUAL SOFTBALL EXTRAVAGANZA
(286 - 8:09am, Jun 28)
Last: Edmundo got dem ol' Kozma blues again mama

NewsblogIndians manager Terry Francona leaves game after not feeling well
(5 - 8:03am, Jun 28)
Last: John DiFool2

NewsblogOTP 26 June 2017: Confederations Cup shows how sports media in Russia differs from politics
(420 - 8:00am, Jun 28)
Last: villageidiom

NewsblogThe Freeze popular with MLB fans and players | MLB.com
(6 - 7:13am, Jun 28)
Last: The Duke

NewsblogAaron Judge or Cody Bellinger?
(18 - 6:56am, Jun 28)
Last: Lance Reddick! Lance him!

NewsblogDodgers rookie Cody Bellinger has never heard of 'Seinfeld,' teammates hilariously respond
(163 - 6:48am, Jun 28)
Last: Hysterical & Useless

Gonfalon CubsMeh
(59 - 5:36am, Jun 28)
Last: Moses Taylor, Unwavering Optimist

NewsblogLive from New York, it's OMNICHATTER! for June 27, 2017
(33 - 3:04am, Jun 28)
Last: Snowboy

NewsblogIs There Still a Place for Ladies Night in Baseball?
(16 - 2:23am, Jun 28)
Last: Tin Angel

NewsblogSprint Speed shows fastest at each position | MLB.com
(35 - 1:45am, Jun 28)
Last: Alex meets the threshold for granular review

NewsblogFox Sports eliminates digital writing staff in favor of promoting their debate shows
(43 - 12:10am, Jun 28)
Last: Pat Rapper's Delight (as quoted on MLB Network)

NewsblogYankee Stadium Charity Is Called Into Question
(1 - 11:45pm, Jun 27)
Last: Cargo Cultist

NewsblogFormer Mets pitcher Anthony Young dead at 51
(9 - 11:41pm, Jun 27)
Last: Cargo Cultist

Page rendered in 0.7599 seconds
47 querie(s) executed