If Eric Byrnes wins the MVP, that'll basically provide me with enough utility to make up for the fact that the Red Sox are going to blow the division lead.

Please god yes.

Please god yes.

Seconded.

I have never understood the mindset that maximizes a closer’s impact during the season and minimizes it in voting for awards and the Hall of Fame.

Four closers winning Cy Young awards and three also winning MVP awards is minimizing their impact? And how ridiculous is it that closers win MVPs but not starters (since Roger Clemens)?

If Eric Byrnes wins the MVP, that'll basically provide me with enough utility to make up for the fact that the Red Sox are going to blow the division lead.

Uh... what? Boston wrapped up the East tonight.

Byrnes doesn’t have the gaudiest numbers, although they are solid, and his career best, but his team is in first place. And he has been their best player as well as their leader.

I have to confess that I hadn't actually looked at any D-Back stats yet this year, so I went to BB-Ref and Holy Crap! that's a BAD offense! Team OPS+ of 88!? Eric Byrnes really has been their best position player.

Of course, just glancing at BB-Ref, it seems pretty clear that the reason the D-Backs are doing well is because of their pitching - Webb and their top 5 relievers mostly.

I like debating MVP awards as much as the next guy, but the NL MVP race really is neither crowded nor wide-open (except compared to the A-Rod coronation in the AL, I suppose). If the Brewers don't win the Central, it's pretty much David Wright's to lose and even if the Brewers do win, the debate will largely be over whether Wright or Fielder <u>will</u> win, not over which of them deserves to (barring some absolutely mind-blowing performance by Fielder over the next two weeks).

barring some absolutely mind-blowing performance by Fielder over the next two weeks

He's off to a .410/.489/.821 September w/ 5HRs so far - in the middle of a tight race - so that certainly doesn't hurt his case with the writers.

Uh... what? Boston wrapped up the East tonight.

This represents the fifth or sixth time that the Red Sox have put away the 2007 division title for good. Congratulations to Boston... now, go get swept this weekend, then recover, then go for the all-time single-season record!

He's off to a .410/.489/.821 September w/ 5HRs so far - in the middle of a tight race - so that certainly doesn't hurt his case with the writers.

Except I think the nation hasn't noticed that. If this was Albert Pujols doing it or Barry Bonds doing it they would be talkimg MVP. But Prince and Milwaukee? I think the press has already written them off. Or I should they have already what that story will be, hot start, collapsed at the end.

For Milwaukee I would rather give it to Braun. Braun pretty much every full month he has played has outhit Prince. Granted Prince was darn good in those months Braun wasn't playing and was darn good when Braun was playing but the story I think is with Braun this year.

I personally would vote for Peavy over any position player. There not going to give it to Jones, he hasn't played enough. Not going to give it to Hanley he plays for Florida. Not going to give it to Bonds. Holliday? Plays for the Rockies though the Rockies are hanging in there and they have an outside shot. If they make it in I could see Holliday getting it.

Of course, just glancing at BB-Ref, it seems pretty clear that the reason the D-Backs are doing well is because of their pitching - Webb and their top 5 relievers mostly.

Here are the Win Shares for all of the 2007 D-backs' players:

2007 Byrnes E 29
2007 Hudson O 21
2007 Webb B 19
2007 Snyder 16
2007 Young C 15
2007 Drew S 14
2007 Jackson C 12
2007 Davis D 11
2007 Valverde J 11
2007 Pena T 11
2007 Lyon B 10
2007 Reynolds M 10
2007 Hernandez L 9
2007 Owings M 9
2007 Tracy C 7
2007 Quentin C 6
2007 Cruz J ARI 5
2007 Slaten D 4
2007 Gonzalez E 4
2007 Johnson R 4
2007 Montero M 3
2007 Hairston S 3
2007 Clark T 3
2007 Ojeda A 2
2007 Salazar J 2
2007 Medders B 1
2007 Nippert D 1
2007 Upton J 1
2007 Petit Y 1
2007 Cirillo J 1
2007 Barden B 0
2007 Murphy B 0
2007 Sadler D 0
2007 Bonifacio E 0
2007 Gonzalez E 0
2007 DaVanon J 0
2007 Peguero J 0
2007 Schultz M 0
2007 Choate R 0
2007 Hammock R 0
2007 Smith J 0
2007 Callaspo A -1
2007 Eveland D -1
2007 Durbin J -1
2007 Kennedy J -1
2007 Kim B -1

And here are the NL leaders in Win Shares:

2007 Wright D NYN 30
2007 Byrnes E ARI 29
2007 Pujols A STL 28
2007 Martin R LAN 25
2007 Beltran C NYN 24
2007 Ramirez H FLA 24
2007 Cabrera M FLA 24
2007 Fielder P MIL 24

** Some have suggested that Byrnes's Win Shares total is over-inflated because the Diamondbacks actual win total (83) has significantly exceded its P-win total (71). It should be noted, however, that the Mets have exceded their P-wins by 5 games. By contrast, Boston (and a few other teams) have done the reverse, winning 5 fewer games than expected. As such, the Red Sox best player, David Ortiz, is perhaps underrated by Win Shares. Of course, Papi Gordo stands no chance to win the MVP against Stray-Rod. But Magglio deserves serious consideration for the AL MVP.

2007 Wright D NYN 30
2007 Byrnes E ARI 29
2007 Pujols A STL 28
2007 Martin R LAN 25
2007 Beltran C NYN 24
2007 Ramirez H FLA 24
2007 Cabrera M FLA 24
2007 Fielder P MIL 24

That says more to me about the weaknesses of Win Shares as a metric than it does about how valuable Eric Byrnes has been.

Kiko, that's empty criticism. Perhaps you could explain it. I suspect you might be undervaluing his defense. Or maybe Win Shares is overvaluing it. I'm not sure. While I realize that there are better metrics for valuing a players' performance relative to others at his position, and hence what his WARP or VORP would be, Win Shares (seems to me) quite good at giving absolute credit for a given performance in a given season. The most valid criticism of WS, AFAIK, is that it is adjusted for actual team wins, not Pythagorean wins. (I don't know that that is the case. Someone suggested it on Primer. And if it is the case, I think it should be modified.)

The most valid criticism of WS, though, is that it is adjusted for actual team wins, not Pythagorean wins. (I don't know that that is the case. It is just what someone said on Primer. And if it is the case, it should be modified.)

It's not adjusted for actual team wins, it's based off of them. I find this to be the single strength of win shares for MVP-type analysis. My conception of value cares about the potentially unreproducible things which lead to a team exceeding its pythagorean record.

Win Shares makes a pretty good case to me that Byrnes should be a legit MVP candidate actually (barring a PBP defensive metric skewering him defensively).

Uh... what? Boston wrapped up the East tonight.

Gonfalon Bubble got at this already, but after the number of times the Yankees have been pronounced dead this year, I'm not wrapping anything up yet. I'm not going to require mathematical elimination first, but more than a five game lead

The most valid criticism of WS, though, is that it is adjusted for actual team wins, not Pythagorean wins.

It's true, and I think it's logical. Mr. James felt he needed a way to account for team success that just wasn't showing up in the run-scoring stats. Now, maybe that's luck, maybe it's not, but I don't think it's wrong to account for it.

EDIT: bibigon beat me to it.

That says more to me about the weaknesses of Win Shares as a metric than it does about how valuable Eric Byrnes has been.

It's the most valuable, not best, player award. If you accept that, Eric is a fine choice. He has really been the spark-plug on an awful offense, without him it's easy to see the DBacks in third or worse. Look at the rest of the outfield. Chris Young, Carlos Quentin, Justin Upton are all way, way below 100 on the OPS+ scale. Byrnes's hitting stands out in AZ more than Pujols does in STL.

And the one offensive strength the DBacks have had this year doesn't show up on OPS+, base running. Their steal percentage is second in the NL, and even guys who don't steal (Upton) are great at taking extra bases. My guess is that EByrnes stealing his last 30 in a row and going 45/51 on stolen bases helps create a lot more runs than his so-so OPS+ would lead you to believe. And those runs are very valuable to this team.

Win Shares (seems to me) quite good at giving absolute credit for a given performance in a given season. The most valid criticism of WS, though, is that it is adjusted for actual team wins, not Pythagorean wins. (I don't know that that is the case. It is just what someone said on Primer. And if it is the case, it should be modified.)

Win Shares are tied to actual wins, not Pythagorean wins, which is part of what's driving Byrnes number here (if you just scale his numbers down to tie to AZ's Pythag wins, his Win Shares drop to 24 or 25).

I don't actually have a problem with tying Player wins to Team wins - that's sort of the point of player performance after all. My problem is that Win Shares doesn't actually do a good job "at giving absolute credit for a given performance in a given season." By relying upon cumulative season-end stats, Win Shares has no way of actually identifying which players are responsible for Arizona's out-performing their Pythag so dramatically. So Win Shares just assumes that the team's best player (who it identifies as Byrnes) is most responsible for the extra wins too. But there's no reason to do that; he could be responsible, but you'd have to look at game-by-game and play-by-play statistics to be able to determine that.

Also, as an absolute measure, "Win" shares make no sense without also calculating "Loss" shares. As Bill James himself I thought first observed, the relevant comparison for a player's value isn't zero it's replacement level. So if you want to have a one-number uber-stat it's going to have to be relative to replacement level. This will also affect Byrnes, who leads the NL in games played and is 8th in plate appearances right now.

As far as defense, the fundamental problem with Win Shares is its failure to use play-by-play data.

So if you want to have a one-number uber-stat it's going to have to be relative to replacement level.

This is what VORP and WARP do. However, they have the problem (to my way of thinking when considering MVPs) that if you happen to be a great third baseman, for example, but play in a league loaded with great third basemen, your absolute performance will be undervalued. You may have the best numbers in your league, but your nearest competitors are all at your same position. Whereas, if you happen to be a great second baseman in a league where all of the other 2Bs suck that season, you will be massively overrated in an absolute sense.

It's the most valuable, not best, player award. If you accept that, Eric is a fine choice. He has really been the spark-plug on an awful offense, without him it's easy to see the DBacks in third or worse. Look at the rest of the outfield. Chris Young, Carlos Quentin, Justin Upton are all way, way below 100 on the OPS+ scale. Byrnes's hitting stands out in AZ more than Pujols does in STL.

Well okay but if they don't have Webb they are in third place as well.

FWIW, here are the top 5 pitchers (Cy Young?) in the NL, based on pitching win shares:

Webb B ARI 23.1*
Peavy J SD 20.4
Penny B LAN 19.7
Harang A CIN 18.5
Oswalt R HOU 18.3

And here are the top 5 pitchers in the NL, based on VORP:

Jake Peavy SDN 68.2
Brad Penny LAN 62.3
Brandon Webb ARI 59.1
John Smoltz ATL 53.2
Tim Hudson ATL 52.1

* Note that Webb has a -3.7 WS for his offense; and because of that, he doesn't rank as high on his team as he does among pitching leaders. And once again, WS likely overrates Webb, as it does Byrnes, due to the non-accounting for P-wins for the (overachieving) D-backs.

But Prince and Milwaukee? I think the press has already written them off. Or I should they have already what that story will be, hot start, collapsed at the end.

But if the Brewers were to squeak out the division, no matter how ugly it ends up being in the end, then all of a sudden the May story of "first Brewer playoff team since Harvey's Wallbangers" gets pulled out of storage again, and their top players will get the feel-good bump from that.

Then the built-in lead-in becomes "Prince Fielder, who led the National League in homeruns, and the Milwaukee Brewers to their first divisional title since [random '82 Brewer namechecks that'll vary by writer], was named the 2007 National League Most Valuable Player today".

Byrnes's hitting stands out in AZ more than Pujols does in STL.

Uhhhhhhh.

Pujols: 154 OPS+
Duncan: 117 OPS+

Next best is Encarnacion, 99 OPS+, unless you count Ankiel's cup of coffee.

Byrnes: 112 OPS+
Jackson: 106 OPS+
Hudson: 106 OPS+
Reynolds: 100 OPS+

Pujols: 154 OPS+
Duncan: 117 OPS+
Byrnes: 112 OPS+

Because of Byrnes's superior baserunning skills, OPS+ is a slightly misleading metric to compare these 3 players offensive productivity. Try EqA (with WARP1 in parentheses):

Pujols .330 (10.1)
Byrnes .289 (7.1)
Duncan .286 (3.5)

Pujols still dominates, but Byrnes surpasses Duncan.

The big story here is that this guy is a "writer." His writing is bizarre. Uses phrases like "there in" and "post the All-Star break" and "while until."

Of Byrnes he says, "Here's another player whose MVP candidacy defies tradition." No other such player was previously named. Later of Saito he says, "Here is the MVP candidate that defies traditional thinking." Well, I thought there was Byrnes and maybe another? But no, Saito is THE candidate who...or is it the candidate that? Or does Byrnes defy tradition and Saito defies traditional thinking? Maybe that's the difference.

Oh never mind.

Pujols still dominates, but Byrnes surpasses Duncan.

I think his point was that Pujols laps all his teammates, while Byrnes has just been a little better than some of his D'back teammates.

PS. Utley has certainly had a great year and I suppose the Phillies will split up some votes among them (some idiot will insist on voting for Howard again, right?), but how can Jimmy Rollins not be in the conversation?

sunny makes an excellent point. Rollins has a lot of markers including leading the NL in total bases. Who was the last shortstop in the NL to do that?

But 190 hits, 35 doubles, 17 triples, 27 homers, 126 runs scored, 32-38 on stolen bases and a pretty decent shortstop? That's a heckuva player................

sunny makes an excellent point. Rollins has a lot of markers including leading the NL in total bases. Who was the last shortstop in the NL to do that?

Ernie Banks?

Misir:

Correct. 1958.

Obviously A-rod in the AL. He did it three times with most recently in 2002.

Updated from previous NL MVP thread. With Byrnes added.

BRAA:
Fielder: 41
Wright: 42
Hanley: 30
Utley: 46
Pujols: 38
Chipper: 40
Byrnes: 22

WPA/LI
Fielder: 4.50
Wright: 4.40
Hanley: 3.7
Utley: 4.42
Pujols: 3.44
Chipper: 3.73
Byrnes: 1.69

Games
Fielder: 142
Wright: 143
Hanley: 138
Utley: 116
Pujols: 142
Chipper: 118
Byrnes: 146

Defense and position
Fielder: negative
Wright: small plus
Hanley: small negative
Utley: plus
Pujols: neutral
Chipper: neutral
Byrnes: neutral

My current thinking is that the above categories are the ones I mostly care about, I also consider how important a players wins are to his teams chances of making the playoffs. Not all things "count" the same, but I think it's important to keep all of them in mind.

I think it's pretty clear Chase Utley has been the best player in the national league this year when he has been on the field. But I'm not comfortable supporting him, based on how much PT he has missed. The same can be said for Chipper, though not to the same degree.

I'd also remove Hanley from contention, as replacement level is not the correct baseline for this discussion and he gets a pretty big hit when switching from replacement to average. Couple that with well his team being not good enough to leverage his wins enough to put him in the running with the other candidates, and of course his defense/position which I don't think is as bad as many do, but certainly isn't a plus.

Eric Byrnes isn't a strong candidate. While he has had a large impact on the pennant race. Comperable to a couple other candidates he is so far behind in batting metrics that it can't be counteracted considering his defensive position value is low, and well frankly he is has one of the lowest "baseball IQ's" i've ever seen from a player. I don't know if it's smarts or instinct or probably some combination, but it costs his teams a few number of runs annually. Then again if he was the smartest player in baseball, it wouldn't counteract his lack of offensive production relative to the other candidates.

Puljos looks to have a better case than I thought. His batting both in context neutral and context specific situations are only a step behind Wright and Prince. A good month, coupled with a Cards surge could put him right up their with the other two.

Wright and Prince are neck and neck offensively at this point, with a very small edge to Wright. Wright gets the head to head checkmark in terms of defense and positioning which puts him ahead of Prince, though Prince has advantage in terms of the value his wins are providing his team. If one player out hits the other by a considerable amount down the stretch that could sway things, the other factor is the Brewers playoff hopes. If they make it, and both players continue to play as is, then I THINK I would give Price a very small nod, but it would be extremely small. If the Brewers don't make it and both players continue to play at the same pace relative to each other then it's Mr. Wrights in a decision that isn't nearly as close.

This is a really interesting season in terms of NL MVP candidates. All stats from fangraphs.

Rollins is having a fine year, but don't forget that he has made even more outs than Juan Pierre -- a tall order indeed ...

You guys are overthinking this. Fielder will win in a walk if the Brewers win the division and he'll probably win in a close one if they don't. Win Shares? OPS+? Please. Count the homers and ribbies and divide by the cuddliness.

Note: I do actually think he will win easily.

Can something really be crowded and wide open?

Does MSNBC employ any baseball analysts whose writing doesn't inspire seppuku?

Here are the Win Shares for all of the 2007 D-backs' players:

Just eyeballing I'd say that the winshare ALLOCATION between offense, defense and pitching is off wrt to the DBacks.

It's the most valuable, not best, player award. If you accept that, Eric is a fine choice. He has really been the spark-plug on an awful offense, without him it's easy to see the DBacks in third or worse.

I don't know if Byrnes is the MVP, but I had completely missed just how good his season has been. He's been really surprising.

As for the D'Backs being in third or worse without him--hell, their RS/RA suggests they should be anyway. I know, I know--Jaffe did an excellent article that provides good tangible reasons for why the D'Backs have outperformed their pythag, and I find a lot of that persuasive. But still--they haven't just outperformed their pythag--they've outperformed it by TWELVE GAMES. And furthermore--they actually have the WORST pythag in the division (half a game worse than San Francisco's!), and they look poised to not only win the division, but to win it without a whole lot of drama. I'd have to think this is unprecedented.

I don't know if Byrnes is the MVP, but I had completely missed just how good his season has been. He's been really surprising.

Byrnes is hitting .296/.365/.485 OPS+ of 112
his career is .269/.331/.456 102

in 2003 he put up an OPS+ of 113
In 2004 he put up an OPS+ of 109

What's so surprising?
(ok the stolen bases- but outside of Roto who cares?)

OK he's hit .310/.417/.471 "late and close"
but here's the kicker- the WHOLE team (as an aggregate) has elevated it's performance "late and close" by the same % so to speak.

Orlando Hudson is hitting .380/.427/.595 "late and close" and he has a 106 OPS+ at 2b- I can't help but think that he's been more valuable than Byrnes

Chris Young is slugging .603 in tie games
Chris Snyder is hitting .304/.377/.500 "late and close"

Is any of this a repeatable skill? Probably not, but if you are an Az fan it must be fun to watch.

Let me echo Rich.

I'm far from a fan of Win shares, but when complaining about method issues it's helpful to point to the specific reason you feel a player is being ranked incorrectly rather than a generic complaint about method flaws.

I'm far from a fan of Win shares, but when complaining about method issues it's helpful to point to the specific reason you feel a player is being ranked incorrectly rather than a generic complaint about method flaws.

The main problem with Win Shares is the defesive Win Shares. Specifically, you get a lot of Win Shares just for showing up at the key defensive positions. Additonally, the actual metrics he uses to allocate defensive Win Shares are nothing short of random, arbitrary and as a result, a little embarrassing. He just assigns 10% to this, and 20% to that, with no reasoned basis for the percentages assigned and little basis for using the metrics he uses.

It's not adjusted for actual team wins, it's based off of them. I find this to be the single strength of win shares for MVP-type analysis. My conception of value cares about the potentially unreproducible things which lead to a team exceeding its pythagorean record.

But Win Shares doesn't do that. It simply divides the credits (or debits) among the players proportional to their contribution to pythagorean wins.

What you want is win shares calculated on pythag wins plus the extra win shares handed out by the players who are doing better by WPA analysis than by runs created. (As a starting point look at the relationship between WPA and BRAA at fangraphs)

Looks to me like the bulk of the extra credit should be going to the bullpen.

I'm far from a fan of Win shares, but when complaining about method issues it's helpful to point to the specific reason you feel a player is being ranked incorrectly rather than a generic complaint about method flaws.

I thought I did in response to him. Eric Byrnes is getting credit for being the primary reason that the D-Backs have outperformed their RS/RA without actually looking at any evidence that could help you determine that. The numbers that JPWF13 show in #35 suggest that Byrnes may be less responsible for these bonus wins than other hitters - e.g., Hudson and Young who appear to have out-hit Byrnes "close and late" and/or in tie games.

I'd also agree with JPWF13 in #33 - Win Shares appears to be giving too little credit to D-Backs pitchers, which translates into too much credit for D-Backs position players (e.g., Byrnes).

What you want is win shares calculated on pythag wins plus the extra win shares handed out by the players who are doing better by WPA analysis than by runs created. (As a starting point look at the relationship between WPA and BRAA at fangraphs)

Looks to me like the bulk of the extra credit should be going to the bullpen.

Precisely.

Kiko, apologies. Posted #36 after reading Rich's post and before reading your response.

Consider it a generic request.

What you want is win shares calculated on pythag wins plus the extra win shares handed out by the players who are doing better by WPA analysis than by runs created.

I've been critical of WPA since day one (especially since some seem to want to use it as the uberstat), but this actually seems to be a reasonable use for it.

(As a starting point look at the relationship between WPA and BRAA at fangraphs)

You really wantto look at the relationship between WPA/LI and BRAA, as WPA doesn't account for opportunity, which if you don't have opportunity is better than nothing, but with LI you have opportunity.

Byrnes is not the MVP of course, and it's ridiculous to suggest that he should be. However I feel that there are a few posters shortselling his defense, and his improvement this year, which is not all captured by OPS+

He is indeed a better player this year than he has ever been. Simply put, he is taking more walks, and he is stealing more bases and still not getting caught very much.

EQA reflects this pretty well, I think.

2003-.268
2004-.275
2005-.273 W/ Oak, .198 and 181 with Colorado and Baltimore
2006-.263
2007-.289

I don't know if this improvement is sustainable. Ask me in September of 2008. But THIS YEAR he has been better than he ever has, by a substantial margin. It is not correct to simply look at OPS+, and then poo poo away the value of his steals.

Also, with regards to his Defense:

THT Zone rating has him as the top rated Left Fielder, with Holliday a distant second

ESPN Zone rating has him tied with Holliday for the top Spot.

So two different Zone Rating systems based on two different Data Gathering outfits both have him atop the heap in Zone Rating.

His FRAR is 25 and FRAA is 13

His WIN Shares Fielding is 4th among all outfielders, and trails only 3 centerfielders.

I realize that all of these defensive metrics have their flaws, but all of them agree that Byrnes is the top rated left fielder this season. So I'd have to see more evidence that he is only "Neutral". I don't disbelieve it could be out there. It's not a challenge. I am just not privy to such information. Based on everything I am able to access, it doesn't seem right to discount his defense the way it's being done here.

I watch him every day. He takes some lousy routes, but he makes the plays. He's fast, has good range, and is sure handed. His arm is strong, though not always the most accurate. The summersaults annoy me to no end, but it is not a factor in evaluation.

Eric Byrnes is no MVP....but his play this year is clearly much better than it has ever been. It may be a one time thing, or he may have found a new level a little later in his career than most players do.

So two different Zone Rating systems based on two different Data Gathering outfits both have him atop the heap in Zone Rating.

Even if he is a + defender in LF, that is only a netrual in the grandscheme of things as the positional value of a LF is so low.

Fielder was interviewed the other day and you have to like the lad's gumption. He said, "it's on me to push us ahead. I gotta get good at bats and drive in those big runs."

Too bad the pitching staff has to be involved in this whole process. Otherwise, the team would have a good shot...........................

You really want to look at the relationship between WPA/LI and BRAA

Yup, makes sense. At least for position players. Not sure about relief pitchers, since the LI is essentially discretionary.

Hmm. This is kind of quick and dirty, but you can estimate predicted wins by RCAA/runs per win. (about 10.7 for hitter, 10.6 for pitchers for the DBacks this year)

So the adjustment is something like ((WPA/LI) - (BRAA/runs per win)) * 3.

Position players who get the biggest adjustment are Drew (+3 -- did surprisingly little damage considering his terrible overall stats) and Byrnes and Snyder (both +1). IOW it ain't the hitters who are pushing them past their pythags.

If you don't adjust for LI, Pena, Lyon and Valverde are worth about 4 extra wins (12 win shares) between them. Drops if you do adjust, but the fact they've done well in high leverage situations is of critical importance.

Davis and Hernandez also end up with a couple of extra win shares. No adjustment for Webb. IOW it's reasonably clear that the extra credit for exceeding their pythag is being assigned incorrectly.

Even if he is a + defender in LF, that is only a netrual in the grand scheme of things as the positional value of a LF is so low.

Ok....but then shouldn't Ramirez be much more than a "small" negative, since all the various metrics seem to pin him as one of the worst defensive shortstops in the game. I have only seen a couple of Marlins games outside the ones they played vs. the D Backs, and I think Hanley was out for some of them, so I have not seen him play enough to form an opinion based on observation. He's DEAD LAST in ZR both at THT and ESPN, -13 FRAA at BP, and ranks very low in Fielding Win Shares.

Since SS is the second most important defensive position, and all the metrics seem to agree that he is just about the worst SS in the NL, then you have to ding him harder than "small" negative to be consistent,...IMHO

Since SS is the second most important defensive position, and all the metrics seem to agree that he is just about the worst SS in the NL, then you have to ding him harder than "small" negative to be consistent,...IMHO

No I think he is being consistent

just being a SS is a defensive positive, being a bad defensive SS would be neutral, being a really bad one a small negative.

Pujols by all accounts a terrific defensive 1B is given a neutral
Fielder a poor defensive 1b is given negative (Rauseo's scale seems to run: negative, small negative, neutral, small pus, plus) Even if Hanley was the worst defensive SS in baseball he still has more defensive value than Prince

Shoewiz - JPWF has it right on. These are the positional adjustments I use:

C : +9
1b: -14
2b: +6
SS: +9
3b: +2
CF: +1
LF: -9
RF: -8
DH: -20

I figure Hanley is something like -10 to -15 defensivly, so when you add it to his position that is something like a 0 to -5. Prince is probably something like a -20 or -25, Albert is probably 0 to -5 which I think I thought have given him a small negative. Maybe you could give Byrnes a small positive for defense, I don't know. But the guy is just such a dunderhead I can't imagine, he adds that many runs with anything but his physical gifts.

JP, you could also see something like a plus plus. Someone like Aaron Rowand, or Mark Ellis or Adam Everett. Prince might be a double negative if I ever gave them!

Ok guys, thanks for the additional explanations. All taken under advisement. Thanks for being patient with me. ;)

I have another question for you guys. I have had this one posed to me before. Let me preface by saying that I am not trying to puff up Byrnes' numbers or make a case for him. I constantly do battle with D Backs fans that overrate him. But since OPS and OPS+ don't account for the steals at all, I am wondering if doing something like this makes sense:

Take his "net steals" or SB - (CS*2), which in this case would be 33, and add that to his total bases count.
He currently has 275 total bases in 567 AB for a .485 slg%. So if you add in the additional bases for the steals you get
308/567 = .543 Slg%, which combined with his OBP would give him a .908 OPS.

I realize there are better "total value" metrics around that allow you to take SB into account, but is this a totally half assed way of doing it, and if so why?

The first thing I thought of was that many of his steals come after a walk, HBP, or OBE.....so it's not as simple as turning 33 singles into doubles. Anyway...thanks for any help in advance.

Not sure about relief pitchers, since the LI is essentially discretionary.

On most teams, the top two or three relievers handle almost all of the high-LI situations. It is rare to see a reliever other than one of the top 2 or 3 pitching when the LI is 2.0 or greater. Most teams use the rest of their bullpen to pitch in low-leverage situations, or middling-leverage situations when the team is behind.

Case in point: 2006 Angels. Relievers faced 345 batters in situations where the LI was 2.0 or greater. 243 of those appearances were against either F-Rod (127, 43% of his BF) or Shields (116, 33% of his BF). The only other relievers who had more than 20 high-LI PA were Donnelly and Carrasco. Top closers will often have over 40% of their BF in situations where the LI is 2.0 or greater - and almost always will lead the team in that respect, as well as leading the team in average LI at the time they entered the game.

-- MWE

The problem with treating steals as *extra* TB is that steals have little or no advancement value. Part of the value of a double is that it not only advances the hitter but also any runners on base at least 2, sometimes three, bases. Adding them to the TB column overstates their value.

EDIT: "them" = SB.

-- MWE

Rollins is having a fine year, but don't forget that he has made even more outs than Juan Pierre -- a tall order indeed ...

Well, he has had over 40 more plate appearances but only four more outs than Pierre.

At any rate, it's not a distinction I'd want to have.

Best Regards

John

Albert is probably 0 to -5 which I think I thought have given him a small negative.

All the PBP metrics I've seen this year have Pujols an absurd amount ahead of every other first baseman in the game as a fielder. Some systems have him 30+ plays above average, with the next guy coming in around 12 or 13.

The idea of Pujols as neutral, if you buy into such metrics, is absurd, let alone at all negative.

The problem with treating steals as *extra* TB is that steals have little or no advancement value. Part of the value of a double is that it not only advances the hitter but also any runners on base at least 2, sometimes three, bases. Adding them to the TB column overstates their value.

-- MWE

DOH! I should have thought of that. Thanks Mike!

So how much value DO they have in these terms? 1/3 of a base? 1/2?

OK, Jimmy Rollins has made some outs. I still think the stuff above the line counts for something. I mean, I haven't made any outs this year. Does that make me the MVP?

Shoe, Steve Mann did something similar quite some time ago, but correctly adjusted the value of the stolen bases down. Don't have the details handy, but it was something like net stolen bases divided by two. (and that might still be over-stating the value)

Byrnes has created a little more than 6 runs with base stealing. (Ish ... It's worth noting that most hitters hit a fair bit worse in PAs when a SB is attempted. But that's the general case. If I was going to argue this point I'd want to know how it's affected his teammates, not the expected impact. In general the guys who do best in PAs when a SB is attempted are guys willing to take a walk)

Jimmy Rollins is probably going to end up coming up to bat more often this year than any other player in major league history. It's unavoidable that he's going to make a lot of outs.

OK, Jimmy Rollins has made some outs. I still think the stuff above the line counts for something. I mean, I haven't made any outs this year. Does that make me the MVP?

Have you played nice with the media?

Best Regards

John

The first thing I thought of was that many of his steals come after a walk, HBP, or OBE.....so it's not as simple as turning 33 singles into doubles. Anyway...thanks for any help in advance.

Think of it this way- in a linear weights type approach -a walk is typically worth approx 2/3 of a single
a steal is worth approx 1/3 of a single.
a CS takes away approx 2/3 of the value of a single.

For each CS take away 2/3 of a hit, Byrnes has 6 CS so take away 4 singles
Byrnes has 45 Sb, add 15 total bases.

His .296/.365/.485
becomes: .289/.359/.511

I'm still waiting for someone in the media to make a case for Gary Sheffield. In their first 100 games, the Tigers were 60-40, and hit at an impressive .290/.351/.472 clip as a team. After Sheff hurt his shoulder July 25th, they've gone just 20-27, and their offense has slumped to .279/.334/.431. If we're just talking about value, doesn't Sheffield deserve a nod when the best offense in baseball (up until the point of injury) depends so heavily on one person?

I don't actually think Sheff should be the MVP, but in any discussion on how much Team A depends on one Player X, he certainly deserves some mention.

Thanks Ron.

OK...so in my completely amateurish way.....here is what I come up with.

He has 45 steals, 6 CS, and I believe a couple of pickoffs. I'll take away 16, that gives him a net 29.
Cut that in half, and round down to 14 additional bases.

His slugging then goes from .485 to to .510, which would up his OPS from .850 to .875

That FEELS about right, in a rough, imprecise, shorthandish kind of way. Not of much value around here with so many professional scholars that do actual work in this area, but good enough for dealing with the average guy on a message board trying to give Byrnes 39 Extra bases to his slugging percentage. :-)

Shoe, i'll run some numbers and get back to you.

JPWF13,

I like the linear weights approach the best. Doing it my half assed way we came out pretty close, but the linear weights approach is the "cleanest"

By the way, is there any place to look up values based on Linear Weights IN SEASON? I know Matt posted his "Poor Mans Super Linear Weights" last November. Does anyone update a similar system and post them somewhere online?

And if not, WHY? It just seems like the time is long overdue for some Website to be posting that. OK...obviously if nobody is paying for it, that would be a good reason, but I would pay for that information, in the form of a subscriber site.

Shoe, where else would you look but baseball-reference.com?

By the way, is there any place to look up values based on Linear Weights IN SEASON? I

Tango might...

I generally go by: 2*(tb + bb + hbp)+ h + sb + sh + sf - cs - gdp -.5*ibb
(if you have reached on error data- BPRo does you can treat those as singles)

subtract .605*outs

and finally multiple by .16

you won't be off by more than 2-3% (average error per team) in any given year.

Just adding "net steals" also fails to recognize that the reward of a SB is one additional base for yourself (but no one else), while the loss is an OUT. A single+steal is worse than a double, while a single+CS should be treated as an out.

So if you're going to just start quick-and-dirty adding net steals in you need to treat the CS as outs, not just lost total bases.

Save your breath, Prince Fielder is going to win the MVP.

Shoe, where else would you look but baseball-reference.com?

Oh...I didn't realize they were added. You are talking about BtRns.....I hadn't seen that for some reason.

Cool...thanks.

What's the consensus (on BTF, on this thread, whatever) about using total bases as a rough estimator of offensive performance? How well do they correlate with team runs?

It's not a stat that gets much talked about, although I'd think it would be a first step up for those who think of batting average as the be-all and end-all of judging hitters.

Save your breath, Prince Fielder is going to win the MVP.

Most likely, yes. But "saving breath" is not something that is done much around here. Ever hear of the Hall of Merit? :-)

What's the consensus (on BTF, on this thread, whatever) about using total bases as a rough estimator of offensive performance? How well do they correlate with team runs?

I don't know about a "consensus" but I would think total bases would correlate well with team runs and would be a decent estimator of offensive performance. Total bases are equal to slugging percentage times at bats and we know slugging percentage correlates well with runs scored (I've seen that it actually correlates slightly better than OBP). Then you're just multiplying by a measure of opportunity (plate appearances would be better, of course).

If you go back to Bill James's simplest Runs Created formula:

RC = (OBP*SLG)*PA = (OBP*(TB/AB))*PA = TB*(OBP*(PA/AB))

So to go from total bases to runs created, all you really need to do is multiply by OBP. Seems decent enough.

When you start comparing hitters, of course, you need to somehow account for opportunities, but that's true with any basic stat, be it counting (which will give too much credit just for showing up) or rate (which won't give enough credit for showing up).

As I said, I don't know if there's really a consensus, but that would be my take on it.

Shoe - I ran a couple numbers. According to my version of baseruns in a modern offensive context: 1 TB is worth .6 runs (.56), 1 SB is .2 runs, and a CS -.3 (-.32) runs.

So for every you can use this formula to "fudge" your Total Bases number. (SB*.33)-(cs*.5)

1 SB is .2 runs, and a CS -.3

This would imply (roughly) a 60% break even rate but don't most people agree that it's more like 70%? Or is this taking into account that SB attempts correlates with speed and other base running? Or is this just rounding and I'm being to picky?

Its rounding since i'm trying to equate to the value of a tb. "my version of baseruns" not mine. It;s the standard one, with average context plugged in.

OK, so in the case of Mr. Byrnes....

(45*.2)- (6* -32) = 7.08 Runs gained by his baserunning.

Using the "fudge" numbers (45*33)-(6*.5) = 11.85 TB

So if you give him the 12 Total bases it bumps his slugging to .506, and gives him a .871 OPS
Am I doing this right?

However I think the example in #62 works pretty good too, because the impact of a CS on BA and OBP is reflected clearly. It all comes out about the same though.

By the way, are you going to post the Poor Mans Super Linear Weights again this year? Any chance we'll get to see them before the post season awards come out? ;)

By the way, are you going to post the Poor Mans Super Linear Weights again this year?

depends when the data comes out.

Ok...again...thanks everyone for taking the time with me on this. All very helpful and edifying.

KIKO, in it's simplest form,

RC = (OBP*SLG)*AB

And simplest form of

RC/G = (OBP*SLG/(1-BA))*25

KIKO, in it's simplest form,

RC = (OBP*SLG)*AB

And simplest form of

RC/G = (OBP*SLG/(1-BA))*25

Thanks. That makes the relation between RC and total bases (TB) even easier then: RC = OBP*TB