Tulo was worth 64 runs more than Braun according to FRAR? I don't think so.

That sounds about right between one of the best defensive players at his position in the league, versus one of the worst.

And when you’ve got two reputable methods that say Tulowitzki was the most valuable rookie, it’s hard to argue that he did not deserve to win.

FRAR isn't really a reputable method, is it?

Braun: His .895 fielding percentage is the second worst -- among third basemen with at least 100 games -- since 1910.

By VORP, the difference between the best offensive player in the league last year (A-Rod, of course) and the worst (Punto) was ~123 runs, so there's some perspective on the other side of ball.

Tulo is pretty close to the best defensive player in the league, if not the best. Braun has got to be the worst, right? 64 FRAR is probably overstating the difference, but I don't think it is by that much, especially when you consider that Braun is reputed to be historically bad.

FRAR isn't really a reputable method, is it?

Didn't we have this discussion a week or so ago? Possibly on a GG thread?

FRAR isn't really a reputable method, is it?

The play by play methods have about the same gap between the two.

Adding BRAR to FRAA gives a more likely difference of 5.4 to 2.5 for Tulo.

The massive difference in WARP is, of course, because of the massive (measured) difference in defensive value.

There's also a huge disparity in playing time. Tulowitzki played in 155 games, Braun 113. That's going to make an enormous difference with these replacement-level measures.

Every year the writers botch these awards. I can't even muster up an emotion anymore. Defense? Obviously, writers dig the long ball.

Okay, playing time might be part of the difference. Still, Braun only had 248 chances at third. How many runs does an error or missed play cost? if you say each is .75 runs, then to be 64 runs better than Braun given his chances, someone would have to make no errors (vs the 26 he made) and make 60 additional plays. That seems ridiculous to me.

Or, to put it another way, if Braun had not made any errors and had made 60 additional plays (given the .75 run/ missed play assumption - which could be a bad one), in that wildly optimistic scenario, his defense would only be equivalent with Tulowitzki's.

Still, Braun is not a good defender and Tulowitzki is, so maybe it's possible. It's just hard for me to fathom that big a difference really exists.

I think it's hilarious that "stat guys" used to be so offense focused and now the switch is to defense.

Looking at fielding win shares, what's amazing about Tulo's 10.9 FWS this year is that only 15 players have had double digit fielding win share seasons over the last decade prior to this season, and of them, seven of them were catchers (Pudge Rodriguez, Jason Kendall, Charles Johnson, Brian Schneider, Mike Matheny, Brad Ausmus, and Paul Lo Duca). The only non-catchers on the list were Orlando Cabrera in 2001 and Neifi Perez in 2000. (Incidental note: Kenji Johjima had a 10.1 FWS.)

According to the Fielding Bible, Tulowitzki was +35 runs (second best in baseball to Pujols), while Braun held the absolute cellar at -41 runs. That's a 76 run difference. Tulowitzki got jobbed.

I think it's hilarious that "stat guys" used to be so offense focused and now the switch is to defense.

I don't see anything wrong with it if you grant that we now have reasonably good measures of defense.

They don't play the same position, and they didn't have the same PT. The apples-to-apples comparison is the following:

1. Using BaseRuns (including things like Net Double Plays), an otherwise league-average team in Coors Field plus Troy Tulowitzki would have scored 835 runs.
2. Using the Fielding Bible's score of +35 plays for Tulowitzki, assuming they were all singles, BaseRuns says an otherwise league-average team in Coors Field plus Troy Tulowitzki would have allowed 797 runs.
3. Using PythagenPat, a team with 835 RS and 797 RA in 162 G would win 84.6 games, making Tulowitzki 3.6 wins above overall league average.
4. According to Nate Silver's Freely Available Talent research, replacement shortstops are 3.0 wins below overall league average per 162 games. Tulowitzki played in 155 games, so 3.0*155/162 = 2.9 wins. Add 3.6 wins above average to 2.9 wins for the gap between average and replacement, and we get Tulowitzki with a very impressive 6.5 wins above replacement. That's enough to get MVP votes, if rarely enough to win one.

Now, doing the same with Braun:

1. Using BaseRuns, an otherwise league-average team in Miller Park plus Ryan Braun would have scored 806 runs.
2. The Fielding Bible has Braun at -41 *bases*, which, assuming marginal plays for 3B are 2/3 singles and 1/3 doubles (which I'm making up--does anyone have the right value?), suggests Braun muffed about 21 singles and 10 doubles. An otherwise league-average team in Miller Park, giving up an extra 21 singles and 10 doubles, would have allowed 797 runs.
3. Using PythagenPat, a team with 806 RS and 797 RA in 162 G would win 81.9 games, making Braun 0.9 wins above overall league average.
4. According to Nate Silver's Freely Available Talent research, replacement third basemen are 1.4 wins below average per 162 games. Braun played in 113 games, so 1.4*113/162 = 1.0 wins. Add 0.9 wins above average to 1.0 wins for the gap between average and replacement, and we get Braun with a stunningly pedestrian 1.9 wins above replacement.

Assuming the defensive statistics are right--and given that the PBP metrics agree with each other and with non-quantitative assessments of fielding, that seems likely--this vote shouldn't have been close. This was a major, major whiff.

Rich Rifkin, those are bases, not runs, as noted in my previous post. Nonetheless, your conclusion is correct.

bibigon, to be clear, I'm not saying there's anything "wrong" with it, just that I think it's funny the way things have changed. Not haha funny, but shake your head and smile funny.

You guys are all way off base.

1) Value over replacement is irrelevant in this question. Assuming a minimum standard playing time should not be considered for the rookie of the year award. The word value doesn't come into play. Dan R's calculations demonstrate the fallacy of this kind of reasoning. According to those calculations Tulowitzki was a more valuable hitter than Braun. But that just proves that the measurement is wrong since Braun was unequivocably the better hitter.

2) The level of certainty with defensive statistics with respect to both accuracy and precision is nowhere near what it is for offensive statistics. While I am confident that Tulowitzki is good and Braun is not it is impossible to quantify the difference with precision. The larger the difference in your measurement the lower your confidence should be in the result.

3) So to sum I am 100% sure that Braun is a much better hitter than Tulowitzki. In order to make up the difference with the glove Tulowitzki would have to be Ozzie Smith and Braun would have to be the worst fielder in history. I am not confident that either, let alone both, are true. Therefore Braun is a very reasonable choice.

In order to make up the difference with the glove Tulowitzki would have to be Ozzie Smith and Braun would have to be the worst fielder in history. I am not confident that either, let alone both, are true.

Ha, but what people are saying is that Tulowitzki is literally the best fielder in the league, and that Braun is literally the worst fielder in the league. Literally, that's what the stats, and many people's observations, tell us. So it's a bit of an outlier situation.

Tulo was worth 64 runs more than Braun according to FRAR? I don't think so.

Still, Braun is not a good defender and Tulowitzki is, so maybe it's possible. It's just hard for me to fathom that big a difference really exists.

Hey Kyle

It's really interesting that you keep giving your opinion, or gut feeling, or whatever it is. It makes about as much sense as saying the world is flat because it looks "flat as hell" from your vantage point. People who have seen the earth from space or who have travelled around the world and ended up where they started, probably have a slightly better vantage point. Do you question what they tell you? But you think your gut feeling should carry more weight than the statistical analysis of people who make a living analyzing and documenting the differences in major league fielders defense. Do you not think that maybe they have a better vantage point than you. After all they see, in a statistical sense, the result of every ball put into play in the majors.

I can't possibly fathom what basis you could possibly have for saying all the defensive metrics are wrong, mis-leading, incorrect, or whatever it is you are claiming.




If you have a better mouse trap then "spring" it on us. Then otherwise let's go with what the all the various metrics are telling us.

In order to make up the difference with the glove Tulowitzki would have to be Ozzie Smith and Braun would have to be the worst fielder in history.

Why?

I happen to think there might be a park effect that makes Tulowitzki look a bit better with the glove than he is

The games I saw the Rockies, they were probably the closest NL team to the Brewers in reliance on defensive positioning (whether the Brewers should be using the fancy football-type schemes they've been employing the last few seasons is a topic Harvey's and I have questioned many times before, but that's not relevant here).

How many OOZ plays did he have? If he's shifting around a lot (like in the Brewers' scheme), he and the Rockies should have a lot of OOZ plays, because they're not playing the same zones everyone else is - they'd be playing more to hitter's tendencies than 'normal' teams would.

Tulo's Home OPS: .960
Tulo's Away OPS: .719

According to those calculations Tulowitzki was a more valuable hitter than Braun.

What calculations?

Tulo's Home OPS: .960
Tulo's Away OPS: .719

Relevance: 0
Irrelevance: 100

Hey Kyle

It's really interesting that you keep giving your opinion, or gut feeling, or whatever it is. It makes about as much sense as saying the world is flat because it looks "flat as hell" from your vantage point. People who have seen the earth from space or who have travelled around the world and ended up where they started, probably have a slightly better vantage point. Do you question what they tell you? But you think your gut feeling should carry more weight than the statistical analysis of people who make a living analyzing and documenting the differences in major league fielders defense. Do you not think that maybe they have a better vantage point than you. After all they see, in a statistical sense, the result of every ball put into play in the majors.

I can't possibly fathom what basis you could possibly have for saying all the defensive metrics are wrong, mis-leading, incorrect, or whatever it is you are claiming.

Hey kwarren,

It's very nice that you believe that statistics are not simply mathematical reconstructions of the world but are actually the world itself. Your analogies are as preposterous as your faith. Your accusations are without merit and your intelligence is profoundly naive. Your silly little numbers aren't the equivalent of the telescope and you haven't proven that the earth revolves around the sun. Your post indicate not only a general metaphysical ignorance but a specific ignorance of what a number is and what it represents. Which is ironic considering the faith you attempt to espouse.

To hopefully forever discredit FRAR (and FRAA), I give you:

Miguel Cabrera
2006 18 FRAR, 4 FRAA
2007 19 FRAR, 5 FRAA

Now there are some who suggest that Braun makes Cabrera look like Brooks Robinson and I suppose it's possible Braun brought the whole league average down so far to explain 2007 :-), but how in the world do you explain 2006?

The General Tone Of This Thread Needs Improvement

The point is that these measurements are at least an attempt at quantifying the difference between Braun/Toluwitzki. You make no such attempt. You simply throw something out there: "Braun was better!" And back it up with absurd statements for which you have no basis.

How do you know that Braun has to be "the worst in history" and Tolu has to be the best in history, to make up for the gap in offense? You don't know that. You just completely pulled it out of your ass.

We all know there is a gap there offensively, and the question at hand is how much the defensive gap is, but instead of contributing to the discussion of how big the gap is, you simply state that the gap isn't big enough because for some arbitrary reason you have decided that the gap is smaller than X, which is some unattainable number.

Maybe the statistics are flawed, but if they are then please point out the flaws instead of making some vague comment about their imprecision and then making some kind of bold statement for which you have no evidence. I will gladly take the statistics over your factoids that you pulled directly from your rectal cavity.

Just because we can't measure something is no reason to ignore it, Gaelan. This is where the concept of "the wisdom of crowds" might be relevant. And the crowd here is pretty objective and should not be ignored either.

Miguel Cabrera
2006 18 FRAR, 4 FRAA
2007 19 FRAR, 5 FRAA

Isn't this the Hanley Ramirez Effect in BPro's numbers? If memory serves, their method for backing into opportunities uses neighboring fielders, and by all accounts, Hanley Ramirez is a bad fielder.

Relevance: 0
Irrelevance: 100

I think it defenitely has some relevance. It's far from the most important point in the argument, and I know some of you guys could spit out some very interesting stats that would have much more relevance and make any argument much more convincing, but it's just a stat, and I'd like to see you tell why it has no relevance, simply because I'm intersted. It's not even a specific "point" I made, just something to ponder. Who knows if it makes a difference or not, it's a random thought. You might feel immensely proud with yourself because there's an idiot saying something completely "irrelevant" who doesn't know quite as much stuff as you, but there's no need to get rude about it.

What calculations?

I assume these:

1. Using BaseRuns (including things like Net Double Plays), an otherwise league-average team in Coors Field plus Troy Tulowitzki would have scored 835 runs...

1. Using BaseRuns, an otherwise league-average team in Miller Park plus Ryan Braun would have scored 806 runs.

I assume those. Of course, those are with the park effects, so the comparison isn't fair. I admit, I would like to know what the park adjusted values are. BBref has Braun as 27 runs better on offense. It would not surprise me much if Tulowitzki made up that gap on defense. What would surprise me is if he were 65 runs better.. that is an immense gap, the difference between Ryan Braun and Kory Casto as a hitter.

I won't respond to kwarren putting words in my mouth, other than to ask him that if someone told him that Willie Harris were better than Manny Ramirez because of a 70 run difference in defense, wouldn't he be somewhat skeptical? I agree with Gaelan in that the imprecision of defensive metrics make me reluctant to ascribe anyone more than 20 runs on either side of the ledger with any confidence; the metrics are too biased by luck for me to say otherwise.

Just because we can't measure something is no reason to ignore it, Gaelan.

I'm not ignoring it. I'm saying that the statistics that have been quoting in this thread are being treated as if they are precise measurements when they are not.

Specifically it is impossible (historically speaking) for the difference between Tulowitzki and Braun to be 60+ runs. This is not something I made up. I'm basing it on diamondmind's research and simulations. Based upon their model the difference between the worst fielder in the league and the best fielder in the league is between 30 and 40 runs. Why should I accept alternative measurements, which are prone to both distortion and measurement error, that require me to accept unprecedented results. That is against reason. It's not the wisdom of crowds it's the tyranny of the majority.

And let's put the offensive difference into perspective. According to baseballprospectus' equivalent runs Braun created 90 runs to Tulowitzki's 86 while using up 130 fewer outs. The positional difference will make up some of that but nonetheless that is a definite, quantifiable, and massive difference that is not subject to error.

I think it defenitely has some relevance. It's far from the most important point in the argument, and I know some of you guys could spit out some very interesting stats that would have much more relevance and make any argument much more convincing, but it's just a stat, and I'd like to see you tell why it has no relevance, simply because I'm intersted. It's not even a specific "point" I made, just something to ponder. Who knows if it makes a difference or not, it's a random thought. You might feel immensely proud with yourself because there's an idiot saying something completely "irrelevant" who doesn't know quite as much stuff as you, but there's no need to get rude about it.

Sorry, I was just being snarky because I thought you were. If you are just posting stats at random without trying to make a point then feel free I guess. Here are some more:

Tulo's OPS vs. Power pitchers: 718
Tulo's OPS vs. Finesse pitchers: 856

I don't see any reason why home/road is more relevant than any other split you can think of. Obviously we have to put his offensive numbers in the proper context, but I don't see why the actual breakdown of where he did his damage matters at all. We take the sum of everything he did, and make adjustments to put it in the right context. But if he had the same OPS on home and road would he be more valuable than he was this season? I don't see how.

Kyle S Posted: November 12, 2007 at 08:44 PM (#2612850)

the imprecision of defensive metrics make me reluctant to ascribe anyone more than 20 runs on either side of the ledger with any confidence; the metrics are too biased by luck for me to say otherwise.

Are you saying that defensive metrics are more biased by luck than BA, ERA, wins, or saves.

And no matter how flawed or luck biased defensive metrics are they are far better than someones gut feeling. And when various different defensive metrics all say the same thing (as in the Tulowizki vs Braun comparison) there is quite likely something to them.

Based upon their model the difference between the worst fielder in the league and the best fielder in the league is between 30 and 40 runs.

Across positions?

BP's fielding stats are garbage to begin with, and setting replacement level on defense so far below average makes their WARP garbage.

I had Tulowitzki at +2 BRAP and +30 BRAR with 16 FR using ZR. Braun was at +25 BRAP and +46 BRAR with -14 FR. This put Troy at just under 4 WARP or 1.5 WARP over Braun.

I would have voted for Troy, but if Braun had been up in the majors all year I'd probably have him ranked pretty close to Troy. I can't really get too upset at the writers for voting for a historically great rookie hitter.

1. Using BaseRuns, an otherwise league-average team in Miller Park plus Ryan Braun would have scored 806 runs.
2. The Fielding Bible has Braun at -41 *bases*, which, assuming marginal plays for 3B are 2/3 singles and 1/3 doubles (which I'm making up--does anyone have the right value?), suggests Braun muffed about 21 singles and 10 doubles. An otherwise league-average team in Miller Park, giving up an extra 21 singles and 10 doubles, would have allowed 797 runs.
3. Using PythagenPat, a team with 806 RS and 797 RA in 162 G would win 81.9 games, making Braun 0.9 wins above overall league average.
4. According to Nate Silver's Freely Available Talent research, replacement third basemen are 1.4 wins below average per 162 games. Braun played in 113 games, so 1.4*113/162 = 1.0 wins. Add 0.9 wins above average to 1.0 wins for the gap between average and replacement, and we get Braun with a stunningly pedestrian 1.9 wins above replacement.

Assuming the defensive statistics are right--and given that the PBP metrics agree with each other and with non-quantitative assessments of fielding, that seems likely--this vote shouldn't have been close. This was a major, major whiff.

"This was a major, major whiff."

I nearly fell out of my chair reading this. Look at all of the hocus-pocus you have to do with data and measures (about 9 lines of math) that only 0.01% of America is familiar with in order to "prove" Tulo is what you say he is with the glove and Braun is as bad as you say he is.

To call that a major whiff is way off. Snooze maybe, not whiff.

Either choice was a good choice.

Based upon their model the difference between the worst fielder in the league and the best fielder in the league is between 30 and 40 runs.



Across positions?

That was exactly my thought. This just can't be true.

Assuming a minimum standard playing time should not be considered for the rookie of the year award. The word value doesn't come into play.

Why? It's clear that the award is for whatever the voters think it should be, and that the voters disagree with you. Seth Smith put up a 271 OPS+ as a rookie but didn't get any votes.

You can argue that value shouldn't be a part of it -- the argument, I guess, would be that ROY is for the best prospect or something -- but you can't argue that value isn't a part of it.

I would have voted for Troy, but if Braun had been up in the majors all year I'd probably have him ranked pretty close to Troy.

If Braun had been in the majors all year, his hitting stats probably would have regressed quite a bit from where they ended up at.

I just took a glance at some of the Denver/Rockies message boards - and some of the messages attribute Braun's win to an 'East Coast, big market' bias (?!?!).

Umm... yeah.

If anymore proof was needed before this that Mike Hampton was FOS about signing with the Rockies due to the Colorado school system, I don't think anymore will be needed :P

Either choice was a good choice.

If you are saying that they are both good rookies, then ya, I guess so.

If you are saying that their rookie seasons were equal and voting for either one of them was reasonable, then you seem to be forgetting that the object of the vote is to select the best rookie, not simply to select one who was good. There is no way their seasons were equal.

If Braun had been in the majors all year, his hitting stats probably would have regressed quite a bit from where they ended up at.

Yeah, I thought about that, but I also think had he been up in the majors the whole year his defense would have improved to make up for some of that difference.

If you are saying that they are both good rookies, then ya, I guess so.

If you are saying that their rookie seasons were equal and voting for either one of them was reasonable, then you seem to be forgetting that the object of the vote is to select the best rookie, not simply to select one who was good. There is no way their seasons were equal.

Wow, you need to get some sleep. Or dedicate your abilities in search of alternative energy.

If Braun had been in the majors all year, his hitting stats probably would have regressed quite a bit from where they ended up at.

Not a lock Braun would regress with more playing time. Braun slid a bit in August and then dominated again in September (.308/.354/.644). It wasn't like he was not playing during the months of April/May, his September was after 7-8 long months of baseball like everyone else. His triple-A stats were off the charts, you can argue his June was a regression after that absurd March/April/May.

Also, as this subject relates to ROY, his counting stats gained, HR, RBI, R, etc...would have vaulted him further up the list of greatest rookie seasons ever.

He also would've gained a lot of errors to help cement his status as worst rookie defensive season ever.

Across positions?

No, within one position, which is to say that the worst fielder will be between -15 and -20 while the best fielder will bet + 15 or +20 as compared to average at that position. But that doesn't make any difference to my point since all of the extannt defensive comparisons between Braun and Tulowitzki use the same method (i.e. they compare them to average at their positions and then use those numbers for a player to player comparison).

Are you saying that defensive metrics are more biased by luck than BA, ERA, wins, or saves?

Speaking only for myself I'm saying that the measurements are further removed from reality. While BA is influenced by luck at least we can see that it is a first order measurement of what actually happened. Braun actually had that number of hits in that number of at-bats. Fielding measurements, on the other hand, are not first order measurements. They are third order calculations (UZR for example) made upon the basis of second order estimations (coding of the observer) of first order events (what actually happened). This doesn't make them useless but it does mean that the numbers, in themselves, don't prove anything and certainly should not be mistaken for a precise picture of reality.

Braun made 26 errors. 21 of them resulted (Braun committing 22 errors in those 21 ROE) in a ROE which for us is easy to find via PI.
8 of the 21 were with the bases empty. 3 of the 8 got to 2nd with one of them Ryan committing 2 errors to let the guy get to second.
4 of the 21 were with a man on first. 1 got a guy to third but also resulted in the trailing runner being out. 1 started this whole chain in which multiple errors resulted and both runners score. So it was two guys to third, two guys score, one guy stays of first, another gets to second, and the other gets thrown out.
5 times there was a man just on second. In only one of those 5 errors did the lead base runner move up a base.
2 times there was a man on first and second. Neither time did the lead runner score on the error.
2 times it was bases loaded. In both cases two runs score
In total 6 runs scored directly because of Braun errors though one of the runs was caused by someone elses error after his error.

4 of the 5 runner on second error had no more value then a walk in that situation, with the 5th probably being slightly less helpful then a single. The 2 times with runners on 1st and 2nd had the same value as a walk. Of the 8 errors with bases empty 5 had the same value as a walk/single and 3 had the same value as a double. Of the 4 at first one is the equivalent of a homer, one a walk, one a two base sac hit, and the other a single. 2 the two bases loaded were equivalent of 2 doubles.

So instead of 21 outs his 22 errors gave the opposing team 12 walks, 5 doubles, 2 singles, and a home run (plus the SH). Using a generic lwts that comes to Braun's 22 errors costing the Brewers 16 runs or roughly the .75 runs that is used as shorthand.

But were they high-leverage errors?

Ha!

I nearly fell out of my chair reading this. Look at all of the hocus-pocus you have to do with data and measures (about 9 lines of math) that only 0.01% of America is familiar with in order to "prove" Tulo is what you say he is with the glove and Braun is as bad as you say he is.

Yes, anything that requires more than one mathematical step and is unknown to the general public must be completely useless.

Tulo's OPS vs. Power pitchers: 718
Tulo's OPS vs. Finesse pitchers: 856

I've noticed that this is consistently true for almost all hitters, that they hit better against finesse no matter the kind of hitter - my guess is that most pitchers who can throw the ball that hard have so much less to rely on. While if you're a finesse pitcher, you need so many things to fall into place - location, movement, speed changes, that it's a lot harder to get all those things working together. While if you're a power pitcher, though you can't always rely on just reaching back, chances are there's less things that need to fall into place, and you can get by with the occasional location miscue or instances when the hitter knows what's coming, because your fastball particularly is already harder to hit.

But if he had the same OPS on home and road would he be more valuable than he was this season?

I'm not sure it would show he's more valuable, but that he's a better overall hitter. He's equally valuable to the Rockies, but perhaps not to another team. That may not make a lot of sense... (This is also assuming that the two OPSs would even out to whatever the average is, so 800 at both instead of 700 or 900.) Let's put it this way: If I had a choice between two players from the Rockies that I was going to sign to a team with a league or below average hitting park, I would rather have the hitter who's consistently shown his numbers haven't been completely affected by Coors. Even if his production is suspiciously lower than what you would expect from Coors, if he's consistently shown he's going to perform at the same level out of it, I would rather have him than someone who tends to rely on hitters parks for the bulk of his production.

3 consistents in that paragraph. Hope I'm not turning into Joe Morgan.

Forget what I said about power/finesse, I've discovered something:

Adam Dunn vs Power: 850
Dunn vs Finesse: 972

Thome vs Power: 942
Thome vs Finesse: 1003

While...
Juan Pierre vs Power: 723
Juan Pierre vs Finesse: 697

Jason Tyner vs Power: 657
Jason Tyner vs Finesse: 616

Forget what it has to do with RoY, but...

There's the Coors field hangover effect to consider, too. I agree you'd want to take park factors into account if you were trading for Tulo, but for a ROY ballot, this year's realized value seems like the appropriate thing to use.

Gaelan:

1. Why on earth should playing time not be considered for ROY? Are you actually suggesting that the fact that there were 43 games when Colorado benefited from Tulowitzki's services and Milwaukee was throwing God-knows-who out there is not germane to this discussion? Now, obviously, it's not Braun's *fault* that he wasn't called up earlier, but this is a major league ROY award, not an MLE award.
2. Depends on what you mean by a more valuable hitter. Braun was obviously a *better* hitter. But in addition to Tulowitzi's PT advantage, there's also the fact that regardless of how *well* they field their positions, Tulowitzki is a shortstop and Braun is a third baseman. It's a lot easier to find a decent-hitting 3B than a decent-hitting SS. Compared to overall league average on a rate basis, of course Braun blows Tulow away, but position-adjusted and accounting for playing time, not necessarily.
3. The gap between the best and worst fielder in the league *at the same position* is indeed usually about 40 runs. The numbers suggest Tulowitzki was meaningfully better than your average Gold Glover--indeed, Ozzie Smith-caliber, as you say--while Braun was meaningfully worse than your average Stone Glover--in fact, a butcher of historic proportions--so it's more like 50 runs in this case. The remainder comes from the fact that playing SS is intrinsically more valuable than playing 3B, and that Tulowitzki played more than Braun did.

By the way, it's interesting that the mere presence of Braun means that every other 3B in the NL is going to look a little better relative to average.


Walt Davis:

No one uses a non-PBP metric like BP FRAA for seasons that have PBP data available. We only rely on them when no PBP numbers exist.


Kyle S:

Park-adjusted, Braun was 3.5 offensive wins above average (so 3.5*162/112 = 5.1 wins per full season played), while Tulowitzki was 1.0 offensive wins above average (so 1.0*162/155 = 1.0 wins per full season played). Braun was thus 2.6 defensive wins below positional average, while Tulowitzki was 2.6 defensive wins above positional average. Limiting them to 2 wins instead of 2.6 each, as you suggest, wouldn't affect the substance of the conclusion as to their relative value in the slightest. The remaining difference comes from the replacement level calculation--the intrinsic value of playing SS vs. playing 3B is a marginal 1.6 wins per season, and since Tulowitzki played more than Braun, that gap increases to 1.9 wins.

Kyle C--Tulowitzki only two batting runs above position? He was an above league average hitter, and SS are most definitely below-average hitters. How did you manage to get him with only two batting runs above position?

abc006--there is So much noise in home/road splits that it seems nuts to me to extrapolate from One year's worth of data that Tulowitzki can't hit outside of Coors. If he shows the same split after 5 years I'd agree with you.

Ooops, thanks Dan, I forgot I had some problems with my NL spreadsheet still. I had already fixed the AL one but had neglected the NL.

That puts Tulowitzki at +9 and +36 on offense.

If this were the MVP award, and Tulo and Braun were the two "finalists" , is there any doubt that Tulo would get the nod from the writers?

Why would the writers treat an MVP decision differently from a ROY decision?

Making the playoffs is definitely weighted more heavily.

If this were the MVP award, and Tulo and Braun were the two "finalists" , is there any doubt that Tulo would get the nod from the writers?

Well yes, but it wouldn't be because of the things we were discussing. Tulo would win the MVP because:

1) He was better "down the stretch"
2) The Rockies had that insane run and made the playoffs while the Brewers "choked"
3) He's a "leader in the clubhouse."

It would have nothing to do with defensive positioning or play...

1) He was better "down the stretch"
2) The Rockies had that insane run and made the playoffs while the Brewers "choked"
3) He's a "leader in the clubhouse."

FYI....Sept
Tulo was .273/.341/.518 in Sept
Braun was .308/.354/.644 in Sept

Sorry, what I meant was that he was better down the stretch than he was before it. His first half OPS was .790, his second half .891.

Being better in the second half than in the first is something that is very important to voters. Just ask Justin Morneau.

But were they high-leverage errors?

Ha!

Some were.
August 5th, top of the 9th up by 2 with two outs and the bases loaded. Ryan muffs a grounder that lets two runs score. Phillies go on to win it in the 11th.
August 14th the Brewers up by one in the 5th inning. 1 out, bases loaded. Ryan fails to get the forceout. Two runs score and the Cardinals go on to score 6 runs in that inning.


Incredibly 23 real runs were scored as a result of Brauns 22 errors.

McCoy - thanks for that. As I pulled the .75 number completely out of, uh, thin air, it's hard to believe it was so close to being accurate.

Dan R, I've never said (on this thread or elsewhere) that I think Braun deserved the award more than Tulowitzki did. I just don't believe the difference between them was 6.5 wins on defense, that's all. It looks like your data show a 5.2 win difference instead (which to me still seems high, but more plausible). I think your method of evaluating their respective contributions is pretty solid.

I've gone 180 degrees from where I was even the 3rd week of September (when I was the one berating the Tulo herd for not recognizing Braun's obvious superiority) -- and while I don't think it was quite the "robbery" I said it was in the original RoY thread, I still find myself in agreement with Neyer.

Having had Braun on roto roster for 2 years (and spending most of my April and a good part of May on MIL message board berating the Crew for not bringing him up) -- I'm well aware of how wonderful his season was at the plate.

It wasn't until I was called to task in a September thread for this opinion -- and really took at just how great Tulowitzki was with the glove (and how atrocious Braun was... I mean, geez... when we can use pure fielding percentage as shorthand?) -- that I flipped.... not to mention, Tulowitzki was hardly inept at the plate himself (though... Braun was 15 of 20 on the bases, while Tulo was only 7 of 13!).

I'm thrilled that my roto team will have Braun for the next 3 seasons at < $20 per annum (I just hope he can keep his 3B eligibility) - but if I had a vote, I'd have gone Tulo-then-Braun.

Why would the writers treat an MVP decision differently from a ROY decision?

The voters would be very unlikely to give the MVP award to a player that only played in 113 games and didn't even qualify for the batting title.

Why would the writers treat an MVP decision differently from a ROY decision?

Because for better or worse while the MVP is almost exclusively a backwards looking award and limited to a single season (with rare quasi-lifetime achievement votes and anti-votes for repeat winners), at least since Willie McCovey, the Jackie Robinson award has always had an element of 'who is more likely to play well in the future' included in the selection. It isn't the sole factor and often the writers are just plain wrong about who will do better in the future. But it is an element that is weighed. Debating this award without taking that into consideration is a fools errand.

B-ref has batting runs +33 for Braun and +6 for Tulo. Give Tulo a +5 position bonus for playing shortstop, which is extremely conservative. At that point Tulo only has to be 23 runs better than Braun defensively, relative to position, to put him ahead. I'm not even giving Tulo credit for playing the whole season, which puts him further above replacement level. Comparing to average gives a bias to Braun.

Every defensive system has the gap between the two larger than 23 runs, most double that. This is not a case where PMR or plus minus says Ichiro is great and UZR says he's terrible. All the fielding systems agree here. They normally do in the infield, we've gotten to the point where infield defense is not that hard to quantify. In this case trying to minimize the defensive gap between these players goes way beyond skepticism, you are just putting your head in the sand.

Saying the best and worst fielders are 15-20 runs above or below average is a good rule of thumb in a normal year, and fine by me if you want to apply it to Tulo, but it doesn't work for Braun, he's historically bad. (Though a +15 for Tulo and -20 for Braun would be more than enough to make Tulo the most valuable rookie) Back to post #4:

Braun: His .895 fielding percentage is the second worst -- among third basemen with at least 100 games -- since 1910.

1910 Dudes. Did they even have webbing in the gloves in 1910? Is Braun using a glove without webbing?

1910 Dudes. Did they even have webbing in the gloves in 1910? Is Braun using a glove without webbing?

:-)

Made my day

Jerry Denny, last barehanded player retired 1894 with a .890% at the hot corner.

If Braun had been in the majors all year, his hitting stats probably would have regressed quite a bit from where they ended up at.

He doesn't have an established level, so you can't say this.

Kyle S, it's 5.2 wins of difference for the *quality* of their fielding--but then you have to factor in that a) playing SS is intrinsically more valuable than playing 3B and b) Tulowitzki played more than Braun. I would say it's about a 7-win difference on defense after accounting for both of those elements.

You can't position-adjust twice - you already did with your "replacement level 3B/SS" calculation.

Do you think that if Tulowitzki had moved to third and played a full season there, the Brewers would have given up 65 fewer runs? That seems quite unlikely to me.

I'm not position-adjusting twice. The replacement level 3B/SS calculation *is* the position adjustment. Tulowitzki about 7 wins better on D (summing fielding quality and position adjustment), Braun 2.5 wins better hitting, Tulowitzki 4.5 wins better overall.

That's assuming that the gap in replacement level between positions (which is mostly hitting) is always equal to the gap in fielding that would be produced by a player actually switching positions (which is, of course, all fielding). That is not at all necessarily the case. Would Tulowitzki be a +40 3B? No, of course not. First, he'd have to learn the position. Second, his skill set might not be best suited for it. Third, he'd get many fewer chances there than at short, reducing his opportunity to exceed average.

I do, however, think that if you put Braun instead of Tulowitzki at short, the Rockies could have given up 100 more runs than they did. :)

I do often think that there might be some opportunity to "arbitrage" the difference between the replacement level gap and the fielding adjustment by moving guys around. For example, I've often heard that Ryan Zimmerman could probably play a decent shortstop. If he would lose no more than 16 runs defensively by moving from 3B to SS (it doesn't matter whether it's +20 to +4, +8 to -8, or -4 to -20), then he should definitely switch positions. The same is true of A-Rod--I suspect he'd be about -10 at SS these days, and a -10 SS is worth more than an average-fielding 3B. Conversely, someone like Hanley Ramírez should be moved off short if he could pick up more than 15 runs in the field by moving to CF. Otherwise, leave him there and suck up the defense.

In short, I do not think all players in MLB play their optimal positions.

I just find it fascinating that some of the same folks who chastised me for stating that Adam Dunn's defense seriously detracts from his value are all over Braun's defense like flies on manure.

BPro and Silver are the worst offenders. The slap Dunn on their cover and later Normandin writes this mash note of an "assessment" and when I write to disagree get told I am "overstating defense's value in player evaluation".

Well, which is it? Defense matters or not?

Snot-nosed punks switching perspectives at their leisure and getting snippy when someone notices.

Troy is a fine player. I enjoy watching him play.

Finally...Harvey is here. I have been waiting for this since the first Braun wins ROY thread popped up and now I can sit back and enjoy. I am in no way being sarcastic. Harvey's arrival to this thread has made my day.

My vote for Rookie of the year would be Troy. But if I was GM of the Angels, and told I could trade for either one with the same package of prospects, I'd pick Braun. Either he improves his D at third, finds a new position (corner OF?) or becomes a league average defender at DH.

My vote for Rookie of the year would be Troy. But if I was GM of the Angels, and told I could trade for either one with the same package of prospects, I'd pick Braun. Either he improves his D at third, finds a new position (corner OF?) or becomes a league average defender at DH.

I don't get this. Why?

The other flawed presumption is that Capuano crashed primarily due to Braun. But for anyone willing to tap a few more keys you will find that CC stunk in the second half of 2006 sans Braun.

So, did the mere THOUGHT of Braun send Capuano into a tailspin? Or maybe, just maybe, there are some other LONG-TERM issues at work.

The half-hearted, half-baked, half-witted nonsense being passed as informed input with respect to Braun lurches from amusing to disturbing.

And how about sample size o quantitative sensei? What of that? Would anyone here be declaring Braun a "historically bad hitter" if he was allowed to stay in the lineup for 100 games and hit .198?
No. There would be a myriad of comparisons to Matt Williams or Mike Schmidt.

Erratic application of analytical methods is the hobgoblin of BBTF.

Using these positional adjustments, 0 for 3b, -5 for LF, -15 for DH:

Lets say Braun is a +35 hitter: (offense, pos, def)

Value at 3B: +35, 0, -30 = +5 total. If he can improve to just "below average" he's +20 or +25.

He's got speed, so what if he can play average in left?
+35,-5,0 = +30

Worst case and his skin is alergic to leather, he's a DH:
+35,-15,0 = +20

Tulo is +5, +5, +20 = +30

But is he a true +20 defender? Maybe he's more like +10 and everything just broke right for him last year. True +20 defenders are extremely rare, I need to see him do it again.

But is he a true +20 defender?

I don't know much about Tulo the minor leaguer, but the guy is a first round draft pick who just put up a 108 OPS+ at 22. Isn't he likely to become a much better hitter?

Would anyone here be declaring Braun a "historically bad hitter" if he was allowed to stay in the lineup for 100 games and hit .198?
No. There would be a myriad of comparisons to Matt Williams or Mike Schmidt.

Hitting .198, especially with the power of Schmidt, would not be historically bad. If somebody hit like Ray Oyler over a full season I'd be confident calling them historically bad.

I have an open mind as to whether Braun can improve enough to stay at third, but he was really bad. Worst fielding percentage since 1910 is pretty serious.

AROM:

Agreed that Ryan was challenged in the field.

I have a modicum of hope due to:

--Braun knowing he can "play" and being able to relax

--braun allegedly has a very good work ethic

--I cannot fathom him playing WORSE

--the bulk of his errors came in two ugly stretches of about 2.5 weeks each

--he has the tools

So I have some hope. The candle is lit.

But it's flickering.....

Isn't he likely to become a much better hitter?

Sure, but Braun is about the same age. I'd expect them both to improve as hitters.

But it's flickering.....

Best of luck on that. Brewers have quite an exciting young infield. It would be a shame if their defense prevents them from playing in a few postseasons.

Sure, but Braun is about the same age. I'd expect them both to improve as hitters.

How much better can Braun get though? Tulowitzki came with similar prospect hype and he had a superior season to Braun by virtue of out of this world fielding whereas Braun's value was in out of this world hitting. If you're going to question whether Tulo is this good defensively, don't you have to raise the same questions about Braun when it comes to hitting? If so, doesn't that still leave Tulowitzki as having done better? If you said Braun had the better year, I would understand choosing him over Tulo, but I don't get how Tulo can have the better year and come with similar clippings and not get chosen over Braun.