Last edited by jojo; 03-11-2013 at 04:08 PM.
"This isn’t stats vs scouts - this is stats and scouts working together, building an organization that blends the best of both worlds. This is the blueprint for how a baseball organization should be run. And, whether the baseball men of the 20th century like it or not, this is where baseball is going."---Dave Cameron, U.S.S. Mariner
Luck. Balls hit in the right spots. Runners taking a few extra chances on his arm that they otherwise wouldn't have the next/previous year. Just happening to play in the right ballpark on certain balls being hit. Changing ballparks entirely because of a trade or free agency. Your team acquiring a new pitcher who is very different than your last pitcher he replaced, giving you more/less chances to field more balls, thus giving you more/less chances to have your range shown to be poor/outstanding. Maybe you wind up playing in more day games and you happen to just see the ball a little better that year. Or worse.
Yep. I would have to think that eventually they will.
And my hypothesis is that one of the first things we'll learn is that, despite the wisdom suggesting otherwise, defense actually does slump. Players have streaks of playing especially well or especially poorly centered around some mean "talent". And just a like a guy can put up an .820 OPS one season and a .700 OPS the next, actual defensive performance will legitimately vary a fair amount as well.
I expect that over the new few years we're going to see some really interesting things happening with player usage and strategies -- more shifts, more defensive platoons, etc.
Games are won on run differential -- scoring more than your opponent. Runs are runs, scored or prevented they all count the same. Worry about scoring more and allowing fewer, not which positions contribute to which side of the equation or how "consistent" you are at your current level of performance.
It can be. But not entirely. For example, adding a pitcher or two who are fly ball pitchers who are taking over for fly ball pitchers can give an outfielder a few extra "out of zone" play chances they may not have otherwise gotten. The skill was there, but they weren't always able to show it off because of limited chances. RBI doesn't really change the situation. Do damage at the plate. While not all scenarios are the same, they are close enough. Take 4 balls and go to first base, or make due with one of the potential three strikes you will see.
But why should the guy who drives in the runner be given an entire run's worth of credit? He's already getting credit for average value of that hit, which includes average base advancement value.
The reason people scoff is that you're suggesting that if we have two players who did the exact same thing while they were at the plate, the one of them should be treated as having produced more because of the performance of his teammate. If you want to do that, you have to make other adjustments or you've be double counting a lot.
I'm not 100% against the idea. In fact, I think it's a very interesting way to look at things, especially if you're talking MVP, for example. I think what you're looking for is called RE24. Basically, it adds up the changes in run expectancy in all of the player's at bats, including the runs that actually get created. Tango explains it here: http://www.insidethebook.com/ee/inde...comments/re24/. Basically, it's runs created above average.
So, for example, Votto comes the plate with 2 outs and the bases loaded. I'm using the 1969-1992 run values from here http://www.tangotiger.net/re24.htm
If he hits a homer:
Start of PA (Bases Loaded, 2 Outs): 0.752
Runs Scored: 4
End of PA (Bases Empty, 2 Outs): .094
Total Change: (4 + .094) - (.752) = 3.342
If hits a single (and 2 runs score)
Start of PA (Bases Loaded, 2 Outs): 0.752
Runs Scored: 2
End of PA (First and Third, 2 Outs): .484
Total Change (2 + .484) - (.752) = 1.732
If he makes out #3
Start of PA (Bases Loaded, 2 Outs): 0.752
Runs Scored: 0
End of PA (First and Third, 2 Outs): 0
Total Change (0 + 0) - (.752) = -0.752 (that's negative value - he essentially cost his team runs because he negated value produced by his teammates, turning possible runs in to definite non-rnus)
So you're giving him credit for actually advancing (or not advancing those runners). It automatically scales for leverage from a run scoring standpoint. Here's the top 10 in wOBA and RE24 per 600 PA from the last 3 years (min 1000 PA). If a guy's RE24 is higher than his wOBA, than it is suggesting that, because of the situations in which he performed how he performed, his actual value was greater than his average expected value. (Go Votto!)
Jay Bruce is ranked 48th and 103rd respectively - he has NOT converted his opportunities. He was one of the worst in MLB over that time period in terms of getting the least out of his production. His .355 wOBA produced the same RE24 as guys with a wOBA around .335, roughly the same as Brandon Phillips, Scott Rolen,Code:Name Team PA wOBA Rank RE24/600 Rank Diff Miguel Cabrera Tigers 2033 .428 1 52.0 2 -1 Joey Votto Reds 1842 .425 2 58.8 1 1 Jose Bautista BluJays 1737 .419 3 46.6 3 0 Ryan Braun Brewers 1991 .405 4 44.0 4 0 David Ortiz Red Sox 1594 .402 5 29.4 18 -13 Josh Hamilton Rangers 1745 .401 6 43.8 5 1 Prince Fielder - - - 2096 .396 7 40.8 6 1 Troy Tulowitzki Rockies 1338 .392 8 24.4 35 -27 Paul Konerko W Sox 1868 .391 9 28.0 24 -15 Carlos Gonzalez Rockies 1757 .390 10 31.7 13 -3 Matt Holliday Cards 1879 .390 11 32.8 11 -1 Albert Pujols - - - 2021 .388 12 33.6 10 2 Buster Posey Giants 1238 .383 15 38.5 7 8 Adrian Gonzalez - - - 2092 .378 17 37.9 8 9 Evan Longoria Rays 1547 .373 22 34.7 9 13
Last edited by RedsManRick; 03-11-2013 at 05:09 PM.
Games are won on run differential -- scoring more than your opponent. Runs are runs, scored or prevented they all count the same. Worry about scoring more and allowing fewer, not which positions contribute to which side of the equation or how "consistent" you are at your current level of performance.
Thanks for the cool stat Rick. You are right - I was discounting the "double counting" aspect of things. I think you do see the gist of my thought though, that someway, somehow, RBIs should be considered in terms of what happened.
I have even crazier theories as well. One says that in terms of MVP (and we usually agree that it is valuable in terms of a player to a team) that positive counting stats in a teams losses should not be included as it did nothing to gain a win. (You could say it burned up more pitches by the other team I guess)
I don't know, I just think there are many interesting ways to look back at the past and we've just scraped the surface, as we mostly focus on projecting forwards.
Some Bill James quotes on defense that are relevant to a couple of points/assumptions brought up earlier in this thread, 1) defensive metrics aren't reliable, 2) defensive production isn't variable:
http://www.freakonomics.com/2008/04/...all-questions/Originally Posted by Bill James;
Take Home: Defensive metrics aren't perfect but they have evolved to a place where they are legitimately useful and it's a mistake to assume a player's defensive value wouldn't fluctuate, sometimes a lot, from season to season.Originally Posted by Bill James taken from the book "The Mind of Bill James: How a Complete Outsider Changed Baseball";
Last edited by jojo; 03-11-2013 at 05:27 PM.
"This isn’t stats vs scouts - this is stats and scouts working together, building an organization that blends the best of both worlds. This is the blueprint for how a baseball organization should be run. And, whether the baseball men of the 20th century like it or not, this is where baseball is going."---Dave Cameron, U.S.S. Mariner
What do you mean, elevate?
The same thing that would elevate a true talent .280 average hitter to a .320 average season. He had a had good year, probably as a combination of simply performing above (or at the peak of) his sustainable ability for a limited period of time and good "luck".
Does zero to +15 seem like a pretty massive jump? Absolutely. But how common is that? And how does the frequency of that occurrence compare to a similarly impressive jump (or dip) in other statistics? A jump of .800 OPS to 1.200 OPS over a season is ridiculous. But that same jump from month to month? That's commonplace. And a jump of say, .950 OPS to .700? That's Ben Zobrist a few years back. It happens from time to time.
I think one reason why UZR fluctuations might seem more frequent is because everybody is put on their positional scale, placing them among a smaller group peers and giving more players a chance to look like an outlier. With the positional adjustment made on the back-end, you don't have one giant pool of defensive performance. You have smaller positional peer group bands in which you might observe highly variable performances that taken in the broader context of all defenders might not jump out. For exaple, A +10 3B season wouldn't stand out if we compared his overall defensive performance/value to 2B, SS and CF too. It would instead look like a .270 hitter hitting 300.
We also have the against baseline & plus/minus aspect due to measuring it as variance that throws our perceptions off a little. Firstly, we have a counting stat where average is 0. Imagine if the average number of RBI were 0 and we reported players RBI as +17 or -10. By contrast, imagine if instead defense were treated like other counting stats, as a sum total of runs prevented where a botched play was 0 instead of negative. A guy might end the season with 88 runs prevented, compared to a league average of 80 and typical observed ranges of 70 to 90 for guys with that many chances.
But because we're looking at these changes around average, we have these small numbers. That's a problem because we have an instinct to compare things using ratios. When we see 8 of one thing and 2 of another, our brain does not say "8 is 6 more than 2", it says "8 is 4 times as much as 2". So the difference in UZR of +8 vs +2 seems like a big deal. But if it were 88 vs 82, we wouldn't have the same gut reaction to the size of the difference.
Secondly, we have the whole plus/minus thing. If feels weird to so obviously see a guy go from "good" to "bad", even though the difference between +2 and -2 isn't that big. If a guy goes from hitting .275 to .250, we don't think of it as going from above to below. And if it were 82 runs prevented vs. 78 runs prevented, we'd think differently.
Of course, one we reason we don't look at defensive performance like this is because it's just not helpful. If a guy scores 80 runs, it doesn't tell us anything we really want to know. If we want to know how good of a run scorer he is, we need to know more context.
Anyway, I find all this stuff fascinating. I'm convinced that part of the reason people resist UZR is simply because they don't have a mental model for processing baseball performance data that is presented like this. If RBI were presented as RBI above or below the average #3 hitter, people would be much more comfortable with it.
Last edited by RedsManRick; 03-11-2013 at 07:05 PM.
Games are won on run differential -- scoring more than your opponent. Runs are runs, scored or prevented they all count the same. Worry about scoring more and allowing fewer, not which positions contribute to which side of the equation or how "consistent" you are at your current level of performance.
dougdirt (03-11-2013)
Not necessarily. If I want to look at what happened for the purpose of projecting what's likely to happen in the future, I don't want to count the stuff that really happened but which was not really a result of a repeatable skill. Think of it like ERA vs. FIP with situations/situational hitting being sort of like defense for pitchers.
Was Votto's plate appearance more more valuable than a normal HR because it drove in 3 extra runs? Without a doubt. But should Votto be given credit for that extra value? That's where we might disagree. From my standpoint, Votto should be given credit for hitting a homer -- the thing he did. And his teammates should get credit for being on base -- the thing they did. But what makes those things worth 4 runs was that, in this case, they happened in a certain order at a certain time.
Now if you take the average value of their 4 plate appearances, it's going to be less than 4 runs. The extra value created by the sequence of the events (the difference between the 4 "real" runs and the 4 "average runs") just vanishes.
All told, it comes out in the wash either way. The difference is that in the real runs model, players get credit for something which does affect game outcomes but which is out of the control of any one person (event sequences). Generally speaking, I prefer to look at just the stuff within the player's control for the simple reason that it will be more reflective of his actual skill and therefore more predictive of future performance.
Of course, it's sort of a false choice. We have and can use both depending on what question we're answering.
I think this is just silly. To follow your logic:I have even crazier theories as well. One says that in terms of MVP (and we usually agree that it is valuable in terms of a player to a team) that positive counting stats in a teams losses should not be included as it did nothing to gain a win. (You could say it burned up more pitches by the other team I guess)
I don't know, I just think there are many interesting ways to look back at the past and we've just scraped the surface, as we mostly focus on projecting forwards.
- Any run scored above 1 more than the opponent shouldn't count either. Those runs weren't needed for a win.
- Negative counting stats in a team win should be ignored. Get torched for 6 runs in 3 innings? If your team scored 10, it doesn't matter.
- A guy who hits a sac fly in a 16-2 game in which his teammates hit 7 HRs should get more credit for being "valuable" than a guy who hits 4 HRs in a 5-4 loss.
Simply put, players contribute to team outcomes. Team outcomes are the aggregation of individual contributions. No player controls nor significantly influences the performance of his teammates such that he should be given credit above and beyond his individual contribution. To give an individual award on the basis of team performance is to undermine the very purpose of the award, which is to recognize individual excellence.
Of course, I'd be willing to bet that once you jumped through all your hoops to discount performance in losses you'd end up coming to such a skewed set of conclusions that you'd give up the idea. I think people are WAY too caught up on the idea of "value". It can make for some interesting conversations, but they're just silly. At the end of the day, why would we want to think about an end of season award that way? Don't we just want to say who had the best season and/or who the best player is?
Games are won on run differential -- scoring more than your opponent. Runs are runs, scored or prevented they all count the same. Worry about scoring more and allowing fewer, not which positions contribute to which side of the equation or how "consistent" you are at your current level of performance.
Board Moderators may, at their discretion and judgment, delete and/or edit any messages that violate any of the following guidelines: 1. Explicit references to alleged illegal or unlawful acts. 2. Graphic sexual descriptions. 3. Racial or ethnic slurs. 4. Use of edgy language (including masked profanity). 5. Direct personal attacks, flames, fights, trolling, baiting, name-calling, general nuisance, excessive player criticism or anything along those lines. 6. Posting spam. 7. Each person may have only one user account. It is fine to be critical here - that's what this board is for. But let's not beat a subject or a player to death, please. |