The fallacy of fielding measurements. The 'common man' is correct, Jeter is better than the stats!
1948_Swell_Robinson
Posts: 1,870 ✭✭✭✭✭
The fallacy of baseball fielding metrics, Derek Jeter, the Yankee dynasty, & some logic
It has always been my contention that the fielding measurement currently used are extremely flawed and not worth anywhere near the validity of the advanced hitting measurements, and that is a big part of the reason why you get crazy results in stats like WAR that make a lot of smart baseball fans scratch their heads.
The 2004 Yankees won 101 games. They were second in league in runs scored. 6th in the league in ERA. Here are the fielding runs saved according to the Fielding Bible that has been talked about in that conversation with Bill james where they measure each of Jeter’s plays and view him so badly.
All 8 of the Yankees starters compared to the best in MLB at that position with each of their runs saved. The – sign means negative runs…which is a recurring theme for the 101 win Yankee team.
SS
Jeter -13
Adam Everett 20
2B
Cairo -8
Hudson 25
3B
Arod 14
Rolen 30
1B
Clark -2
Helton 12
C
Posada -12
Schneider 24
LF
Matsui -4
Crawford 14
CF
Williams -20
Hunter 18
RF
Sheffield -9
Ichiro 30
So the Yankees won 101 games with the most atrocious fielding ever(according to that metric), especially up the middle, and still managed to win 101 games and still have the sixth best ERA in the league?? Either the old saying, “strong up the middle” means nothing thus rendering the defenisve value as minimial, or the fielding results are mostly a product of random chance and what type of pitchers you have on your staff, also rendering the defensive value as minimal…or both.
The Yankees also won three world series with the same absolute crap playing defense up the middle?? Knoblauch was viewed as crap defensively too according to the metric when they had him.
I don’t buy the validity of the defensive measurements. Nobody should.
In order for each of those top defenders in the league to replicate their numbers, if they all played on the Yankees, would mean they would need the exact same amount of chances. There aren’t enough balls in play to even make that possible.
There is difference of 227 Runs saved between those top defenders above and the Yankee starting eight..
So is someone going to tell me with a straight face that if you replaced all those Yankee defenders with the league leaders at each position that the Yankees ERA would drop all the way down to 3.27 when everyone else in the league was over 4.00???
Give. Me. A. Break.
Top Teams in ERA:
4.03
4.17
4.18
4.28
4.53
Yankees 4.69
As for John Dewan’s viewing every play in his fielding measurements that are hailed as accurate?
1). When he measures Jeter Vs. Everett. Does he have their spot where they are standing on the infield measured to the exact inch before the ball is hit? If they are positioned 8 inches differently and the ball is hit to the exact same spot in the infield, then it completely renders his findings as meaningfess.
2). Does he have the exact velocity of each batted ball? If he has number one covered above, unless he has the exact velocity of the batted ball, then a difference of 4 MPH on the ball can(and is) the difference between it being caught or not. He says he classifies them as soft, medium, hard. That is hogwash. There is no way he can decipher between 5 MPH or even 10 MPH. A 10 MPH difference would be asrtronimical in the difference between it being able to being caught of not.
3). Does he measure the height of the last hop? Not all ground balls are created equal. The last hop bounching even four inches higher than another similarly hit ball is often the difference between it being caught of not.
The reality is that 90% of defensive plays in MLB are of the routine variety. As long as someone can convert the routine play, then the more balls hit to them the better their fielding metric will appear when compared to someone else who simply did not have as many chances.
Now it is absolutely true that a person has to be able to prove that he can convert those routine plays and that is most definitely a skill that the average human cannot do, and people like Ryne Sandberg have shown an ability to do it even a little better than any other 2B in MLB history(better than Mazeroski whom is also viewed ridiculously too high with the faulty fielding measurements).
However, once that ability is proven, then the wild differences in fielding measurements between players is mostly the result of simply having more balls hit your way from either pure random chance OR because of the type of pitchers you have that are providing the balls. Also, the surface of the field can be a factor.
We have seen players valued low as making several non-routine plays. So the notion that the other 10% of non routine plays are only converted by just a few people is not accurate either. Jeter has shown that ability to make more than a fair share of non routine plays…of the few that are made around MLB. People think that circus plays are the norm because that is what the highlights always show. They do not happen as often as they appear.
Take Jeter and compare him to himself, Derek Jeter.
Fielding measurements have a fatal flaw that has never been rectified, especially for older players before the year 2000 or so. The easiest way to have a high fielding rating is to simply have more balls hit your way. Since you are already proven you can handle the routine play in MLB, then you will handle those balls at a rate high enough to make you look better than a league mate who simply may not have had the same amount of chances.
Take Jeter and compare Jeter to himself:
1996 he had 444 assists in 1,370 innings.
1997 he had 457 assists in 1,417 innings.
2001 he had 343 assists in 1,278 innings.
2002 he had 367 assists in 1,383 innings.
2005 he had 455 assists in 1352 innings.
Did he somehow lose his range in 2001 and 2002 and only to somehow regain it a few years later at an older age in 2005?
2001 he lost 112 assists compared to 2005. Yes, 74 less innings, so if we equalized the innings it would still be around a 100 assist difference.
1996 to 2002 he lost 77 assists and played 13 more innings.
So what is the driving force in that swing of 77 to 100 assists by the same player on the same team?
Imagine a hitter losing 100 singles from one year to the next. One does not simply walk into Mordor. That simply does not happen to all-star caliber hitters in the middle of their prime.
The driving force is sheer luck of a ball being hit your way or not. In baseball, 90% of the defensive opportunities are of the routine variety. Of those 90%, if one player is simply getting more chances his way, then he will have a better defensive rating in any defensive metric.
This is a guy on the same team with sometimes the same pitchers and look at those wild jumps. Now imagine comparing Jeter to a league mate who has a completely different set of pitchers...it only compounds it more.
Whatever fielding metrics Jeter has, you can apply an 80% discount to them. There is no way on earth that the validity of his hitting ability should be cancelled out by invalid or faulty fielding measurements.
So take Jeter's hitting ability at 100% face value. Take his current defensive metrics and give them a 20% value....and then combine them. Then you will get a more realistic valuation of him.
When you see Jeter make ridiculous acrobatic plays and heads up plays, there is no way he becomes that much of a stiff on balls hit seven steps to his left or right.
Add Jeter's excellent baserunning skills in both stolen bases and standard baserunning. Jeter is better than what the defensive metrics are doing to reduce how great he really was..
Comments
Are defensive measurements truly accurate? Should they be relied upon to the same degree as offensive measurements are?
Here are the 1985 cardinals Total Zone defensive runs saved above average and where they ranked in the league.
1.Pendleton 24
3.Ozzie smith 20
14.Van Slyke 9
16.Coleman 7
29.Mcgee 5
42.Landrum 5
49.Porter 3
The Cardinals had seven defenders in top 50 in league in defensive runs saved. Notably, their third baseman and shortstop were first and third.
The 1985 Mets defensive runs saved above average at those two positions were:
Shortstop -8
Third base -5
The Cardinals had a team ERA of 3.10. The Mets had a team ERA of 3.11. The league average ERA was 3.59.
The defensive run differential between Pendleton/Smith compared to the team total to the Mets SS and 3B were 57 defensive runs saved.
So if we are to take those defensive TOTAL Zone runs at face value as a measurement of their true defensive value, then putting Pendleton and Smith on the Mets would make their team ERA go from 3.10 all the way down to 2.76.
From 1982-1986 no team has had a team ERA under 3.00. In fact, during that time, the 3.10 and 3.11 ERA's put up by the cardinals and Mets already represent the first and second best marks.
Let's go further. If we took the top position players at each position in MLB where the Mets 'struggled' on defense.
2B Hubbard 17
3B Pendleton 24
SS smith 20
LF Henderson 13
CF Pettis 12
RF Barfield 23
Total runs saved 109 runs above league average.
And then put them on defense instead of who the Mets had out their on defense.
Let's see what happens.
Keep in mind the Mets already won 98 games and had a fantastic 3.11 ERA that was second in the league(3.10 was first).
According to the stats the Mets defenders were only good enough to accumulate ONE defensive run saved. One. The funny thing is that Hernandez and Carter were credited with 26 defensive runs saved, meaning the rest of their defenders were only good enough to save NEGATIVE 27 runs. I don’t buy it.
So simply trading out those six defenders and putting them in for the Mets defenders saves an extra 136 runs!!
So based on that defensive measurement, all you have to do is replace those six defenders and the Mets team ERA Goes down to 2.28!
So what is more feasible as the reason why those defenders are credited with such wide degree of defensive prowess? That they actually were that good, and all you had to do was get those six defenders on one team and you are instantly a WS contender? Or they simply played on teams that provided a high amount of balls hit in their direction?
Add the fact that the Mets pitchers struck out more batters than everyone else, and they simply didn't give as many defensive opportunities to every position other than catcher and first.
The following year the Meta defense still only were credited with 3 runs saved above average. This time it was Hernandez and Dykstra 'carrying the load' with 25 runs saved, and the rest of the team at negative 22. The chances simply flipped to other spots. Pure chance at work there.
The reality is that luck and the type of pitching staff are the biggest determining factors of how many balls that defenders will get...and those will pump up a players defensive 'range factor, making them appear far superior to other defenders in the league.
Defensive skill plays a role too, but not nearly as much as the defensive runs say, otherwise it would be quite easy for teams to save 130 runs by grabbing those defenders with such high runs saved totals.
Individually, Ozzie Smith averaged 9.7 fielding runs per 1,200 innings per year in his four years with San Diego.
Then his next four years with St Louis he averaged 16 defensive runs saved per 1,200 innings per year
From 1985-1990 on St. Louis, Pendleton averaged 16.4 defensive runs saved per 1,200 innings per year.
Then his next four years in Atlanta Pendleton averaged 5 defensive runs saved per 1,200 innings per year.
In his five full seasons with St Louis, Gary Templeton averaged 5.1 runs saved per 1.200 innings per year.
In his five full seasons after leaving St Louis, Templeton averaged negative 2 runs saved per 1.200 innings per year.
But more pressing with Templeton was the immediate impact. In his last two years with St Louis he averaged 13.5 runs saved per 1,200 innings per year. In his next two years with San Diego he averaged 0.5 runs saved per 1,200 innings per year.
So the defensive carry over when plopped in a new environment isn’t exactly something to be counted on. The true value of the defender will remain the same, but the measurement used to express that value will be wrong as it typically is.
I don't understand the simplest thing about fielding......the error.
I've seen guys run hard to get into position to catch the ball and fumble it when they surely should have made the play, it gets called a hit because the fielder made a strong effort to get there. Isn't he supposed to? I've also seen line drives hit to third base, pretty much right to the fielder, and he can't catch it, again called a hit. Finally, a lazy fly ball that the player totally misplays and it's a hit because he didn't touch it.
Bottom line is, these calls boil down to someone's opinion.
Gold Gloves have been deemed worthless by many here. Now I have seen a player win one and you know that was a gift, but Vizquel wins one year after year and gets called a lousy fielder. Personally, I don't know if he was average, good or great.
Next is Puckett who I do know about. Superb fielder and people have called him overrated.
The fielding debate is a minefield, stay out.😁
During the Brett-Schmidt debates they always tried to tell me how Schmidt was far superior on defense when no adjustment was ever made for the pitcher hitting in the nl compared to the dh in the al.
All anyone ever said was well Schmidt had to come in and field all those tough bunts from the pitcher.
😂
Yup, the Puckett example is exactly the point.
Official scorers, I have no explanation for that. They are worse than the NFL replay officials. They must get them from the same academy.
Defense should only be 10% of the equation in that debate. I have jested about Brett's miscues in the playoffs, but that was some old fashioned sports talk needling, and used to explain a different point.
or the fielding results are mostly a product of random chance
This single thing is the most important.
In light of removing players for PED use, do we need to bump up the players who didn't take them but are still being compared to them sabermetrically?
What is Jeter's OPS+ in his era compared to clean players only?
When you combine that aspect and then combine the aspect in this thread where the sabermetrics incorrectly degrade Jeter on his defense, what you end up with is...
Jeter has a claim for the GOAT SS
The whole you have to touch it thing has always been dumb. If an outfielder falls down or turns the wrong way and misses it that should be an error. Just like if 2 or 3 guys are just staring at each other and the ball falls at their feet at the very least that should be a team error.
Some former scorers and insiders have talked about it, but basically for the last 20 years or so MLB has been pressuring the scorers to give more hits and less errors. Its looks better on the stat lines for the casual fan not to see 4 or 5 errors. They basically want anything within spitting distance of being debatable to be a hit instead of an error.
Exit velo is currently the narrative for those at them ones. There are some special times where that is valid but more than not are plays that should be made
Missouri 14 OSU 3
Back to the 1985 comparison of the Cardinals and Mets:
That fact that the Mets pitching staff had approximately 230 more strikeouts than the Cardinals must have had some impact as the Cardinals had to field those outs.
The Mets' fielders can't control that. It definitely seems that the Cardinals' players are rewarded for it. Who knows if they were difficult or easy plays?
Quantifying defensive stats is nowhere near an exact science. Chances may imply range, but i) you can only field the balls hit to you; and ii) not all chances are the same. These stats have to be combined with the eye test.
All I can say is that after watching a decent amount of baseball over 50 years, I can identify players who are liabilities. Guess what? So can baseball professionals. There's a reason players get replaced for defensive reasons. To defend one player, Jeter never needed to be replaced for defense.
Also, there are those that you see do things that you don't see others do. Mike Schmidt was one of those players. Multiple times, with a runner on 2nd I saw him dive to field the ball and then have the awareness and ability to lunge and tag out the runner trying to go to third. I haven't seen anyone else do that. Also, in 85 when he played first base, with runners on first and second, he had a set play with Rick Schu, that if the ball was hit to Schmidt, Schu would act like a first baseman and Schmidt would throw to third for the force. IIRC, I think they did this at least a couple times and got at least one double play. While it's probably been done other times, I don't recall seeing it done myself.
I generally agree with all your points, but Jeter the GOAT? I'm going to still disagree if you move him to #2, but to say that Jeter was better than Honus Wagner means you have to be not just downgrading baseball from before 1920, but denying its existence entirely. By some measures, Wagner dominated baseball more than Ruth did. I'd disagree less with an argument that Wagner was the GOAT baseball player than an argument that he wasn't even as good as Derek Jeter.
Lots of defensive stats, and virtually all of them on baseball-reference, do not take this into account and as a result are completely worthless. There are also pitching staffs that get an inordinate amount of their outs via ground balls or via fly balls, and the fielding stats of the infielders vs. outfielders on those teams are completely worthless since the stats don't consider that, either.
Derek Jeter was not as good a fielder as Ozzie Smith; I think that's obvious. But WAR rates Ozzie Smith as the 9th best SS in history and Jeter the 13th best, and I think that's nonsense. Jeter was a more valuable player than Smith, hands down. WAR way overstates Smith's defensive value because the Cardinals didn't strike out many hitters, and allowed a lot more ground balls than fly balls. Well, WAR overstates Smith's value and understates Jeter's value for lots of reasons because WAR is pure, unadulterated crap in evaluating fielders, but the point you raised is one of those reasons.