General 2020 Blue Jays Discussion Thread

[Spanky99]

Why is peak Olerud better than Peak Alomar according to fangraphs?

 

A .350 hitter with 110 walks 40 double 20ish homers or so and gold glove 1b defense (93 and 98)

 

vs A .320 hitter with 80 walks 40 doubles, 20ish homers or so and the stats didn't read his defense (93, 96, 99, 01 I think)

 

Olerud, your original post was money, this follow up doesn't make sense, bro. It answers itself.

[Olerud363]

Because Alomar is a victim of oddly false negative defensive value from a time it wasn't tracked very well

 

Why did the metrics work for Devon White and Ozzie Smith??

 

If you look at D-leaders from 88 to 93 it's Ripken, Ozzie Smith, Devon White, Benito Santiago, Dick Scholfield, Mike Gallego, Scott Fletcher... some of those names make a lot of sense... some of them I don't remember enough about.

 

The worst were an older Winfield, an older George Brett, Danny Tartubal, Goerge Bell... kind of who you'd expect (some of it is a penalty for DHing)

 

Goerge Brett is interesting. If the formula isn't complete crap you'd expect him to be positive early in his career, then fall off. That is exactly what happened.

 

https://www.fangraphs.com/players/george-brett/1001400/stats?position=3B

 

I can accept that Alomar was a bit better... but I doubt he was a 25 run a year defensive player and the formula missed it but found the other 25 run a year defensive players.

[Jimcanuck]

There must be a flaw in the formula. Alomar was regarded as a top defensive 2B - 10 gold gloves.

 

I wonder if the astroturf at the Dome at the time, crazy fast, dampened his chances and the algorithm doesn't account for it.

[Olerud363]

There must be a flaw in the formula. Alomar was regarded as a top defensive 2B - 10 gold gloves.

 

I wonder if the astroturf at the Dome at the time, crazy fast, dampened his chances and the algorithm doesn't account for it.

 

The formula seems to work for Tony Fernandez and for Frank White. Manuel Lee had good defensive numbers when playing beside Alomar on the same turf.

 

It could be a combination of things. Lee and Olerud were also good defenders, so maybe the formula is thrown off if they occasionally take some of Alomar's plays.

 

Did Alomar hustle all the time?? I seem to remember him turning it on and off. He did not play quite as well in 94 and 95 when the team wasn't winning.

 

What about positioning? They didn't shift like they do now, but some players and teams may have been better at positioning.

[connorp]

I see Olerud is apparently a negative defender at 1b as well, even though he’s historically considered a good defender

[Grant77]

I see Olerud is apparently a negative defender at 1b as well, even though he’s historically considered a good defender

 

What metric are you looking at? Total Zone has him as a good defender.

 

Looking at these old metrics is like using errors to judge fielders. Sometimes they will fit in with our preconceived notions, but in reality it's just random chance.

 

Sometimes the eye test can fool you, but Robbie is like watching Arenado, Kiermaier, or Simmons today. You just know they are the best without looking at a stat sheet.

[Spanky99]

What metric are you looking at? Total Zone has him as a good defender.

 

Looking at these old metrics is like using errors to judge fielders. Sometimes they will fit in with our preconceived notions, but in reality it's just random chance.

 

Sometimes the eye test can fool you, but Robbie is like watching Arenado, Kiermaier, or Simmons today. You just know they are the best without looking at a stat sheet.

 

Bingo! Also Olerud363, Robbie got to balls he had no f***ing business getting to, maybe that's the flaw in the algorithm.

[wamco]

I can’t believe I’m finally agreeing with spanky. Any stat that shows Robbie as a negative or even average defender should be thrown out and never mentioned again.

[Orgfiller]

Bingo! Also Olerud363, Robbie got to balls he had no f***ing business getting to, maybe that's the flaw in the algorithm.

 

That's not how defensive metrics work. Jeter didn't get to anything and was punished heavily. You don't get penalized for getting to balls that others don't either and not making the play.

 

The most likely explanation here is Alomar had weird starting position so he didn't get to balls the average 2B would get to, or would have to put in more effort in getting to those balls as a result. But that doesn't mean he wasn't an elite defender. The same happened to Adeiny Hechavarria recently, the eye test saw an elite defender but the stats didn't agree, so one year they fixed his positioning and suddenly the stats agreed with the eye test.

[Grant77]

That's not how defensive metrics work. Jeter didn't get to anything and was punished heavily. You don't get penalized for getting to balls that others don't either and not making the play.

 

The most likely explanation here is Alomar had weird starting position so he didn't get to balls the average 2B would get to, or would have to put in more effort in getting to those balls as a result. But that doesn't mean he wasn't an elite defender. The same happened to Adeiny Hechavarria recently, the eye test saw an elite defender but the stats didn't agree, so one year they fixed his positioning and suddenly the stats agreed with the eye test.

 

The most likely explanation here is that the stats are not correct. I think you're giving far too much credit to defensive metrics from 30 years ago. We've only very recently learned to evaluate fielders properly most of the time, not even all of the time today.

[Laika]

If you don't think the way TotalZone is calculated (or maybe more correctly, not adjusted) for the 1991 to 1995 seasons unfairly maligns Alomar, consider the following:

 

Ages 20-22 with SDP

+13 TZ runs (total across all years)

 

Ages 23-27 with TOR

-26 TZ runs

 

Ages 28-33 with BAL and CLE

+13 TZ runs

 

For his age 34 season we start getting UZR. His UZR/150 was -1.6 and -6.7 in his age 34 and 35 seasons. Age 36 was lower but that's a very small sample.

 

It just does not pass the sniff test that he was as bad at 23-27 with Toronto as he was in his mid 30's when UZR kicks in. The fact that he was above average before and after Toronto makes me think something funny was going on with Toronto. Maybe it was the turf, and maybe it was positioning (which could have been his fault I guess, and then it would be fair to slap the negative runs on him).

 

But even if you forgive some of the negative Toronto years and assume he was above average, you're only moving the needle by a few career WAR. If you take the -26 from Toronto and turn it into a +14, that's only four more career wins.

 

"the stats are not correct" is not an argument that holds much water with me. These stats are more or less just looking at catch/out rates... certain years can be wonky but you can't hide being an elite defender for an entire career; if he was truly elite it would have shown up in the out rates. Elite defenders by definition get outs. It really looks like he was just a good defender. Org is hitting the nail on the head with the Hech analogy. It's possible to be an "elite" defender by tools/actions and not by results, for various reasons. It's kind of like having bat speed and power but a hole in your swing, or elite stuff on the mound but being one ingredient short of a good pitcher. If Alomar's career began in 2020 perhaps the +15.8 defensive runs on his Fangraphs dashboard would be more like +158 runs.

[keggy]

The amount of blind faith casual fans put into defensive metrics, especially for historical players, is why the old school guys hate analytics so much.

 

Look up how Total Zone is calculated.

[Jimcanuck]

Let's call it turf and positioning, blame the former on the Dome and the latter on Cito Gaston.

[Laika]

The amount of blind faith casual fans put into defensive metrics, especially for historical players, is why the old school guys hate analytics so much.

 

Look up how Total Zone is calculated.

 

Where are the casual fans?

 

You think we don't know how TZ is calculated?

[Olerud363]

I see Olerud is apparently a negative defender at 1b as well, even though he’s historically considered a good defender

 

Fangraphs defensive runs are adjusted per position, 0 or slightly negative is good for a 1b. Bad 1st basemen are -20.

[Olerud363]

The most likely explanation here is that the stats are not correct. I think you're giving far too much credit to defensive metrics from 30 years ago. We've only very recently learned to evaluate fielders properly most of the time, not even all of the time today.

 

Why did it work for Tony Fernandez? And Ozzie Smith? And Ryne Sandberg? And Frank White? And Devon White?

 

The guys that won lots of golds in Jays history are Barfield, Fernandez, Devon White and Alomar. Fangraphs top 3 Blue jays defensively are Fernandez (91), Barfield (83) and White (81), but Alomar is -4.

 

What is more likely??

 

a) A s***** random system correctly scored three of the top 4 but was 100 runs off on the fourth?

The system isn't perfect but also isn't totally s***** or random and there was something off about Alomar (Positioning, effort, pitching staff, dynamics with other defenders)

[Olerud363]

My above post is a bit flawed because Fernandez and Barfield played a lot more games as Jays then say Devon White and Kevin Pillar

 

What is the best way to summarize peak defensive value that can be applied to players going back to the 70s?? It seems the fangraphs system is scoring the top 3 seasons of guys like Ozzie Smith and Devon White correctly. Or am I just cherry picking?? What other players from the 80s and 90s are rated low statistically but considered great defensive players?

[connorp]

I didn’t remember Barfield being silky smooth like White. I remember the cannon I think. I was only 8 in 1987 when I can remember following George Bell and the Jays close, though I was Jays fan couple years before.

 

Just looked at his FG page for first time. If he was around today he probably would’ve been set for huge payday if he made it to FA but would’ve been disappointing outcome seems like.

 

I came across him at some minor league game I think when I was around 14 or so and I talked to him for a bit asking him about Japan. I remember him being a really cool dude and he was happy to talk about it

[Laika]

I see Olerud is apparently a negative defender at 1b as well, even though he’s historically considered a good defender

 

Fangraphs defensive runs are adjusted per position, 0 or slightly negative is good for a 1b. Bad 1st basemen are -20.

 

Yeah, Olerud was +82 runs in his career by Total Zone and then +16.8 runs at the end of his career by UZR. That's +6.63 runs per 150 games in his career. For reference, Keith Hernandez and Daric Barton are like +9 runs per 150 games.

[Laika]

My above post is a bit flawed because Fernandez and Barfield played a lot more games as Jays then say Devon White and Kevin Pillar

 

What is the best way to summarize peak defensive value that can be applied to players going back to the 70s?? It seems the fangraphs system is scoring the top 3 seasons of guys like Ozzie Smith and Devon White correctly. Or am I just cherry picking?? What other players from the 80s and 90s are rated low statistically but considered great defensive players?

 

All you can really do is look at Total Zone, which is TZ under standard fielding on Fangraphs or I think Rfield on BR. Self calculate their TZ per 150 games let's you compare it directly to UZR/150 of a modern player

[Abomination]

The most likely explanation here is Alomar had weird starting position so he didn't get to balls the average 2B would get to, or would have to put in more effort in getting to those balls as a result. But that doesn't mean he wasn't an elite defender. The same happened to Adeiny Hechavarria recently, the eye test saw an elite defender but the stats didn't agree, so one year they fixed his positioning and suddenly the stats agreed with the eye test.

 

It's almost certainly just the turf. I mean, that old stuff had regular ground balls skipping all the way to the wall with regularity. Olerud also played way off 1st if I remember right, so it's possible that affected his numbers too.

[Jimcanuck]

It's almost certainly just the turf. I mean, that old stuff had regular ground balls skipping all the way to the wall with regularity. Olerud also played way off 1st if I remember right, so it's possible that affected his numbers too.

 

Yes, and it's amazing on that rock hard turf at Exhibition Stadium that Tony Fernandez put up the numbers he did. Same turf that wrecked his knees.

[Grant77]

Why did it work for Tony Fernandez? And Ozzie Smith? And Ryne Sandberg? And Frank White? And Devon White?

 

The guys that won lots of golds in Jays history are Barfield, Fernandez, Devon White and Alomar. Fangraphs top 3 Blue jays defensively are Fernandez (91), Barfield (83) and White (81), but Alomar is -4.

 

What is more likely??

 

a) A s***** random system correctly scored three of the top 4 but was 100 runs off on the fourth?

The system isn't perfect but also isn't totally s***** or random and there was something off about Alomar (Positioning, effort, pitching staff, dynamics with other defenders)

 

I never said it was random, but just picking players whose stats match with your perceptions doesn't mean that it is bang on either.

 

Let's just look at the gold glove winners for 1990 for a more random sample:

 

Sandy Alomar -3

Mark McGwire 12

Harold Reynolds 20

Kelly Gruber -4

Ozzie Guillen 13

Ellis Burks -8

Ken Griffey Jr 2

Gary Pettis 4

 

A couple of hits, a couple of misses. Kind of what you would expect from an extremely flawed stat. Voting matches up a lot closer with the statistics today.

[Olerud363]

I never said it was random, but just picking players whose stats match with your perceptions doesn't mean that it is bang on either.

 

Let's just look at the gold glove winners for 1990 for a more random sample:

 

Sandy Alomar -3

Mark McGwire 12

Harold Reynolds 20

Kelly Gruber -4

Ozzie Guillen 13

Ellis Burks -8

Ken Griffey Jr 2

Gary Pettis 4

 

A couple of hits, a couple of misses. Kind of what you would expect from an extremely flawed stat. Voting matches up a lot closer with the statistics today.

 

Your stats are off, or at least don't match with fangraphs defense. Mark Mcquire was 2, Sandy Alomar 4, and Ozzie Guillen 20ish in 1990.

 

And including 1b men in there biases it a bit because they are always lower

 

That being said the system has identified Kelly Gruber and Ellis Burks as guys that might also be over-rated. It identified Ozzie Guillen (and Ripken the next year) as being really good.

 

I think the gold glove voters were idiots for a while. Didn't Pailmero win won at DH?? To do this properly you have to identify a set of "awesome" defenders, guys that won multiple gold gloves, at a premium defensive position and were well regarded by the eye test.

 

Do the stats miss on a lot of them? Or only on Alomar?

[Grant77]

Your stats are off, or at least don't match with fangraphs defense. Mark Mcquire was 2, Sandy Alomar 4, and Ozzie Guillen 20ish in 1990.

 

And including 1b men in there biases it a bit because they are always lower

 

That being said the system has identified Kelly Gruber and Ellis Burks as guys that might also be over-rated. It identified Ozzie Guillen (and Ripken the next year) as being really good.

 

I think the gold glove voters were idiots for a while. Didn't Pailmero win won at DH?? To do this properly you have to identify a set of "awesome" defenders, guys that won multiple gold gloves, at a premium defensive position and were well regarded by the eye test.

 

Do the stats miss on a lot of them? Or only on Alomar?

 

It seems obvious by that post that you aren't aware how defensive statistics work on fangraphs, so I don't think it's worth my time to craft a response. My statistics are correct.

[Ryu In My House]

Yes, and it's amazing on that rock hard turf at Exhibition Stadium that Tony Fernandez put up the numbers he did. Same turf that wrecked his knees.

 

That is such a great point. I forgot that he had knee problems. If he has played in St. Louis his whole career I wonder if he may be in the HOF.

[Olerud363]

It seems obvious by that post that you aren't aware how defensive statistics work on fangraphs, so I don't think it's worth my time to craft a response. My statistics are correct.

 

What don't I know?? Fan graphs has a collumn called defense. The numbers you reported do not match fangraphs defense statistic.

 

https://www.fangraphs.com/players/ozzie-guillen/1005125/stats?position=SS

 

For example you reported that Ozzie Guillen had a defense of '13' in 1990, no clue what that means. Fan graphs had him at 21 runs above average that year.

 

Stop being a jerk. 13 doesn't equal 21, if you are getting your numbers somewhere else, or looking at a different stat, just say where and what.

[Olerud363]

What don't I know?? Fan graphs has a collumn called defense. The numbers you reported do not match fangraphs defense statistic.

 

https://www.fangraphs.com/players/ozzie-guillen/1005125/stats?position=SS

 

For example you reported that Ozzie Guillen had a defense of '13' in 1990, no clue what that means. Fan graphs had him at 21 runs above average that year.

 

Stop being a jerk. 13 doesn't equal 21, if you are getting your numbers somewhere else, or looking at a different stat, just say where and what.

 

Maybe you are looking at stats with the position adjustments taken out? Where did you find that? If you are doing that, it is fine, but we have to do the same for Alomar, in which case he would be even a little worse each year.

 

The entire time I've just been looking at the fangraphs defense raw totals as presented on the player page just before the WAR column. If your stats have an adjustment what is it and how do we get Alomar's stats with the adjustment? **

 

** this is confusing so the best thing to do is link to the stat you are using, or give enough information so someone else can look up the stat. I personally am comfortable with the "adjusted" fangraphs totals. I know 30 is a great score for a short stop... I know 0 is actually a good score for 1b. I know -20 Vlad like numbers are bad. Other stats have different scales and adjustments, so just tell us what stat you use and point us to it.

[Laika]

What don't I know?? Fan graphs has a collumn called defense. The numbers you reported do not match fangraphs defense statistic.

 

https://www.fangraphs.com/players/ozzie-guillen/1005125/stats?position=SS

 

For example you reported that Ozzie Guillen had a defense of '13' in 1990, no clue what that means. Fan graphs had him at 21 runs above average that year.

 

Stop being a jerk. 13 doesn't equal 21, if you are getting your numbers somewhere else, or looking at a different stat, just say where and what.

 

Yeah the "defense" number in Fangraphs dashboard is defensive skill + positional adjustment. If you want to see those two figures separated, click on the Value tab and you can see Fielding + Positional as two columns.

 

Example with John Olerud... he has +99 fielding runs in his career because he was an exceptional first basemen, but first basemen have the worst positional adjustment done to their value so in his career he suffered -144 runs from the positional adjustment. Hence his -45 number on the Fangraphs dashboard.

 

It was only five years ago, or so, that Fangraphs changed their dashboard to show this adjusted "Defense" number instead of raw UZR. I remember hating it at the time because of this exact reason. What I want to see on the dashboard is how good the player was at their position. No reason they can't just show both on the dashboard (defense + positional adjustment). I said at the time that this would confuse lay-people and it clearly does.

 

(Note that the defensive runs number on Fangraphs will be either UZR, or Total Zone 1, or Total Zone 2, or Total Zone 3, depending on the year. There are three different formulas for TZ based on how retrosheet boxscores changed over time).

[Olerud363]

Yeah the "defense" number in Fangraphs dashboard is defensive skill + positional adjustment. If you want to see those two figures separated, click on the Value tab and you can see Fielding + Positional as two columns.

 

Example with John Olerud... he has +99 fielding runs in his career because he was an exceptional first basemen, but first basemen have the worst positional adjustment done to their value so in his career he suffered -144 runs from the positional adjustment. Hence his -45 number on the Fangraphs dashboard.

 

It was only five years ago, or so, that Fangraphs changed their dashboard to show this adjusted "Defense" number instead of raw UZR. I remember hating it at the time because of this exact reason. What I want to see on the dashboard is how good the player was at their position. No reason they can't just show both on the dashboard (defense + positional adjustment). I said at the time that this would confuse lay-people and it clearly does.

 

 

Thanks, I understand now that Grant was using TZ. Alomar is negative career with TZ, so without the positional adjustment he is even worse.

 

The 1991 gold glovers rated by TZ are still +35 total, and 4ish on average. The system is somewhat hit and miss, but on average it rates gold glovers as positive. There aren't many other guys like Alomar, who were rated very high defensively, but had a long run of negative TZ.

 

Craig Biggio is similar to Alomar I guess (in terms of defensive and offensive stats). Whitaker, White, Reynolds, and Sandberg all did well with TZ.

 

Biggio was on turf, but so was Frank White. Biggio had his best season on Turf though, and didn't seem to improve when moving to grass.

 

I think Laike mentioned that Alomar's Toronto stats are an outlier as compared to San Diego and after Toronto...

 

It looks to me that Alomar's run of bad TZ went from 91 to 97, his first 2 years in Baltimore were bad.

 

Maybe a lot of it is positioning and for some reason Alomar didn't get the positioning down until 98.