Fun with Career Metrics and Graphs

[Tyler]

CAREER VORP BY AGE (PITCHERS)



I took the top players by age for 18-26, a total of five players. I then graphed their careers by cumulative VORP by age. If I had continued, you'd see a bunch of four-year lines like Kojima and Thompson, an artifact of our limited 3.5-history. Also note that all fourth-year data is incomplete.

[Tyler]

CAREER VORP BY AGE (BATTERS)



Same deal as the pitchers.

Fun Fact: McDonald is also the career leader at ages 25-29, even though he is currently in his age 24 season. This is a player who will retain his age records even as PEBA develops a more complete history.

[klewis]

Yeah the career metrics thing is great. I'm still happy with Saunders production considering how young he is but sometimes I do wonder what kind of beast he would been without all those injuries.

[Tyler]

Top 25 Single-Season Walks



Pretty self-explanatory. Note that the bar in red indicates an active season.

[John]

Love the graphs, Tyler. Those are unbelievable walk totals for Riley and Miyata. For reference, Miyata's 168 walks in 2009 would rank 5th all-time in MLB, while Riley's 170 BB in '09 would tie him for 4th in MLB history with Babe Ruth (the top three single-season walk totals all belong to Barry Bonds). Those are just fascinating numbers.

I've been playing around quite a bit with the Career Metrics module as part of my testing. There's an unbelievable amount of interesting facts that pop up when you look through it. There are also a few quirks in the system.

For instance, try running this search in the PEBA utilities: Progressive Sovereign League RBI Leaders by Designated Hitters through age 26. Note that Kuemon Kiyomizu is the progressive career leader for every season the PEBA has been in play. However, he's not listed as the 2010 Active leader despite having more RBI than the player who is listed as the leader, P.J. Thomas (Kiyomizu has 363 to Thomas's 340).

The reason for this is that the utilities don't consider a player "active" until he's played at least one game in the parent league (PEBA or LRS, depending on which version of the utilities are being used) during the current season. Kiyomizu, of course, is still recovering from a devastating injury and won't be healthy for several more months.

I'm wondering what anyone thinks about this behavior? I kind of view "active" as meaning "not retired", though I do agree I wouldn't list a player who is now playing in a league other than the utilities' parent league.

[Tyler]

For instance, try running this search in the PEBA utilities: Progressive Sovereign League RBI Leaders by Designated Hitters through age 26. Note that Kuemon Kiyomizu is the progressive career leader for every season the PEBA has been in play. However, he's not listed as the 2010 Active leader despite having more RBI than the player who is listed as the leader, P.J. Thomas (Kiyomizu has 363 to Thomas's 340).

The reason for this is that the utilities don't consider a player "active" until he's played at least one game in the parent league (PEBA or LRS, depending on which version of the utilities are being used) during the current season. Kiyomizu, of course, is still recovering from a devastating injury and won't be healthy for several more months.

I'm wondering what anyone thinks about this behavior? I kind of view "active" as meaning "not retired", though I do agree I wouldn't list a player who is now playing in a league other than the utilities' parent league.

Hmm. I think this is the correct behavior. Imagine that Kiyomizu never recovers from his injury, or that during his time on the DL he discovers his true calling as a TV chef. The list would then be correct. If he comes back late this season as expected, then he'd return to the top of the chart as the utilities are currently coded, I believe? This seems appropriate. I assume the utilities are counting "last game played" as the date of a player's "end of active status", and although it will produce quirks "in the moment" I think historically the results will make sense.

[klewis]

What happens if hypothetically.....

1. Kiyomizu does not return this season but returns next season
2. Thomas "slows" down this season and has less career RBIs than Kiyomizu
3. Kiyomizu continues next season to lead Thomas in career RBIs

Would the 2010 season show Thomas as the career active RBI leader since Kiyomizu did not play? Then in 2011 revert back to showing Kiyomizu as the career leader?

[John]

What happens if hypothetically.....

1. Kiyomizu does not return this season but returns next season
2. Thomas "slows" down this season and has less career RBIs than Kiyomizu
3. Kiyomizu continues next season to lead Thomas in career RBIs

Would the 2010 season show Thomas as the career active RBI leader since Kiyomizu did not play? Then in 2011 revert back to showing Kiyomizu as the career leader?

That's a good question. I don't actually know the answer.

[Tyler]

2010 Pan-Atlantic Team SB



Partial season totals, of course, as of last Friday's sim.

[John]

1. That's a real eye-opener there with the Pan-Atlantic steals. I had no idea Connecticut was so far ahead of the pack. And wow, Manchester doesn't like to run at all. Aurora has four players with as many or more steals than the Maulers!
2. How did you get Career Metrics to show you those team totals? Or did you get that info from in-game?
3. You're going to let us know how to make those beautiful graphs, right?

[Tyler]

1. That's a real eye-opener there with the Pan-Atlantic steals. I had no idea Connecticut was so far ahead of the pack. And wow, Manchester doesn't like to run at all. Aurora has four players with as many or more steals than the Maulers!
2. How did you get Career Metrics to show you those team totals? Or did you get that info from in-game?
3. You're going to let us know how to make those beautiful graphs, right?

B. That's just in-game. After I had fun creating the first graphs from Career Metrics I just started making fun graphs. I'll edit the title.
C. They're just basic Excel templates.

[klewis]

You should do a X-Y scatter correlation graph. I bet that would really impress John

1. That's a real eye-opener there with the Pan-Atlantic steals. I had no idea Connecticut was so far ahead of the pack. And wow, Manchester doesn't like to run at all. Aurora has four players with as many or more steals than the Maulers!
2. How did you get Career Metrics to show you those team totals? Or did you get that info from in-game?
3. You're going to let us know how to make those beautiful graphs, right?

B. That's just in-game. After I had fun creating the first graphs from Career Metrics I just started making fun graphs. I'll edit the title.
C. They're just basic Excel templates.

[John]

What happens if hypothetically.....

1. Kiyomizu does not return this season but returns next season
2. Thomas "slows" down this season and has less career RBIs than Kiyomizu
3. Kiyomizu continues next season to lead Thomas in career RBIs

Would the 2010 season show Thomas as the career active RBI leader since Kiyomizu did not play? Then in 2011 revert back to showing Kiyomizu as the career leader?

Frank answers:

...the way the current util works, Thomas is displayed for 2010 until Kiyomizu plays another game, at which point Kiymizu would again be recognized as an active player and would replace Thomas as the 2010 active leader.

That said, what I was thinking about on the way home was a way to get Kiyomizu to show up even though he's injured, which I've now implemented. The new way it'll work is that Kiyomizu will show up instead of Thomas. If Kiyomizu then retires without ever playing again, Thomas will replace him as the 2010 active leader. So the definition of active changes from "actively playing" to "not retired". While Baseball Reference doesn't have progressive leaders broken down any more specifically than all of MLB, this approach is consistent with what they do.

Nice job, Kevin: your question prompted a change in the behavior of the utilities!

[Maulers]

1. That's a real eye-opener there with the Pan-Atlantic steals. I had no idea Connecticut was so far ahead of the pack. And wow, Manchester doesn't like to run at all. Aurora has four players with as many or more steals than the Maulers!
2. How did you get Career Metrics to show you those team totals? Or did you get that info from in-game?
3. You're going to let us know how to make those beautiful graphs, right?

My players barely believe in getting on-base, let alone stealing!