There have been three distinct parts to Harrison Bader‘s season.
Part 1. The hits falling in
From the beginning of the season through May 19, a span of 39 games and 140 PA, Bader slashed .280/.321/.333 thanks to a .371 BABIP
Part 2. Hitting for power
From May 20 to July 12, a span of 43 games and 140 PA, Bader slashed .276/.314/.528 – which gave him a .252 ISO
Part 3. Nothing going right
From July 13 to August 20, a span of 25 games and 76 PA, Bader slashed .186/.250/.229
In Part 1, he was below-average offensively. In Part 3, he’s been downright bad. Which leaves us with that middle stretch, where he was a plus. So, what happened in the middle? It’s a little bit longer than usual but he essentially had a six-week hot streak.
When the six-week hot streak was first introduced here back in 2018, here was the description — “In rough terms, these unusual streaks comprise six weeks or 30 games or 120 PA.” If we start Part 2 on May 31, we get a span of 33 games and 113 PA where Bader slashed .286/.301/.543 – which is even more productive than the Part 2 listed earlier.
At the time that the Mets signed Bader, there was virtually no reason to be optimistic that he was going to be a good addition to the team. But thanks to that six-week stretch – even though he wasn’t good before or after – Bader has a 97 OPS+. Of course, there are still 36 games to be played. It will be very surprising to me if he ends the year with a 97 OPS+. Instead, the question is how much further will it drop from here to the end of the season.
It’s not fair to talk about Bader and not discuss defense. The problem there is that the three advanced systems have a pretty wide split on how they view his play in the field. The good news is that Statcast has Bader with a +8 FRV. But UZR has him with a +0.5 and DRS has him with a (-5) with the Mets. My subjective view would put him somewhere between UZR and Statcast – and closer to Statcast.
Yet even if you give him full Statcast value on defense, that’s just a 1.4 fWAR for the season. A 2.0 fWAR is an average player and it’s highly unlikely Bader will reach that threshold this year. And unless he picks it up with his hitting here down the stretch, it’s possible Bader could finish with a mark lower than what he has currently.
Don’t have much to say about Bader, except that he was a factor in today’s victory over the O’s
Twice, he grounded out on a 3-2 pitch
Further, Winker PH for him, and thereby blasted the 2nd walk-off HR in 3 days against Seranthony Domínguez
The following article appeared at FG earlier today (8/21/24)
“Uneven Progress as the Mets Try to Escape Their Early-Season Hole”
I found it OK, but it provided close to nothing of which I wasn’t previously aware
There was, however, one remark which I found noteworthy:
Since [May 31st], [Lindor has] played his way into MVP contention yet again, and his 6.0 WAR is tied with Shohei Ohtani for the NL lead, though he still has to be considered a long-shot to bring home the hardware.
This must be referring to fWAR, because I checked BR and found that for Lindor in NL bWAR, he’s
5th among all
4th among position players
2nd in offensive
Back in the Mesozoic ERA of baseball stats, some silly rationale was offered – usually under the guise of “intangibles” – for giving postseason awards to players with awful numbers
As much as I appreciate the Modern Era of stats, I think that, this year, the non-popular choice would be best.
Consider the 4 NLers whose total bWAR is equal to or surpassing Lindor’s 5.4:
Ohtani (LAD) – 6.3
Marte (ARI) – 6.2
Chapman (SFG) – 5.6
Greene (CIN) – 5.4
Now, ask yourself, would any of these teams be doing worse – in any important way – without these four dudes?
Note that:
– Ohtani is a DH
– Marte, despite his marvelous season, is currently on the IL, and the Snakes haven’t skipped a beat
– Chapman’s team is less like to reach the postseason than NYM
– Greene is a SP, his team is doing worse than SFG, and the last two pitchers to win an MVP were Kershaw (NL 2014) and Verlander (AL 2011)
And then, please explain your reasoning if you disagree that, without Lindor, the 2024 Mets would be dead meat
Conclusion: Lindor deserves to be the Mets’ 1st NLMVP, especially if the Mets make it to the postseason
Why would someone write an article for FanGraphs and not use fWAR?
Thanks, Bruh… I mean Bri
With all due respect, sir
ESPN often refers to WAR without indicating whether it is bWAR or fWAR to which they refer
And in my shameful ignorance, I didn’t realize – until a few months ago – that there are 2 kinds of WAR, both referred to quite often, yet which are computed in significantly different ways
Bader has an unexpected , good year. It is a positive player on the team. He current slump has level off his season to where he stasticly should be career-wise. The Mets have three centerfielders and should used them equally. Bader his slumping and Taylor isn’t doing so well either, so when Nimmo starts in CF have McNeil play LF. Iglesias has force a platoon with McNeil because his .961 OPS vs LHP. Iglesias has a .811 OPS vs RHP in 105 PA. He deserves more playing time so therefore McNeil should get some reps in LF when Iglesias starts vs RHP. The bottom line is Bader should get less starts.
Since July 1, Taylor has a .713 OPS while Bader has a .604 OPS. And that includes a chunk of Bader’s hot streak, including his 2-HR game.
Taylor hasn’t hit LHP this year and his bad stretch came when they were essentially platooning Taylor and Stewart. But he’s done just fine since the beginning of July.
In August, Taylor has a .596 with 43 PA. Bader and Taylor aren’t getting the job done offensively. That is why Iglesia should get more playing time and McNeil should more time in left field. McNeil, in the second half has a .992 OPS. Iglesia has a .866 OPS this season.
I don’t want to make judgments based on 43 sporadic PA
In August , Taylor had a stretch of 7 of 8 complete games and then another 3 complete games in a row. For a fourth outfielder that is pretty good playing time and chances show what he can do. Right now, Iglesias and McNeil are going good.
From ESPN today (8/22/24)
How MVP candidate Francisco Lindor has led Mets’ turnaround
From Tim Britton
“4. Don’t rule out a return for Kodai Senga in the final week.
Senga is on the 60-day injured list and ineligible to return until the final week of the season — Sept. 25 to be exact. But the Mets think there’s a chance the right-hander could make an appearance in those final five games against Atlanta and Milwaukee.
Because Senga’s injury is to his lower body, he’s been able to keep his arm relatively strong by throwing from a seated position. He should begin throwing while standing this week. The biggest hurdle will be the recovery in his strained calf, which he has yet to test.”
https://www.nytimes.com/athletic/5715290/2024/08/22/mets-jose-quintana-christian-scott/
Very interesting
Did a minor analysis, looking at the stretch between 7/13/24 and 8/22/24, during which the Mets’ record is 17-16
During that time, they played 5 teams with winning record and 5 teams with losing records (overall seasonal records, not including any of the 33 games Mets played between 7/13 and 8/22)
Note: For extra clarity, I will refer to the 5 teams with winning record as Studs
And the 5 teams with losing record as Slobs
In this analysis I did not include STL for… reasons (which I will provide if asked)
Against the 5 Studs
(BAL, MIN, ATL, SDP and NYY whose combined record between 7/13/24 and 8/22/24 – but not including any games played against Mets in this period – is 85-65)
Mets’ record is 9-4 against the Studs
Against the 5 Slobs
(SEA, LAA, OAK, COL, and MIA whose combined record between 7/13/24 and 8/22/24 – but not including any games played against Mets in this period – is 61-78)
Mets’ record is 8-12 against the Slobs
Notice that the Studs’ positive sum is mostly due to SDP’s 22-8 record from 7/13 to 8/22
And that the Slobs’ negative sum is mostly due to SEA’s 9-19 record from 7/13 to 8/22
Remove SDP and the Studs go from 85-65 to 63-57
Remove SEA and the Slobs go from 61-78 to 52-59
Not as pronounced, but still quite clear
Remove both SDP and SEA and the Mets go from 17-16 to 16-13
And from 9-4 to 8-4 versus Studs
And from 9-12 to 8-9 versus Slobs
Thence, against 4 Studs and 4 Slobs (excluding SDP and SEA)
Mets have same number of wins against the Studs as against the Slobs
Yet 5 more losses against the Slobs
Final note: most of the time used in devising this essay was due to the writing
The arithmetic (and data collection) part of it took about one fifth the time
On 8/24/24, Footballhead posted the following elsewhere
“A comment in today’s sport page noted that this was the 11th time that the Mets were shutout this season.”
First of all, if I’m not mistaken, shutout, as written, is a noun
In the above sentence, FBH should have used shut out, which can be identified as a verb
What?
OK, that wasn’t my relevant point
Anyway, I was thinking about this baseball… stuff recently
What I plan to do is look at the individual game run distribution in 2024 for three teams – ARI, CHW and NYM, what I believe are the best and worst offensive teams, plus the Mets
I plan to look at the distribution and determine if it is a normal distribution centered at the mean per game run amount, look at standard deviation and other statistical thingies of which I am aware but of which I generally don’t care.
I am not as interested in the actual totals, but rather the distributions, to see if I can discern a pattern
Some preliminary stuff first, just basic observational and not any deep number gnashing
In the 11 shutouts inflicted upon the NYM this year, the opposing team scored every number between 1 and 10 at least once, except for 2 and 9.
Also, the vanquishers of the NYM scored 1 twice and 4 thrice
And obviously, their W/L record in those 11 games is 0-11
First point, everybody seems to attach great relevance to scoring zero in a game. What if NYM had scored precisely one more run in each of those 11 games?
Then they still would have lost 9 of those games, but these wouldn’t have drawn attention to themselves the way in which a big fat hole does
We do have a name for when a team scores no runs
We say they have been shut out
Do we have a name for the situation when a team scores exactly one run?
Yet the result is hardly much better, and possibly no better
In those pair of 1-0 losses, if they had scored 1 run instead of zero, the most likely outcome would have been to split this pair of (what would have become) extra-inning games, since one was played at Citi, the other at Nationals Park
So, we can justify saying that if they had scored exactly 1 more run in the first 9 innings of those 11 games in which they were shutout, then their record most likely would have been 1-10 instead of 0-11
In truth
Do the following – tell a (non-NYM fan) that “The Mets scored 1 run last night”
Then ask them, “Would you say they lost, or they won?”
Well, just about everyone would say, “They suck”
That is, “They (almost certainly) lost”
Maybe because there is only one way to win a game in which you only scored one run, and that’s if you blanked the other team
So
I can infer that scoring just one run in a game is almost as bad as scoring zip
Not quite, but almost as bad
For the next step in preliminaries, I decided to look at all the games this year in which they scored precisely one run – there were exactly ten of these
I found that the opposing team scored 0, 5, 8 and 12 runs exactly once each
But also, their opponents scored 4 twice and 3 four times
The Mets record in these 10 games was 1-9
Now, let’s add one run to the Mets in each of the 10 games
Their record is still 1-9
Summary of part one:
If you look at the 21 offensively limpest Mets’ game in 2024, their record was 1-20
If you add exactly 1 run to their (first 9 inning run) total, their record most likely improves to 2-19
But it’s also possible that their record remains unchanged, or improves to 3-18
Conclusion, if you score 1 or fewer runs in a game, you really suck
And there really doesn’t seem to be a difference – anyone gives AF about – between scoring one run, or no runs
Now, of course, this sample size is much too miniscule to have any true statistical relevance, but I am trying to build some intuition/expectation from what comes next
So
The three teams I investigated, to determine some relationship between how runs scored in and per game are distributed/dispersed, are ARI, NYM and CHW. In turns out, through 8/24/24, all three have played 130 games
Total runs scored
ARI – 694
NYM – 631
CHW – 404
Total runs allowed
ARI – 604
NYM – 590
CHW – 675
W/L record
ARI – 74-56
NYM – 68-62
CHW – 31-99
Pythagorean expected W/L (using 1.88 as exponent)
ARI – 73-57
NYM – 68-62
CHW – 37-93
So, Pyth yields that
ARI has overperformed by 1 win
NYM’s result are exactly what is expected
CHW has underperformed by 6 wins
In other words, for a team which has scored 404 runs and allowed 675 run through their first 130 games, Pyth predicts W/L record of 37-93
Next, average runs scored per game
ARI – 5.34
NYM – 4.78
CHW – 3.11
Now I want to investigate how these individual game run scored values are dispersed
Standard deviation (SD) of runs scored data
ARI – 3.64
NYM – 3.33
CHW – 2.37
Looking at these raw SD values, it seems that the comparative
dispersion between different data sets is significant
SD is computed by taking the difference between each data point and the data set average, and squaring each amount (which makes each difference a positive number) and summing among all data points
The problem is that SD will be larger for one set if the actual data values for said set are larger than a second set, even if they have the same “dispersal”
The way to adjust for this, in order to be able to compare the dispersal of data sets with different average values, is to use something called coefficient of variation, which is obtained by dividing SD by average
CV
ARI – 1.47
NYM – 1.44
CHW – 1.31
This indicates that the dispersal of ARI’s individual game run amounts is very close to NYM’s dispersal, whereas CHW’s values more closely bunched
More soon