One of my concerns with The Churn is how much cumulative playing time it gives to sub-replacement players. Each one might not stick around long but if you add them all up it can have a significant impact, or so it goes in theory. I decided to try to quantify The Churn effect by adding up all the PA, IP and fWAR of all the "churned" players for a whole season.
To get started, I had to define a "churned" player. I looked at players who were acquired as minor league FA's, waiver-wire claims or minor league trades since the end of last season who at some point made the Active Roster. I started with those players who are now no longer on the roster but realized that may exclude the more successful ones who stuck and and are still on the roster, so I added those. Here's what I found:
There were a total of 23 players who met the criteria. 15 were off the Active Roster by the end of the season* with 8 remaining, mostly pitchers. Here's the list followed by PA or IP and fWAR.
Position Players:
Lewis Brinson OF: 39, 0.1
Yermin Mercedes UT: 83, -0.2
Donovan Walton IF:78, -0.7
Willie Calhoun OF: 9, -0.2
Austin Dean UT: 9, 0.1
Stuart Fairchild OF: 8, -0.2
Mike Ford UT: 4, 0
Andrew Knapp C: 7, 0
Dixon Machado SS: 17, 0
Kevin Padlo UT: 12, -0.3
Mike Papierski C: 10, -0.2
Kai Tom OF: 1, 0
Luke Williams OF: 12, 0.1
Ford Proctor C/IF: 22, 0.0
Austin Wynns C: 177, 0.4
Total PA's: 498. Total fWAR: -1.0
Pitchers:
Mauricio Llovera RHP: 26.1, 0.1
Alex Young LHP: 26.1, 0.4
Shelby Miller RHP: 7 IP, 0.4
Scott Alexander LHP: 17.1, 0.3
Jharel Cotton RHP: 8, 0.2
Luis Ortiz RHP: 8.2, 0.1
Andrew Vasquez LHP: 2, 0.1
Total IP: 94. Total fWAR: 1.6.
*Walton and Llovera are on the injured list but I counted them as off the roster as I highly doubt they will be back.
As you can see, The Churn worked better for pitchers than for position players. FZ impressively rebuilt the bullpen, particularly the left side by adding Young and Alexander, using The Churn. On the position side, Austin Wynns was the only added player who accumulated more than 0.1 fWAR. Overall the average "churned" position player was worse than replacement.
As for why "churned" pitchers produced better than position players, it may have just been luck, but I think it's easier to tweak a pitcher's delivery or grip and see immediate results than swing adjustment with hitters who often need to accumulate some PA's to get their timing down.
This is admittedly data for just one season and a down one at that. It would be interesting to do the same study for all 4 years of FZ's tenure. My impression is while The Churn produced a handful of productive players, most notably YtY, it also gave a whole lot of cumulative playing time to sub-replacement level players. That impression would appear to be correct based on the above data.
Thanks for the painstaking research and interesting findings.
ReplyDeleteGreat work Doc!
ReplyDeleteYou also cannot calculate the angst and psychological impact on the teams clubhouse.
I mean new guy comes in you meet etc; then 2 days later he's gone.
As a regular; one would be saying we are all just a piece of meat to Farhan!
I could be over stating it, but if we the fans dont like it >> How must the players feel.
Anyway Doc again great work>> After checking the weather forecast >> your page is my next view in the morning.
Richard in Winnipeg
P.S. lets finally end the Belt Bromance!
Yes, the churn might be annoying to some, but as Doc pointed out, Farhan remade the left side of the bullpen on the fly. They stabilized the bullpen, other guys got into their roles and the bullpen was a lot better after that. You don't get that without the churn. As you know, whatever psychological effect, winning solves all things. When the team won 107 games, there wasn't much, if any, complaints about the churn.
DeleteAnd the Belt bromance has been one of the best relationships over the past decade. Maybe one more year is in the cards.
As with everything in life, I think there has to be a balance. Yes, the Giants should be active in the minor league free agent and waiver market. No, they should not grab everybody that comes available just because.....they are available.
DeleteAs for Belt/Longoria, to me it's a numbers game. If they are going to acquire all these younger, more athletic players they keep talking about, then some of the older, more fragile, less athletic players are going.....well, there just aren't an unlimited number of roster spots.
You know Doc that while Farhan is probably the most active PBO, he does not grab everyone just because they're available. Every one of his pickups I think has had a rationale behind it. Sure, sometimes it's trying to catch lightening in a bottle but at the end of the day, it's always about talent acquisition. And Farhan hasn't gotten the reputation he has over the years by just being random.
DeleteFZ has his types, that's for sure, but I have to say it's kind of fun to look at names of players DFA'd by other teams and say to myself, "wait!" FZ is going to grab that player! He can't help himself!" I see the potential in some of those players myself but then please, fix whatever you think can be fixed but then give them a large enough sample size to make a true judgment.
DeleteI believe many talent evaluators in baseball have their diamond-in-the rough candidates. They often don't have the opportunity or roster spot (in contention), or they don't believe they have enough time (games) to accurately evaluate any of them.
DeleteThought of a couple of players who went thru the churn were let go and had some success with another team. Remember Connor Joe, after 15 Abs with the Giants, he's gone on to hit 15 hrs 60 RBI 267 BA spanning 2 seasons and 600 abs. He started 2022 on fire but cooled off in the end. Nobody has mentioned (LHP) Jose Quintana ended 2021 with the Giants, was sent to AAA and the Pirates signed him, Cardinals traded for him and he started game 1 of their playoff series. Although I wonder if they regret not trading for Carlos Rodon instead. Did the Giants move on too quickly from Connor Joe and Jose Quintana?
DeleteAdd Tyler Anderson LHP to the list of ones who got away.
DeleteThis actually bodes well for Giants evaluators because other teams seem to always pick up those the churn left behind. So it's not just the Giants who see something in them. You can include Joe McCarthy and Tyler Anderson. There's only so many roster spots and if we kept Connor Joe and Jose Quintana, maybe you don't get LaMont Wade and Alex Cobb.
DeleteGreat work!
ReplyDeleteShould less than 10-12 ABs count? Depending on where the batter came from, is he adjusted to the ML's, especially to ML pitchers?
Villar looked awful in his first 63 AB with just 11 Hs and 28 Ks. Of course, Villar was not part of the Churn and he was coming to the ML's for the first time. He's included only to show how short experiences aren't necessarily predictive. Even when Villar came back for his 2nd bite he was 0-7 before his first hit. Proctor is an example of the Churn and his numbers aren't great, 2-18, but he won a game with a grand slam and caught 2 innings.
Although the schedule was relatively soft (did include series with LA, SD, and Atl) the Giants finished the season going 16-8 - wasn't that the end product of the Churn?
Wasn't FZ the producer of much of the 107-win season? Wasn't he right to let those guys, minus a few, plus a few, go after 2022? Didn't he have to try something - anything - to right the ship?
This has been the precursor: judge FZ and the Churn with the results of the 2023 team. This one is all his.
So, yes, 10-12 PA's do both count and add up to almost 500 PA's. That's almost 1 full player-season. The small samples are an integral part of the churn philosophy so they count when doing a cumulative analysis.
DeleteDoc, you're right to count those 10-12 PA's. The brute-force method of trying every possibility, or a close imitation of it, means the team can't give all those in the churn many PA's. If it were based on 'insights' by, say, Sabean, there would have been fewer candidates* and they could have gotten more PA's. Imagine if those 500 PA's had been given to two, three or four candidates whom you believed who were more likely to succeed and you put your reputation on the line for them. I think this is what you mean by 'the small samples are an integral part of the churn philosophy.
DeleteI remember one definition of elegance from way back was that with it, there are fewer waste motions.
Doc, Could you put up your ten bold predictions and anyone else's that were made before the season started? No problem if not interested.
ReplyDeleteHa! I want to bury those as far underground as I can get! How embarrassing!
DeleteThe left side of the bullpen was not remade on the fly. The innings pitched sample size is way too small to consider these guys performances repeatable and reliable for a whole season. It’s super easy for pitchers to do solid over 10-30 innings. All it takes is for a few bad games in a row for that ERA to jump from good to bad, and that’s exactly what always happens to these churn pitchers. The GM thinks he may have found a bullpen diamond in the rough based off of a small sample size, brings him back the next year for a bullpen role and they falter. If you actually want a good bullpen year after year, you need to acquire or develope some studs, not hedge your hope that these AAAA pitchers will carry your bullpen.
ReplyDeleteWhat do you think on the fly means? You can't get some studs on the fly in the middle of a season. They had issues on the left side and Farhan got guys who did their jobs. Are they world-beaters? Probably not. But they should be in the mix for jobs next year.
DeleteIn the National League, the Dodgers, Mets, and Braves won 100+ games and were out of reach of the Giants even had they had a good year.
ReplyDeleteCards (93 wins), Padres (89), and the Phillies (87) made the playoffs. Plus the Brewers (86 wins), those 4 teams weren't SO far ahead of the Giants that, had SF had a "good" year, they well could have finished ahead of all 4, at least 3.
And, of course, been a Playoff team.
Is there ONE STAT that distinguishes them? One stat that sunk the Giants?
How about total "unearned runs" given up for the season?
Cards 32
Padres 49
Phillies 55
Brewers 73
Giants 84
Hmm, that's the way they finished.
SF finished 3rd in the NL in most errors committed, the 6 Playoff teams finished it the top 8 (in fewest errors).
Like errors, Giants Fielding Pct was also 13th in the NL - it was 27th in MLB.
Is that worth addressing this off-season?
If it's a priority, for UZR, up-the-middle, YtY and Estrada are the best rated of the Giants and Slater isn't bad. Crawford is not well-rated. (See Fangraphs.)
For B-R's dWAR, those above 0.0 (50 or more games) are YtY (0.3), Bart (0.3), Wynns (0.3), Villar (0.2), and Crawford (0.1).
By the "eyeball test" errors absolutely killed the Giants this season. Coaches/GM needs to figure out how and why that happened and make sure it doesn't again.
DeleteAnother factor when it comes to defense is our preference for 'versatile' players.
DeleteWhen we say a player can play C, 1B, 2B, OF, for example, what do we mean by that? How good are they at all the positions those players are asked to play? "Jack of all trades, master of none," is one extreme. Could it become 'too many trades, average-or-above at none?"
Thank you again for the grinding work throughout the season as well as thoughtful analyses like the above. It is always my first read, providing a brief insulation from the less pleasant events of the day.
ReplyDeleteOn the Churn, what you say seems corroborated by the collective viscera. What's more, from a marketing standpoint, the Churn seems to exhaust fans at all levels and tends to make the game offputting. No matter how much somebody tries to keep up, turning the game on TV usually results in a quizzical "who's that guy?" I don't think it's good for the popularity of the game.
As to what they'll do, I'll predict (a) they DO sign Judge, (b) they DON'T re-sign Belt, (c) they DO re-sign Longoria, but for less than the option, and (d) they DON'T sign a shortstop, at least to play shortstop. Just hunches, of course.
Thanks for the kind words. I think your predictions are fairly solid. Hope you are right about Judge. It's not my money and he would be a great reason to watch games.
DeleteLess than option - is the equation like that: (buy out option) + (new contract) < (option to retain)?
DeleteThe first option (to buy out) is $5 million I think. The second option (to retain) is $13 million, I believe...or something like that. But the equation above is correct, I assume.
Rearranging, we get: (new contract) < (option to retain) - (option to buy out)
I have no statistical proof of this idea but just tossing it out: I think the Churn contributes to the Giants defensive woes. Defenses as a group work better if the defenders have experience working together regularly, being very well aware of each others' range, tendencies, etc. With frequent shifting, even more uncertainty is introduced defensively, on top of unfamiliarity with their teammates. I think back to the Giants stellar defenses during their championship years. There was no Churning then... Again, this is just a gut feeling of one who has observed the Giants many years through thick and thin and statistics may prove me to be totally off base.
ReplyDeleteThey entered the year having penciled going around the diamond: Belt, Estrada, Crawford, Longoria. Duggar would play a lot of CF, Yaz a lot of RF, Ruf in LF. Due to a spate of injuries they had to scramble. Belt was hurt, Crawford was not 100%, Longoria got injured. Turns out, Estrada is a slightly below average defender.
ReplyDeleteDuggar and Dubon were traded to try to fill roles. So then Joc had to play a lot more LF than was anticipated, although overall Z believes LF defense can be sacrified for offense. Luis Gonzalez was pretty bad in RF or wherever he played in the OF.
So yeah, defensive cohesiveness is preferred. But Giants D being worst in the league happened primarily because of injuries. Injuries led to more churn than wanted, i.e., the epitome being Donovan Walton being so bad at SS that they had to go out and trade for Dixon Machado.