Helps set a market. This got done quickly, sorta odd they couldn't get it sorted before season's end. The most important aspect I think is the 7 years, through age 37. Teams have generally shied away from extra long-term deals for pitchers.

I'll save the boring back-of-the-envelope math and say my Ohtani offer would something in the neighborhood of

2024: $30
2025: $30 + $5 + $750 K per start > 10 (max about $50)
2026: $55 (if 25+ starts in 2025) or repeat 2025
2027-33: team/vesting/player options (depending on 2025-26 pitching) maxing out around 7/$350.

Max healthy Ohtani would get about 10/$485. That might not get it done but either (a) that's the role of the player options or (b) happy to tweak incentives around 2025-26 or boost the later options for a CYA-style season.

I know Nola and Ohtani are hardly natural pitching comps but the key factor with Ohtani the pitcher is that he hasn't even reached 500 career innings yet. They've been 500 outstanding innings, far better quarlity than Nola, but pgragmatically speaking, even before the injury, it wasn't clear how many innings you could expect out of Ohtani. (I know, some of that low IP total is due to the Angels keeping him in a 6-man rotation most of the time.)

here's what i posted in one of the postseason chatters, when the topic of nola's contract came up:

501. a brief article regarding 57i66135 Posted: October 18, 2023 at 07:11 PM (#6144838)

Nola had kind of a "meh" year and there was certainly a vibe that maybe he had started to decline. But he puts up some great post -season starts which is the second long Phillies season and I think maybe it does allay some concerns about him. Is he worth 8/200? Idk but he's worth more to the Phillies than anyone else. One team guy, HOVG to HOF stats. He feels like a wainwright one team guy to me.

Anyway, pitchers don't get 8 year contracts, especially not the $25 M AAV types. Nola probably "deserves" to beat Wheeler's 5/$118 from 2020 quite easily but I doubt he'll do much better than that. You never know of course.

for 2 of the last 3 years, nola has been a below average innings eater.

i think he's likely to be better than that moving forward, but i sure as hell wouldn't bet 8 years on it. and i wouldn't be overly eager to give him a 5/125 contract, either.


nola is "only" 30. if i were him, i might look for a high paying 2-3 year deal, and then try to hit free agency again while i'm still in my "theoretical" "prime", maybe coming off a better season.

yeah...i don't love this.

@JonHeyman

Nola met with other teams and turned down more $ to remain in Philly. Teams love Nola for his dependability, but Philly players love Philly.

20 NL SPs qualified for the ERA title this year

ERA - Nola was 18th at 4.46, beating only Mikolas (4.78) and Corbin (5.20)

WHIP - Nola was 7th at 1.15, better than CYA winner Snell 1.19 and ROY runnerup Senga 1.22, among others

Nola's an interesting pitcher. His one amazing year, his ERA was much better than his FIP. Just about every other year, his ERA has been higher. The past three seasons, he's averaged an excellent 3.31 FIP, but only a 4.09 ERA for a 105 ERA+.

The Phillies have not ranked as a good defensive team so it's probably a combination of that and some bad luck thrown in.

So yeah, this does seem like a big contract for an average-ish innings eater, but there's at least some reason to believe Nola is more than that.

7 years, $172M

Heyman had projected 6 years, $168M
Expert 1 had projected 5 years, $175M
Expert 2 had projected 5 years, $140M

I'm surprised at the years, if not the AAV. My pet theory is that teams overvalue a pitcher's most recent season and especially their last postseason starts. Nola had a mediocre year (96 ERA+) and was particularly bad in September (5.47 ERA). And he had a couple of very good October starts, but the lasting image is a stinker at home in Game 6 in a chance to clinch the pennant.

The long years are to manage the cap. If you look at Howies numbers they come out at the same money, different years. Nola got his money and the Phillies spread it over 7 instead of 6 or 5. The only question is whether Nola feels like he's giving up more money for age 38-39. Maybe but not likely

Nola's an interesting pitcher. His one amazing year, his ERA was much better than his FIP. Just about every other year, his ERA has been higher. The past three seasons, he's averaged an excellent 3.31 FIP, but only a 4.09 ERA for a 105 ERA+.

The Phillies have not ranked as a good defensive team so it's probably a combination of that and some bad luck thrown in.

I would have been curious to see the results if Nola had signed with a good defensive team, but so it goes.

Philly players love Philly.

If you're good, they love ya; if you're bad, a lifetime supply of batteries.

That was a Heyman piece? Dare one suggest that "Expert 1" might have the initials "SB" and that maybe, just maybe, a few of his (or her!) predictions are a wee bit on the high side?

Now I won't suggest that "Expert 2" has the initials "SB's assistant" but only because those numbers seem ... well, not sensible but far enough below SB (err, E1) to be believable.

2024: $30
2025: $30 + $5 + $750 K per start > 10 (max about $50)
2026: $55 (if 25+ starts in 2025) or repeat 2025
2027-33: team/vesting/player options (depending on 2025-26 pitching) maxing out around 7/$350.

Max healthy Ohtani would get about 10/$485. That might not get it done but either (a) that's the role of the player options or (b) happy to tweak incentives around 2025-26 or boost the later options for a CYA-style season.

This is exactly the kind of contract I was thinking but would he sign it? Everyone is predicting that he gets $500 mil. guaranteed.

7 years, $172M

Heyman had projected 6 years, $168M
Expert 1 had projected 5 years, $175M
Expert 2 had projected 5 years, $140M

Adding to this:

FG - 6 yr/$150M
MLBTR - 6 yr/$150M

FWIW, there's a huge discrepancy between FG and BBRef on Nola's value.

2021 4.4 fWAR, 2.5 rWAR
2022 6.3 fWAR, 5.9 rWAR
2023 3.9 fWAR, 2.1 rWAR
Avg: 4.9 fWAR, 3.5 rWAR

Philly's defense, particularly in the OF/1B, has been pretty questionable in the past two years. I know both systems attempt to adjust for defense but for some reason I trust a FIPS approach more in the case of extremely bad defense.

Per Scott Lauber, the Braves offered 6/$162M and the Dodgers offered $165M.

I trust a FIPS approach more in the case of extremely bad defense.

No idea. But the Phils' had a low BABIP allowed in 2023 (291 vs 299 lg). But they were 10 points worse in 2022 and 8 worse in 2021.

By DRS they were -23. -30 and -50 runs. Statcast had them at +1, -29 and -18 runs. Given the closeness of 2022 fWAR v bWAR and DRS v statcast, I'll WAG that 10 points of BABIP and 30 runs is about the same and the 3 systems are in pretty good agreement in 2022. DRS clearly disagrees a lot with BABIP and statcast in 2023 and it seems quite extreme in 2021 as well.

But even at -50, assuming Nola has pitched about 15% of all Phils' innings, that should be only 3 runs or so below average. For the last 3 years, the big difference seems to be FG's "LOB wins" aka "sequencing" (also reliever effectiveness?). Sequencing I have always understood as, roughly, BB then HR vs HR then BB has a big effect on RA9 but probably usually isn't the pitcher's fault. Relievers stranding (or not) runners is obvious but I'm not sure if that's in here or not.

bWAR and fWAR take a completely different approach to how they handle pitchers’ defensive support, as far as I remember. Aaron Nola was a 3.31 FIP pitcher from 2021-2023, so fWAR treats him like a 3.31 RA/9 pitcher (there’s probably some adjustment for UER as well, so it’s a little higher). bWAR says that he was a 4.27 RA/9 pitcher, then makes an adjustment of 0.29 R/9 because the Phillies had a bad team defense in general (they use the same adjustment for every Phillies pitcher who pitched in those seasons; it’s not specific to the innings Nola threw), so it treats him like a 3.98 RA/9 pitcher.

193 IP per season with a difference of 0.67 R/9 (3.98-3.31) equates to a difference of 14 runs per season, which matches the 1.4 difference in WAR almost perfectly.

The Baseball Reference approach (assuming every pitcher on the team received identical defensive support) has always struck me as nuts, especially when looking at a single season. I don’t completely buy into the Fangraphs approach either, but for projecting future performance it’s almost certainly better.

The Baseball Reference approach (assuming every pitcher on the team received identical defensive support)

But fangraphs/FIP does exactly the same thing, roughly assuming every pitcher in baseball gets the same defensive support, at least on all their BIPs. (Obviously the proportion of all PAs that are in-play varies by pitcher.)

It's also not quite true to say that bWAR treats every pitcher on a team the same, I assume for the same reason that they don't all rely on defense the same amount. Reliever Jose Alvarado has a RA9def of just -0.16 while starter Bailey Falter is at -0.26 ... and poor Josh Harrison at -0.59. :-) The Phils' main starters all come out about the same but they all have fairly simillar IP rates around 2/3. Alvarado's IP rate is 50%, Falter's was 75% and poor Josh Harrison was 87%.

But also, the way fangraphs presents the data, it is NOT FIP. They give the RA9-war for Nola which are reasonably close to the total bWAR numbers: These are then adjusted for two things: BIP-wins and LOB-wins. The BIP-wins adjustmets are fairly small -- he picks up 0.7 wins in 2021, loses 0.1 wins in 2022 and loses 0.6 wins in 2023. The LOB-wins adjustments are rather large -- he picks up 1.8 wins in 2021, 0.9 wins in 2022 and 2.1 wins in 2023. Totaled over three years, the BIP-wins adjustment is zero and the LOB-wins adjustment is 4.8 wins.

From fangraphs:

BIP (balls in play) -Wins is an estimate of how many wins a pitcher has added by having a BABIP above or below league average. It adjusts for park and league.

LOB (left on base) -Wins is an estimate of how many wins a pitcher has added as a result of stranding runners on base. This is just a general name for a component that involves different aspects of a pitcher’s game with respect to baserunners, for example: controlling the running game, pitching differently with runners on, or pitching out of the stretch.

So the upshot seems to be that, on average, a pitcher "strands" 72% of his baserunners (see LOB% in the fangraphs glossary for the formula) but in 2 of the last 3 years, Nola has been around 66%. He is "rewarded" for this in fWAR, fangraphs claiming this will usually regress over a career and, in fact, for his career he is at 73%. He allows give or take 10 runners per 9 so that's about 0.5 runs/9 in those down years. (In the other year he was average but he still gets a plus, I don't know why. Similarly 0.5 R/9 is about 10 runs per full season not 15-20 that he's been getting credit for -- take it up with fg. :-)

Hi Walt: sorry to interrupt your very interesting discussion:

But even at -50, assuming Nola has pitched about 15% of all Phils' innings, that should be only 3 runs or so below average.

isnt this 7.5 runs? I know you are good at math but perhaps an error here. Does that help to make any more sense out of it?

WAG that 10 points of BABIP and 30 runs is about the same and the 3 systems are in pretty good agreement in 2022. DRS clearly disagrees a lot with BABIP and statcast in 2023 and it seems quite extreme in 2021 as well.

Just to be clear DRS is not measuring the same thing. There's a lot of little things that go into defense and DRS is at least trying to measure them with its Good Fielding Plays measure which I think is based on people making observations. Statcast for OFers is measuring both range and assists (and maybe holding runners). But for infielders it is only measuring range on GBs, and not baserunner kills on other bases or double plays, nor errant throws.

I dont think Good Fielding Plays measures all of that either, but at least its trying to add something to range per se. You probably know all this but just pointing it out.

Again great discussion. I hope to study this a bit more and be able to make some intelligent comments. Very interesting.

The Baseball Reference approach (assuming every pitcher on the team received identical defensive support)

But fangraphs/FIP does exactly the same thing, roughly assuming every pitcher in baseball gets the same defensive support, at least on all their BIPs. (Obviously the proportion of all PAs that are in-play varies by pitcher.)

My comment probably wasn't clear, sorry. BB-Ref takes the actual runs allowed by the pitcher, assumes every pitcher on the team received the same quality of defensive support and corrects for that when calculating WAR. Fangraphs takes the opposite approach; it starts from FIP when calculating WAR, so it doesn't even need to take into account the number of actual runs allowed by the pitcher or the defensive support a pitcher received.

It's also not quite true to say that bWAR treats every pitcher on a team the same, I assume for the same reason that they don't all rely on defense the same amount. Reliever Jose Alvarado has a RA9def of just -0.16 while starter Bailey Falter is at -0.26 ... and poor Josh Harrison at -0.59. :-) The Phils' main starters all come out about the same but they all have fairly simillar IP rates around 2/3. Alvarado's IP rate is 50%, Falter's was 75% and poor Josh Harrison was 87%.

This is a good point -- it treats all pitchers on a team as though they received the same defensive support per BIP. Pitchers who allow a greater percentage of BIP are going to receive a more extreme defensive adjustment, either positive or negative, on a per inning basis.

Aaron Nola in 2018 is actually my textbook case of why it doesn't make sense to calculate value this way. I was trying to figure out why Nola, with an ERA+ of 173 in 212 IP, had more pitching bWAR than Jacob DeGrom, who had an ERA+ of 218 in 217 IP. (DeGrom also had a significant lead in FIP, 1.98 to 3.01.)

But the Phillies' defense was bad that year, ranked last in the league with -82 Rdrs (51 runs behind the next-worst team, DeGrom's Mets). They allowed a .306 BABIP compared to a league average of .295. So I guess it makes sense that bWAR gives Nola .41 RA/9 of extra credit for pitching in front of such a lousy defense, right? That amounts to an extra 10 runs of value or about 1 WAR.

However, unlike the rest of the Phillies' staff, Nola actually had a BABIP much lower than league average -- .254, more than 50 points better than the Phillies' team average. Maybe he achieved that despite getting really terrible defensive support from his teammates, or maybe he got better support than some of the other guys on his team. We know that offensive run support and bullpen support can vary from pitcher-to-pitcher on the same team, so why not defensive support as well?

(It's worth noting that DeGrom also got 4 runs of credit for pitching in front of a bad defense even though his BABIP was also below league average -- .283 compared to .295. And Nola gave up 6 fewer UER that season, so they were closer than you would think just based on ERA+. But still, without the RA9def adjustment, DeGrom would have had more pitching bWAR than Nola.)

But also, the way fangraphs presents the data, it is NOT FIP. They give the RA9-war for Nola which are reasonably close to the total bWAR numbers: These are then adjusted for two things: BIP-wins and LOB-wins. The BIP-wins adjustmets are fairly small -- he picks up 0.7 wins in 2021, loses 0.1 wins in 2022 and loses 0.6 wins in 2023. The LOB-wins adjustments are rather large -- he picks up 1.8 wins in 2021, 0.9 wins in 2022 and 2.1 wins in 2023. Totaled over three years, the BIP-wins adjustment is zero and the LOB-wins adjustment is 4.8 wins.

I think you may be misreading Fangraphs (not your fault, they don't explain it well). On their pitching WAR explainer, they are pretty clear that they base it on FIP:

Calculating WAR for pitchers is conceptually straightforward, but there are many steps and a lot of notation to follow. Generally speaking, the first thing you need is some estimate of your pitcher’s value relative to league average. There are all sorts of different approaches to selecting this number. FanGraphs uses Fielding Independent Pitching (FIP), with a few adjustments, but you could use RA9, DRA, or any other metric related to pitcher performance. This post does not address the merits of choosing FIP for use in WAR. (emphasis added)

(The adjustments have to do with things like how they handle infield fly balls, park effects, etc. - not stranded runners).

The way I read their pitcher pages is as follows:

RA9-WAR - self explanatory. It's the pitcher's value using his actual RA/9 allowed that season.

WAR - the pitcher's value using his FIP, rather than his actual RA/9 allowed.

BIP-Wins - the number of wins a player got, relative to his FIP, by allowing a lower BIP than league average. This is part of the difference between RA9-WAR and WAR.

LOB-Wins - a simple plug to account for the rest of the difference between RA9-WAR and WAR, such that RA9-Wins = WAR + BIP-Wins + LOB-Wins.

(They show that equation at this link, although they combine BIP-Wins and LOB-Wins together into one number called FDP-Wins (Fielding Dependent Pitching).)

In other words, WAR isn't a calculation that is derived from RA9-Wins and those two adjustments. Rather, WAR, RA9-WAR, and BIP-Wins are calculated independently (with WAR being based on FIP) and then LOB-Wins is derived from the other three to keep the equation in balance.

I'm speculating a little bit here that it's a simple plug, but unlike many of their other metrics where they give detailed explanations of how they're calculated, Fangraphs give no such calculation for LOB-Wins. And if you pick any pitcher and check the numbers, that equation always perfectly balances each year (give or take 0.1 for rounding).

I'm always annoyed when I find that a site like Fangraphs has a number like this, where instead of admitting that it's just a catch-all for "stuff we don't have an explanation for", they try to pretend that it's actually measuring something like "controlling the running game, pitching differently with runners on, or pitching out of the stretch."

This is a great discussion and can we get a summary of sort of where we're at here?

Like after reading through the Inge post, are we back to square one with fWAR? I.e. is the difference between fWAR and bWAR for pitchers coming back to fWAR starts with FIP? Or is there something in there that Im missing?

#22 That's the essence of Bill James' complaints about the slope correctors in Linear Weights.

That’s funny, because James did pretty much the same thing to get his player Win Shares to add up to the total number of team wins (there could be a big discrepancy if the team had over or underperformed their Pythagorean win total). James admitted that he didn’t know what caused the difference, but he still had a formula for assigning that value to individual players, which I always had an issue with.

Ron can you elaborate? is that an aside or is that relevant to FG pitching WAR?

#26, Pete Palmer claimed a really high level of accuracy for Linear Weights. Much better than anything that James could get with Runs Created.

So he checked. What it boils down to is a method that is actually less accurate (at the team level -- Linear Weights will always be better for individual players than any multiplicative method) and then a fudge factor that's generated each year to make things work. Or what amounts to a separate formula each year. As James pointed out, he could cut the standard error for runs created in any given year with a little tinkering.

And James' complaint about Palmer is essentially Dave's about FG. It happens to give desired results, but that's weak. Where are the studies that show what FG is including actually measures something.

Takes me back to the days of rec.sport.baseball. We'd see all sorts of formulas for various parts of the game and the response was always "show your work" or "I checked and it doesn't hold up"

With Fangraphs it’s not actually an issue because LOB-Wins isn’t actually used for anything. But they appear to be misrepresenting what the stats measures.

With James and Win Shares, it’s frustrating because after a lot of effort to be precise in the various components, he introduces a relatively big source of potential error at the end.

Yes. Endless topic on RSB back in the day.