Chris Sale for MVP
Aaron Judge’s monster first half made him an obvious MVP candidate, even as he’s slowed down in the second half of the year. Jose Altuve leads qualified hitters in wRC+, as a second baseman who also happens to steal a bunch of bases, so he’s an obvious MVP candidate. Mike Trout is within +1 WAR of both of them despite spending two months on the disabled list, and on a per-game impact, is again obviously making the biggest impact of anyone on the planet, so he’s a less-obvious MVP candidate, but he should be in the mix by season’s end.
But if the voting were held today, there would be a pretty clear choice for the American League’s Most Valuable Player, and it would be Chris Sale.
Sale pitched again last night, facing the Yankees on Sunday Night Baseball. It was a typical Chris Sale in 2017 start, as he threw seven innings, allowed one run, and struck out 12. It was his seventh start this year with 12 or more strikeouts; he has nine more with 10 or 11 strikeouts, giving him 16 double-digit strikeout games this year. He’s made 24 starts.
Sale’s 37% strikeout rate is the third-best any starter has ever posted in a season with at least 160 innings pitched, behind only 1999 Pedro Martinez and 2001 Randy Johnson. That ’99 Pedro season is maybe the single best season any pitcher has ever had. Johnson’s 2001 season was maybe the best year of the best left-handed pitcher in baseball history.
Sale already has racked up 241 strikeouts, and he has seven or eight starts left in the regular season, so getting to 300 is simply a matter of staying healthy at this point. There have only been 34 seasons in history where a starter struck out 300 or more batters, and most of those came in the days when starters threw 300+ innings.
So, yeah, Sale is having himself some kind of strikeout season. Of course, you don’t win an MVP just for having a high strikeout rate. But Sale’s overall numbers make it clear that he’s been the best player in the American League this year.
By FIP-based WAR, Sale’s +7.4 WAR is the best in baseball, and it’s not close. By runs allowed, he’s at +6.7 WAR, so he’s less far ahead of Altuve, Judge, and Trout, but still has some distance between them. And despite the persisting belief that ERA or runs allowed measure “what the pitcher actually did”, as usual, crediting the pitcher for everything that happens while he’s on the mound obscures the truth.
Yes, Sale’s 2.51 ERA is a lot higher than his 1.92 FIP, and yes, sometimes FIP overestiamtes the value of pitchers who dominate the strike zone but give up hard contact on balls in play. By ignoring all batted ball results besides home runs, FIP can give a misleading picture of a pitcher who throws meatballs in the zone that just don’t happen to be hit for home runs.
But that’s not Chris Sale. His BABIP is just .286, better than league average mark. And that’s not just good defensive support masking a contact problem; his average exit velocity allowed of 86.2 mph is also well below the league average. He’s allowed a .235 wOBA this year; by MLB’s xwOBA calculation using Statcast data, his expected wOBA is .233.
FIP isn’t missing Sale giving up line drives all over the place here. When you add batted ball data into the mix, Sale doesn’t get any worse. So why is his ERA over half a run higher than his FIP?
| Season | Situational | BB% | K% | HR/9 | BABIP | FIP | xFIP |
|---|---|---|---|---|---|---|---|
| 2017 | Bases Empty | 5% | 34% | 0.57 | 0.273 | 1.99 | 3.05 |
| 2017 | Men on Base | 5% | 42% | 0.93 | 0.315 | 1.79 | 1.80 |
| 2017 | Men In Scoring | 5% | 47% | 0.68 | 0.400 | 0.92 | 0.95 |
With runners in scoring position, Sale’s allowed a .400 BABIP this year, so he’s stranding a lower percentage of his runners than you’d expect for a guy who has been this dominant. But there’s nothing in the data to suggest that the .400 BABIP he’s allowed in those situations is the result of bad pitching. His average exit velocity with RISP is actually lower (85.0 mph) than his overall average EV allowed. His xwOBA with RISP is .219, second-best among MLB starters, quite a bit lower than the .246 wOBA he’s allowed in those situations. And, of course, the sub-1.00 FIP and xFIP in RISP situations shows that he’s not really crumbling in high-pressure situations. When men are in scoring position, he’s been his most dominant self.
Defensive support isn’t static, and sometimes, your defenders fail to convert outs more often when you need them most. That appears to be what has happened with Sale this year, as balls have fallen in at the least opportune time, despite no evidence to support the notion that he’s pitched worse in those situations. And that’s not something you want to hold against Sale, as the performance of his teammates is beyond his control.
Of course, that all goes to explain why Sale is the easy choice for AL MVP right now. Even if we judged by runs allowed, ignoring everything in the paragraphs above, he’d still be the best player in the American League. But we shouldn’t judge Sale by runs allowed, and by the measures he’s able to control, his lead is even larger at this point.
There’s six weeks for things to change. The way Mike Trout is playing, we might end up having a really interesting debate about how many games a position player needs to participate in to be considered for the award. Maybe Sale gets bombed in his next few starts and comes back to the pack a bit. The race isn’t over yet.
But right now, Chris Sale is winning, and winning rather handily. Forget the notion that pitchers have their own award. Sale is a player, and the award is for the Most Valuable Player. To this point, in the American League, that’s been Chris Sale.
Dave is the Managing Editor of FanGraphs.
It would be odd to say Mike Trout didn’t play enough games so instead let’s give the award to a starting pitcher.
Except that playing time matters. That’s why WAR is a counting stat and not a rate stat. Sale provided more value than Trout has thus far because he’s been producing excellence while Trout was recovering.
Trout’s not on the leaderboards because they have a filter, but Cameron’s argument isn’t that he missed an arbitrary cutoff, so therefore he shouldn’t be considered. That Trout is going to be right near the top despite the injury is an achievement in and of itself.
It’s quite disappointing that Trout got hurt this year. Given his insane slash line before *and* after the injury, he could have had a season for the ages.
You missed my point. I’m not saying Trout deserves some extra credit because he missed time and has been so good otherwise. I’m saying that just arbitrarily saying “he didn’t play enough games,” if you believe he has been the most valuable player, and instead giving the award to a guy who plays every fifth day, would be an extremely weird way of going about the process.
If Sale and Trout are very similarly valuable at the end of the day, by whatever normal metrics you want to use, how many games Trout missed shouldn’t matter, just as the fact that Sale is a starting pitcher shouldn’t matter.
(ETA: The rest of this is not addressed to you, but to the comments below)
As far as how many “plays” you are a part of, that seems just as silly as counting games. We have metrics that show the value of the outcomes of your plays. Why don’t we use those?
Games missed do matter since he is providing zero value on the bench and WAR is a counting stat. MVP is not best player its most valuable (in terms of helping your team win actual and not theoretical games)
The leader in PA last year was Springer at 744. The leader in Total Batters faced by a pitcher last year was Price at 951. The top individual pitchers impact the outcome of more plate appearances than the top hitters do. Maybe it’s time to rethink your logic.
Fielding matters.
This year, Sale has faced 654 batters. The leader in plate appearances is Inciarte at 530.
Sale: 654 batters faced, 23 fielding plays, 3 PA’s (as a batter)
Trout: 329 PA’s, 169 fielding plays
Consider that Sale has ended 241 at bats via strikeout. Those may not be ‘fielding plays,’ but they’re the outs he’s personally responsible for.
Stop trying to sell the same old notion that pitchers aren’t as valuable as fielders just because they pitch once every five days. In the context of ‘most valuable,’ consider the fact that the Red Sox would very likely be challenging for last place without him. Could the same be said for Altuve and the Astros this season? The Astro’s recent skid suggests he can’t carry a team when he plays like a pitcher can once every five days.
I’m not trying to sell anything. I just cited some relevant stats.
I think the stats support the fact that starting pitchers are just as valuable as position players.
Sorry, I read that as support for fielder > pitcher. That was obviously me projecting that argument on to you.
It appears like you did not understand TKDC’s point. He’s pointing out, correctly, that it doesn’t make any sense to disqualify Trout for missing games and then give the award to a pitcher, who by definition will miss far more games. He’s not saying that Sale can’t be deserving, he’s saying that the games or lack thereof shouldn’t disqualify Trout or Sale. At the end of the season, give the award to the guy who was the most valuable period.
Does anyone know of any examples of a pitcher winning despite the presence of good position player candidates?
Ok I just answered my own question. In 2014, Kershaw beat Andrew McCutchen who had a classic .300/.400/.500 season. In 2011 Verlander beat Ellsbury’s 30/30 year and Bautista’s 1.000+ OPS.
What did Kershaw and Verlander have in common those years? They had outstanding W/L records, with more than 20 wins and no more than 5 losses.
Hate to say it, but Sale’s MVP chances may come down to whether he can get a W in most of his remaining starts, and whether the Red Sox can hold onto the division.
Those are the two stupidest considerations I can think of for MVP voting, but you’re right that voters would probably look at them. After this weekend’s Series vs the Yankees, the Sox playoff chances are up to 99.3%. Predicting pitcher W-L record is a fool’s errand, but both Zips and Steamer have him finishing at 19-6.
Brilliant seasons-worth of play by elite players and the MVP to award the *best* could come down to W-L records.
I am triggered.
If I recall correctly, Kimbrel has 4 blown saves this season. And Sale was the starter in 3 of those. Sale has also gotten very poor run support. It feels like every pitcher he faces turns into Chris Sale. He should easily have another 5 wins, but that’s what you get with a simplistic statistic like wins.
Felix won the Cy Young with a sub-.500 record a few years ago (I believe?) so I think the voters are starting to wise-up to the uselessness of this statistic.
Voters have gotten better at looking past win totals, but if someone wins 20 games it gives them a big boost. Just look at last year’s voting. Sale, Porcello and Verlander all had 5.2 WAR. Porcello won with 137 points, Verlander second with 132, Sale fifth (?) with 40 (?!!?!?) points.
I think it is safe to say that Porcello got a big boost for winning 20+ games.
Porcello got a big boost playing for Boston (large Eastern market) and being on a division winner. Worst choice for Cy Young in a long time. Verlander should have won IMO
I honestly think the two awards have different standards now. For the Cy Young, the writer is thinking “how well did he pitch?”, a question most people realize is tangential to W/L. For MVP, he is thinking “how valuable was he?”, something most of them find to be inseparable from team results. From a narrative perspective, it makes the most sense, even though it is null analytically.
This is correct. Kimbrel has blown 4 saves, 3 were on C. Sale starts (2017-04-20, 2017-07-15, and 2017-08-01). The other was a Porcello start (2017-07-03).
However, Sale was not in line for the win on the 2017-08-01 start, Blaine Boyer was in line for that one.
Why don’t people start thinking of FIP-based WAR as “what the pitcher actually did”? It’s not as if someone else gave up those strikeouts, walks, and HRs.
I think because most recognize that FIP isn’t perfect, as pitchers can and do out/underperform on a long term basis.
It is also a bit in conflict with how we evaluate hitters. We don’t adjust the hitter’s stats based on what they ‘should have been’, depending on the defense they were facing or their luck on BABIP.
But this is the disagreement. FIP isn’t not ‘what should have happened’. It is equally as descriptive as ERA, its just describing something different than earned runs. The fact that it can be predictive doesn’t change the fact that it a perfectly descriptive measure of K,BB,HR.
Even something like xwOBA, which you could argue is trying to show what should have been by trying to scale results closer to wOBA is still essentially a descriptive statistic of exit velocity and launch angle.
‘What should have been’ would be taking something where a pitcher tries pitching through an injury for a month, loses a tick off his fastball and you just say, I am going to throw that data out because clearly he was injured. It may be useful analytically, but it is also discarding actual events on the field. But anything based on data from on field events is descriptive of something
K, BB, and HR are perfect descriptors. FIP is an amalgamation of them put on a scale to resemble ERA.
FIP is excellent, but it isn’t as perfect a descriptor as the raw data itself. So while ERA is less useful, and it includes the subjective idea of an “error”, what it is actually measuring is much more concrete.
And I think that is what people get hung up on with FIP. RA and ERA include factors outside the pitcher’s control, but the actual number is a representation of “what really happened”. While FIP is more useful, it is incomplete as a measurement of pitching ability, and the actual number itself doesn’t “mean” anything. The units are much more abstract.
It being more abstract may make people more hesitant to accept FIP, its understandable but its also wrong. Neither FIP nor ERA is a perfect representation of a pitchers performance, that’s why neither should be the sole focus of any analysis.
But FIP is still what happened. It’s like wOBA. wOBA is on a scale similar to OBP, but no one says wOBA is a measure of what should have happened. It just takes singles, doubles, triples and homers, weighs them and puts them in one number. It’s still a completely descriptive measure, albeit one that includes multiple inputs instead of a simple, single input like hits or runs allowed.
I’m not saying its wrong to put more stock in runs allowed than FIP (I clearly fall on the FIP side of this debate if we must reduce it an either/or), but I think the criticism of FIP that its trying to show what should have been vs. what actually happened is wrong. They’re just two different types of results.
I feel like wOBA is more analagous to ERA (treating a broken bat single as better than a HR that gets robbed) and xWOBA would be more like FIP (where you get credited ONLY for what you control, hitting the ball hard and at a good angle).
That’s fair. I think for all of these stats there is a very legitimate debate about the distribution of individual credit for team events no matter what you’re metric of choice is. I want to see a few more years of statcast data, but as a common place to begin analysis xwOBA/xwOBA allowed are going to be the go to’s before too long
I mostly agree. I see both sides. It depends on perspective I think. I think you have the (not incorrect) perspective that someone’s FIP is his FIP. It is it’s own metric, and it is representative of exactly what happened. This is true. Those BBs, Ks, and HRs are exactly what happened.
But FIP is also an estimate of a pitcher’s skill at preventing runs. It is, by design, an approximation of ERA that excludes balls in play. In a sense, an estimate of “what should have been” is exactly what it is. I don’t think it’s wrong to have either perspective.
It only shows what happened in 50% of a pitchers opp PA. Assuming that BIP are beyond a pitchers control, an assumption proven wrong long ago.
ERA is flawed because it excludes runs that are UER, but thats a relatively minor flaw, and of course a teams defense and park can influence ERA
That is pretty far from the only flaw in ERA
Proven wrong? That assumption has been mostly proven right. Or at least closer to right than what was available at the time. I agree that we can do a little better now with more batted ball data available.
for the same reason people don’t give out awards based on Pythagorean wins or base runs. In the end it’s what resulted in terms of runs and wins. If a player hits 60 HRs just barely over the fence over the shortest parts of the fences he may have been extremely lucky but he still hit 60 homers, even if he will never be able to do anything close to that ever again. On the flip side while Sale low strand percentage isn’t indicative of his ability to pitch those runs still scored and his team still lost.
In the end it’s what people think MVP is, best performance or best results? Sale winning would be more on the former than the latter.
What I’m trying to say is that there are opponents of FIP who place little to no weight on it because they say it’s not based on what “actually happened”. I think this view is flawed. FIP still provides you with a measurement based on what “actually happened”, it’s just that this measurement is based on strikeouts, walks, and HRs, not earned runs. It’s scaled to ERA so it’s easier to interpret (similar to how wOBA is scaled to OBP).
If discussing someone hitting an abnormally high number of HR due to luck in a season, the number if 50, not 60. Brady Anderson only hit 50 HR in 1996.
Good analogy, baubo. FIP doesn’t account for good fortune on balls in play or home runs per fly ball.
Best response I’ve seen so far.
Baseball is a game that is determined by actual runs and not estimated runs on some theoretical team
I noticed that there was just over a run difference between xFip and Fip with the bases empty, doesn’t that point to at least slight regression?
If the angels make playoffs…
“If the Angels win the pennant…then we can be a family again…”
WPA makes a pretty good case for Sale as well:
Sale: 3.83 WPA, tops among all pitchers
Nelson Cruz: 3.50 WPA, tops among AL hitters
Altuve: 3.08 WPA
Judge: 1.61 WPA
Trout: 5.07 WPA, tops among AL players (incl. non-qualified)
…woah
I really hope Trout gets to qualified status by the end of the year so that no voters go looking at leaderboards, see his name not near the top, and assume he doesn’t measure well.
Fangraphs currently has Sale projected to finish with 319 strikeouts. Over the last 50 years, the only pitchers with more strikeouts than that in a season were Nolan Ryan and Randy Johnson.
Ryan and Johnson each did it 5x. I’m suddenly less impressed with Sale’s numbers than by the singularity of those other two.
Obligatory mention here that Unit would probably own the single-season K record if his final start in 2001 wasn’t scrubbed. He was supposed to start the final day of the regular season against the Brewers who had just set the team single-season K record. But the game meant nothing, not even home field, so Johnson was scrubbed. He finished with 372, just 11 short of Ryan’s 383 and Johnson basically averaged 10.6 K/start that year.
(I say basically because he made 1 relief appearance – a suspended game where a transformer in the lights blew in the 2nd inning and they resumed the game the next day. The Dbacks went ahead and just brought in Johnson to finish the suspended game since the next day was his regular scheduled start. The Padres had loaded the lineup with lefties to face Schilling so Johnson had some easy pickings. He struck out 16 in relief to set a MLB record. I’ll count that as basically a start.)
As long as pitchers are eligible, Sale should win provided something weird doesn’t happen. Personally, I would prefer MVP to be for position players.
To what extent should we discount Sale’s historic strikeout rate due to the “age of the strikeout” we’re currently in?
His 36.9% K% is eye-popping, but Kluber and Scherzer are right on his heels.
Is Chris Sale nasty? Yes. But strikeouts are at an all-time high, 21.6% according to Fangraphs. In 1999 they were 16.4% and in 2001 they were 17.3%. So let’s keep things in perspective.
Agreed. Really important.
Right now we have 3 qualified starters at 12+ K/9, 2 @ 11+ K/9, 7 @ 10+, and 11 at 9+. That’s a total of 23 qualified starters with at least one K per inning pitched.
In 2001 (Randy Johnson insane year) there was Johnson at 13.4 K/9, Kerry Wood at 11.2, Schilling and Nomo right above 10, and then no one until Clemens at 8.7. That’s a total of 4 qualified starters that averaged more than a K/inning.
Oh … and in 1999 there were just 2 pitchers that were 9+ K/9 … Johnson (12.1) and Martinez (13.2). Sterling Hitchcock and Chuck Finley were next closest with 8.5 and 8.4 respectively.
This isn’t to take anything away from Sale. It also has nothing to do with voting for MVP. However, it is really important when we talk about the historical context of this season.
I know this ground has been covered, but I think that as important as it is to adjust, it’s equally important not to over-adjust. Strikeouts are up across baseball, but so is velocity. And it’s been shown before that hitters don’t adjust to higher velocity. In other words, hitters are limited by human reaction time and more velocity means less reaction time.
If there were no underlying changes across baseball between now and 2001 other than the observed velocity changes, then we would expect league K% to increase a good deal, due to velocity alone. In this scenario, we should not adjust Sale’s K% rate at all; rather, we should just acknowledge there are even more good, flame-throwing pitchers than there used to be.
Of course, there are other changes too, some of which probably have also contributed to higher K rates, but the takeaway is: don’t just blindly adjust for league K rate. You need to figure out which changes are because of changes to the pitcher population and which are due to changes in hitter tendencies.
This doesn’t make any sense. Why would you need to figure out which changes are from pitching population and which are from hitters? Both are applicable to adjusting for different eras. If velocity has increased for the general pitching population, then that makes Sale’s velocity, and his K’s, less impressive compared to a pitcher from a different era. If Sale were pitching in the 80s, maybe he’d be averaging 91 instead of 94 because of the differences in bullpens, training, etc.
The point is that you shouldn’t penalize a player just because there are simply more good players. A true era-to-era comparison would take the baseline skill of players in the league into account, and potentially compare a worse “relative” player favorably to earlier players if the league as a whole has improved.
You’ve illustrated another reason why it is hard to do, given that we don’t even know how much the velocity boost has come from training and how much from better player selection and identification. In the end, if Sale was pitching 20 years earlier, I don’t know what the default assumption should be: would he have a similar K rate to now (like Pedro and Johnson did, illustrating it is possible), or would he have a lower K rate, in line with the relative league K rate? Neither feels like a safe assumption to me.
I disagree, I’d say that’s a fundamental aspect of comparing different eras, that you are taking into account how good relative players are (also, I don’t know that lots of Ks and more velocity always means “better”, it’s not like it’s a pitching-dominant league at the moment). That’s the whole point of replacement level being variable season to season, you’re comparing players the general skill level.
Indeed. I always think of this chart whenever I see talk of current pitchers’ K rates. Context matters.
Only slightly on topic but am I the only one that basically just averages bWAR and fWAR for pitchers? I know it’s far from perfect but usually leaves me feeling pretty good about the result.
I use the 50% FIP based, 50% RA9-WAR on the combined WAR leaderboard here. RA9-WAR and rWAR (aka bWAR) should be close as both are based on run prevention.
Using this method, Sale is still about a win better than Altuve this season.
Very little would make me happier than seeing Trout win the MVP on the basis of the Angels making the playoffs. It’d be nice to get one of those early career MVPs back that he deserved and didn’t get, and it would be amazing for it to happen because of an old-school, traditional media reason. It’d feel like justice, for Trout, if not for Sale.
The last 3 starting pitchers to win the MVP went:
– 21-3 in a year where nobody hit over .319, hit more than 37 HR or drove in more than 116 RBI (Kershaw in 2014)
– 24-5 (Verlander in 2011)
– 24-4 (Clemens in 1986)
Sale will be lucky to get to 20 wins.
Meanwhile:
– Altuve is going to hit over .350 with 220 hits and 110 runs (plus a handy 36 SB) on a 100-win team
– Judge is going to hit 45 HR with 110 runs and 100+ RBI
– Trout’s counting stats are going to come up a little short but if the Angels make the playoffs, Trout will draw a ton of votes because his slash stats are outrageous.
I know the voting bloc has become more saber-friendly, but I don’t see Sale winning the MVP unless those 3 fall apart ans Sale gets over 20 wins. To me, it’s Altuve’s to lose right now.
The NL is actually more interesting. With Harper going down, Goldschmidt is probably going to put up the best stats and for a likely playoff team, but how do you not give the MVP to someone from a 110 win team?
Don’t count out Rendon.
Maybe. I just think he comes up a little short in the traditional stats. ZIPS right now has Goldy and Rendon finishing:
– Goldschmidt: .313/.428/.584 with 37 HR, 118 R, 121 RBI and 6.7 WAR
– Rendon: .293/.393/.531 with 28 HR, 85 R, 99 RBI and 6.3 WAR
I can’t see Rendon beating out Goldschmidt in the voting. There is no doubt that the sabermetric movement is causing a lot more people to look beyond the traditional stats but I think they still hold significant sway. For example, Trout’s 2 MVPs came in the only 2 years that he topped 100 RBI.
It is hard to give the MVP to one player when there are five legitimate contenders with no standout on a single team.
Hasn’t Kluber been almost as good as Sale but for the injury? Sale has a small lead in K/9 and BB/9, but the rest all comes from an unsustainably low HR/FB rate of 7%.
Sale has certainly been better than Kluber, but the difference is not so large that I would say Sale is so much better than the second best pitcher that he also deserves the MVP.
I’s say it’s less “so much better than the 2nd best pitcher” and more “been better than all of the other AL players(either a little or a lot)”
Perhaps the best argument for Sale is the team record when he is not on the mound. Red Sox are a 500 team without him and its not like he has got great run support (esp earlier in the year and all 3 Yankee games)
I’d say he is as deserving as Kersaw in 2014 and Verlander in 2011. Lot of hitters with great stats this year though but thats actually an argument for Sale. He is doing this in a juiced environment like Pedro did in a park known as being tough on LHPer
Well, no, the Sox are not a .500 team without Sale. They are 67-50 and Sale is 14-4 with the team 18-6 overall in games that he starts. So they are over a .500 team.
Furthermore, this analysis applies to lots of pitchers every year. The ROyals are 59-58 and 16-7 when Jason Vargas starts (14-6 actual record). The Dbacks are 65-52 and 16-7 when Grienke starts.
IN fact, the SOx are 15-8 when Pomeranz starts which isn’t much worse than Sale’s 18-6.
I used to tell my friends about how Chris Sale is today’s Randy Johnson (lanky build, high-velo fastball, wipe out slider, pitch delivery), and I always took slack for making the comparison. Then I remember tweeting about Sale about a month ago, and how he’s on pace to outperform Randy Johnson’s 2001 season from a WAR-based perspective. Reading this post makes me a proud cub, to say the very, very least.
Even if Sale loses out to Altuve or Judge or Trout (all of which are well deserving in their own respective rights), it’s impossible to ignore the parallels between the era Johnson pitched in to do what he did that year, and the “juiced” ball 2017 season Sale is practically coasting through.
This is an historic campaign Sale is amassing here.
“how he’s on pace to outperform Randy Johnson’s 2001 season from a WAR-based perspective.”
No. According to prediction models,
he is projected to end between 9.1 WAR (Steamer) and 9.4 WAR (ZiPS).
Even if we simply prorate his outings he would end with 10.2 WAR
which is great but not better than Johnson’s 2001 season (10.4 WAR)
I see Sale as the leader right now. But he does have a pattern of fading down the stretch. In his 5 prior years as a starter:
— Thru team game #117: 57-29, 2.86 ERA, 1.01 WHIP, 0.78 HR/9 (106 starts)
— From team game #118: 13-18, 3.52 ERA, 1.18 WHIP, 1.32 HR/9 (42 starts)
— September: 7-13, 3.86 ERA, 1.24 WHIP, 1.36 HR/9 (27 starts)
The one year he finished strong was 2014, when he missed time early, ending with 174 IP. So, we’ll see.