The MLB pitch tracking database goes back to 2008, and from the start of that season through Sept. 3 of this year, we've seen 2,384 different pitchers appear in the big leagues. That's a group that ranges from future Hall of Famer Justin Verlander (2,553 innings) to Zack Weiss, who got into one game for Cincinnati last year without recording an out.

On average, those pitchers have thrown four different pitch types. Verlander, for example, has thrown a four-seam fastball, a slider, a curveball, and a changeup this season. (He's previously, and rarely, experimented with a sinker and a cutter as well.) Clayton Kershaw throws the same four pitches. Max Scherzer does, too, plus a cut fastball.

An overwhelming number of pitchers in the database -- 92% -- have thrown between 3 and 6 pitch types in their careers. Only 99 pitchers, or about 4%, have been tracked as throwing seven or more pitch types. Just 11 have thrown eight or more.

And then there's Yu Darvish, who stands alone at the top of the mountain. Darvish has been tracked with nine different pitch types in 2019, the only pitcher in our database with that many. (Remember: he's thrown that many just this year. No one's even had nine over multiple years. Only three other pitchers, including Zack Greinke this year, have made it to eight in a season.)

But! Even that doesn't even tell the full story, because Darvish is throwing variations of pitches we don't even have classifications for. As confirmed with the Cubs organization, and in some cases by Darvish himself, he's actually thrown 10 different pitches this year. He's unlike anyone we've ever seen -- and, as you may have noticed, over his last 14 starts, he's struck out 105 batters ... and walked just eight. (While he skipped his Sunday start, he's expected to start on Saturday as scheduled.)

(It's worth remembering here that pitch classification is something of a moving target, especially since the label defers to what the pitcher wants it to be called. As an example, Zack Britton throws a breaking ball at 80.7 mph, with 51.7 inches of vertical drop and 11.5 inches of horizontal break. He calls it a slider. Meanwhile, Scherzer throws a breaking ball at 78.3 mph, with 51.7 inches of vertical drop and 10.7 inches of horizontal break. He calls it a curveball. Corey Kluber throws a curve/slider hybrid that's probably just "a breaking ball," but has to be termed as something. This is why different sites may have pitches named differently.)

We've confirmed Darvish's types with the team, and when MLB.com's Jordan Bastian listed out the 10 pitch types in a tweet, Darvish himself re-tweeted it, which we'll take as tacit endorsement.

Here are the 10 different pitches Darvish is throwing this year, roughly ordered from most used to least.

1. CUT FASTBALL (soft version) (32.8%, 86.3 MPH)

See, already this gets difficult. Were you to look at Darvish's Baseball Savant page, which displays Statcast information, you'll see that he throws a cutter 35.9% of the time, most of any pitch. Except, as Bastian noted in August, there's now two cutters.

"In June, Darvish started featuring a pitch that fell between four-seamer and cutter in velocity and movement," Bastian tweeted. "Internally, Cubs call it a hard cutter. He manipulates grip/pressure of cutter to increase velocity. Has better command of it than 4-seamer."

Fine, make this complicated. OK, let's just take them both together ...

2. CUT FASTBALL (hard version) (3.0%, 91.2 MPH)

... because we don't have separate classifications available for "hard cutter" and "soft cutter," so you'll usually find them combined into one pitch, but if you look at the game-by-game velocity on it, it's pretty clear that it's as advertised: Before June, the cutter rarely went above 90 mph, but starting on June 5 against the Rockies, now it does.

We have somewhat arbitrarily split the two at 90 mph for simplicity's sake -- Bastian astutely suggested the "Darpitch" as a better name than "hard cutter" -- and first-year Cubs pitching coach Tommy Hottovy explained how it works.

¡°You can call it what you want,¡± Hottovy told Bastian. ¡°It¡¯s him adjusting the cutter grip or it¡¯s him adjusting the fastball grip.

"If you have a fastball and you throw it like 96 or 97 and you hit one -- you feel like you threw it exactly where you want -- and then you throw two that are close, but not exactly there ¡­ you can have a 93¨C94 mph pitch that has a little late movement like a heater and you can throw it exactly to a location you want to throw it, too, that¡¯s going to be a pitch that he has more confidence throwing in certain situations. And to certain lefties that we need a ball in a certain area.¡±

When Darvish throws the hard cutter, 76.2% of the time it's been to a lefty hitter, so that tracks. Combining the two together, Darvish's cutter has allowed just a .183 average and a .324 slugging -- overall, that's the third-best cutter of any pitcher with 100 plate appearances ending on a cutter. No wonder he's inventing new versions of it.

There's still eight more pitches here.

3. FOUR-SEAM FASTBALL (27.9%, 94.0 MPH)

The cheese, the "high stinky cheddar," as it were. Almost every pitcher, more than 93% of them in our decade-plus sample, has a four-seam fastball, and Darvish is no different. This is mostly his get-ahead pitch, because he throws it on the first pitch 34.6% of the time, his most-used 0-0 offering.

The velocity (94.0 mph) is somewhat above-average for a starter (92.9 mph), and the spin rate (2520 RPM) is top-10 elite (average is 2274 RPM), but that hasn't translated to particularly notable movement or outcomes. The pitch has about one inch more rise than other four-seamers at his velocity and release point, and it's relatively straight, with 4.2 inches less break than average.

Despite throwing it just under 28% of the time, nearly 39% of his 31 home runs allowed have come off the four-seamer, and the .622 slugging on it is 10th-highest among starters. That this one exists serves mostly to set up the breaking pitches that come along with it; as far as pitches go, this is definitely one of the ones that Darvish throws.

4. SLIDER (14.1%, 82.5 MPH)

One man's cutter is another man's slider, in the sense that so often there's little difference in these pitches beyond what the pitcher chooses to call them. In Darvish's case, we can argue about the labels, but this is definitely a different pitch. The slider is slower than either of his cutters, at 82.5 mph, and it comes with far more horizontal break, moving 16.7 inches as opposed to the 5.8 inches of his cutter(s).

In fact, that break is nearly nine full inches more than the average slider at Darvish's velocity, and that puts him in the top five (minimum 250 thrown), along with some pretty legendary breakers.

Most horizontal break above average, sliders

+15.3 inches vs. average -- Chaz Roe, TB
+10.5 inches vs. average -- Sonny Gray, CIN
+10.4 inches vs. average -- Kyle Crick, PIT
+9.9 inches vs. average -- Adam Ottavino, NYY
+8.9 inches vs. average -- Darvish, CHC
+8.4 inches vs. average -- Trevor Bauer, CIN

Darvish gets a swing-and-miss or a called strike on 38.6% of his sliders, sixth-best among starters, a list topped by Scherzer.

5. TWO-SEAM FASTBALL (12.4%, 93.5 MPH)

Darvish has had this one in his arsenal for years, but he didn't actually throw it in his first three starts of the season, until breaking it out on April 15 in Miami.

"I remember during the game [our staff] brought it up and then it showed up," Cubs manager Joe Maddon said at the time. "Yu looked right in the dugout and [he and the staff] smiled at each other because it was pretty successful."

You probably remember his final two-seamer of the day, because he managed to hit the batter, catcher and umpire on a single pitch.

Later, he used it as a way to deal with four-seam inconsistency -- "I'm still struggling with my four-seam to throw strikes ... but the sinker feels better," Darvish said on May 5, a month where he threw it a season-high 23.8% of the time -- though it was used just 6.4% of the time in August.

6. CURVEBALL (4.8%, 76.2 MPH)

This is, as you'll see, just one of three different Darvish curveballs. This is completely unfair.

In fact, let's just group these all together...

7. KNUCKLE-CURVE (0.5%, 81.3 MPH)

8. SLOW CURVE (0.1%, 63.7 MPH)

... and wow, that's a lot of curveballs. As you can see, most of this is about speed, because his primary curve is five miles slower than his knuckle-curve and nearly 13 miles faster than the "slow" curve.

The regular curve is solid enough, though it's never been a primary pitch of Darvish's; he's never thrown it more than 9% or less than 2% in a season. It gets about two inches more drop than other curves at a similar velocity and release point, and it often looks pretty, but it's rarely been a notable weapon of Darvish's.

But the knuckle-curve is perhaps more interesting because it almost literally came out of nowhere. Darvish watched Craig Kimbrel throw a knuckle-curve in mid-August, decided he liked it, brought into a game a mere week later ... and it was good. Who does that?

"He walked up to me the other day and he was like, 'Hey, I've been working on that,'" Kimbrel said on Aug. 28. "I was like, 'Cool.' I didn't know he was going to go out and throw it. I thought that was pretty cool."

"Cool" somewhat undersells it, as manager Joe Maddon accurately summed up.

"It doesn't happen," Maddon said. "As a Major League pitcher, I know they experiment with stuff all the time, but to take a pitch like that that requires a lot of feel, from a teammate, and just bring it into a game that quickly and effectively, that doesn't happen."

The knuckle-curve is thrown harder than the regular curve, but it's only been seen 12 times so far, all in starts Aug. 21 and 27. One of those did turn into an opposite-field Mike Yastrzemski home run, but the other 11 have all either been strikes, fouls or in-play outs.

And then there's the slow curve, also sometimes called an eephus, which he'll break out on occasion, as if doing so just for fun. He does this from time to time; you might remember what he did to poor Torii Hunter with it back in 2014. This one doesn't come out a lot, just 15 times in Darvish's career, and only twice so far in 2019, but it's there. It's always there. It drops 84 inches. That's seven feet. Good luck.

But wait! There's still two more pitches -- and we're not talking about the fact that he can throw his slider and curveball left-handed. (He can.)

9. SPLIT-FINGER FASTBALL (3.7%, 88.6 MPH)

It's relatively rare for a pitcher to throw both a splitter and a changeup, because those are two pitches that share many of the same characteristics. But we know he has a splitter, because he's said as much, and we know that he's got a somewhat slower version that's a changeup.

The split-finger is interesting, because it essentially disappeared after his 2015 Tommy John surgery. (As the Dallas Morning News detailed in 2017, Darvish attempted to bring it back in Spring Training last year, but threw it just six times.) Last year, he barely threw it at all, and through the end of July this year, he'd thrown only 44, or about 3%.

But in August, that went up to 10%, the highest splitter month of his career. As The Athletic reported, that's due to changes in grip and confidence; coach Mike Borzello compared it to that of Roger Clemens.

This one's still a work in progress, but there may be something to it. We have 91 of them in the database for this year, and Darvish has allowed three hits, racking up 14 strikeouts.

10. CHANGEUP (0.6%, 84.1 MPH)

We don't see this one often, but it's there, a few miles slower than the splitter, almost like it's a changeup to a pitch that's already like a changeup. Obviously, Atlanta's Ozzie Albies didn't have much of a chance against the one shown in the video above. But since he barely throws this pitch, what if we took a lesson from the knuckle-curve? We already saw Darvish eyeball Kimbrel and come away with a lesson. What if next, he looked at Cubs reliever Brandon Kintzler, who throws the sixth-most effective reliever changeup over the last two seasons? Darvish probably doesn't need any help or new pitches, sure. But why stop now?

Break out the forkball. See about a screwball. Maybe we'll be so lucky that Darvish will turn out to be the savior of the knuckleball someday. After all, there aren't pitches that Darvish can't throw. There are just pitches he hasn't thrown yet.