#TBT: How fast was Bob Feller's fastball?

Bob Feller celebrates his 1946 no-hitter with catcher Frankie Hayes. AP Photo

Pitchers are throwing harder than ever before. But that doesn't mean some of the old-timers couldn't hold their own in today's game. Take Bob Feller, the legendary fireballer of the Cleveland Indians, and owner of maybe the fastest fastball the game has ever seen.

We'll get to that in a minute.

Let's start with this date in history. April 16 was a good day for Feller. In 1940, the Hall of Fame right-hander threw the only Opening Day no-hitter in MLB history. In 1946, he started the season with a three-hit, 10-strikeout shutout, kicking off what he called his greatest season. "That was my best year," he told the Cleveland Plain Dealer in 2006. "I won the All-Star Game in Boston. I pitched a no-hitter in Yankee Stadium. We had a sixth-place ballclub, but Bill Veeck bought the ballclub and made some trades, and in 1948, we won the World Series."

Feller's numbers from that season are mind-boggling from a modern perspective. He went 26-15 with a 2.18 ERA and 348 strikeouts, but that's not even the fun stuff. He started 42 games and completed 36 of them, throwing 10 shutouts. He pitched 371.1 innings. He even saved four games for good measure. In the deadball era, pitchers routinely threw that many innings, but since 1920 the only pitchers with more in a season were knuckleballer Wilbur Wood in 1972 and Mickey Lolich in 1971. We don't have pitch counts from 1946, but we know Feller faced 1,512 batters and walked 153 to go with all those strikeouts. Using a basic pitch count estimator, we guess that Feller threw 5,848 pitches that year. That's 2,118 more than the 3,730 that MLB leader David Price threw in 2014.

Somehow his arm remained attached.

Going through Feller's game logs produces more fascinating numbers. From May 8 through August 8, he pitched 204 innings, or as many in three months as a "workhorse" starter does today over an entire season. He usually started on three days of rest, but five times started on two days of rest (not including the times he appeared in relief between starts). On July 28, he threw a shutout against the Senators, striking out 10. He started again on July 31 against the Red Sox. He pitched a one-hitter -- with nine walks and nine strikeouts. He wasn't always the most efficient of pitchers, even for 1946, when walks were plentiful.

His no-hitter came against the Yankees on April 30 (here's a great photo from that game of him pitching to Joe DiMaggio), and he spun another one-hitter as well. He struck out at least 10 batters 12 times, which sounds even more impressive when you realize all other major league starters combined to do that just 20 times (nine of those by Hal Newhouser). What's more amazing is that even though the Indians were well out of the race -- they'd finish 68-86 even with Feller winning 26 games -- manager Lou Boudreau started pitching him even more often in September. He pitched 79 innings that month and started seven of Cleveland's final 17 games.

Why? Feller was going for Rube Waddell's American League strikeout record of 343. In his book, "Now Pitching, Bob Feller," Feller writes about that chase:

As the strikeouts continued to mount in '46, I stayed with the pitch that brought me there. I had two or three different kinds of curve balls and a lot of confidence in each of them. I was never afraid to throw my curve when I was behind in the count, 2 and 0 or 3 and 1, or even late in a pressure situation.

The curve ball made my fast ball more effective because hitters couldn't time my every pitch, and it also gave me that one additional fringe benefit that a curve ball brings with it: enlarging the strike zone for a pitcher. When you have a good curve ball, it makes your strike zone larger because the hitters swing at pitches that are outside.

But the fast ball was my bread-and-butter pitch. As long as hitters kept missing it, I was going to keep throwing it. On August 13, in a night game at home against the Tigers, I recorded my 262nd strikeout, breaking my career high of 261 in 1940. I reached 261 in 320.1 innings, but I reached 262 in only 260.2 innings.

Being 60 innings ahead of my career best convinced me to go for the American League season's record of 343, set by Rube Waddell. The team wasn't going anywhere anyhow. We finished in the second division, 18 games below .500 in sixth place. Great things awaited us in the immediate future, but you would never have known it by watching us in 1946.

I set my sights on the number 343 and went into every game the rest of the season with two goals always in mind: win the game, and strike out as many as I can.

So Feller started on Sept. 19, Sept. 22, Sept, 25, pitched five innings of relief on Sept. 27 and then started the season finale on Sept. 29, tied with Waddell at 343 strikeouts. Pitching for the Tigers: Newhouser, who had 26 wins. Feller was sitting at 25. Feller won 4-1, striking out five to get his record.

The historical record suggests Feller paid a price for that record chase. He was never as dominant after that. In 1947, his K's per 9 dropped from 8.4 to 5.9 (although that still led the league). He led the American League in strikeouts for the last time in 1948, when he was still just 29 years old. Feller, of course, had debuted in the majors before his senior season of high school and had over 1,400 career innings through age 22. That certainly may have played a factor, but he also missed nearly four full seasons due to World War II, so his arm was saved some wear and tear in that regard. Feller has also written that he slipped from the mound in a June 13, 1947 game against Philadelphia and injured his back, saying his fastball was never the same after that. That sounds plausible, but he was averaging 6.6 K's per nine before that game, well below his 1946 figure. His rate did decline even more the rest of that season, however.

Back to how fast Feller threw. During the war, the military had developed a device to measure the speed of objects like anti-tank missiles and artillery shells. Somebody got the idea to test Feller's velocity. At Griffith Stadium on Aug. 20, the machine was set up and Feller would fire in some fastballs.

Except nobody told Feller about the plan.

From Robert Weintraub's "The Victory Season," his book on the 1946 season:

"Rapid Robert" was not one to be trifled with when it came to cash. Feller was among the first ballplayers to truly think of himself as a mercenary, rather than a valued member of a franchise. He was loyal to the Indians, but only because they paid him. Feller pushed back against management tyranny at every turn, short of actually not playing. ...

So Feller refused to take part, staying under the stands until (Senators owner Clark) Griffith was forced to come down to him and negotiate. A few minutes later, Feller walked out to a loud ovation, $700 richer.

The machine, called the "Lumiline Chronograph," used photoelectric cells to clock the object that passed through the device's opening. Feller's second pitch was the fastest one, clocked at 98.6 mph. Here's a photo of Feller throwing into the machine. Weintraub writes:

However, since the machine measured the speed of the ball as it passed through its sensors, unlike modern radar guns that clock the ball as it leaves the pitcher's hand, it actually flew much harder. Some estimates put the fastball at 101-103 mph, others as high as 107.6 mph.

Feller then started the game (and lost).

I have serious doubts about the 107 mph estimate. For one thing, if Feller threw that hard, or something close to it, I think that would be reflected more in the numbers. He didn't even have the highest strikeout rate in the league in 1946; that belonged to Newhouser, who struck out 8.46 batters per nine innings compared to Feller's 8.44. Now, Newhouser was no slouch, a two-time MVP who also finished second in the voting in 1946 (Feller was sixth). He probably had the second-best fastball of that era. But nobody asked him to throw into the Lumiline Chronograph. Plus, it's hard to know, 70 years later, how precise the Lumiline Chronograph was. On the other hand, it's also possible that Feller didn't throw as hard in 1946 as he did in 1938 or 1939. Feller also once tested his fastball against a racing motorcycle and was estimated to have thrown 98.6 mph that time as well (or 104 mph by some modern estimates).

There seems little doubt that Feller had one of the hardest fastballs of all time. I'll buy that he could throw 100 mph, although it's impossible to know whether he did that consistently within games.

But maybe he did. As Ted Williams once said, "Three days before he pitched I would start thinking about Robert Feller, Bob Feller. I'd sit in my room thinking about him all the time. God I loved it. Allie Reynolds of the Yankees was tough, and I might think about him for 24 hours before a game, but Robert Feller: I'd think about him for three days."